jea cover (i).

qxd 30/07/2012 13:12 Page 1

ISSN 0957-8234
Volume 50 Number 5 2012

Celebrating 50 years 1963-2012

Journal of

Educational
Administration
Accountability and school
leadership
Guest Editor: Karen Seashore Louis

www.emeraldinsight.com
Journal of Educational ISSN 0957-8234

Volume 50
Administration Number 5
2012

Accountability and school leadership

Guest Editor
Karen Seashore Louis

Access this journal online __________________________ 534

CONTENTS
Editorial advisory board ___________________________ 535
Guest editorial _____________________________________ 536
School administration in a changing
education sector: the US experience
James P. Spillane and Allison W. Kenney _____________________________
_ 541

At-risk student averse: risk management

and accountability
Julian Vasquez Heilig, Michelle Young and Amy Williams _______________ 562

Contrasting effects of instructional leadership

practices on student learning in a high
accountability context
Moosung Lee, Allan Walker and Yuk Ling Chui_______________________
_ 586

The strength of accountability and teachers’

organisational citizenship behaviour
Eyvind Elstad, Knut-Andreas Christophersen and Are Turmo ____________ 612

External mandates and instructional

leadership: school leaders as mediating agents
Karen Seashore Louis and Viviane M. Robinson _______________________ 629

Managing the intersection of internal and external

accountability: challenge for urban school leadership
in the United States
Michael S. Knapp and Susan B. Feldman _ ___________________________ 666

Bridging accountability obligations, professional

values and (perceived) student needs with integrity
This journal is a member of and
Heinrich Mintrop________________________________________________
_ 695 subscribes to the principles of the
Committee on Publication Ethics
www.emeraldinsight.com/jea.htm

As a subscriber to this journal, you can benefit from instant, Structured abstracts
electronic access to this title via Emerald Management eJournals. Emerald structured abstracts provide consistent, clear and
Your access includes a variety of features that increase the value of informative summaries of the content of the articles, allowing faster
your journal subscription. evaluation of papers.

How to access this journal electronically Additional complementary services available

Our liberal institution-wide licence allows everyone within your When you register your journal subscription online you will gain
institution to access your journal electronically, making your access to additional resources for Authors and Librarians, offering
subscription more cost-effective. Our web site has been designed to key information and support to subscribers. In addition, our
provide you with a comprehensive, simple system that needs only dedicated Research, Teaching and Learning Zones provide
minimum administration. Access is available via IP authentication or specialist ‘‘How to guides’’, case studies, book reviews,
username and password. management interviews and key readings.

To benefit from electronic access to this journal, please contact E-mail alert services
[email protected] A set of login details will then These services allow you to be kept up to date with the latest
be provided to you. Should you wish to access via IP, please additions to the journal via e-mail, as soon as new material enters
provide these details in your e-mail. Once registration is completed, the database. Further information about the services available can
your institution will have instant access to all articles through the be found at www.emeraldinsight.com/alerts
journal’s Table of Contents page at www.emeraldinsight.com/
0957-8234.htm More information about the journal is also available Emerald Research Connections
at www.emeraldinsight.com/jea.htm An online meeting place for the world-wide research community,
offering an opportunity for researchers to present their own work and
Emerald online training services find others to participate in future projects, or simply share ideas.
Visit www.emeraldinsight.com/help and take an Emerald online Register yourself or search our database of researchers at
tour to help you get the most from your subscription. www.emeraldinsight.com/connections

Key features of Emerald electronic journals Choice of access

Electronic access to this journal is available via a number of
Automatic permission to make up to 25 copies of individual channels. Our web site www.emeraldinsight.com is the
articles recommended means of electronic access, as it provides fully
This facility can be used for training purposes, course notes, searchable and value added access to the complete content of the
seminars etc. within the institution only. This only applies to articles journal. However, you can also access and search the article
of which Emerald owns copyright. For further details visit content of this journal through the following journal delivery
www.emeraldinsight.com/copyright services:
EBSCOHost Electronic Journals Service
Online publishing and archiving
ejournals.ebsco.com
As well as current volumes of the journal, you can also gain access
to past volumes on the internet via Emerald Management Informatics J-Gate
eJournals. You can browse or search these databases for relevant www.j-gate.informindia.co.in
articles. Ingenta
www.ingenta.com
Key readings Minerva Electronic Online Services
This feature provides abstracts of related articles chosen by the www.minerva.at
journal editor, selected to provide readers with current awareness of OCLC FirstSearch
interesting articles from other publications in the field. www.oclc.org/firstsearch
Non-article content SilverLinker
Material in our journals such as product information, industry trends, www.ovid.com
company news, conferences, etc. is available online and can be SwetsWise
accessed by users. www.swetswise.com
Reference linking
Direct links from the journal article references to abstracts of the Emerald Customer Support
most influential articles cited. Where possible, this link is to the full For customer support and technical help contact:
text of the article. E-mail [email protected]
Web http://info.emeraldinsight.com/products/
E-mail an article subs/customercharter.htm
Allows users to e-mail links only to relevant and interesting articles to Tel +44 (0) 1274 785278
another computer for later use, reference or printing purposes. Fax +44 (0) 1274 785201
EDITORIAL ADVISORY BOARD Editorial
Ken Avenell Ken Leithwood advisory board
Brisbane Catholic Education Centre, Australia University of Toronto, Canada
Ibrahim Bajunid John McCormick
Inti Laureate University, Malaysia University of Wollongong, Australia
Les Bell Joseph Murphy
University of Leicester, UK Vanderbilt University, USA 535
Narottam Bhindi Anthony H. Normore
University of Wollongong, Australia California Lutheran University, USA
Susan Bon Wendy Pan
George Mason University, USA National Taiwan Normal University, Republic of
Carol Cardno China
UNITEC Institute of Technology, New Zealand Keith F. Punch
Judith D. Chapman University of Western Australia, Perth, Australia
Australian Catholic University, Sydney, Australia Brian Roberts
Pam Christie University of Hull, UK and Shenyang Normal
University of Cape Town, South Africa University, PR China
Neil Cranston Viviane Robinson
University of Tasmania, Australia University of Auckland, New Zealand
Larry Cuban Zehava Rosenblatt
Stanford University, USA University of Haifa, Israel
Clive Dimmock Pamela Sammons
NIE-Nanyang Tech University, Singapore University of Oxford, UK
Stephen Dinham Karen Seashore Louis
University of Melbourne, Australia University of Minnesota, USA
Deirdre J. Duncan Peter Sleegers
Australian Catholic University, Sydney, Australia University of Twente, The Netherlands
Patrick B. Forsyth Jim Spillane
University of Oklahoma, USA Northwestern University, USA
Frances Fowler Jacqueline A. Stefkovich
Miami University, USA Pennsylvania State University, USA
Ellen Goldring Cynthia Uline
Vanderbilt University, USA San Diego State University, USA
Peter Gronn T. Velayutham
Cambridge University, UK University of the South Pacific, Fiji
Alma Harris Wang Hong
Institute of Education, University of London, UK South China Normal University, China
Ron Heck Charles F. Webber
University of Hawaii-Manoa, USA Thompson Rivers University, Canada
Olof Johansson Yang Xiao Wei
University of Umea, Sweden East China Normal University, China
Gabriele Lakomski
University of Melbourne, Australia

Journal of Educational
Administration
Vol. 50 No. 5, 2012
p. 535
# Emerald Group Publishing Limited
0957-8234
JEA
50,5 Guest editorial
Special issue on accountability and school leadership
This special issue consists of a collection of seven papers that deal with the increasing
536 pressures for school accountability and its effects. The special issue grew out of
a roundtable at the American Educational Research Association in 2011, in which three
of the paper authors (Mintrop; Knapp and Feldman; and Louis and Robinson)
interacted with approximately 40 participants whose enthusiasm for the topic required
us to move into a hall and steal chairs from less well-attended sessions. When we raised
the possibility of a special issue to the editors of the Journal of Educational
Administration, they noted that another paper on the same topic (Spillane and Kenney)
had been presented at the 2nd Asian Leadership Roundtable in Bangkok, and that they
had recently accepted two related papers (Elstad, Christophersen and Turmo; and
Vasquez Heilig, Young and Williams). As we looked at the accumulating manuscripts,
the editors and I felt that it was insufficiently international in its focus, and we solicited
an additional manuscript (Lee, Walker and Yuk). The final line-up thus consists of
an overview essay, four empirical papers from USA, and two from other countries.
Three of these papers are multi-method, two are predominantly survey-based and one
is based entirely on case study data.
Although the papers do not reflect an integrated framework, together they address
critical but unanswered questions:
(1) How do external accountability policies affect the development of school-based
initiatives for change and improvement?
(2) How do leaders influence teacher responses to external accountabilities?
(3) How do schools develop consensus around norms and values that sustain their
efforts at improvement under conditions of increasing external pressure?
(4) In what ways (if at all) do school leaders reconcile external accountability
demands with the school’s internal accountability system and improvement
goals?
(5) What are some of the intended and unintended consequences of the growing
emphasis on accountability for professional behaviors and the achievement of
educational goals?
All of the papers examine both formal leaders (administrators) and the social and
professional context in which they work. In particular, the emphasis is on the role of
school leaders in interpreting external mandates, including the extent to which they
are able to integrate them with their internal priorities and values in ways that affect
teachers and students. Given the investment in standards-based accountability-driven
reform (and the public faith placed in this strategy for improving the quality of public
education) the field deserves better answers to questions about the school-level
responses to these policies. The collection thus adds insights into a central set of
Journal of Educational
Administration
variables that affect the impact of local leadership on schools’ responses to emerging
Vol. 50 No. 5, 2012 policy definitions of the common good.
pp. 536-540
r Emerald Group Publishing Limited
Each paper focusses on tensions or dilemmas in the accountability movement.
0957-8234 Spillane and Kenney set the stage, and address all the questions posed above through a
review of the literature. Most of the research on the effects of the accountability Guest editorial
movement has, until recently, emphasized its impact on classroom practices,
instruction and student achievement, but they point out that “While these foci make
sense, they often ignore other aspects of the school organization, potentially critical
to understanding the implementation process of this new genre of education policy”
(p. 543). Spillane and Kenney’s essay emphasizes the need to examine how
accountability has changed the dynamic of how schools must manage their need to 537
maintain external legitimacy in the eyes of the public (and particularly in the political
arena), while at the same time maintaining an internal focus on integrity and value-
driven practices.
This issue of how the tensions implicit in the questions outlined above are perceived
and managed emerges in the six empirical papers as well:
. Vasquez Heilig, Young and Williams (“At-risk student averse: risk management
and accountability”) use a “risk management” framework to analyze how
administrators and teachers in low-performing and high-poverty high schools in
Texas respond to several decades of high-stakes accountability. The authors
suggest that the risk of “failure” on the state tests were real because they
undermined legitimacy, and risk is managed by explicitly undermining
perceived integrity objectives (teaching all students well, providing support for
immigrant students who may not be “legal”) in order to “game” the system and
increase test performance.
. Lee, Walker and Yuk (“Contrasting effects of instructional leadership practices
on student learning in a high accountability context”) show that when school
leaders in Hong Kong do what is asked (closely monitor the classroom work of
teachers), their role as positive instructional leaders is undermined. In addition,
the use of close supervision of teachers (which is increasingly emphasized as a
policy in a number of countries) has a negative impact on students’ engagement
with school, presumably because teachers own engagement and sense of
professionalism is undermined.
. Elstad, Christophersen and Turmo’s investigation of three different
accountability systems in Norway (“The strength of accountability and
teachers’ organizational citizenship behaviour”) suggests a different problem
in managing external legitimacy and internal integrity. In this setting, adults in
school settings are expected to engage in a wide variety of citizenship behaviors,
which are largely voluntary efforts directed toward improving the culture and
performance of the school. Their findings suggest that there is a positive
relationship between higher-stakes accountability and stronger leadership
among the formal leaders/higher levels of organizational citizenship behavior.
Thus, their finding is that external legitimacy/accountability and internal
integrity may be, at least in their setting, less problematic than either the Hong
Kong or Texas data suggest.
. Louis and Robinson’s analysis of US elementary schools (“External mandates
and instructional leadership: principals as mediating agents”) focusses on the
way in which accountability may affect the legitimacy of the formal school
leader. Their data suggest that administrators who have more questions
about the legitimacy and value of external accountability initiatives from either
the state or district are less likely to be viewed as instructional leaders by their staff.
JEA High instructional leadership, in their qualitative data, is associated with the capacity
50,5 of school leaders to negotiate perceived tensions between external legitimacy
demands and the need for internal autonomy in ways that balance the two.
. Knapp and Feldman (“Managing the intersection of internal and external
accountability: challenge for urban school leadership in the United States”) also
examine the role of principals in mediating policy messages and integrating
538 them into an internal school agenda. Their analysis of 15 urban schools suggests
that principals in these settings were generally successful in creating an
environment that focusses on professional responsibilities while establishing
structures that reflected external accountability demands. Principals and staff
members created more reciprocal school cultures and mutual accountability that
seemed to increase both leaders’ and staff capacities to pursue high expectations.
. The issue of legitimacy and integrity is central to Mintrop’s contribution (“Bridging
accountability obligations, professional values, and (perceived) student needs with
integrity”). Using in-depth data from nine California schools, he is able to show that
where goal integrity is high, responsiveness to external accountability (and success
in meeting accountability standards) was also higher. In schools where principals
operated as conduits of accountability pressures without integrative narratives that
included integrity, defensiveness and an adversarial climate tended to ensue.

While these papers emerge from studies in different accountability contexts (even
within the USA there is considerable variation in accountability policies between
states), there are thus, common themes in response to the questions outlined above.
First, external accountability policies have an effect on internal school leadership in
all of the contexts, but have both positive and negative effects on the development of
a coherent internal story about improvement. Second, where positive effects on internal
cohesiveness and a focus on improving outcomes for students are observed, there is
evidence of an active role for school leaders in creating coherence. Third, but less
conclusively, the papers point to a variety of ways in which leadership effects occur,
but in general they point to coherence around goals, a sense of an internal “story” about
school improvement, and integrity in addressing both external demands and internal
conditions. The story of how this is done differs between the papers, suggesting an
additional need to study how leadership, both from individuals and that which is more
broadly distributed among the professional staff, contribute to managing the tension
outlined in Spillane and Kenney’s initial paper.
One additional cross-cutting theme in the papers is the degree to which external
accountability policies have both the anticipated effect of increasing student learning
and unintended effects of various kinds. The issue of unintended effects is raised most
explicitly by Vasquez Heilig, Young and Williams (At-risk student averse: risk
management and accountability), who point to many unintended negative effects of the
Texas accountability system on schools with large numbers of at-risk older adolescent
students. However, scattered less explicitly through the other empirical papers is
evidence that external accountability policies have a wide variety of unanticipated
effects, some that are harmful to students or the school’s internal adult culture, but
others that have unanticipated positive effects. Without intervention by school leaders,
the papers suggest that the anticipated positive consequences may be less likely to
occur (or occur only in some places), while the opportunities for unintended negative
consequences to emerge are increased. Table I is not intended to be comprehensive, but
 “Debatable,” variable or
Guest editorial
“Positive” effects mixed effects “Negative” effects

Unintended Increase in district Focus on “bubble kids” Loopholes/“gaming”

cohesiveness and reform (Vasquez Heilig, Young and inauthentic compliance
(Louis and Robinson) Williams) vs educational (Vasquez Heilig, Young and
Increase in OCB (Elstad, “triage” and pluralism Williams; Louis and 539
Christopherson and (Mintrop) Robinson)
Turmo) Increasing system Teachers view students as
Focus on internal complexity (Vasquez Heilig) liabilities (Vasquez Heilig,
coherence within the vs effort to simplify Young and Williams)
school (Knapp and “Fearfulness” and
inconsistent with the reality
Feldman) of children’s needs unwillingness to take risks
(Mintrop) (Vasquez Heilig, Young and
Moral purposes are Williams; Louis and
challenged (Vasquez Heilig, Robinson; Mintrop)
Young and Williams) vs Disruption of student-
supported (Mintrop) teacher relationship (Lee,
Walker and Yuk; Mintrop)
Intended Focussing teachers on “High reliability “Generally an empty cell
learning; internal achievement” vs risk [y]. policies are not
coherence within management (Vasquez typically intended to make
schools (Mintrop) Heilig, Young and Williams) schools worse”
Increasing collective Weak distribution of
responsibility, especially pressure – focussed on low-
in “high risk” schools performing schools
(Knapp and Feldman) (Vasquez Heilig, Young and
Increasing leader focus Williams) or low-
on core school activities performing teachers
and functions (Lee, Walker and Yuk)
Making inequitable Table I.
practices and outcomes Effects of accountability
visible creates moral policies (as implemented)
dissonance (Mintrop) on leadership

only to draw the reader’s attention to the fact that the empirical papers consistently pay
attention to the issue of intended and unintended policy consequences. A careful read
will find additional implications for the effects of policy on leadership and schools
which, collectively, point to the need for more information about the types, range and
incidence of policy effects.
In sum, this special issue sheds some light on the often debated but still ambiguous
question of how school-based leaders affect the implementation of accountability
systems “on the ground.” Even though accountability systems are intended to issue
clear, simple, authoritative and incontrovertible performance demands for schools, the
studies show that local leaders’ interpretations of these demands strongly shape
schools’ internal responses. The alternatives mechanisms explored in these papers
include:
. perceiving the system as enabling or, alternatively, constraining;
. exploring vs ignoring contradictions between professional values and external
equity expectations; and
JEA . forging a sense of value integrity that merges internal standards of good practice
50,5 with external system tools (e.g. use of performance data).
While this special issue, like most, raises as many questions as it answers, as a group
the papers shed considerable light on these issues, particularly on how schools are able
to manage a tension between internal and external accountabilities that continue to
trouble the profession.
540
Karen Seashore Louis
Guest Editor
 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0957-8234.htm

School
School administration in a administration
changing education sector:
the US experience
James P. Spillane and Allison W. Kenney 541
School of Education and Social Policy, Northwestern University, Evanston,
Received 25 October 2011
Illinois, USA Revised 21 February 2012
Accepted 18 April 2012
Abstract
Purpose – Research, spanning half a century, points to the critical role of school administration and to
the successful implementation of US government policies and programs. In part these findings reflect
the times and a US educational governance system characterized by local control, a constitutionally-
constrained federal government, resource-poor state governments, and an overall system of segment
arrangements for governing education. However, the US education policy environment has changed
dramatically over the past several decades, with standards and high stakes accountability becoming
commonplace. The purpose of this paper is to examine the entailments of shifts in the policy
environment for school administrative practice, focusing on how school leaders manage in the middle
between this shifting external policy environment and classroom teachers.
Design/methodology/approach – The paper’s focus is on how school administration manages the
dual organizational imperatives of legitimacy and integrity in a changing institutional environment.
This paper is an essay in which the authors reflect on the entailments of shifts in the education sector
for school administration over the past quarter century in the USA.
Findings – While considerable change for school administrative practice is suggested, the authors
argue that organizational legitimacy and organizational integrity are still central concerns for school
leaders.
Originality/value – Although the paper’s account is based entirely on the US education sector,
several aspects of the framing may be relevant in other countries.
Keywords United States of America, Educational administration, Schools, Government policy,
Leadership, Administration
Paper type Research paper

Over several decades, local, state, and federal policy makers in the USA have directed
their attention and policy initiatives on classroom teaching, specifying what teachers
should teach, in some cases how they should teach, and acceptable levels of student
achievement. They have done so by mobilizing policy instruments – rewards and
sanctions – for compliance with externally imposed performance standards. As a
result of the dramatic change in the institutional environment of US schools over the
last 25 years, curriculum standards and test-based accountability have become staples,
perhaps even taken for granted, in the educational sector. Policy makers are not the
only ones implicated in this transformation. Extra-system agents and agencies
(e.g. comprehensive school reform designs, charter school networks, philanthropic
institutions) have also played a prominent role, albeit with government support and
incentives, in transforming the American education sector. These shifts in the
institutional environment of America’s schools represent a considerable departure for Journal of Educational
business as usual inside schools. Administration
Vol. 50 No. 5, 2012
Though commentators often associated the transformation with the federal pp. 541-561
r Emerald Group Publishing Limited
“No Child Left Behind” (No Child Left Behind Act of 2001 (NCLB), 2001) legislation, 0957-8234
these institutional shifts pre-date NCLB, as several state and local governments DOI 10.1108/09578231211249817
JEA introduced standards and accountability mechanisms prior to NCLB. Careful,
50,5 empirical, analysis suggests that the press for standardization and accountability in
US education dates back at least to 1983, and more than likely earlier, with the
publication of “A Nation at Risk” (Mehta, revise and resubmit, under review; National
Commission on Excellence in Education, 1983). These shifts in the education sector in
the USA are not historically novel, nor are they unique to the education sector. As Jal
542 Mehta points out, there were two other periods of rationalization efforts in the USA,
one in the early 1900s and again in the late 1960s and early 1970s (Mehta, revise and
resubmit, under review). The emergence of minimum competency testing in several US
states in the 1970s might be seen as a precursor for the standards and accountability
movement at the local, state, and federal levels in the 1980s and 1990s (Fuhrman and
Elmore, 2004; Pipho, 1978). Federal or national policy making in the USA often builds
on, extends, and galvanizes local and state policy making initiatives (Fuhrman and
Elmore, 2004). Further, these institutional shifts are not unique to education, more
broadly reflecting the emergence of an “audit culture” across institutional sectors in the
USA and indeed globally (Strathern, 2000, p. 2). In fields from health care to human
service and higher education, we see a press for standardization, efficiency, and
accountability (Colyvas, 2012; Espeland and Sauder, 2007; Power, 1994).
Though our paper focusses on the USA, the shifts in the educational sector we
describe are not unique to the USA – these are global trends. For a quarter century,
educational reform initiatives have spanned national boundaries as several countries,
despite different political arrangements, borrow reform ideas from one another (Ball,
1999; Davies and Guppy, 1997; Whitty and Power, 2003). Some combination of
standards, high-stakes accountability, and school performance metrics based on
student achievement can be found in education policy making and more broadly in
educational reform discourses in several countries spanning several continents over
the past several decades. These reform themes and policy levers are part of policy
discourses, and policy texts, that are transnational. In Singapore, for example, an
accountability system implemented in the 1990s uses national rankings and rewards
for high-performing schools (Ng, 2010). Since the Education Reform Act of 1988, school
accountability based on student performance has been part of the education system
in the UK enabling cross-school comparisons through “league tables” (Burgess et al.,
2010; Ranson, 1994; Tomlinson, 2001). In New Zealand, the early 1990s saw the
emergence of national standards for school practice, curriculum content, student
examination, and teacher qualification, as well as the publication of national “league
tables” similar to those in the UK. These national standards were accompanied with
the creation of new agencies to monitor performance and compliance and to grant
accreditation to compliant institutions (Broadbent et al., 1999). Policy initiatives in
New Zealand also forced primary schools to create Boards of Trustees consisting
mainly of elected parent representatives to monitor student progress against the
national curriculum, though standards and targets were not widely used for student
evaluation (Robinson and Timperley, 2000).
While the press for standardization, performance metrics, and accountability in the
education sector can differ in terms of form, focus, and function between countries,
there are many similarities. Indeed, supra-national organizations, such as Organization
for Economic Cooperation and Development (OECD) and its PISA project, promote
such standardization and the use of performance metrics tied to external tests (OECD,
2004). Though organizational and governance arrangements differ between countries,
as do the broader societal culture and social arrangements in which schools operate,
careful cross-country comparisons can inform how school administration operates in a School
radically changing institutional sector. We leave the comparative work in this paper to administration
the reader, though we do offer a particular framing to guide and focus that work with
respect to relations among school administration and the institutional environment.
With respect to the USA, the evidence suggests that these shifts in the educational
sector, especially in government policy, increasingly make it beyond the schoolhouse
door and even inside classrooms (Au, 2007; Clotfelter and Ladd, 1996; Herman, 2004; 543
Mintrop and Sunderman, 2009; Valli and Buese, 2007). Research suggests, among other
things, that these educational policy pressures influence what teachers teach – thereby
marginalizing low-stakes subjects, diverting resources to students based on their
likelihood of passing the test, and increasing the time devoted to teaching test-taking
skills (Amrien and Berliner, 2002; Booher-Jennings, 2006; Darling-Hammond, 2004;
Diamond and Spillane, 2004; Firestone et al., 1998; Jacob, 2005; McNeil, 2002; Nichols
and Berliner, 2007; Smith, 1998; Valenzuela, 2004; Wilson and Floden, 2001). At the
same time, there is some evidence that high-stakes testing has increased student
achievement, though variation between states is tremendous and the evidence with
respect to narrowing the achievement gap is weak (Jacob, 2005; Lee, 2007; Mintrop and
Sunderman, 2009; Neal and Schanzenbach, 2007; Wong et al., 2009).
Much of the research attention has focussed on policy effects, typically student
learning outcomes, as measured by standardized tests. There is also a growing
literature on how this shifting policy environment is influencing, for worse and for
better, classroom instruction. While these foci make sense, they often ignore other
aspects of the school organization, potentially critical to understanding the
implementation process of this new genre of education policy. In this paper we focus
on one such aspect – school administration. For a half century, research on policy
implementation has consistently identified the critical role of school-level leadership in
the successful implementation of externally and internally initiated policies (Berman
and McLaughlin, 1977; McLaughlin, 1990). There is good reason then to consider
school administration in this shifting policy environment. Our focus in this manuscript
is on how school administration manages the dual organizational imperatives of
legitimacy and integrity in a changing institutional environment.
This paper is an essay in which we reflect on the entailments of shifts in the
education sector for school administration over the past quarter century in the USA.
While we draw selectively on the extant literature and use examples from empirical
work, including our own research, to develop our argument, the paper is neither a
literature review nor a report on the findings from an empirical study. Our essay is
organized as follows: we begin with a retrospective, briefly and broadly considering
how things once were by focussing on popular portrayals in the research literature of
school administrative practice. By administrative practice we mean more than the
school principal’s work; though, consistent with several decades of research, we afford
the principal a prominent place in school administration. Next, we consider the shifting
policy discourses and policy texts in the USA over the past several decades identifying
several central tendencies. We then consider the entailments of these shifts in the
policy environment for school administrative practice. Specifically, we examine how
school administration manages in a shifting US policy environment – how it manages
external policy pressures that increasingly target classroom instruction. Exploring
school administrative practice in a shifting policy environment, we look at how school
leaders’ respond in their day-to-day work. Getting inside the black box of the
schoolhouse to look at school administrative practice up-close, we uncover how school
JEA leaders manage in the middle between external policy and classroom teachers as they
50,5 work to increase cooperation with external policy. We conclude by pondering changes
in school administrative practice in response to a changing institutional environment
and suggesting some directions for cross-national work.

Educational policy and school administrative practice: retrospective

544 Most consumers of the US education literature will be familiar with popular portrayals
of the role of school administration in education policy implementation, where school
administration is mostly though not always equated with the work of the school
principal. The literature often depicts a system in which policy, school administration,
and classroom instruction are loosely coupled or decoupled from one another on
matters of the core technical work – instruction. School administrators, for example,
are depicted as responding to environmental pressures by making symbolic or
ceremonial changes to their schools’ formal organizational structure, preserving the
organization’s legitimacy by conforming to institutional pressures, but avoiding any
close internal coordination or external scrutiny of classroom instruction. Classroom
instruction is portrayed as loosely coupled or decoupled from both the institutional
environment (e.g. government policy) and from the school’s administrative structure
(Deal and Celotti, 1980; Firestone, 1985; Gamoran and Dreeben, 1986; Malen and
Ogawa, 1988; Malen et al., 1990). In this way, institutional conformity can take
precedence over technical efficiency as schools strive for legitimacy and resources from
their institutional environment. Consistent with these decoupled or loosely coupled
portrayals, scholars also present the public schoolhouse, based on empirical research,
as an “egg carton structure” where teachers practice mainly as isolates (Lortie, 1975).
School leaders’ work, despite their best intentions to focus on instruction, is shaped by
a managerial imperative constraining their time on instructional matters (Cuban,
1988)[1].
Some more recent empirical literature offers another image of how schools might
organize, portraying the school as a “professional community” in which teachers
engage in instructionally focussed conversations, collaborate to develop and refine
collective norms of work practice, and where classroom practice is de-privatized (Bryk
and Schneider, 1996; Lewicki and Bunker, 1996; Louis et al., 1995; McLaughlin and
Talbert, 2001; Mishra, 1996). Related work points to the critical role of school leaders
as instructional leaders in bringing about improvement in instruction (Bullard and
Taylor, 1994; Darling-Hammond and Wise, 1985; Eubanks and Levine, 1983; Hallinger
and Murphy, 1986; Purkey and Smith, 1985). While the available evidence suggested
that “strong” professional communities were the exception rather than the norm, and
that instructional leadership was weak at best, among American public schools, this
literature did offer an alternative image of school administration and its relation to
classroom instruction.

Changing policy discourses and texts

Framing policy as both “text” and “discourse” assists with thinking analytically about
the term (Ball, 1994, 2006, p. 44). When identified as text, policy involves both the
policy makers’ encoding of representations of ideas and the actors across the system
then decoding these representations (Ball, 1994; Coburn, 2001; Cohen, 1990; Cohen and
Weiss, 1993; Spillane, 2006). Policy discourses create the frameworks in which policy
texts are situated. Based on the work of Foucault, Stephen Ball (2006) argues that
policy discourses “produce frameworks of sense and obviousness with which policy is
thought, talked, and written about” (p. 44). Thus, policy discourses are systems of School
practice, beliefs, and values outlining what is acceptable, “obvious, common sense, and administration
‘true’” (Ball, 2008, p. 5). In and through these discourses, policy is developed, worked
out, made sense of, negotiated, and disputed. Meanwhile, policy discourses in
education systems are both enabled and sometimes constrained by policy texts as
instantiated in practice. Personnel in education sector government agencies and extra-
system agencies use policy texts to negotiate for resources, jockey for status, argue for 545
a particular prognosis or solution to a problem and so on. Hence, policy texts both
reflect policy discourses and contribute to the definition of those discourses by
validating some ideas at the expense of others. It is important to remember that these
policy discourses pertain not just to education practice in schools but also to education
policy making practice at all levels of the system – local, state, and federal, and to
educational research funding.
Over the past several decades, government agencies at all levels in the USA have
become increasingly confident about flexing their policy muscles with respect to
education. More important still, government agencies have gradually concerned
themselves more with influencing the core work of schools, classroom teaching, and
student learning, albeit in often very narrow ways. For example, a few subjects –
typically English language arts and mathematics – have consumed most of policy
makers’ attention. These developments can be traced back beyond key federal
legislation such as No Child Left Behind (NCLB, 2001), to state and indeed local
government policy initiatives (Lipman, 2004).
Most of these key themes in the US policy discourse were popular in state and local
government policy texts prior to playing on the national stage (Smith and O’Day, 1990).
Before NCLB, for example, Chicago and other school districts held schools accountable
for student performance using sanctions (Lipman, 2004). Regardless of origins, these
are discourses that have become more established over the past several decades,
gaining prominence and conspicuousness (Fuhrman et al., 2007):
. articulating student learning and performance standards centrally;
. aligning standards with state assessments of student learning;
. holding schools accountable for student performance on state assessments
through sanctions and rewards;
. evidence-based practice using rigorous research and better testing data; and
. using markets to improve schools through competition.
While these ideas figure prominently in the current policy discourses in the USA, other
themes feature less prominently or increasingly at the fringes (e.g. teacher
professionalism, decentralization or local control, democratic goal of schooling).
Of course, while the federal government in the USA remains constitutionally
constrained in matters of education, as well as administratively segmented and
resource poor, neither the Bush nor Obama administrations have shied away from
trying to influence education policy and practice with respect to America’s schools.
This federal optimism is admirable considering the constraints they work under;
however, history suggests caution is in order. The successful Soviet launch of Sputnik,
at the height of the Cold War, prompted increasing attention to and more investment in
education by a federal government fearful about American pre-eminence in the
international arena. More federal education policy making activity ensued and over
JEA time contributed to more education policy activity at the state and local government
50,5 levels (Spillane, 1996). New federal programs, such as the Elementary and Secondary
Education Act of 1965 (ESEA) and the NCLB, defined new responsibilities and
provided new resources for state and local government agencies, enabling them to
expand (Cohen, 1982; Meyer et al., 1987; Rowan, 1982). The federal government
depends on state and local governments to develop policies and programs that support
546 the goals and requirements of NCLB. State governments, for example, hold
responsibility for student assessment, a key element of the NCLB legislation. Indeed,
while some states in response to NCLB implemented assessments that measured more
ambitious student learning goals, other states developed assessments that centered on
more basic learning goals (Wong et al., 2009). In this way, while state policy makers for
the most part complied with “the letter of the law” in terms of assessing students
annually in core subjects and reporting student achievement for different student
groups, between-state differences in state assessments suggests considerable
variability in compliance at the state level with the “spirit of the law.” Local school
districts and schoolhouses, despite the ramped up federal and state incentives and
sanctions, continue to be where the rubber of education policy meets the road of school
improvement, though incentives and support do vary by state. Moreover, despite
increased federal policy making on matters of instruction, there has been no decline in
state and local district instructional policy making – policy making is not a zero-sum
game, at least not in a fragmented federal system such as the USA (Fuhrman and
Elmore, 1990; Spillane, 1996).
But, government instructional policy must be analyzed not simply in terms of its
instrumental goals but also in terms of its broader entailments for the education
system and extra-system. In the USA, government education policy making has
contributed to the development of a sprawling and mostly unregulated extra-system of
non-governmental agencies, including publishers, testing companies, professional
associations, and private consultants (Burch, 2009; Cohen, 1982; Hill, 2007). Recent
federal policies such as NCLB have opened up access to local education markets for
firms in the for-profit and non-profit sectors (Burch, 2009; Hill, 2007). Lacking the
administrative capacity, state and federal government agencies relied on these extra-
system actors to provide many of the services required under their policies.
So while federal, state, and local government policy makers have gone to
considerable lengths over the past several decades to target their policies at the
technical core of schooling – specifying what teachers should teach, at times how they
should teach, and acceptable levels of mastery for students – their initiatives, which
represent a considerable shift in the policy environment of schools, ultimately depend
on school administration for their successful implementation. These instructionally
focussed policies are also transforming educational governance arrangements by
changing the relations among existing agencies and creating opportunities for new
providers to emerge (Burch, 2009; Cohen, 1982). Increasing federal, state, and local
government policy activity does not always result in more streamlined arrangements
for governing classroom instruction, but often more segmented and unwieldy
arrangements (Fuhrman et al., 2007; Spillane, 2004; Tyack and Tobin, 1994).
Still, local schools are left to figure out the entailments of policy makers’ externally
imposed measures of success for school and classroom practice, and they are left to do
so in a fragmented education system where instructional guidance is often weak and
inconsistent. Though schools are held accountable for student learning outcomes
as measured by state standardized tests in selected subjects, and these tests vary
tremendously between states, they are left to a great extent to their own devices to School
figure out the particulars, albeit with more or less support from district and state administration
depending on their situation (Wong et al., 2009). Moreover, NCLB and state policy
requirements place differential pressures on school districts and schools depending on
their student populations. As a result, the institutional environment is experienced
differently by school administrators and teachers depending on how they are
positioned vis-à-vis the broader institutional sector. Again, these circumstances may be 547
especially pronounced in the USA due to the fragmented infrastructure to support
instruction (Cohen and Moffitt, 2009).

Organizational legitimacy and organizational integrity in a changing

institutional environment
Even in the face of tremendous change in the institutional environment of America’s
schools, school leaders still have to deal with the dual organizational imperatives of
legitimacy and integrity. Schools, like some other institutions such as universities, are
situated in a “pluralistic” institutional environment, marked by “persistent internal
tensions” that arise in response to potentially “contending logics” and the tendencies
among diverse stakeholders to “project different identities and purposes upon it”
(Kraatz, 2009, p. 71). In the USA, market, bureaucratic, and professional logics
increasingly compete in policy discourses often creating tensions within schools as
teachers and administrators struggle to manage these competing logics (Hallett, 2007;
Spillane et al., 2011).
The demands placed on US schools by external stakeholders may be more diverse
than in most other countries, especially countries where the education system
developed as part of the nation state or was imposed by a colonial power (Cohen and
Spillane, 1994). In contrast, the US education system grew up from below as part of a
common school movement resulting in a system, and vast extra-system, where matters
of authority and jurisdiction over education have been unsettled (Cohen and Spillane,
1994; Confrey and Stohl, 2004). While state governments have the constitutional
authority on educational matters in the USA, they have delegated a large part of
the administrative responsibility for schooling to local government and for much of the
last century federal involvement was confined to specialized categorical programs.
Educational governance and instructional guidance in the US education system is both
vertical and horizontally segmented and in a constant state of flux. To complicate
matters, there is considerable disagreement among Americans on the means and
ends of schooling. Under these organizational, political, and cultural arrangements,
US schools are left to “manage” diverse and sometimes competing demands on their
attention, demands that in other educational systems are “managed” at the national
level.
Dealing with pluralistic institutional environments requires institutional work that
falls, broadly, into two main categories (Kraatz, 2009). First, there is organizational
legitimacy as school leaders strive to gain the support of diverse stakeholders by
demonstrating to those stakeholders their school’s “cultural fitness.” In pluralistic
organizations such as schools, school leaders have to convince diverse stakeholders
that the organization is legitimate – a “real” school – as stakeholders expect it to be.
As policy makers work increasingly to define this cultural fitness in terms of student
learning in a few core school subjects and as measured by state mandated
standardized achievement tests, it shifts the metric for legitimacy. Indeed, the core
work of schools, long buffered from external scrutiny by school administrators, is now
JEA exposed to such scrutiny. To the extent that various stakeholders, not just policy
50,5 makers, take to these new metrics, school administrators have to attend to them in
order to preserve the legitimacy of their school. Moreover, under NCLB, schools in the
most challenging circumstances, charged with educating students who traditionally
have been disenfranchised by the system, have more opportunities not to meet
adequate yearly progress (AYP) as defined by NCLB and states. In the era of high-
548 stakes accountability tied to student performance, the threats to legitimacy are greatest
in schools enrolling poor students, students of color, and students for whom English is
not a first language. Other policy developments such as the emergence of charter
schools and turnaround schools may also threaten a school’s organizational legitimacy
if enrollment declines as the school being to lose students to a neighboring charter
school.
“Organizational integrity” is also important. School leaders must work at knitting
together the expectations of diverse stakeholders in order to create an “organizational
self” that is minimally coherent, integrated, and self-consistent (Kraatz, 2009; Mead,
1934; Selznick, 1992). Addressing organizational functions such as setting direction for
the school and developing short and long terms goals to realize this direction are
critical when it comes to organizational integrity. The appearance of self-consistency,
integration, coherence, and reliability are critical for school leaders as they strive for
organizational integrity. School leaders play an important role in helping their schools
manage the dual imperatives of legitimacy and integrity (Kraatz and Block, 2008).
The shifting policy environment in the USA puts pressure on school administrators
to attend to instructional matters as measured by student performance metrics in core
school subjects and to engage in efforts at recoupling the external policy environment
with administrative practice and with classroom instruction. Indeed, scholars argue
that as the institutional environment of schools “becomes more unitary and as rules
about work in the technical core become more specific” and “get attached to outcomes
or other inspection systems,” they would have a stronger effect on work activity
in schools (Rowan and Miskel, 1999, p. 373). These scholars hypothesized that the
emergence of a more elaborate technical environment in the education sector
(e.g. standards and high-stakes testing) would lead to schools facing much stronger
environmental pressures on their core technical work – teaching and learning. As
discussed in the introduction, there is some empirical evidence to support this
hypothesis, with several studies documenting that government policies influence
school leaders and classroom teachers for good and bad (Booher-Jennings, 2006;
Diamond and Spillane, 2004; Firestone et al., 1998; Jacob, 2005; Lee, 2007; McNeil, 2002;
Mintrop and Sunderman, 2009; Neal and Schanzenbach, 2007; Wong et al., 2009). We
discuss this more as follows.
Though the institutional environment of schools has changed in dramatic ways,
most notably with a very definite focus on the technical core of schooling, the dual
imperatives of organization legitimacy and integrity remain – though the challenge of
meeting them has likely changed with student achievement on state tests now being
the key performance metric. How do school administrators manage organizational
integrity and legitimacy in this changing institutional environment?

Managing in the middle: administrative practice in a shifting policy

environment
High-stakes accountability levers that are directly tied to instruction, if they are
to work, operate in and through particular school administrative arrangements.
Of course, school leaders are not passive receptors of their environments. Rather, they School
enact their environments; that is, they “construct, rearrange, single out, and demolish administration
many objective features of their surroundings” (Weick, 1979, p. 164). School leaders as
mid-level managers (Harris, 2002; Hatch, 2001; Leithwood et al., 2004; Louis et al., 2010;
Spillane et al., 2002) occupy a somewhat unique situation: their work focusses
in at least two directions in the organizational hierarchy. On the one hand, school
leaders are dependent on their institutional environment for the legitimacy of their 549
organization – local school council, school district, state, parents, and local community.
On the other hand, they are also dependent on classroom teachers and students for the
organizational integrity of their buildings. Without the cooperation of teachers and
students, the coherence, integration, and self-consistency of their school is likely to
fall apart. Moreover, organizational integrity and organizational legitimacy are
interdependent: in a changing institutional environment, legitimacy is increasingly tied
to student achievement as measured by standardized tests aligned, more or less, with
district and state standards. The standardization advances by these policy initiatives
demand a particular sort of organizational integrity that is tied to externally imposed
standards. This externally imposed standardization of instruction flies in the face of
business as usual for most US schools where isolated teacher practitioners with
considerable professional autonomy over instructional matters was the dominant
operating procedure (Lortie, 1975).
Relations between school administrators and teachers are characterized by
interdependency and conflict (Lipsky, 1980). The objectives of school administrators
and teachers, despite the rhetoric of what is best for students, are often in conflict with
teachers having a strong desire to maintain their professional autonomy over
instructional matters (Hallett, 2007; Spillane et al., 2011). School leaders seek to achieve
results that they see as consistent with federal, state, and school district objectives and
thus have to work to constrain teachers’ autonomy and discretion. Still, teachers have
resources critical to student achievement with which they can resist school leaders’
desires, even in the era of high-stakes accountability policy. Teachers’ expertise,
their willingness to change, and to engage seriously in the work of instructional
improvement are all critical resources. At the same time, teachers are to some extent
dependent on school leaders who allocate resources including funding, curricular
materials, and class assignments. The nature of teaching as a practice also contributes
to this interdependency between school administrators and teachers (Cohen, 1988;
Lipsky, 1980). Overall, while teachers depend on administrators, administrators also
depend on teachers. And of course, school administrators depend on district
administrators who evaluate their performance and decide if they are to continue in
their positions.
How do school leaders manage in such circumstances? How do they juggle the dual
demands of organizational integrity and legitimacy, especially in an institutional
environment that challenges norms of classroom privacy and teachers autonomy?
As one might expect considering the threat to organizational legitimacy posed by
external performance metrics tied to student achievement on state tests, school leaders
appear to have paid attention to standards and test-based accountability and
responded in strategic ways that often involved gaming the system, though this
varies across schools and school systems depending at least in part on the student
population. Among other things, school leaders emphasize tested subjects and
instruction in test-taking strategies as well as reclassifying students and not actively
preventing students from dropping out (Anagnostopoulos and Rutledge, 2007;
JEA Diamond and Spillane, 2004; Figlio and Getzler, 2002; Ladd and Zelli, 2002). Further,
50,5 school administrators targeted low-performing students, particularly Hispanic and
black students, through expanding exemption rates by classifying more students as
special needs, encouraging absences, and aiming instruction toward these subgroups
(Cullen and Reback, 2006; Lauen and Gaddis, 2012). Still, school leaders also worked
to try and create conditions that might contribute substantively to improvement
550 in teaching and learning. Such efforts included transforming the organizational
infrastructure, including initiating weekly staff meetings, department-wide curriculum
development, designing organizational routines that were tied directly to instruction
and its improvement in tested subjects, meeting with teachers, and creating leadership
teams to supervise their schools (Anagnostopoulos and Rutledge, 2007; Ladd and
Zelli, 2002; Mintrop, 2003; Spillane et al., 2004). These efforts not only focussed on
organizational legitimacy but also organization integrity as they strove to create a
more coherent instructional program, at least in tested subjects. Of course, schools,
depending in important measure on their student populations, experience state,
federal, and local government accountability policies differently. For schools that enroll
mostly students of color and students living in poverty, for example, there are more
ways to fail to meet AYP under NCLB requirements.
These accounts, though we are not questioning their empirical accuracy in any way,
fail to capture the dilemmas US school administrators must manage in light of recent
shifts in the institutional environment. Moreover, we believe that many accounts
may underplay the “good faith” efforts of school administrators in the USA to manage
organizational legitimacy and integrity in a changed institutional sector where
externally imposed performance metrics – student achievement on standardized test –
has become the coin of the realm. Specifically, we want to examine efforts by school
leaders to transform their schools’ administrative infrastructure. What follows is an
attempt to articulate a series of hypotheses about the school administrative response to
high-stakes accountability in the USA that goes beyond gaming the system, though
fully acknowledging that this happens, based on empirical work in a handful of
schools. Future empirical work will have to test these hypotheses though in some
respects it may be too late if the institutional environment becomes more settled,
though that remains to be seen.
New government instructional policy did not walk into schools and invoke itself.
Federal, state, and local policy makers depended on school leaders to invoke and frame
the new policy directives and their entailments for local practice especially with respect
to the school’s instructional program – including what content to teach and how to
teach it (Spillane et al., 2011). Our ongoing work across several empirical studies
(Spillane and Anderson, under review; Spillane and Hunt, 2010; Spillane and Kim,
under review; Spillane et al., 2011) suggests several ways in which school
administrators manage in the middle in these changing times.
Our work in schools, early in these recent institutional shifts and prior to NCLB,
suggests that school leaders engaged in elaborate efforts to design the formal structure
of their schools in an effort to transform school administrative practice so it was more
responsive to and less distinctive from external policy directives and classroom
instruction (Spillane et al., 2011). More specifically, they engaged in extensive efforts to
build local school infrastructures that supported tighter connections between external
policy and classroom instruction, facilitated by the school’s formal organizational
structure. These efforts were not mere fiddling with the existing structure, but
represented extensive redesign efforts. By formal structure or infrastructure here we
mean aspect of the formal organization including but not limited to formally School
designated leadership positions and their responsibilities, formal organizational administration
routines (e.g. grade-level meetings, leadership team meetings) and tools (classroom
observation protocols). Formal structure or infrastructure is both constitutive of and
constituted in practice – without structure there is no practice. Thus, to understand
administrative practice and changes in that practice attention to formal structure or
infrastructure is critical. School leaders, for example, designed organizational routines 551
to standardize their instructional program both vertically and horizontally, working to
align classroom practice with the content covered in state and district standards
and student assessments. School leaders intended these organizational routines to
standardize curricula, monitor student and teacher performance, and make classroom
practice more transparent.
Veteran staff in the schools in our studies reported that these transformations of the
formal structure represented a dramatic shift in ways of doing business at their
schools. Our analysis of these formal organizational routines in practice showed that,
rather than buffering instruction from external regulation, these routines in practice
promoted recoupling of government regulation and classroom teaching (Spillane et al.,
2011). These routines promoted recoupling because school leaders used state and
district regulation as templates and rubrics in performing key technical efficiency
functions including standardizing the instructional program, setting and maintaining
direction, identifying and addressing needs including professional development, and
monitoring instruction. Formal school organizational routines facilitated recoupling of
government regulation with the technical core by making classroom instruction more
transparent, albeit some aspects of instruction and some school subjects more than
others. Moreover, these formal organizational routines focussed almost exclusively on
those school subjects tested by external agencies – mathematics and language arts.
Further, our analysis suggests that the implementation of these changes to the formal
organizational structure of schools met with considerable resistance from school staff
as they contended with a taken for granted professional logic (e.g. teacher autonomy)
that many veteran teachers cherished (Hallett, 2007; Spillane et al., 2011).
These local infrastructure-building initiatives may be uniquely American in that
most US schools reside in a system where the infrastructure to support instruction is
impoverished and underdeveloped (Cohen and Moffitt, 2009). As a result, school
administrators are left to design the infrastructure to support the sort of instructional
changes pressed by the external policy environment. This is a huge undertaking – one
that we suspect is not delegated to school administrators in other education systems.
Many observers of the system fail to recognize this, assuming that all that it takes to
implement high-stakes accountability policies is for individual school leaders, typically
the school principal, to change their behavior vis-à-vis teachers. In reality, school
administrators are left, in this changing institutional environment, to design entirely
new formal organizational structures in their schools that support tighter coupling
between policy, administration, and instruction. And, they design these structures
for a few selected tested subjects. This is a departure from a time when school
administration buffered classroom instruction from external scrutiny with myth and
ceremony (Meyer and Rowan, 1978).
In implementing this new formal organizational structure and in performing
organizational routines, school leaders did not rely solely on their own positional
authority or on the authority of government agencies to get teachers to cooperate
in performing the new routines and with their sense of policy makers’ directives
JEA (Spillane and Anderson, under review). Instead, school leaders worked at framing
50,5 policy so as to appeal to teachers’ interests, values, goals, and norms, reflecting their
position in the middle between an increasingly demanding external institutional
environment on matters of instruction on the one hand, and teachers accustomed to
professional autonomy and discretion on the other. In their efforts to convince teachers
to comply with external government policy related to instruction, school
552 administrators relied on various persuasion tactics, reflecting their positions in the
middle between external stakeholders (e.g. policy makers) and internal stakeholders
(e.g. teachers). School leaders deployed persuasion tactics (Lindblom, 1977): to
persuade teachers and compel their cooperation with external policy, school leaders
worked to frame policy in ways that would appeal to teachers’ interests, values, goals,
and norms using agenda-setting, aligning, asserting their in-group identity as
teachers, other-oriented dispositions, and brokering of information and policy framing
(Spillane and Anderson, under review).
These infrastructure redesign efforts were not the only action that school leaders
took. As discussed above, school leaders also promoted tested subjects, the teaching
of test-taking strategies, the reclassification of students to improve their school’s
performance, and targeting students who they thought could perform better on the test
(Anagnostopoulos and Rutledge, 2007; Cullen and Reback, 2006; Diamond and
Spillane, 2004; Figlio and Getzler, 2002; Ladd and Zelli, 2002; Lauen and Gaddis, 2012).
School leaders also sought out programs and professional development that would
help teachers improve their teaching in tested subjects.
These changes in infrastructure also highlight the role of other school leaders, in
addition to the school principal, in school administrative practice. While the literature
often portrays the school principal as a “lone ranger” who operates as a solo
practitioner in the schoolhouse, increasingly scholarship points to the principal as one
of several individuals involved in the work of leading and managing (Camburn et al.,
2003; Copland, 2001; Gronn, 2000; Hargreaves and Fink, 2004; Harris, 2005; Leithwood
et al., 2007; MacBeath et al., 2004; Portin et al., 2003; Spillane et al., 2007; Spillane and
Diamond, 2007; Timperley, 2005). Moreover, principals and these other school leaders,
such as assistant principals and part-time and full-time administrators and specialists,
are especially central to leading and managing instruction. In one study, for example,
of 30 elementary schools in a mid-sized urban school district we found that, while
the principal’s day is largely spent performing administrative duties, principals
nonetheless devote between 20 and 30 percent of their time to instruction and
curriculum work (Spillane and Hunt, 2010). In this same district, we found that part-
time formally designated school leaders (e.g. coaches, mentor teachers) were key advice
givers and brokers in the English language arts and mathematics instructional advice
networks (Spillane and Kim, under review). This is not a US phenomenon; empirical
work in several countries captures the distribution of responsibility for school
leadership and management over administrators and teachers leaders (Bennett et al.,
2003; Day, 2005; Gronn, 2000; Harris, 2005; Leithwood et al., 1999, 2007; MacBeath
et al., 2004; Moolenaar et al., 2011; Notman and Henry, 2011; Sleegers et al., 2002;
Timperley, 2005), though there are likely important differences between countries in
the distribution of leadership and management responsibilities that have yet to be
explored by sound empirical work. Though policy makers tend to focus their attention
mostly on the school principal, our account suggests that the work of leading and
managing the schoolhouse, especially with respect to the core technical work of
instruction involves an array of other full-time and part-time leaders. Further, the
design and redesign of the school’s infrastructure is a core component of school-level School
efforts to transform school administrative practice. administration
Conclusion
Federal, state, and local government policy makers’ desire to regulate instruction
in US public schools, albeit in select school subjects, has steadily grown over recent
decades. These policy initiatives have increasingly put pressure on school leaders to 553
adapt their organization to meet new demands from the institutional environment in
order to maintain their schools’ organizational integrity and legitimacy. School leaders
manage these dual imperatives in this changing institutional environment on a
day-to-day basis by working to transform the school’s infrastructure so that
administrative practice is more tightly coupled with policy and with classroom
instruction. This design and redesign work is intended to make the school’s technical
core less private and more transparent. Further, rather than rely solely on political
authority, school leaders also use persuasion in an effort to convince teachers to heed
and respond to a shifting policy environment.
While our account is based entirely on the US education sector, we suspect that
several aspects of our framing may be relevant in other countries. As noted in the
introduction, the US education sector does not have a monopoly on standards and high-
stakes accountability based on common metrics – typically student achievement on
standardized tests; these policy levers are relatively commonplace in policy discourses
in several countries. Of course, the particulars of these policy levers and their
deployment vary by country. Moreover, government arrangements, political culture,
and social arrangements also differ across countries and these in turn have
implications for how standards and accountability are played out in schools (Cohen
and Spillane, 1994).
Still, organizational integrity and organizational legitimacy are likely to be concerns
for schools cross-nationally, though their relative import will depend in some measure on
the pluralism of the institutional environment and the prevalence of tensions as diverse
stakeholders make different and often conflicting demands on schools. For example, in
some respects the legitimacy imperative may be especially pronounced in the USA due
to the education system’s local origins (as distinct from an arm of the nation state) and
the sprawling and continually shifting school governance system that spans locally
elected school boards, a state government apparatus, and the federal government. Add to
this a vast and ever expanding extra-system of textbook publishers, lobbyists, testing
agencies, charter school networks, professional development providers and so on, that
operate at each level of the system, and the diversity of demands placed on schools is not
at all surprising. In countries where the education systems are a product of the nation
state (e.g. France) or former colonial powers (e.g. India, Ireland), stakeholders’ demands
may be less diverse or their influence confined by law or tradition to one level of the
system (e.g. collective bargaining among the various partners at the national level).
Under these arrangements, school leaders may be less susceptible the demands of
diverse stakeholders. At the same time, it is probable that both the legitimacy and
integrity imperatives become more pronounced at times of major change in the
institutional environment as new ideas and institutional logics become prevalent in both
policy discourses and texts. Such has been the case in the USA and several other
countries over the past few decades.
The organizational legitimacy and organizational integrity imperatives provide a
potentially powerful framework for cross-national work on school administrative
JEA practice for a few reasons. First, the framework situates the work of school leadership
50,5 and management firmly in the institutional sector pressing scholars to systematically
examine how school leaders make sense of, notice, and respond to their institutional
environment, in both settled and unsettled times, as they work to lead and manage
schools. As a result, researchers are pressed to not give accounts of autonomous school
leaders working in an institutional vacuum but to take account of the broader institutional
554 environment that informs and infuses their work is both constituted of and constitutive in
school administrative practice. Second, systematic attention to both organizational
legitimacy and integrity imperatives in studies across several countries would enable the
field to understand how differences in educational governance arrangements, social
arrangements, and culture interact with school administrative practice.
Note
1. Our focus on the school level should not be construed as negating the role of other levels of
the school system including the Local Education Agency or school district, as well state and
federal agencies. School district offices, for example, are critical in understanding the
implementation of state and federal policies and programs (Anderson, 2003; Anderson
and Togneri, 2005; Firestone, 1989; Honig, 2003, 2006; Spillane, 1994, 1996, 1998, 2004). Still,
school leaders and teachers are the final brokers of education policy, especially in a federal
system where authority is still segmented vertically and horizontally.

References
Amrien, A.L. and Berliner, D.C. (2002), “High-stakes testing, uncertainty, and student learning”,
Education Policy Analysis Archives, Vol. 10 No. 18, pp. 1-70.
Anagnostopoulos, D. and Rutledge, S.A. (2007), “Making sense of school sanctioning policies in
urban high schools: charting the depth and drift of school and classroom change”,
Teachers College Record, Vol. 109 No. 5, pp. 1261-302.
Anderson, S.E. (2003), The School District Role in Educational Change: A Review of the Literature,
International Centre for Educational Change, Ontario Institute for Studies in Education,
Toronto, ON.
Anderson, S.E. and Togneri, W. (2005), “School district-wide reform policies in education”, in
Bascia, N., Cumming, A., Datnow, A., Leithwood, K.A. and Livingstone, D. (Eds),
International Handbook of Education Policy, Vol. 13, Springer, Cornwall, pp. 173-94.
Au, W. (2007), “High-stakes testing and curricular control: a qualitative metasynthesis”,
Educational Researcher, Vol. 36 No. 5, pp. 258-67.
Ball, S.J. (1994), Education Reform: A Critical and Post-Structural Approach, Open University
Press, Philadelphia, PA.
Ball, S.J. (1999), “Global trends in educational reform and the struggle for the soul of the teacher!”,
paper presented at the British Educational Research Association Annual Conference
University of Sussex, Bringhton, September.
Ball, S.J. (2006), Education Policy and Social Class: Selected Works, Routledge, London.
Ball, S.J. (2008), The Education Debate, The Policy Press, Bristol.
Bennett, N., Wise, C., Woods, P.A. and Harvey, J.A. (2003), Distributed Leadership: A Review of
Literature, National College for School Leadership, Nottingham.
Berman, P. and McLaughlin, M.W. (1977), Federal Programs Supporting Educational Change:
Factors Affecting Implementation and Continuation, RAND, Santa Monica, CA.
Booher-Jennings, J. (2006), “Rationing education in an era of accountability”, Phi Delta Kappan,
Vol. 87 No. 10, pp. 756-61.
Broadbent, J., Jacobs, K. and Laughlin, R. (1999), “Comparing schools in the UK and School
New Zealand: individualizing and socializing accountabilities and some implications for
management control”, Management Accounting Research, Vol. 10 No. 4, pp. 339-61. administration
Bryk, A.S. and Schneider, B. (1996), “Social trust: a moral resource for school improvement”, paper
presented at the 1996 Meeting of the American Sociological Society, Washington, DC.
Bullard, P. and Taylor, B.O. (1994), Keepers of the Dream: The Triumph of Effective Schools,
Excelsior Foundation, Chicago, IL. 555
Burch, P. (2009), Hidden Markets: The New Education Privatization, Routledge, New York, NY.
Burgess, S., Wilson, D. and Worth, J. (2010), A Natural Experiment in School Accountability:
The Impact of School Performance Information on Pupil Progress and Sorting, Centre for
Market and Public Organisation, Bristol Institute of Public Affairs, Bristol.
Camburn, E., Rowan, B. and Taylor, J.E. (2003), “Distributed leadership in schools: the case of
elementary schools adopting comprehensive school reform models”, Educational
Evaluation and Policy Analysis, Vol. 25 No. 4, pp. 347-73.
Clotfelter, C.T. and Ladd, H.F. (1996), “Recognizing and rewarding success in public schools”, in
Ladd, H.F. (Ed.), Holding Schools Accountable: Performance-Based Reform in Education,
The Brookings Institution, Washington, DC, pp. 23-64.
Coburn, C.E. (2001), “Collective sensemaking about reading: how teachers mediate reading policy
in their professional communities”, Educational Evaluation and Policy Analysis, Vol. 23
No. 2, pp. 145-70.
Cohen, D.K. (1982), “Policy and organization: the impact of state and federal educational policy on
school governance”, Harvard Educational Review, Vol. 52 No. 4, pp. 474-99.
Cohen, D.K. (1988), Teaching Practice: Plus c¸a Change, National Center for Research on Teacher
Education, East Lansing, MI.
Cohen, D.K. (1990), “A revolution in one classroom: the case of Mrs. Oublier”, Educational
Evaluation and Policy Analysis, Vol. 12 No. 3, pp. 311-29.
Cohen, D.K. and Moffitt, S.L. (2009), The Ordeal of Equality: Did Federal Regulation Fix the
Schools?, Harvard University Press, Cambridge, MA.
Cohen, D.K. and Spillane, J.P. (1994), “National education indicators and traditions of
accountability”, in OECD (Ed.), Making Education Count: Developing and Using
International Indicators, Centre for Educational Research and Innovation, Paris, pp. 323-37.
Cohen, D.K. and Weiss, J.A. (1993), “The interplay of social science and prior knowledge in public
policy”, in Redner, H. (Ed.), Studies in the Thought of Charles E. Lindblom, Westview,
Boulder, CO, pp. 210-34.
Colyvas, J.A. (2012), “Performance metrics as formal structures and through the lens of social
mechanisms: when do they work and how do they influence?”, American Journal of
Education, Vol. 118 No. 2, pp. 167-97.
Confrey, J. and Stohl, V. (2004), On Evaluating Curricular Effectiveness: Judging the Quality of
K-12 Mathematics Evaluations, National Academies Press, Washington, DC.
Copland, M.A. (2001), “The myth of the superprincipal”, Phi Delta Kappan, Vol. 82 No. 7,
pp. 528-33.
Cuban, L. (1988), The Managerial Imperative and the Practice of Leadership in Schools, SUNY
Press, Albany, NY.
Cullen, J.B. and Reback, R. (2006), “Tinkering toward accolades: school gaming under a
performance accountability system”, Advances in Applied Microeconomics, Vol. 14 No. 1,
pp. 1-34.
Darling-Hammond, L. (2004), “Standards, accountability, and school reform”, Teachers College
Record, Vol. 106 No. 6, pp. 1047-85.
JEA Darling-Hammond, L. and Wise, A.E. (1985), “Beyond standardization: state standards and
school improvement”, The Elementary School Journal, Vol. 85 No. 3, pp. 315-36.
50,5
Davies, S. and Guppy, N. (1997), “Globalization and educational reforms in Anglo-American
democracies”, Comparative Education Review, Vol. 41 No. 4, pp. 435-59.
Day, C. (2005), “Principals who sustain success: making a difference in schools in
challenging circumstances”, International Journal of Leadership in Education, Vol. 8
556 No. 4, pp. 273-90.
Deal, T.E. and Celotti, L.D. (1980), “How much influence do (and can) educational administrators
have on classrooms?”, Phi Delta Kappan, Vol. 67 No. 7, pp. 471-3.
Diamond, J.B. and Spillane, J.P. (2004), “High stakes accountability in urban elementary
schools: challenging or reproducing inequality?”, Teachers College Record, Vol. 106 No. 6,
pp. 1145-76.
Espeland, W.N. and Sauder, M. (2007), “Rankings and reactivity: how public measures recreate
social worlds”, American Journal of Sociology, Vol. 113 No. 1, pp. 1-40.
Eubanks, E.E. and Levine, D.U. (1983), “A first look at effective schools projects in New York City
and Milwaukee”, Phi Delta Kappan, Vol. 64 No. 10, pp. 697-702.
Figlio, D.N. and Getzler, L.S. (2002), Accountability, Ability and Disability: Gaming the System,
National Bureau of Economic Research, Cambridge, MA.
Firestone, W.A. (1985), “The study of loose coupling: problems, progress, and prospects”,
in Kerckhoff, A. (Ed.), Research in Sociology of Education and Socialization, Vol. 5,
JAI Press, Greenwich, CT, pp. 3-30.
Firestone, W.A. (1989), “Using reform: conceptualizing district initiative”, Educational Evaluation
and Policy Analysis, Vol. 11 No. 2, pp. 151-64.
Firestone, W.A., Mayrowetz, D. and Fairman, J. (1998), “Performance-based assessment and
instructional change: the effects of testing in Maine and Maryland”, Educational
Evaluation and Policy Analysis, Vol. 20 No. 2, pp. 95-113.
Fuhrman, S.H. and Elmore, R.F. (1990), “Understanding local control in the wake of state
educational reform”, Educational Evaluation and Policy Analysis, Vol. 12 No. 1, pp. 82-96.
Fuhrman, S.H. and Elmore, R.F. (2004), Redesigning Accountability Systems for Education,
Teachers College Press, New York, NY.
Fuhrman, S.H., Goertz, M.E. and Weinbaum, E.H. (2007), “Educational governance in the United
States: where are we? How did we get here? Why should we care?”, in Cohen, D.K.,
Fuhrman, S. and Mosher, F. (Eds), The State of Education Policy Research, Lawrence
Erlbaum Associates, Mahwah, NJ, pp. 41-61.
Gamoran, A. and Dreeben, R. (1986), “Coupling and control in educational organizations”,
Administrative Science Quarterly, Vol. 31 No. 4, pp. 612-32.
Gronn, P. (2000), “Distributed properties: a new architecture for leadership”, Educational
Management Administration Leadership, Vol. 28 No. 3, pp. 317-38.
Hallett, T. (2007), “Between deference and distinction: interaction ritual through symbolic power
in an educational institution”, Social Psychology Quarterly, Vol. 70 No. 2, pp. 148-71.
Hallinger, P. and Murphy, J.F. (1986), “The social context of effective schools”, American Journal
of Education, Vol. 94 No. 3, pp. 328-55.
Hargreaves, A. and Fink, D. (2004), “The seven principles of sustainable leadership”, Educational
Leadership, Vol. 61 No. 7, pp. 8-13.
Harris, A. (2002), “Effective leadership in schools facing challenging contexts”, School Leadership
& Management, Vol. 22 No. 1, pp. 15-26.
Harris, A. (2005), “Distributed leadership”, in Davies, B. (Ed.), The Essentials of School
Leadership, Paul Chapman Publishing, London, pp. 160-72.
Hatch, T. (2001), “Incoherence in the system: three perspectives on the implementation of multiple School
initiatives in one district”, American Journal of Education, Vol. 109 No. 4, pp. 407-37.
administration
Herman, J.L. (2004), “The effects of testing on instruction”, in Fuhrman, S.H. and Elmore, R.F.
(Eds), Redesigning Accountability Systems for Education, Teachers College Press, New York,
NY, pp. 141-66.
Hill, H.C. (2007), “Learning in the teaching workforce”, The Future of Children, Vol. 17 No. 1,
pp. 111-27. 557
Honig, M. (2003), “Building policy from practice: central office administrators’ roles and capacity
in collaborative education policy implementation”, Educational Administration Quarterly,
Vol. 39 No. 3, pp. 292-338.
Honig, M. (2006), New Directions in Education Policy Implementation: Confronting Complexity,
SUNY Press, Albany, NY.
Jacob, B.A. (2005), “Accountability, incentives and behavior: the impact of high-stakes testing
in the Chicago Public Schools”, Journal of Public Economics, Vol. 89 Nos 5-6, pp. 761-96.
Kraatz, M.S. (2009), “Leadership as institutional work: a bridge to the other side”, in Lawrence, T.B.,
Suddaby, R. and Leca, B. (Eds), Institutional Work: Actors and Agency in Institutional
Studies of Organizations, Cambridge University Press, Cambridge, pp. 59-91.
Kraatz, M.S. and Block, E.S. (2008), “Organizational implications of institutional pluralism”,
in Greenwood, R., Oliver, C., Suddaby, R. and Sahlin-Andersson, K. (Eds), Handbook of
Organizational Institutionalism, Sage Publications, London, pp. 243-75.
Ladd, H.F. and Zelli, A. (2002), “School-based accountability in North Carolina: the responses of
school principals”, Educational Administration Quarterly, Vol. 38 No. 4, pp. 494-529.
Lauen, D.L. and Gaddis, S.M. (2012), “Shining a light or fumbling in the dark? The effects of
NCLB’s subgroup-specific accountability on student achievement”, Educational Evaluation
and Policy Analysis, Vol. 34 No. 2, pp. 185-208.
Lee, C.D. (2007), Culture, Literacy, and Learning: Taking Bloom in the Midst of the Whirlwind,
Teachers College Press, New York, NY.
Leithwood, K.A., Jantzi, D. and Steinbach, R. (1999), Changing Leadership for Changing Times,
Open University Press, Buckingham.
Leithwood, K., Louis, K.S., Anderson, S. and Wahlstrom, K. (2004), How Leadership Influences
Student Learning: A Review of Research for the Learning from Leadership Project,
The Wallace Foundation, New York, NY.
Leithwood, K., Mascall, B., Strauss, T., Sacks, R., Memon, N. and Yashkina, A. (2007),
“Distributing leadership to make schools smarter: taking the ego out of the system”,
Leadership and Policy in Schools, Vol. 6 No. 1, pp. 37-67.
Lewicki, R.J. and Bunker, B.B. (1996), “Developing and maintaining trust in work relationship”,
in Kramer, R.M. and Tyler, T.R. (Eds), Trust in Organizations: Frontiers of Theory and
Research, Sage Publications, Thousand Oaks, CA, pp. 114-39.
Lindblom, C.E. (1977), Politics and Markets: The World’s Political Economic Systems, Basic
Books, New York, NY.
Lipman, P. (2004), High Stakes Education: Inequality, Globalization, and Urban School Reform,
RoutledgeFalmer, New York, NY.
Lipsky, M. (1980), Street-Level Bureaucracy, Russell Sage Foundation, New York, NY.
Lortie, D.C. (1975), Schoolteacher, University of Chicago Press, Chicago, IL.
Louis, K.S., Kruse, S.D. and Bryk, A.S. (1995), “Professionalism and community: what is it and
why is it important in urban schools?”, in Louis, K.S. and Kruse, S.D. (Eds),
Professionalism and Community: Perspectives on Reforming Urban Schools, Corwin Press,
Thousand Oaks, CA, pp. 3-22.
JEA Louis, K.S., Leithwood, K.A., Wahlstrom, K., Anderson, S.E., Michlin, M., Mascall, B., Gordon, M.,
Strauss, T., Thomas, E. and Moore, S. (2010), Learning from Leadership: Investigating the
50,5 Links to Improved Student Learning, The Wallace Foundation, New York, NY.
McLaughlin, M.W. (1990), “The Rand change agent study revisited: macro perspectives and
micro realities”, Educational Researcher, Vol. 19 No. 9, pp. 11-6.
McLaughlin, M.W. and Talbert, J.E. (2001), Professional Communities and the Work of High
558 School Teaching, University of Chicago Press, Chicago, IL.
McNeil, L.M. (2002), “Private asset or public good: education and democracy at the crossroads”,
American Educational Research Journal, Vol. 39 No. 2, pp. 243-8.
MacBeath, J., Oduro, G.K.T. and Waterhouse, J. (2004), Distributed Leadership in Action:
A Study of Current Practice in Schools, National College for School Leadership,
Nottingham.
Malen, B. and Ogawa, R.T. (1988), “Professional-patron influence on site-based governance
councils: a confounding case study”, Educational Evaluation and Policy Analysis, Vol. 10
No. 4, pp. 251-70.
Malen, B., Ogawa, R.T. and Kranz, J. (1990), “What do we know about school-based management?
A case study of the literature – a call for research”, in Clune, W.H. (Ed.), Choice and Control
in American Educations: Vol. 2. The Practice of Choice, Decentralization and School
Restructuring, Falmer, London, pp. 289-342.
Mead, G.H. (1934), Mind, Self, and Society: From the Standpoint of a Social Behaviorist,
The University of Chicago Press, Chicago, IL.
Mehta, J. “The penetration of technocratic logic into the educational field: rationalizing schooling
from the progressives to the present”, Teachers College Record, (revise and resubmited).
Mehta, J. “How paradigms create politics: the transformation of American educational policy
1980-2001”, American Educational Research Journal, (under review).
Meyer, J.W. and Rowan, B. (1978), “The structure of educational organizations”, in Meyer, M.W.
(Ed.), Environments and Organizations, Vols 78-109, Jossey-Bass, San Francisco, CA.
Meyer, J.W., Scott, W.R. and Strang, D. (1987), “Centralization, fragmentation, and school district
complexity”, Administrative Science Quarterly, Vol. 32 No. 2, pp. 186-201.
Mintrop, H. (2003), “The limits of sanctions in low-performing schools: a study of Maryland
and Kentucky schools on probation”, Education Policy Analysis Archives, Vol. 11 No. 3,
pp. 1-30.
Mintrop, H. and Sunderman, G.L. (2009), “Predictable failure of federal sanctions-driven
accountability from school improvement – and why we may retain it anyway”, Educational
Researcher, Vol. 38 No. 5, pp. 353-64.
Mishra, A.K. (1996), “Organizational responses to crisis: the centrality of trust”, in Kramer, R.M.
and Tyler, T.R. (Eds), Trust in Organizations, Sage Publications, Thousand Oaks, CA,
pp. 261-87.
Moolenaar, N.M., Daly, A.J. and Sleegers, P.J.C. (2011), “Ties with potential: social network
structure and innovative climate in Dutch schools”, Teachers College Record, Vol. 113 No. 9,
pp. 1983-2017.
National Commission on Excellence in Education (1983), A Nation at Risk: The Imperative for
Educational Reform, United States Department of Education, Washington, DC.
Neal, D.A. and Schanzenbach, D. (2007), Left Behind by Design: Proficiency Counts and
Test-Based Accountability, National Bureau of Economic Research, Cambridge, MA.
Ng, P.T. (2010), “The evolution and nature of school accountability in the Singapore
education system”, Educational Assessment, Evaluation and Accountability, Vol. 22 No. 4,
pp. 275-92.
Nichols, S.L. and Berliner, D.C. (2007), Collateral Damage: How High-Stakes Testing Corrupts School
America’s Schools, Harvard University Press, Cambridge, MA.
administration
No Child Left Behind Act of 2001 (NCLB) (2001), in 20 U.S.C. y 9101, Vol. 20 U.S.C. y 9101.
Notman, R. and Henry, D.A. (2011), “Building and sustaining successful school leadership in
New Zealand”, Leadership and Policy in Schools, Vol. 10 No. 4, pp. 375-94.
Organization for Economic Cooperation and Development (OECD) (2004), Learning for
Tomorrow’s World: First Results from PISA 2003, OECD, Paris. 559
Pipho, C. (1978), “Minimum competency testing in 1978: a look at state standards”, Phi Delta
Kappan, Vol. 59 No. 9, pp. 585-8.
Portin, B., Schneider, P., DeArmond, M. and Gundlach, L. (2003), Making Sense of Leading
Schools: A Study of the School Principalship, Center on Reinventing Public Education,
Washington University, Seattle, WA.
Power, M. (1994), “The audit society”, in Hopwood, A.G. and Miller, P.M. (Eds),
Accounting as Social and Institutional Practice, Cambridge University Press, Cambridge,
pp. 299-316.
Purkey, S.C. and Smith, M.S. (1985), “School reform: the district policy implications
of the effective schools literature”, The Elementary School Journal, Vol. 85 No. 3,
pp. 353-89.
Ranson, S. (1994), Towards the Learning Society, Cassell, London.
Robinson, V.M.J. and Timperley, H.S. (2000), “The link between accountability and
improvement: the case of reporting to parents”, Peabody Journal of Education, Vol. 75
No. 4, pp. 66-89.
Rowan, B. (1982), “Organizational structure and the institutional environment: the case of public
schools”, Administrative Science Quarterly, Vol. 27 No. 2, pp. 259-79.
Rowan, B. and Miskel, C.G. (1999), “Institutional theory and the study of educational
organizations”, in Murphy, J. and Louis, K.S. (Eds), Handbook of Research in Educational
Administration, Jossey-Bass, San Francisco, CA, pp. 350-84.
Selznick, P. (1992), The Moral Commonwealth: Social Theory and the Promise of Community,
University of California Press, Berkeley, CA.
Sleegers, P.J.C., Geijsel, F.P. and van den Berg, R.M. (2002), “Conditions fostering educational
change”, in Leithwood, K.A. and Hallinger, P. (Eds), Second International Handbook of
Educational Leadership and Administration, Kluwer, Dordrecht, pp. 75-102.
Smith, B. (1998), It’s About Time: Opportunities to Learn in Chicago’s Elementary Schools,
Consortium on Chicago School Research, Chicago, IL.
Smith, M.S. and O’Day, J. (1990), “Systemic school reform”, in Fuhrman, S.H. and Malen, B. (Eds),
The Politics of Curriculum and Testing: Yearbook of the Politics of Education Association,
Falmer Press, Philadelphia, PA, pp. 233-67.
Spillane, J.P. (1994), “How districts mediate between state policy and teachers’ practice”, in
Elmore, R.F. and Fuhrman, S.H. (Eds), The Governance of Curriculum, Association for
Supervision and Curriculum Development, Alexandria, VA, pp. 167-85.
Spillane, J.P. (1996), “School districts matter: local educational authorities and state instructional
policy”, Educational Policy, Vol. 10 No. 1, pp. 63-87.
Spillane, J.P. (1998), “State policy and the non-monolithic nature of the local school district:
organizational and professional considerations”, American Educational Research Journal,
Vol. 35 No. 1, pp. 33-63.
Spillane, J.P. (2004), Standards Deviation: How Local Schools Misunderstand Policy, Harvard
University Press, Cambridge, MA.
Spillane, J.P. (2006), Distributed Leadership, Jossey-Bass, San Francisco, CA.
JEA Spillane, J.P. and Anderson, L.M. (under review), “Policy in practice: school administrative
practice & policy implementation.
50,5
Spillane, J.P. and Diamond, J.B. (2007), Distributed Leadership in Practice, Teachers College Press,
New York, NY.
Spillane, J.P. and Hunt, B. (2010), “Days of their lives: a mixed-methods, descriptive analysis of
the men and women at work in the principal’s office”, Journal of Curriculum Studies, Vol. 42
560 No. 3, pp. 293-331.
Spillane, J.P. and Kim, C.M. (under review), “Organizing for instruction: instructional advice and
information seeking in elementary schools for mathematics and English Language Arts”,
American Journal of Education.
Spillane, J.P., Camburn, E.M. and Pareja, A.S. (2007), “Taking a distributed perspective to the school
principal’s workday”, Leadership and Policy in Schools, Vol. 6 No. 1, pp. 103-25.
Spillane, J.P., Parise, L.M. and Sherer, J.Z. (2011), “Organizational routines as coupling
mechanisms: policy, school administration, and the technical core”, American Educational
Research Journal, Vol. 48 No. 3, pp. 586-620.
Spillane, J.P., Diamond, J.B., Sherer, J. and Coldren, A.F. (2004), “Distributing leadership”, in
Coles, M. and Southworth, G. (Eds), Developing Leadership: Creating the Schools of
Tomorrow, Open University Press, New York, NY, pp. 37-49.
Spillane, J.P., Diamond, J.B., Burch, P., Hallett, T., Jita, L. and Zoltners, J. (2002), “Managing in the
middle: school leaders and the enactment of accountability policy”, Educational Policy,
Vol. 16 No. 5, pp. 731-62.
Strathern, M. (Ed.) (2000), Audit Cultures: Anthropological Studies in Accountability, Ethics and
the Academy, Routledge, New York, NY.
Timperley, H.S. (2005), “Distributed leadership: developing theory from practice”, Journal of
Curriculum Studies, Vol. 37 No. 4, pp. 395-420.
Tomlinson, S. (2001), Education in a Post-Welfare Society, Open University Press,
Buckingham.
Tyack, D. and Tobin, W. (1994), “The ‘grammar’ of school: why has it been so hard to change?”,
American Educational Research Journal, Vol. 31 No. 3, pp. 453-79.
Valenzuela, A. (Ed.) (2004), Leaving Children Behind, How “Texas-Style” Accountability Fails
Latino Youth, SUNY Press, Albany, NY.
Valli, L. and Buese, D. (2007), “The changing roles of teachers in an era of high-stakes
accountability”, American Educational Research Journal, Vol. 44 No. 3, pp. 519-58.
Weick, K.E. (1979), The Social Psychology of Organizing, McGraw-Hill, New York, NY.
Whitty, G. and Power, S. (2003), “Making sense of education reform: global and national
influences”, in Torres, C.A. and Antikainen, A. (Eds), International Handbook on the
Sociology of Education, Rowman & Littlefield, Lanham, MD, pp. 305-24.
Wilson, S.M. and Floden, R.E. (2001), “Hedging bets: standards-based reform in classrooms”, in
Fuhrman, S.H. (Ed.), From the Capitol to the Classroom: Standards-Based Reform in the
States, University of Chicago Press, Chicago, IL, pp. 193-216.
Wong, M., Cook, T.D. and Steiner, P.M. (2009), An Evaluation of No Child Left Behind Using Short
Interrupted Time-Series Analyses with Non-Equivalent Comparison Groups, Northwestern
University, Evanston, IL.

About the authors

James P. Spillane is the Spencer T. and Ann W. Olin Professor in Learning and Organizational
Change in the Human Development and Social Policy program in the School of Education and
Social Policy at Northwestern University. He also serves as a Professor of Management and
Organizations and a Faculty Fellow at Northwestern University’s Institute for Policy Research. School
His research interests include education policy, policy implementation, organizational change,
school reform and school leadership. James P. Spillane is the corresponding author and can be
administration
contacted at: [email protected]
Allison W. Kenney is the Research Program Coordinator for the Distributed
Leadership Studies at Northwestern University. She received her BA in Public Policy from
Duke University and her research interests include education policy, school leadership and 561
school reform.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints
 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0957-8234.htm

JEA
50,5
At-risk student averse: risk
management and accountability
Julian Vasquez Heilig
Department of Education Administration, The University of Texas at Austin,
562 Austin, Texas, USA, and
Received 8 December 2010 Michelle Young and Amy Williams
Revised 3 April 2011 The University of Texas at Austin, Austin, Texas, USA
2 May 2011
18 May 2011
Accepted 24 May 2011 Abstract
Purpose – The prevailing theory of action underlying accountability is that holding schools and
students accountable will increase educational output. While accountability’s theory of action
intuitively seemed plausible, at the point of No Child Left Behind’s national implementation, little
empirical research was available to either support or critique accountability claims or to predict
the long-term impact of accountability systems on the success of at-risk students and the schools that
served them. The purpose of this paper is to analyze the work and perceptions of school teachers and
leaders as they seek to meet the requirements of educational accountability.
Design/methodology/approach – Interviews with 89 administrators, staff and teachers revealed a
variety of methods utilized to manage risks associated with low test scores and accountability ratings.
Findings – The findings reported in this paper challenge the proposition that accountability
improves the educational outcomes of at-risk students and indicates that low-performing Texas high
schools, when faced with the press of accountability, tend to mirror corporate risk management
processes, with unintended consequences for at-risk students. Low-scoring at-risk students were often
viewed as liabilities by school personnel who, in their scramble to meet testing thresholds and
accountability goals, were at-risk student averse – implementing practices designed to “force kids out
of school.”
Originality/value – In this paper, the authors use theory and research on risk management to
analyze the work and perceptions of school teachers and leaders as they seek to meet the requirements
of educational accountability. This paper is among the first to use this particular perspective to
conceptualize and understand the practices of educational organizations with regards to the treatment
of at-risk students attending low-performing high schools in the midst of accountability.
Keywords United States of America, Secondary schools, Urban areas, Rural areas, Students,
Educational administration, Ethnic groups, Minorities, Urban education, Accountability, Risk analysis
Paper type Research paper

In 1983, “A Nation at Risk” made the case that the US school system was not
competitive globally and that US students’ achievement was in decline (National
Commission on Excellence in Education, 1983). Prepared by an influential committee
created by Secretary of Education, Terrell Bell, and subsequently endorsed by
President Ronald Reagan, this report indicted the American public education system
for failing to adequately prepare the country’s children to be competitive in the
twentieth century. Citing declining standardized test scores, unfavorable international
comparisons, and waning public support for America’s schools, the report argued that
American students’ low performance was a direct result of weak curriculum, poor
Journal of Educational
Administration educational programs, and an untalented educator workforce.
Vol. 50 No. 5, 2012
pp. 562-585
r Emerald Group Publishing Limited
0957-8234 The authors thank Roberta Rincon, Suyun Kim, Kori Stroub and Linda McNeil for their feedback
DOI 10.1108/09578231211249826 and assistance with the work.
 Soon after the release of the Nation at Risk report, there was a push by business At-risk student
leaders in Texas to reform the state’s schools. Ross Perot, a prominent Dallas averse
businessman and former candidate for president of the USA, and his allies were
“influential actors” and proponents of accountability and testing in Texas (Carnoy and
Loeb, 2003). The Perot Commission, and later the Texas Business-Education Coalition,
united corporate leaders in Texas in an effort to promote a business perspective in
education reform (Grissmer and Flanagan, 1998). Codified in this reform effort was a 563
determination to inculcate reform measures in the public consciousness that increased
efficiency, quality and accountability in a push for schools to perform more like
businesses (Grubb, 1985). Salinas and Reidel (2007) argued that the Texas business
elite utilized “the policy process, power relationships and educational value conflicts”
to promote accountability as the “paradigm for educational policy reform” (p. 1). As a
result, Texas was one of the earlier states to develop statewide testing systems during
the 1980s, and the state adopted minimum competency tests for school graduation
in 1987.
Texas business leaders were (and continue to be) interested in the efficient use of
funding to public schools. Additionally, the state had been challenged to craft school
finance legislation that would survive the state’s supreme court (Vasquez Heilig et al.,
2010). Former Lt Governor Bill Ratliff, Republican senate sponsor of Texas Senate Bill
7 (SB 7) (1993), related that the inclusion of the accountability system in SB 7 was
necessary to gain passage of the proposed school finance system, also known as the Robin
Hood plan because it redistributed funds from richer districts to poor districts. He stated:
If you are going to pass a school finance bill, it’s almost inevitable to get the votes to pass one
that you have to put considerably more money into the system, because if you don’t you have
school districts who are winners and you have school districts who are losers, and the losers,
the members who represent losing districts can’t vote for it [y]. But many members were
skeptical about putting that much new money in unless we required some kind of an
accountability for the money. That is [y] are we getting the bang for our buck? (personal
communication, November 14, 2007).
Thus, in addition to reforming school finance, SB 7 modified the existing public school
accountability system from a diagnostic to a performance-based system. Signed into
law by Democratic Governor Ann Richards in 1993, SB 7 represented a bipartisan
solution to the state’s educational woes as it was passed by a wide margin in both the
Texas House and Senate. When asked about whether there were any legislators against
the accountability system at the time that SB 7 was considered, Lt Governor Ratliff
stated the following:
Well, there were some that were against the accountability, but frankly the accountability was
sort of overshadowed by the school finance [y]. Most of the votes in the Senate [y] were
votes against the Robin Hood plan, not against the accountability system. The accountability
system, except for some members who wanted some accountability, sort of flew under the
radar because the school finance bill was so controversial (personal communication,
November 14, 2007).
Despite the fact that accountability “flew under the radar,” SB 7 mandated the creation
of the Texas public school accountability system to rate school districts and evaluate
campuses. The first Texas accountability system was enacted in 1994, under the
leadership of Governor Ann Richards, was an information forum that utilized test
scores and other measures of student progress to determine whether school districts
should remain accredited by the state[1]. The Texas accountability system was
JEA undergirded by the Public Education Information Management System (PEIMS) data
50,5 collection system, a state-mandated curriculum, and a statewide standardized test to
measure student proficiency in core subjects.
From 1995 to 1999, Texas test-based accountability commenced under Governor
George W. Bush[2]. During this period, educational policy in the state evolved beyond
district-level consequences to applying a variety of sanctions on teachers, principals
564 and schools[3]. Achievement gains across grade levels conjoined with increases in high
school graduation rates and decreases in dropout rates brought nationwide acclaim to
the Texas accountability “miracle” (Haney, 2000, p. 1). Citing the success of Texas-style
high-stakes testing and accountability rating formulas, former President George W.
Bush chose Rod Paige, the Houston Independent School District Superintendent, as his
first Secretary of Education. During his first day on the job, Rod Paige declared that
Texas-style accountability had made a difference for at-risk students during his tenure
in Houston. He stated, “I personally witnessed in the last seven years schools where
most would say these students had all the at-risk characteristics associated with
failure, and they shouldn’t grow. In fact they did.” He explained that accountability had
highlighted “islands of excellence” and that the Texas system of sanctions and rewards
would be integrated into a national Bush education plan (Suarez, 2001).
According to McNeil (2005), Texas-style high-stakes testing and accountability
policy, by force of federal law, became the driving education policy for the entire nation
through the reauthorization of the Elementary and Secondary Education Act (ESEA)
of 2002 – also known as No Child Left Behind (NCLB)[4]. NCLB (2002) replicated the
Texas model of accountability, injecting public rewards and sanctions into education
policy for states, districts and schools nationwide. Eight years after the passage of
NCLB, accountability policies are influencing the process of public schooling
nationwide. High-stakes standardized tests have become the foundation for decisions
determining the progression of children through school, access to education, student
achievement progress and the amount of resources a school receives to educate its
student body (Darling-Hammond, 2003).
The prevailing theory of action underlying Texas-style testing and accountability
ratings is that holding schools and students accountable through measures of
achievement will increase the quality of education in the USA because educators will
try harder, schools will adopt more effective methods and students will learn more
(Vasquez Heilig, 2011). In sum, the argument holds that identification of student
success by disaggregating test scores by demographic groups will pressure schools
and educators to improve test scores (Hanushek and Raymond, 2003). While
accountability’s theory of action intuitively seemed plausible, at the point of NCLB’s
national implementation, little scientifically based research was available to establish
the efficacy of accountability claims or to predict the long-term effects of accountability
systems on the success of poor and ethnic/racial minority students and the schools that
served them (Ravitch, 2010).
Indeed, the effects of high-stakes testing policies in Texas have been debated
(Carnoy et al., 2001; Haney, 2000; Klein et al., 2000; McNeil et al., 2008; Linton and
Kester, 2003; Toenjes and Dworkin, 2002; Vasquez Heilig and Darling-Hammond,
2008). Researchers, educators, parents and policy makers alike have asked whether
policies that reward and sanction schools and students, based on average school-level
test scores and disaggregated by student demographic groups, can improve
achievement and the quality of education for all or most students. Also debated is
the question of whether accountability and high-stakes testing policies institutionalize
other, less helpful, school practices, such as narrowing the curriculum and gaming the At-risk student
system – actions that systematically manipulate student populations to achieve averse
accountability goals (McNeil and Valenzuela, 2001; Valencia and Bernal, 2000).
Evidence on the effects of high-stakes testing and accountability policies on school
responses suggests that high-stakes testing systems that reward or sanction schools
based on average student scores may create incentives for schools to boost scores by
manipulating the population of students taking the test. In addition to retaining 565
students in grade so that their relative standing will look better on “grade-equivalent’
scores, schools have been found to label large numbers of low-scoring students for
special education placements so that their scores are not factored into school
accountability ratings (Allington and McGill-Franzen, 1992; Figlio and Getzer, 2002),
exclude low-scoring students from admission to “open-enrollment” schools, and
encourage poorly performing students to leave school, transfer to GED programs or
dropout (Darling-Hammond, 1991; Haney, 2000; Booher-Jennings, 2006; Vasquez Heilig
and Darling-Hammond, 2008). This paper focusses on developing a framework that
conceptualizes these events, often called accountability “gaming.”
A growing body of literature indicates that many educational professionals under
the pressure of accountability are engaging in practices that enable their schools to
survive but do not adequately support high-quality education. In this paper, we
examine the responses of educational personnel to the current high-stakes
accountability environment in Texas, paying particular attention to risk-
management behaviors, what we are calling at-risk student averse.
Specifically, in this project we use theory and research on risk management (Ericson
and Leslie, 2008; Power, 2007; Stowe and Jeffery, 2008; Stulz, 1996; Williams and Heins,
1989) to analyze the work and perceptions of teachers and school leaders as they seek
to meet the requirements of educational accountability in Texas. We believe this paper
is among the first to use this particular perspective to conceptualize and understand
the practices of educational organizations with regard to the treatment of at-risk
students attending high schools serving a majority of at-risk students in the midst of
accountability. Specifically, this study asks the following questions: how might risk-
management theory inform our understanding of the effects of accountability pressure
on teachers and school leaders? To what extent has the monitoring and measuring of
student and school performance led to the treatment of at-risk students as liabilities for
meeting the objectives of accountability policies? In their efforts to meet the objectives of
accountability policy, have school personnel implemented practices that deemphasize
quality education? To answer these questions we interviewed 89 administrators, school
staff members and teachers and used risk-management theory to analyze their
responses. Our paper begins with a review of research revealing the unintended negative
outcomes of high-stakes accountability. This is followed by an overview of risk-
management literature and a delineation of how this literature is applied in our paper.
Following a description of our methods, we share and discuss our findings.
Unintended consequences of high-stakes accountability
Carnoy et al. (2001) proposed two scenarios about the possible relationship between
test-based accountability – then underpinned by the Texas Assessment of Academic
Skills (TAAS) test – and student success in Texas. They stated:
In the first, an emphasis on increasing TAAS scores increases the overall quality of schooling,
leading to gains in student learning on multiple levels and decreases in the dropout rate. In an
alternative scenario, however, increased emphasis on TAAS comes at the expense of other
JEA learning or leads to efforts to screen students before they take the TAAS. This may lead to
increases in the dropout rate, either as low performing students are forced out of schools in
50,5 order to increase school average TAAS scores or as students choose to leave (p. 18).
Research reveals that Carnoy et al. (2001) adequately predicted some deleterious
effects of high-stakes accountability. Indeed, these scholars suggested that while
accountability legitimated complex technocratic responses throughout the educational
566 system, that these responses might not have elicited the increase in educational output
originally envisioned by school reformers. Instead, such technocratic responses may
have escalated historical inequities by legitimating problematic practice. To illustrate,
the press of accountability has led to schools identifying and “pushing” students out of
school that are seen as liabilities due to their low-performance on standardized tests
(Gotbaum, 2002; McNeil et al., 2008; Vasquez Heilig and Darling-Hammond, 2008).
Studies have also revealed evidence of specific gaming actions in Texas schools.
For example, Booher-Jennings (2006) identified teachers who sought to create the
appearance of test score improvement by using “educational triage” practices. In some
classrooms, this was accomplished by diverting resources to elementary students
believed to be on the threshold of passing the TAAS (bubble kids) and to “accountable”
students (those affecting the school’s accountability rating). She also found that
teachers sought to affect their school’s accountability rating by referring non-bubble
students for special education, though this was not necessarily the teachers’ own idea.
Booher-Jennings (2006) detailed the advice given by a consultant to a San Antonio
Elementary School on how to improve test scores using educational triage:
“Using the data, you can identify and focus on the kids who are close to passing. The bubble
kids. And focus on the kids that count [y]” (p. 4). “Take out your classes’ latest benchmark
scores”, the consultant told them, “and divide your students into three groups. Color the ‘safe
cases’, or kids who will definitely pass, green. Now, here’s the most important part: identify
the kids who are ‘suitable cases for treatment’. Those are the ones who can pass with a little
extra help. Color them yellow. Then, color the kids who have no chance of passing this year
and the kids that don’t count – the ‘hopeless cases’ – red. You should focus your attention
on the yellow kids, the bubble kids. They’ll give you the biggest return on your investment”
(p. 5).
Booher-Jennings (2005) reported that educational triage has become an increasingly
widespread response to accountability systems documented in Texas, California,
Chicago, Philadelphia, New York and England. Ironically, accountability appears to be
creating an educational environment that is fostering risk management rather than
excellence. More importantly, schools serving a majority of at-risk students appear to
be becoming at-risk student averse due to the press of high-stakes testing and
accountability. Several decades of research and business management literature
suggest that these organizational responses to NCLB are both predicable and rationale.
Business ideology and schools
Business ideology is often used in educational research and practice to outline potential
courses of action (Ravitch, 2010) and to present preferred approaches to educational
leadership (Tyack and Hansot, 1982). Gumport (2000), for example, conceptualized
organizational change in institutions of education as fertile territory for private-sector
inspired managerialism, though she cautioned that wholesale adaptation to managerial
rationales could subsume the discourse of educational institutions’ “logic of economic
rationality at a detriment to the longer-term educational legacies and democratic
interests that have long characterized US public education” (p. 1).
 Empiricist scholars have studied the use of management theory and practices in At-risk student
education by direct examination (i.e. Murphy and Beck, 1995), while others, like averse
Gumport (2000) and Cuban (2004) have used business ideology to intuit what was and
is likely to occur. Our use of management theory reflects the latter approach. We
consider the emergence of school-based risk-management practices as a response to the
recent entrenchment of long-standing business ideology – data management practices,
accounting measurement formulas and high-stakes testing – in educational policy. 567

Risk management
Risk analysis emerged four decades ago as a field that drew on the statistical sciences
to measure risk, expressing results as probability statements (Ericson and Leslie,
2008). However, this approach suffered from the fact that it is rarely directly applicable
to the situations defined by decision makers on the ground (Power, 2007). As a result
there has been a rush to invent new technologies that emphasize the management and
governance of risk. Crockford (1986) defined the main features of risk management as
the identification, evaluation, minimization and mitigation of risk. Risk management
involves a holistic evaluation of an organization to measure the potential loss that an
organization faces if an event occurs (Gorrod, 2004). According to this perspective,
decision makers, who remain objective, can use risk management to reach
organizational goals, while minimizing potential negative outcomes (Flyvbjerg,
2006). In a context of concern over organizational effectiveness and the possibility of
blame if things go wrong, Power (2007) explains this shift in focus to governing the
processes of risk management has led to a rise in regulatory mechanisms that
emphasize audits and accountability.
Risk “objectification, rationalism and standardization are crucial in process of risk
governance” (Ericson and Leslie, 2008, p. 616). To illustrate, the risk-management
process typically encompasses six courses of action:
(1) determining the objectives of the organization;
(2) identifying exposures to loss;
(3) measuring those same exposures;
(4) selecting alternatives;
(5) implementing a solution; and
(6) monitoring the results (Williams and Heins, 1989).
An institution will modify the strategy for managing risk based upon the primary
objective of institutional success (as seen in Figure 1). In business, the chief objective of
a public company may be increasing profit; for schools a chief objective may be
increasing achievement scores and accountability ratings. The risk-management
process in both cases would involve identifying and measuring exposures, selecting
alternatives, implementing solutions and monitoring results.
As one might imagine, such processes may produce valid and useful information;
however, as Ericson and Leslie (2008) assert, information is only as good as its
interpretation and use:
Knowledge is always subject to mediation and massaging as part of political strategies
“to govern unruly perceptions [of risk] and to maintain the production of legitimacy in the
face of these perceptions.” Reputation protection and the management of public perceptions
JEA
50,5
Objectives Exposure
to loss

568
Monitoring Measuring exposure

Solutions Alternatives
Figure 1.
Process of risk
management
Source: Williams and Heins (1989)

have become primary goals of the re-invented risk analysis and its redeployed scientific
authority (Ericson and Leslie, 2008, p. 616).
Although consideration of risk can be traced back to ancient Greece (Bernstein, 1998),
contemporary practices most closely reflect the content of government reports (e.g. the
Treadway Committee Report, 1991) that examine fraudulent financial reporting and
legislation, such as the Sarbanes-Oxley Act of 2002 (SOX). Such reports quickly
became blueprints for risk management in business, government and non-
governmental organizations, advancing the importance of internal controls and
processes to improve compliance.
Risk management in education
The titular usage and focus on accountability in the SOX legislation is conspicuously
similar to the theory of action underlying the re-authorization of the ESEA as NCLB a
mere six months prior. SOX focussed on accuracy and accountability in auditing – in
parallel, NCLB required a new system of accounting in every state to improve schools
by measuring student outcomes and progress through school. George W. Bush, the
first US president to hold an MBA, declared in the signing ceremony for NCLB that
accountability was his “first principle” for schools (White House, 2002).
Although there are a growing number of risk-management processes and strategies,
public companies seeking to comply with SOX 404, a section of the federal law that
requires assessments of internal controls and risks (Thomson, 2007), typically choose
either a SWOT (Strengths, Weaknesses, Opportunities and Threats) analysis or
enterprise risk management (ERM). Each of these risk management processes
encompass the conceptual logic delineated by Williams and Heins (1989), but with a
greater specificity of processes. SWOT analyses indentify internal and external
opportunities and threats that the company can either embrace or prepare to respond
to through the development of a risk-management plan designed to mitigate
weaknesses and potential threats to the organization. ERM is a risk-management
approach wherein regulatory mechanisms are strategically developed throughout
an organization to bring the activities of all organizational members under the
jurisdiction of management, thus integrating concepts of internal control (Ericson and At-risk student
Leslie, 2008). ERM is used by businesses to determine the risk capacity of an averse
organization, including norm violation and system irregularities – then objectives are
set to comply with the firm’s goals and risk.
Similar to SOX for the business community, NCLB created accountability
requirements for schools such as reporting adequate yearly progress (AYP) measures
to allow the US Department of Education to examine how every school is performing 569
academically according to results on standardized tests. And similar to public
companies, schools are hyper-aware of the attention being paid to their performance as
well as the consequences for missteps. Moreover, like public companies seeking to
meet the accountability demands of SOX, some schools are using risk-management
strategies. How these two pieces of legislation differ is in the requirement to
assess risks, whereas SOX requires that public companies assess and mediate risks,
NCLB has no such explicit requirement. However, considering the data collection
and accounting mechanisms institutionalized by NCLB, the environment is ripe for
risk-management logic and subsequent behaviors, particularly in high schools serving
a majority of at-risk students in Texas.
Applying the literature on organizational risk-assessment practice and theory
within educational organizations, this paper seeks to understand whether and how
Texas high schools have assessed the risk of at-risk students when faced with the
pressure of NCLB-inspired policies. This analysis aims to examine if Texas high
schools mirror corporate risk-management processes as a means to improve school
performance measures. Furthermore, this paper seeks to understand whether risk-
management practices have led to the treatment of certain students as liabilities and
what deleterious practices, such as developing an aversion to at-risk students, have
emerged as school personnel have attempted to meet the objectives of accountability
policies.

Method
This paper coalesces data from two studies that examine the effects of Texas-style
high-stakes accountability from 1995 to 2008. The analysis seeks to understand how
schools assess and respond to the risk associated with at-risk students for test-driven
accountability. The research presented here is a précis of several years of in-depth
qualitative interviews with 89 administrators, staff and teachers from seven high
schools in Brazos City (pseudonym for a large, urban Texas school district) and four
rural, small city and suburban public high schools located along the USA-Mexico
border region of South Texas.
Each of the 11 high schools in this analysis had student populations with more than
50 percent of their students categorized as at-risk. The Texas Education Agency
(TEA), the state governmental agency responsible for overseeing education in Texas,
calculates an at-risk indicator for each student in the state of Texas based on 13 state-
determined criteria (e.g. retention, poor performance on state assessment, pregnancy,
etc.) to identify whether a student is “at-risk” of dropping out of school (TEA, 2007).
Ten high schools where the majority of students were ethnic/racial minorities were
contacted in Brazos City; seven high schools agreed to participate. The sample
includes Latina/o majority high schools and African American majority high schools.
The selection of the sample high schools in the Rio Grande Valley began by randomly
selecting six candidate schools based upon locality (rural, small city and suburban).
JEA Once schools were identified, four high schools agreed to participate in the research.
50,5 One school declined and the sixth did not respond to repeated e-mail or phone calls.
To understand the achievement distribution of the schools, we present the schools
results on state and federal accountability measures (see Tables I and II). As discussed
above, the state of Texas has had an accountability system in place since the early
1990s. The Texas accountability system uses formulas to measure a variety of student
570 outcomes such as test scores, graduation rates and dropout rates. The indices are then
disaggregated by student groups (i.e. race/ethnicity, English Language Learner, etc.)

High school 2005-2006 2006-2007 2007-2008 2008-2009 2009-2010

Camino Academically Academically Academically Recognized Recognized

acceptable acceptable acceptable
Del Oro Academically Academically Academically Academically Recognized
acceptable acceptable acceptable acceptable
Palma Academically Academically Academically Academically Academically
acceptable acceptable acceptable acceptable acceptable
Tierra Academically Academically Academically Recognized Recognized
acceptable acceptable acceptable
Edgeview Academically Academically Academically Recognized Recognized
acceptable unacceptable acceptable
Clearbend Academically Academically Academically Academically Academically
acceptable acceptable unacceptable unacceptable unacceptable
Crockett Academically Academically Academically Recognized Acceptable
unacceptable unacceptable unacceptable
King Academically Academically Academically Recognized Recognized
acceptable acceptable acceptable
Carver Academically Academically Academically Academically Academically
unacceptable acceptable acceptable acceptable acceptable
Douglas Academically Academically Academically Academically Academically
unacceptable acceptable acceptable unacceptable unacceptable
Table I. Lincoln Academically Academically Academically Academically Academically
Respondent high schools’ unacceptable acceptable unacceptable acceptable acceptable
state accountability Unacceptable
ratings (2005-2010) rating 4/11 2/11 3/11 2/11 2/11

High School 2005-2006 2006-2007 2007-2008 2008-2009 2009-2010

Camino Missed AYP Missed AYP Missed AYP Meet AYP Missed AYP
Del Oro Meet AYP Missed AYP Missed AYP Meet AYP Meet AYP
Palma Meet AYP Missed AYP Meet AYP Missed AYP Meet AYP
Tierra Meet AYP Missed AYP Missed AYP Meet AYP Meet AYP
Edgeview Missed AYP Missed AYP Missed AYP Meet AYP Meet AYP
Clearbend Meet AYP Missed AYP Missed AYP Missed AYP Missed AYP
Crockett Missed AYP Missed AYP Missed AYP Missed AYP Meet AYP
Table II. King Meet AYP Missed AYP Missed AYP Meet AYP Meet AYP
Respondent high Carver Missed AYP Missed AYP Missed AYP Missed AYP Missed AYP
schools’ federal adequate Douglas Meet AYP Missed AYP Missed AYP Meet AYP Missed AYP
yearly progress ratings Lincoln Missed AYP Missed AYP Missed AYP Meet AYP Missed AYP
(2005-2010) Did not meet AYP 5/11 11/11 10/11 4/11 5/11
and entered into formulas to determine ratings for districts and schools. For example, At-risk student
the Texas accountability system ratings for 2010 are: “unacceptable,” “acceptable,” averse
“recognized” and “exemplary.”[5]
The federal NCLB act requires that each state establish an AYP timeline for
ensuring that all students in the USA would meet or exceed state standards within
12 years of implementation of the federal education law. Although AYP is defined by
individual states, the measurement of AYP is similar across the states. Specifically, 571
high-stakes tests and other measures of student progress (i.e. graduation rates and
attendance rates) are used to determine whether every public school and school district in
the USA is making yearly progress toward all students meeting state standards by 2014[6].
The randomly selected high schools examined in our work have had a range of
accountability results over the past five years. Table I shows that the high schools in
the sample were ranked academically unacceptable (the lowest accountability rating)
12 times over a five-year period – about 22 percent of the time. Since 2005, the 11 high
schools in the sample performed somewhat worse on the federal accountability
measure. The sample high schools did not meet their AYP goals 35 times, about
64 percent of the time (see Table II). In summary, the schools in the sample are not
consistently “below the line.” In fact, during the past two years, six of the 11 schools
received a “recognized” Texas accountability rating.
The studies utilized a key informant strategy, wherein principals and counselors in
each high school were asked to identify veteran school staff that could share their
institutional memory about the evolution of accountability in their schools (El Sawy
et al., 1986). These key informants were then used to snowball sample other individuals
who could provide a variety of perspectives on the affect of accountability and high-
stakes testing on their schools (Goodman, 1961). Interviews with high school personnel
were conducted on campus over several days and were spread throughout the
academic year. Table III shows the total numbers of teachers and administrator/staff
members interviewed at each school in both studies. To ensure privacy and
confidentiality, all high schools are referred to by pseudonyms.

Data analysis
Glaserian comparative analysis method involves concurrent data gathering and
analyzing and emphasizes induction and emergence during the research process
(Strauss and Corbin, 1990). Thus, analysis begins the moment one enters the field.

High school Teachers Administrators/staff Total

Camino 3 7 10
Del Oro 8 4 12
Palma 12 4 16
Tierra 10 3 13
Edgeview 2 6 8
Clearbend 3 1 4
Crockett 8 3 11
King 2 2 4
Carver 3 0 3 Table III.
Douglas 3 1 4 Summary of qualitative
Lincoln 3 1 4 interviews by Texas
Total 57 32 89 high schools
JEA School visitations during data collection facilitated the analysis process by enabling
50,5 the contextualization of the experiences of students and school staff in relation to
high-stakes testing and accountability.
Comparative analysis method is grounded in flexibility as the research evolves
(Glaser, 1992). As a result, the interview and focus group discussions in both studies
drew from a bank of open-ended questions based on high-stakes exit testing and
572 accountability. As the research proceeded, most but not all questions in the bank were
used, and several were amended to capitalize on relevant issues and ideas that arose
during the research process. In essence, to gather richer data, the research was
grounded in emerging themes revealed by the participants.
The transcripts in both studies were analyzed using the constant comparative
method ( Janesick, 1994; Patton, 1990). Once the interviews were transcribed, several
individuals coded phrases that had meaning in relation to the main purposes of the
study. These categories were modified as each new interview transcript was analyzed.
This process is described by Lincoln and Guba (1986) as the “saturation of categories”
or the “emergence of regularities” (p. 350). Comparative pattern analysis was used to
illuminate recurring patterns in the data.
To triangulate the research process, research team members (a mix of faculty and
graduate students) from Rice University, Stanford University and the University of
Texas at Austin independently conducted axial coding that identified consistent themes
within the phrase coding. Informant data were triangulated with fieldnotes, archival
materials provided by schools and local press reports. For synthesis, informant counts
by category were conducted to understand the representativeness of the dominant
themes generated in the field interviews. The full set of themes were detailed and
published in two prior peer-reviewed articles. Subsequently, an additional overarching
theme was revealed through the analysis of both data sets: the idea that schools were
developing an aversion to at-risk students. This, third paper is dedicated to this theme.

Findings
Advocates of test-based accountability hold that educators should be held accountable
for student achievement and that student achievement should be measured through
standardized tests (Stecher et al., 2003). In theory, test-based accountability inspires
educators to feel personal and collective responsibility for how much students learn.
Supporters of test-based accountability point to successful private-sector management
practices as a model for schools and argue that “student achievement will improve
when educators are judged in terms of student performance” and when these
“judgments carry some consequences for educators” (p. 3). However, our research
provides a different picture. Rather than fostering increased focus and commitment to
improving student achievement and to higher-quality teaching, test-based
accountability in the schools we studied is undermining these important goals.
Specifically, the press of accountability has at least three unintended and negative
consequences. It is putting intense pressures on educators and fostering an
environment of fear in many schools, it has resulted in the search for and use of
loopholes for navigating accountability, and it has led many educators to view students
a liabilities. Each of these consequences is delineated below.

The press of accountability

The pressure of accountability in Texas high schools was readily apparent in every
interview. In each of the high schools studied, accountability placed pressure on
teachers and administrators to increase test scores and ratings. A math teacher, with At-risk student
almost 20 years of experience teaching in Carver and King high schools, described the averse
pressure on teachers. He related, “Unfortunately, [accountability] puts the pressure on
the teachers. We are in a tough situation as teachers. If you are low-performing, you
lose your job.” A teacher, with 20 years of experience at Douglas and Lincoln, also
described the pressure on teachers and students:
A lot of students say to me [y] these tests are making me just not want to be here. These 573
tests are making me just want to give up. And then throw the pressure in on top of that. The
principal says, “If our scores don’t go up, I’m going to be fired. And if I’m going to be fired,
you’re going to pay for that.” And the teachers say, “if these scores don’t go up, we’re going to
get fired.”
The principal at Edgeview described the impact of at will employment on
school leaders (Rothstein et al. (1987) related that “at will” employment in US case
law is any hiring where the employer is free to discharge individuals “for good
cause, or bad cause, or no cause at all”. She related, “All of us are on at will
contracts. So if we – we can be let go at the end of the year. And its – it’s a lot of
pressure, and its – its not even subtle pressure. Its just hard pressure put on you to get
those scores up.”
The pressure of job loss was also underscored during fieldwork at Camino High
School. The visit coincided with the announcement from TEA that the high school was
under threat of school reconstitution due to the AYP results of students in the school.
The principal provided the research team access to five-hour-long conversations that
the principal convened with faculty from each department. Essentially, the principal
told the teachers assembled in her conference room that she was serious about
replacing staff if high-stakes test scores results did not improve. During her interview,
the principal related that the threat of reconstitution and the pressure of accountability
had necessitated these ominous conversations.

Seeking and using loopholes

Although the theory of action for test-based accountability holds that the pressure on
educators associated with accountability would increase positive student outcomes
such as test scores, lower dropout rates and higher graduation rates, other unintended
outcomes are common. To investigate the nature and prevalence of such unintended
outcomes, we asked school staff at urban, suburban and rural schools in Texas to
discuss how the pressure of accountability has led to organizational behavior
that the proponents of accountability may not have originally envisioned. School
administrators, staff and teachers revealed that they have sought loopholes to mediate
the risk associated with accountability. The testing coordinator at Edgeview High
School related:
I think each year we get a new set of regs, and we try and figure out how is the best way to use
it to our advantage [y] I mean, the game changes [y] it’s [y] like any – like a game that has
a set of instructions. And everybody gets the same set of instructions, and everybody follows
the same set of instructions [y] If you’re really savvy, and if you’re really into everything as a
principal you may see a problem [y] you may give your campus an advantage that another
campus doesn’t have.
Edgeview’s testing coordinator suggested that staff in other Brazos City high schools
studied the details of accountability to take advantage of loopholes. She revealed that
Fine Oak, a peer high school in Brazos City, had dramatically raised their test scores by
JEA not allowing students that had failed ninth grade core courses to test on the high
50,5 school exit exams. When the scores came out each year, Edgeview’s staff was
bewildered about how their peer high school had continued to dramatically
increase their exit test scores until in the Brazos City high schools gathered at the
district’s central administration building to discuss strategies to raise test scores.
Fine Oak shared their “waiver” policy that systematically excluded at-risk students
574 from exit testing. Edgeview’s testing coordinator related, “Why did we have
to wait two years to find out Fine Oak was doing it [excluding students from exit
testing]? [y] Fine Oak’s [test scores] were up here, and I kept saying, ‘How the hell is
that possible? Their kids aren’t any different than ours are’.” The testing coordinator
suggested that Brazos City high school staff viewed excluding students from exit
testing to manage the pressure and risk associated with accountability ratings as a
rationale choice when faced with the policy constraints. The vice principal at
Clearbend high school echoed the approach that Brazos City high schools have taken
toward accountability:
It’s human nature to [y] look at your game plan and to look at the rules of the game [y] You
know, and to say that using a loophole is not right or is a bad thing to do, I don’t necessarily
agree with, because it could be a good thing. It depends on the loophole [y] schools, yes, are
under pressure to look for creative ways to be successful, okay, that’s obvious.
A conversation with a Clearbend senior revealed how loopholes were utilized
at the high school. A Latina student related that her brother was suspended for
truancy by a Brazos City middle school. The student’s mother asked her to stay
home to ensure that the sibling avoided trouble during his school suspension.
She explained that she was having a difficult time passing the Texas Exit
examinations and that missing a week or more of school was going to be an issue for
her academically. Yet, she wanted to assist her working mother with her brother’s
disciplinary situation and stayed home with the sibling during his suspension.
She related that a few weeks later a letter arrived from Clearbend high school.
In this letter, the school detailed that she was going to be ticketed, fined and sent to
court for her absences. Notably, Clearbend offered her another option – she could
skip the ticket and fine by signing and returning an enclosed form to officially dropout
of school.
In this same conversation, the student mentioned that Clearbend would allow
students to make-up failed courses through “service” credits. These service credits
were determined by the school and were earned by weekend remediation activities. By
completing these service credits, her friend skipped from the tenth grade into the 12th
grade. Unfortunately, she also skipped over the Texas exit testing. Her friend was
classified as a senior, but she could not graduate because she never tested or passed the
exit exams, which are required for graduation in Texas (a quantitative analysis of the
large-scale exclusion of low-performing students in Brazos City School District testing
is also detailed in Vasquez Heilig and Darling-Hammond, 2008).

At-risk students as liabilities

School administrators revealed why many Texas high schools have sought to
marginalize low-performing and at-risk students. It seems that the risk associated with
job loss and public embarrassment for leaders whose schools do not perform as
expected under the Texas’s present accountability system – currently underpinned by
the Texas Assessment of Knowledge and Skills (TAKS) test – has resulted in the
identification of at risk students as liabilities. The principal of Edgeview high school At-risk student
provided an illustrative quote: averse
There is no incentive to keep low-performing students [y] because a lot of pressure is put on
by all the administration in this district. For instance, I was told by my immediate superior,
“I don’t care what other good things you’re doing, the only thing that matters is how your kids
do on the TAKS test. That is all that matters. That’s all the public cares about. That’s all that
Brazos City cares about [y].” 575
An apparent dilemma of accountability emerges. One the one hand, the disaggregation
of test score data is intended to encourage technical assistance for at-risk youth. While
disaggregated data is suppose to be used to design technical interventions, interviews
revealed that urban, rural and small city high schools serving majorities of at-risk
students in Texas have also used data to assess their exposure to risk and to make
decisions concerning the exclusion of students based on whether they were likely to
exhibit low test scores. A counselor at Edgeview related:
Schools were [not allowing transfers] just before the time for the standardized tests – because
they didn’t want students to come in. They might be students that were low-performing
[and] [y] would affect our scores [y] when we tested for TAKS last week, there
were students that needed to enroll. The test coordinator emailed me with the principal’s
sanction to let me know that we weren’t going to enroll until after the 28th [y] we were
allowed to ask those students to not enroll until the 28th. That means they did not take
that test.
The vice principal at Clearbend high school echoed that schools see no benefit, whether
they be exemplary or low-performing, in enrolling at-risk students. He stated:
I don’t think that schools that are blue ribbon schools in the state of Texas or that are
exemplary schools necessarily will take students like we have at Clearbend [y] I’ve heard
stories of schools in our district that turn our kids away. They find a way. They find a way,
and that’s wrong. That’s morally wrong, but they get away with it [y] and it starts at the top.
Similarly, the vice principal at Edgeview identified that risk associated with students’
low test scores adversely affected, in particular, the enrollment of mobile students and
the recovery of dropouts. Many high schools were not willing to enroll students that
they deemed to be at-risk and low-performing due the potential impact of such
students on their school’s test scores:
I encountered a student just a week ago, and he is 16 years old, this is his first year in the 9th
grade. His chances of graduating are slim. [y] Most of the 9th grade kids are like this, he is
going to give up by the time, if he does not make it to the 10th grade, he is going to be 17 years
old and he is going to be a dropout [y].

No school is going to want to take him. They are not going to want him. He is going to screw
up their test scores [y] There are no incentives [to keep him in school] [y] These kids move
from school to school and then dropout. Would you go to school and be 17 or 18 in the 9th
grade and sit with 15year-olds? There is something like 60% or 70% of our 9th graders [who]
are 16 years old.
The principal at Edgeview shared that enrolling at-risk students was in direct
competition with the incentives set up by the accountability system. Educators, he
explained, had to ponder the risks involved for their school when enrolling low-
performing students under the press of accountability. For most educators, this is in
direct opposition to their belief in the importance of educating all students.
JEA There are all these kids [in dropout recovery program] that are going to struggle on the
TAKS. That’s going to pull our scores way down. [But] what I want is to get those kids to
50,5 march across that stage. But, you know, it’s going to make us not look that good. We may be
on everybody’s list of low-performing or whatever [y] Are we about looking good in the
newspaper? Are we about really, really teaching kids and saving kids and bringing them in
and helping them to achieve?

576 One of the more surprising examples of a high school’s management of the liability of
at-risk students was revealed in a Tierra high school teacher focus group. Focus group
members revealed that their principal had used school TAKS test data and returned
mail to identify low-performing Latina/o students in the school and then to accuse
them of being illegal aliens. He called a meeting of the students and threatened to
report them to US immigration authorities regardless of whether they were actually
illegal aliens or not with the intention of encouraging their departure. Understandably,
Tierra teachers were very hesitant to discuss the details of this meeting. A teacher
described the situation:
The administration threatened the kids [y] I thought that was really [expletive]. If they
reported an address that they did not live at and they got back the letter, the school would
drop them and they would call immigration. What kind of an effect do you think that has on
kids? [y] It was coercion – it was a threat. And I see it that way, and I found it, I mean it
offended me. I spoke to our principal about it. He justified it completely, he did.
When asked, administrators and staff tended to attribute at-risk students’ lack of exit
test scores and increased risk of dropping out to characteristics of high school. The
principal at Edgeview underscored the obstacles created by high schools in response to
high-stakes testing and accountability:
I think that the kids are being forced out of school. I think that what has happened at Fine
Oaks is what happens at many schools. I had a kid who came here from Fine Oaks High
School and said, “Miss, if I come here could I ever take the TAKS?” And I said, “What do you
mean, if you come here you must take the TAKS.” And he said, “Well, every time I think I’m
going to take the TAKS, they either say, ‘You don’t have to come to school tomorrow, or you
don’t have to [take the test]’ [y] we’re told different things.” That’s when kids drop out [y]
when you never give them a chance [y] I think we’ve done a lot to force kids out of school
[y] When you talk the company talk, you forget what honesty is. And my fear is that [y]
we’ve forgotten what honesty is. I think that what has happened is that we’ve gotten all
caught up in [TEA accountability] that we don’t know what honesty is anymore [y] And I
think that we’ve gotten so caught up [y] that we’ve lost sight of what is the essence of what
we should be doing. And that is truly educating all these kids [y].
This principal provided a rich description of the ethical dilemmas associated with
accountability – such as deciding between potential school closure or forcing at-risk
and low-performing students out of school in order to improve test scores and
accountability ratings. Such searches for loopholes to foster short-term gains, rather
than investments of time and resources on longer-term substantive educational
changes, are a direct result of the environment of fear created by accountability, an
environment in which students are viewed as either assets or liabilities and where
flawed understandings of what is truly at risk thrive.

Discussion
Economic turmoil in the USA and beyond has been tied to private sector risk practices
gone awry. For example, during the mortgage boom, risk-management formulas
determined how many mortgages could be generated and how quickly they could be
bundled into securities and sold to another financial institution (Wyman, 2008). To At-risk student
generate short-term profit and bonuses, firms moved from the bounds of standard averse
banking practice and worked to generate more mortgages and increase earnings.
These mortgage lenders, working within a competitive environment, deemed the
greatest risk to be the loss of market share to other companies. Unfortunately, the real
risk was an impending economic downturn that would deplete home values, making
many mortgage loans worthless. Clearly, the financial industry was operating from 577
risk assessments that were based on faulty notions of what was truly at risk.
In our research, we explored whether a similar flawed notion of risk might be
operating in K-12 education. As explained in an earlier section of this paper, research
on risk management describes how organizations have come to think of, reform and
govern themselves through the vague but powerful notion of risk (Ericson and Leslie,
2008). Given this understanding, we explored the affect of risk management on the
behaviors of educational personnel, particularly those working with at risk students,
within the context of accountability policy. We found that schools in Texas had turned
to risk-management strategies in response to the accountability requirements
of NCLB. We also found that the adoption of risk-management systems has had several
unintended and negative consequences, including the development of an environment
of fear for educators in many schools, the search for and use of loopholes to foster
short-term gains, the definition of students as either assets or liabilities, and the
development of faulty notions of what is truly at risk.
Fear, insecurity and concern are seen more and more frequently in the discourse
surrounding public education and accountability (Young and Brewer, 2008). Put
simply, an uncertainty has developed around the “core technology” of education,
questions have been raised about whether educators know what they are doing and
systems have been put in place to embed organizational routines that manage risk
(Power, 2007). Many educators feel as if their work has changed dramatically, from a
focus on curriculum and instruction to one on assessment and intervention.
Furthermore, the intense focus on test results and how those results are used and
shared with the public has left many feeling disillusioned, anxious and uncertain. As
one teacher stated “I don’t care what other good things you’re doing, the only thing that
matters is how your kids do on the TAKS test. That is all that matters. That’s
all the public cares about.” Although working under such pressure is unfortunate for
teachers, the outcomes, as we have demonstrated in this paper, for students are
of greater concern. In our research we found that fear, fostered by the press of
accountability, has led to at-risk student aversion. To make matters worse, for many
educators, the exclusion of at-risk students from school appeared to be a rational
response, given the goals, pressures and constraints they faced.
The exclusion of at-risk students from school in order to manage impressions
regarding the effectiveness of a teacher or an entire school demonstrates a bias toward
short-term rather than long-range solutions. This is not uncommon risk-management
behavior. Indeed, pro-active firms in market-driven environments utilize risk-
management tools to identify potential threats and determine how to manage those
risks to exceed the performance of competitors in their current environment (Beasley
et al., 2008). Thus, there is a bias toward positive short-term outcomes, which may be
leading educators to emphasize the benefits of excluding student rather than the risks
of an action (Flyvbjerg, 2006). Returning to the process of risk management detailed in
Figure 1, the NCLB accountability environment has operationalized the objectives of
schools serving majorities of at-risk students by integrating standardized testing into
JEA formulas such as accountability ratings and AYP. There appear to be tangible benefits
50,5 for Texas high schools to systematically enact procedures that minimize certain risks,
such as school closure, in an environment of high-stake testing and accountability by
modifying the student management process and utilizing clear strategies that reduce a
school’s overall risk profile (Bowling and Rieger, 2005).
Considering Williams and Heins’ (1989) risk-management framework, the high
578 schools included in this study, which served a majority of at-risk students, exhibited
classic risk-management behaviors. Educational personnel measured exposure to risk
through the analysis of test score data and based on their results they made decisions
about how to minimize risk for the school, which involved excluding low-scoring
at-risk students, a behavior we have labeled at-risk student averse practice. School
leaders and their staff managed their exposure to loss by investing in the education of
students that had a positive material impact on the NCLB-defined objectives – better
test scores and ratings. As evidenced by the interviews in the urban, small city and
rural high schools, the pressure of accountability stirred school leaders to view
high-achieving students as assets and at-risk students as liabilities in the pursuit of
testing and accountability goals. In an effort to increase assets and decrease liabilities,
a number of administrators actively sought loopholes in accountability requirements
that would enable them to exclude low-performing students. As noted by the testing
coordinator at Edgeview, as student demographics change, so do the policies that high
schools implement to manage and measure risk. She posited that “savvy” schools
would develop management procedures that are constantly monitored and updated to
meet the changing rules and interpretation of accountability. Although some
administrators described this risk-management behavior as “human nature,” others
viewed these gaming responses as an attack on the honesty of the profession.
What is notable about the at-risk student aversion described in this paper is that
staff in a majority (eight of 11) of the urban, suburban and rural Texas high schools
reported activities in response to high-stakes testing and accountability that mirror
private-sector risk-management processes. More than two-thirds of all school
administrators and staff provided confirming examples of schools eliminating
‘accountable kids’ (i.e. those who would likely bring down the school accountability
ratings due to low test scores) in response to the current accountability system.
Although many educators seemed saddened that patterns of at-risk student aversion
had flourished in their schools, few questioned if whether schools had accurately
identified what was truly at risk in these schools. The at-risk student aversion
techniques documented here (i.e. seeking loopholes, denying enrollment to low-scoring
students, threatening Latina/o students with deportation) appear to conflict with the
traditional task of these Texas high schools – educating and graduating at-risk
students.
Yet, audits by TEA and careful empirical scrutiny of individual-level district data
throughout the more than 15 years of accountability reveal a lack of student progress
and low graduation rates for large numbers of at-risk students in Texas (Peabody, 2003;
Vasquez Heilig and Nichols, 2011). Recognizing problems in the data, TEA has
modified the PEIMS data system codes that identify student leaving from Texas
schools over the past 15 years (author). Furthermore, in 2005, Texas began to use the
National Center for Education Statistics (NCES) dropout definition for leaver reporting.
When the new standard was phased in, the yearly dropout count instantly tripled for
Latina/os and quadrupled for African Americans (author). Clearly, Latina/os and African
Americans had been over-represented in the underreporting of yearly dropouts.
 The Intercultural Development Research Association (IDRA) has argued that At-risk student
adopting the NCES national dropout definition for Texas has provided a more accurate, averse
yet still understated representation of the magnitude of the dropout problem in Texas
(Johnson, 2008). More than two decades of IDRA’s annual analysis of PEIMS high
school attrition data suggest that TEA has consistently and severely undercounted
student leaving in publicly reported dropout and graduation rates. For example, while
IDRA found the overall student attrition rate of 33 percent to be the same in 2007-2008 579
as it was more than two decades ago ( Johnson, 2008), TEA reported that annual
dropout rates had declined from 5 to 1 percent and longitudinal cohort dropout rates
that declined from about 35 percent to around 5 percent over the same time frame
(author). IDRA also posited that the high school attrition rates for Latina/o and African
American students accounted for more than two-thirds of the estimated 2.8 million
students lost from Texas public high school enrollment since the 1980s ( Johnson, 2008).
Thus, two important question arise: are the costs of high-stakes accountability
greater than the benefits, and are at-risk students best served by high schools enacting
risk-management processes similar to those utilized in the private sector? For schools
involved in our work, the costs of high-stakes accountability are clearly greater for
at-risk students. It is important to point out that the purpose of a public school is not to
produce a snapshot test score, or a school rating, or a gold star for a principal or
superintendent, but to educate children. Thus, any operationalization of student
outcomes should foster that collective goal as a public good, rather than fomenting an
environment in which students are viewed as either assets or liabilities. The reliance on
standardized accountability indicators based on arbitrary thresholds reduces the
capacity of creative schools and educators to act in complex, flexible ways, ways
that are essential to running a stable, yet nimble public enterprise, and for educating
children.

Conclusion
US educators are not alone – tests and accountability are intercontinental policies.
England, in fact, was one of the first countries to enact a test-based accountability
system, beginning in the late 1980s. Its high-stakes accountability system could
be considered most similar to NCLB in the USA. The English system utilized a
combination of national testing, curriculum, inspection and school choice policies
(Rustique-Forrester, 2005). With the exclusion of the original English system, most
international accountability systems, such as MySchool in Australia, do not have
entrenched sanctions attached to testing and other school-related information. Rather
the vast majority of international accountability systems are conduits to “learn” about
local schools[7]. Countries in Asia and Europe – such as France, Hong Kong, China,
Japan, Korea – have used national assessments to measure student and school progress
and to make decisions about each (Anderson, 2005). In South America, Chilean laws
have required schools to produce test results for increased funding (Garcı́a-Huidobro
and Bellei, 2006). A few countries in the Middle East and North Africa, such as Jordan
and Tunisia, have recently implemented low-stakes tests, incentives and accountability
measures (Shafiq, 2011), while Israel has contemplated revisions to national education
indicators (Justman and Bukobza, 2010).
The creators of Texas’ system of accountability originally envisioned the policy
as an information exchange very similar to accountability systems that exist in
practice in countries around the world. However, the evolution of Texas testing and
accountability under successive gubernatorial administrations has fomented
JEA disillusionment amongst many former supporters of accountability (see e.g. Ravitch,
50,5 2010). When asked about the current state of Texas’ educational accountability system,
Lt Governor Bill Ratliff (personal communication, November 14, 2007) shared several
disconcerting issues related to its evolution:
[Accountability] is so complicated. Back at the time we initiated it, we wanted a simple report
card to the public as to what each school is doing. Well, if you tell a parent, “Here’s our
580 accountability system on the school that your child goes to,” the parent looks at that mess
[y]. You have 36 different factors that a school district is rated on. And if they fall down on
one, then they get classified as a poor school district. The system has just kind of gone
berserk [y] I’m concerned that this animal that I helped create is turned around and
devouring us, and in particular devouring our students and our teachers [y].

The problem with the system as it has morphed today is that [y] there are proposals to use
the system to punish teachers, to punish school districts, to deny the school districts funding,
to fire teachers if their students are not making certain grades on the test, to punish students
[y] The stakes now are so high that you have teachers that are under enormous pressure
[y]. It’s what brings on the accusation of teaching to the test because the stakes have gotten
so high that there’s the sort of irresistible temptation to do those kinds of things in order to
survive.

We concur with Lt Governor Ratliff and posit that teachers and leaders are victims of a
broken system, responding to high-stakes testing and accountability in ways that
under other circumstances they would not do, and under the current circumstances, are
often rather uncomfortable doing. The research described in this paper and the
departure from the creators’ original intent of accountability raises many questions
about the test-driven educational policy espoused by NCLB.
We, of course, do not believe that any human or organizational theory is ubiquitous
and predicts behavior in all situations or contexts. The purpose of our work is not to be
tendentious, instead our analysis provides a counter-conceptual narrative to the
numerous accounts that already exist in the literature and popular media purporting
that high-stakes testing and accountability has had spectacular success in US schools
without causing disconcerting unintended consequences. For future research, of
interest is the external validity of the current findings beyond the 11 urban, small city
and rural Texas high schools visited in this study. For example, do high schools
serving at-risk students in other states where accountability is a fairly recent policy
prescription exhibit similar responses to Texas schools where the pressure of high-
stakes testing and accountability has been institutionalized over the past 15 years?
Would these types of responses occur in other countries if current accountability
systems segue from information systems for the public to test-based accountability
systems undergirded by high-stakes exams and sanctions? There is existing evidence
in the literature that the gaming and exclusion practices identified in this paper already
occurred in the midst of England’s test-driven accountability policy (Rustique-
Forrester, 2005). Future research in heterogeneous contexts over time will begin to shed
light on these questions.
The 2002 reauthorization of the ESEA (NCLB) was hailed by many as a milestone
for civil rights for poor and ethnic/racial minority students in the USA. In theory,
accountability should spur high schools to increase education output for all students,
especially low-performing students who have been historically underserved by US
schools. Conversely, voices from Texas high schools that have experienced test-driven
accountability for more than 15 years have revealed that the majority of the high
schools included in this study have responded to Texas-style test-based accountability At-risk student
as would be expected from an accounting and risk-management paradigm. The averse
long-term implications of high school’s at-risk student aversion will be dire, both for
the US schools and the students they serve. Considering the précis of research
presented here examining the current US form of test-based accountability,
incentivizing the exclusion of students from high schools is not in the best interest
of world societies and is an American export best left on the shelf. 581
Notes
1. For more about Ann Richard’s approach to Texas-style accountability, go to http://ritter.
tea.state.tx.us/perfreport/account/94/manual.pdf
2. For more about George W. Bush’s approach to Texas-style accountability, go to http://
ritter.tea.state.tx.us/perfreport/account/99/manual/manual.pdf
3. The state also saw the promulgation of higher-stakes for students such as the abolition
of automatic grade progression. For example, in Houston, Superintended Rod Paige
utilized TAAS and Stanford 9 test scores to determine whether students advanced to the
next grade.
4. ESEA was the first large-scale federal legislation aimed at equalizing educational
opportunities for all of America’s students. President Johnson posited that a significant goal
of ESEA was to address resource allocation inequities among US schools serving wealthy
and poor students (Johnson, 1965). Of its many provisions, ESEA was the first large-scale
federal effort to federal dollars to schools serving large populations of students of poverty – a
goal that had been sought in the US since 1870 ( Johnson, 1965).
5. For more about the current Texas accountability system go to http://ritter.tea.state.tx.us/
perfreport/account/2010/manual/index.html
6. For more about NCLB go to www2.ed.gov/nclb/landing.jhtml
7. For more information about the Australian system go to www.myschool.edu.au/

References
Allington, R.L. and McGill-Franzen, A. (1992), “Unintended effects of educational reform in
New York”, Educational Policy, Vol. 6 No. 4, pp. 397-414.
NCLB (2002), An Act To Close The Achievement Gap With Accountability, Flexibility, and
Choice, So That No Child Is Left Behind (NCLB), 107 U.S.C. y 110.
Anderson, J.A. (2005), Accountability in Education, International Institute for Educational
Planning (IIEP); International Academy of Education (IAE), Brussels.
Beasley, M., Branson, B. and Hancock, B. (2008), “Rising expectation”, Journal of Accountancy,
Vol. 205 No. 4, pp. 44-50.
Bernstein, P. (1998), Against the Gods: The Remarkable Story of Risk, John Wiley, New York, NY.
Booher-Jennings, J. (2005), “Below the bubble: ‘Educational triage’ and the Texas accountability
system”, American Educational Research Journal, Vol. 42 No. 2, pp. 231-68.
Booher-Jennings, J. (2006), “Rationing education in an era of accountability”, Phi Delta Kappan,
Vol. 87 No. 10, pp. 756-61.
Bowling, D. and Rieger, L. (2005), “Success factors for implementing enterprise risk
management”, Bank Accounting and Finance, Vol. 18 No. 3, pp. 21-6.
Carnoy, M. and Loeb, S. (2003), “Does external accountability affect student outcomes? A
cross-state analysis”, Educational Evaluation and Policy Analysis, Vol. 24 No. 4, pp. 305-31.
JEA Carnoy, M., Loeb, S. and Smith, T. (2001), “Do higher state test scores in Texas make for better
high school outcomes?”, CPRE Research Report No. RR-047, ERIC ED478984, Consortium
50,5 for Policy Research in Education, Philadelphia, PA.
Crockford, N. (1986), An Introduction to Risk Management, 2nd ed., Woodhead Faulkner,
Cambridge, MA.
Cuban, L. (2004), The Blackboard and the Bottom Line: Why Schools Can’t be Business, Harvard
582 University Press, Cambridge, MA.
Darling-Hammond, L. (1991), “The implications of testing policy for quality and equality”, Phi
Delta Kappan, Vol. 73 No. 3, pp. 220-5.
Darling-Hammond, L. (2003), “Standards and assessments: where we are and what
we need”, Teachers College Record, available at: www.tcrecord.org/Content.asp?Content
ID¼11109 (accessed July 1, 2012).
El Sawy, O., Gomes, G. and Gonzalez, M. (1986), “Preserving institutional memory: the
management of history as an organizational resource”, paper presented at the 46th Annual
Meeting of the Academy of Management, Chicago, IL, August 13-16.
Ericson, R. and Leslie, M. (2008), “The architecture of risk management”, Economy and Society,
Vol. 37 No. 4, pp. 613-24.
Figlio, D.N. and Getzer, L.S. (2002), Accountability, Ability, and Disability: Gaming the System?
National Bureau of Economic Research, Cambridge, MA.
Flyvbjerg, B. (2006), “Nobel prize to project management: getting risks right”, Project
Management Journal, Vol. 37 No. 3, pp. 5-15.
Garcı́a-Huidobro, J.E. and Bellei, C. (2006), “Remedio para la inequidad? La Subvención escolar
preferencial”, Revista Mensaje, Marzo-Abril, pp. 102-5.
Glaser, B. (1992), Basics of Grounded Theory Analysis, Sociology Press, Mill Valley, CA.
Goodman, L.A. (1961), “Snowball sampling”, Annals of Mathematical Statistics, Vol. 32 No. 1,
pp. 148-70.
Gorrod, M. (2004), Risk Management Systems: Technology Trends in Finance and Capital
Markets, Palgrave Macmillan, Basingstoke.
Gotbaum, B. (2002), Pushing Out at-Risk Students: An Analysis of High School Discharge Figures,
Advocates for Children of New York, New York, NY.
Grissmer, D. and Flanagan, A. (1998), Exploring Rapid Achievement Gains in North Carolina and
Texas: Lessons from the States, National Education Goals Panel, Washington, DC.
Grubb, N. (1985), The Initial Effects of House Bill 72 on Texas Public Schools: The Challenges of
Equity and Effectiveness, Lyndon B. Johnson School of Public Affairs, Austin, TX.
Gumport, P. (2000), “Academic restructuring: organizational change and institutional
imperatives, higher education”, The International Journal of Higher Education and
Educational Planning, Vol. 39 No. 1, pp. 67-91.
Haney, W. (2000), “The myth of the Texas miracle in education”, Education Policy Analysis
Archives, Vol. 8 No. 41, pp. 1-323.
Hanushek, E.A. and Raymond, M.E. (2003), “Improving educational quality: how best to evaluate
our schools?”, in Kodrzycki, Y. (Ed.), Education in the 21st Century: Meeting the Challenges
of a Changing World, Federal Reserve Bank of Boston, Boston, MA, pp. 193-224.
Janesick, V.J. (1994), “The dance of qualitative research design: metaphor, methodology, and
meaning”, in Denzin, N.K. and Lincoln, Y.S. (Eds), Handbook of Qualitative Research,
Sage, Thousand Oaks, CA, pp. 209-19.
Johnson, L. (1965), “Presidential remarks following enactment of the elementary and secondary
education bill”, in Woolley, J.T. and Peters, G. (Eds), The American Presidency Project,
Santa Barbara, CA, available at: www.presidency.ucsb.edu/ws/?pid¼26883
Johnson, R. (2008), “Texas public school attrition study, 2007-08: at current pace, schools will lose At-risk student
many more generations”, IDRA Newsletter, October, available at: www.idra.org/
newsletterplus/October_2008/ (accessed July 1, 2012). averse
Justman, M. and Bukobza, G. (2010), Guidelines to Revise the System of Education Indicators in
Israel: Summary, Conclusions and Recommendations, Committee to Revise the System of
Education Indicators in Israel, Israel Academy of Sciences and Humanities, Jerusalem.
Klein, S.P., Hamilton, L.S., McCaffrey, D.F. and Stecher, M.B. (2000), What Do Test Scores in Texas 583
Tell Us? RAND, Santa Monica, CA.
Lincoln, Y. and Guba, E. (1986), “But is it rigorous? Trustworthiness and authenticity in
naturalistic evaluation”, in Williams, D.D. (Ed.), Naturalistic Evaluation, New Directions for
Program Evaluation, Jossey-Bass, San Francisco, CA, pp. 73-84.
Linton, T.H. and Kester, D. (2003), “Exploring the achievement gap between white and
minority students in Texas: a comparison of the 1996 and 2000 NAEP and TAAS
eighth grade mathematics test results”, Education Policy Analysis Archives, Vol. 11 No. 10,
pp. 1-19.
McNeil, L. (2005), “Faking equity: high-stakes testing and the education of Latino youth”, in
Valenzuela, A. (Ed.), Leaving Children Behind: How “Texas-Style” Accountability Fails
Latino Youth, State University of New York Press, New York, NY, pp. 57-111.
McNeil, L. and Valenzuela, A. (2001), “The harmful impact of the TAAS system of testing in
Texas: beneath the accountability rhetoric”, in Kornhaber, M. and Orfield, G. (Eds), Raising
Standards or Raising Barriers? Inequality and High Century Foundation, Century
Foundation Press, New York, NY, pp. 127-50.
McNeil, L.M., Coppola, E., Radigan, J. and Vasquez Heilig, J. (2008), “Avoidable losses:
high-stakes accountability and the dropout crisis”, Education Policy Analysis Archives,
Vol. 16 No. 3, pp. 1-48.
Murphy, J. and Beck, L.G. (1995), School-Based Management as School Reform: Taking Stock,
Corwin Press, Thousand Oaks, CA.
National Commission on Excellence in Education (1983), A Nation at Risk: The Imperative for
Educational Reform, US Government Printing Office, Washington, DC.
Patton, M.Q. (1990), Qualitative Evaluation and Research Methods, 2nd ed., Sage, Newbury Park, CA.
Peabody, Z. (2003), “Sharpstown had ‘breakdown’”, Houston Chronicle, available at:
www.chron.com/disp/story.mpl/topstory2/2073994.html (accessed October 15, 2011).
Power, M. (2007), Organized Uncertainty: Designing a World of Risk Management, Oxford
University Press, Oxford.
Ravitch, D. (2010), The Death and Life of the Great American School System: How Testing and
Choice are Undermining Education, Basic Books, New York, NY.
Rothstein, M., Knapp, A. and Liebman, L. (1987), Cases and Materials on Employment Law,
Foundation Press, New York, NY.
Rustique-Forrester, E. (2005), “Accountability and the pressures to exclude: a cautionary
tale from England”, Education Policy Analysis Archives, Vol. 13 No. 26, pp. 1-41.
Salinas, C. and Reidel, M. (2007), “The cultural politics of the Texas’ educational agenda:
examining who gets what, when and how”, Anthropology and Education Quarterly, Vol. 38
No. 1, pp. 42-56.
Sarbanes-Oxley Act of 2002, 107 U.S.C. y 204 (2002).
Shafiq, M.N. (2011), “Do school incentives and accountability measures raise skills in the Middle
East and North Africa? The cases of Jordan and Tunisia”, Review of Middle East
Economics and Finance, (Forth coming), available at: http://ssrn.com/abstract¼1547562;
http://dx.doi.org/10.2139/ssrn.1547562
JEA Stecher, B., Hamilton, L. and Gonzalez, G. (2003), Working Smarter To Leave No Child Behind,
RAND, Santa Monica, CA.
50,5
Stowe, D. and Jeffery, C. (2008), “Risk management in isolation is risky”, Financial Executive,
Vol. 24 No. 2, pp. 26-9.
Strauss, A. and Corbin, J. (1990), Basics of Qualitative Research: Grounded Theory Procedures
and Techniques, Sage, Newbury Park, CA.
584 Stulz, R. (1996), “Rethinking risk management”, Journal of Applied Corporate Finance, Vol. 9
No. 3, pp. 8-25.
Suarez, R. (2001), “News hour with Jim Lehrer”, Television broadcast, PBS, Washington, DC,
January 24.
Thomson, J.C. (2007), “SOX 404 and ERM: Perfect partners y or not?”, Journal of Corporate
Accounting and Finance, Vol. 18 No. 3, pp. 29-36.
Texas Education Agency (TEA) (2007), Glossary for the Academic Excellence Indicator System
2006-07, TEA, Austin, TX.
Texas Senate Bill 7. (1993), 73rd Texas Legislature, Education Code y 16.007.
Toenjes, L.A. and Dworkin, A.G. (2002), “Are increasing test scores in Texas really a myth, or is
Haney’s myth a myth?”, Education Policy Analysis Archives, Vol. 10 No. 17, pp. 1-11.
Tyack, D.B. and Hansot, E. (1982), Managers of Virtue: Public School Leadership in America,
1820-1980, Basic Books, New York, NY.
Valencia, R.R. and Bernal, E.M. (2000), “An overview of conflicting opinions in the TAAS case”,
Hispanic Journal of Behavioral Sciences, Vol. 22 No. 4, pp. 423-43.
Vasquez Heilig, J. (2011), “Understanding the interaction between high-stakes graduation tests
and English language learners”, Teachers College Record, Vol. 113 No. 12.
Vasquez Heilig, J. and Darling-Hammond, L. (2008), “Accountability Texas-style: the progress
and learning of urban minority students in a high-stakes testing context”, Educational
Evaluation and Policy Analysis, Vol. 30 No. 2, pp. 75-110.
Vasquez Heilig, J. and Nichols, S. (2011), “A quandary for school leaders: equity, high-stakes
testing and accountability”, in Tillman, L.C. and Scheurich, J.J. (Eds), Handbook of
Research on Educational Leadership for Diversity and Equity, American Educational
Research Association, New York, NY.
Vasquez Heilig, J., Williams, A. and Jez, S. (2010), “Inputs and student achievement: an analysis
of Latina/o-serving urban elementary schools”, Association of Mexican American
Educators Journal, Vol. 10 No. 1, pp. 48-58.
Williams, C. and Heins, R. (1989), Risk Management and Insurance, McGraw-Hill, New York, NY.
White House (2002), “President signs landmark No Child Left Behind education bill”, press
release, January 8, available at: http://georgewbushwhitehouse.archives.gov/news/
releases/2002/01/20020108-1.html (accessed October 15, 2011).
Wyman, O. (2008), The Future of Risk Management in Microfinance, The Fletcher School at Tufts
University, Medford, MA.
Young, M.D. and Brewer, C. (2008), “Fear and the preparation of school leaders: the role of
ambiguity, anxiety and power in meaning making”, Journal of Education Policy, Vol. 22
No. 1, pp. 106-29.

About the authors

Julian Vasquez Heilig is an award-winning researcher and teacher. He obtained his PhD in
Education Administration and Policy Analysis and a Master’s degree in Sociology from Stanford
University. He also holds a Master of Education Policy degree and a Bachelor degree in History
and Psychology from the University of Michigan. Currently he is Assistant Professor of
Educational Policy and Planning at the University of Texas at Austin and his current research At-risk student
includes quantitatively examining how high-stakes testing and accountability-based reforms
and incentive systems impact urban minority students. Additionally, his qualitative work
averse
considers the mechanisms by which student achievement and progress occur in relation to
specific NCLB-inspired accountability policies in districts and schools for students of different
kinds. His research interests also include issues of access, diversity and equity in higher
education. He is a Faculty Affiliate of the Warfield Center for African and African American 585
Studies and the Center for Mexican American Studies at the University of Texas at Austin. He
also serves as the Associate Director for the University Council of Education Administration
(UCEA). Julian Vasquez Heilig is the corresponding author and can be contacted at:
[email protected]
Michelle Young is the Executive Director of the University Council for Educational
Administration (UCEA) and an Associate Professor in Educational Leadership and Policy at the
University of Texas. Currently, her scholarship focuses on educational leadership and policy,
with a particular emphasis on the politics, policy and practice of leadership preparation. She is
the recipient of the William J. Davis award for the most outstanding article published in a volume
of the Educational Administration Quarterly. Her work has been widely published and is
internationally recognized for its quality and impact. The organization she leads, UCEA, is an
international consortium of research institutions with doctoral programs in Educational
Leadership and Administration. As Executive Director of UCEA, she works with universities,
practitioners, professional organizations and state and national leaders to improve the
preparation and practice of school and school system leaders and to create a dynamic base of
knowledge on excellence in educational leadership.
Amy Williams is in the PhD Educational Policy and Planning program at the University of
Texas at Austin. She holds a Master’s degree in Business Administration and a Bachelor’s degree
in Management from Texas State University-San Marcos. She is currently the fundraising chair
for an educational non-profit organization. Her research uses business concepts in an educational
context to improve school practices. She has focused this research on urban school districts in the
areas of school finance, risk management and the policy decision-making process.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints
 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0957-8234.htm

JEA
50,5
Contrasting effects of
instructional leadership practices
on student learning in a high
586 accountability context
Received 25 October 2011
Revised 9 April 2012
Moosung Lee
Accepted 18 April 2012 Faculty of Education, University of Hong Kong, Hong Kong, P.R. China, and
Allan Walker and Yuk Ling Chui
Department of Education Policy and Leadership,
Hong Kong Institute of Education, Hong Kong, P.R. China

Abstract
Purpose – The purpose of this paper is to examine the effects of different dimensions of instructional
leadership on student learning in Hong Kong secondary schools, whose broader institutional contexts
are critically characterized by high accountability policy environments.
Design/methodology/approach – This study utilizes standardized test scores collected from
(n ¼ 2,037) students in 42 secondary schools and data collected from key staff’s perceptions of
leadership practices, to investigate two dimensions of instructional leadership, which are conceptually
interdependent but distinctive – i.e. instructional management and direct supervision of instruction.
A cross-level interaction analysis of hierarchical linear modeling was employed to investigate the
effects of the two dimensions of instructional leadership on student learning.
Findings – Leadership practices focused on instructional management were found to enhance student
learning by boosting the positive effect of students’ attachment to their school on academic
achievement. In contrast, leadership practices related to direct supervision of instruction were found to
undermine student learning by weakening the positive effect of student perceptions of school
attachment on academic performance when other school- and student-level characteristics are held
constant.
Originality/value – The paper reveals the contrasting effects of instructional leadership as a
multi-dimensional construct which is central in the current education reform agenda, rooted in
accountability-oriented policy of Hong Kong. It draws a number of implications for principal
instructional leadership in Hong Kong Schools as they deal with demands for external accountability.
Keywords Hong Kong, Secondary schools, Leadership, Students, Academic staff,
Principal instructional leadership, Instructional management, Direct supervision of instruction,
Accountability, Student outcomes
Paper type Research paper

Investigation into the shape, place and effect of principal instructional leadership has
followed numerous pathways. Among these is the role and impact of school leadership
in a policy environment that demands increased school accountability for student
outcomes (Cooley and Shen, 2003; Heck and Hallinger, 2009; Linn, 2003; Vanderhaar
et al., 2006). For example, centrally driven school accountability policies increasingly
hold a prominent place in government education reform agendas internationally
Journal of Educational
Administration
(Ingram et al., 2004; Lee et al., 2012a,b; Linn, 2003; O’Day, 2002). While these policies
Vol. 50 No. 5, 2012
pp. 586-611
r Emerald Group Publishing Limited
0957-8234 The authors wish to acknowledge the funding of the Research Grant Council (RGC) of Hong Kong
DOI 10.1108/09578231211249835 for its support through the General Research Fund (GRF 451407).
differ in form and emphasis both across and within national boundaries, there is little Instructional
doubt that they impact the context in which school leadership is exercised. Generally leadership
couched within the framework of a broader “quality education” agenda, these
policies typically aim to devolve some decision-making power and educative practices
responsibility for student outcomes to the school level (Carnoy and Loeb, 2002; Ng and
Chan, 2008).
A consequence of this policy trend has been to return “instructional leadership” to a 587
central position within reform discourse (Hallinger, 2005; Wiseman, 2004), often under
the label “leadership for learning” (Hallinger, 2003, 2005). While scholarly interest in
instructional leadership has endured since the early 1980s (e.g. Bossert et al., 1982) it
has returned to the limelight by virtue of an increasing global emphasis on school
accountability measures linked directly at improving student learning (Hallinger,
2005). This has in turn been accompanied by substantial empirical evidence of the
positive impact of instructional leadership on teacher practices and student outcomes
(Blase and Blase, 1999; Hallinger and Heck, 1996; Leithwood et al., 2004; Louis et al.,
2010; Marks and Printy, 2004; O’Donnell and White, 2005; Robinson et al., 2008).
Drawing on data from 23 countries involved in the Teaching and Learning
International Survey (TALIS), a recent Organization for Economic Co-Operation and
Development (OECD) report indicated that greater instructional leadership contributes
significantly to a wide range of teacher and school outcomes (OECD, 2009). Similarly,
a study on school leadership across eight different societies highlighted instructional
leadership as a key characteristic of high-performing principals in those societies
(Barber et al., 2010).
This paper focusses on principal instructional leadership in Hong Kong within a
high accountability environment. Stricter accountability mechanisms first appeared
in Hong Kong during the early 1990s. The most pervasive of these were embedded in
school-based management (SBM) reforms (Cheng, 2009; Walker, 2004). The often
technocratic policy prescriptions adopted within SBM incorporated globally validated
language such as “performance indicators” and “quality assurance.” Taken together
these gradually came to comprise a key foundation for the Hong Kong Government’s
ambitious accountability framework; a framework clearly aimed at the quality of
student outcomes (Education Bureau (EDB), 2008).
Externally imposed accountability policy requires principals to simultaneously
respond to the specific needs of their schools while adhering to common benchmarks
and complying with new reporting mechanisms. Previous research suggests that these
requirements have steered Hong Kong principals toward instructional leadership
practices (Walker and Ko, 2011). However, empirical studies have not explicitly
examine the link between principals focussing their instructional practices directly
on student outcomes in direct response to externally imposed accountability policies in
the context of Hong Kong.
This paper reports a study which investigated the impact of principal instructional
leadership on student learning outcomes in Hong Kong secondary schools operating
in a high accountability context. Given that principal instructional leadership is a
multidimensional construct we investigated the effects of different dimensions on
student achievement (Hallinger, 2005; Hallinger and Murphy, 1985; Heck et al., 1990;
OECD, 2009). Our assumption was that different dimensions of instructional leadership
would have dissimilar impacts on student achievement within an accountability
context (Cheng, 2009; Ho, 2005; Walker, 2004, 2006). Our study was driven by the
following question: how do two different dimensions of principal leadership practices
JEA (i.e. instructional management and direct supervision of instruction) impact student
50,5 achievement?
In the following section we discuss instructional leadership as a multidimensional
construct and its implications for accountability in the Hong Kong context.

Theoretical perspective
588 This section consists of three parts. First, we discuss the conceptual framework (i.e. the
multidimensionality of instructional leadership) that guided data collection and
analysis. Second, we discuss the implications of this multidimensionality in relation
to the accountability policies facing Hong Kong school principals. Third, we justify
the analytical model underpinning our investigation of the effects of instructional
leadership on student learning.

Instructional leadership as a multidimensional construct

Research into instructional leadership began in earnest in the early 1980s as part of
the school effectiveness movement (Hallinger, 2005). Bossert et al. (1982) argued for the
importance of understanding the relationship between instructional leadership
and principalship behaviors. As such, Bossert et al.’s research led to the emergence of
empirical studies into the enactment of instructional leadership (Hallinger, 2005). Since
then researchers have used multiple conceptual models and methodologies to
investigate the concept. Regardless of the variety of models and subsequent debate
in the area, at last two common agreements cut across studies. The first is the positive
impact of instructional leadership on school improvement; and the second, that
instructional leadership is a multifaceted construct.
Research has documented the indirect impact of instructional leadership on student
learning through how it is applied to shape school learning environments and teacher
practices (e.g. Hallinger and Heck, 1996; Leithwood et al., 2004). The indirect impact of
instructional leadership on student learning is especially salient in elementary schools
(Heck and Hallinger, 2009; Louis et al., 2010). Other empirical research suggests that
the impact of transformational leadership on school performance can be enhanced
by instructional leadership (Marks and Printy, 2004). A recent study conducted in
Chicago public schools found that instructional leadership plays a key role as “the
driver for change” (Bryk et al., 2010, p. 61) for school improvement and student
learning. Robinson et al. (2008) reaffirmed that, in general, instructional leadership
practices have a greater impact on student learning than those associated with
transformational, transactional or other types of leadership. In summary, the
international literature provides general agreement of the contribution of instructional
leadership to school improvement.
There is also agreement that instructional leadership is a multifaceted construct.
Hallinger and Murphy’s (1985) model (Principal Instructional Management Rating
Scale (PIMRS)) presents instructional leadership as comprising multidimensional
features. Their model has three dimensions: defining the school’s mission, managing
the instructional program and promoting a positive school learning climate (Hallinger,
2005; Hallinger and Murphy, 1985). Using a similar set of leadership practices to
PIMRS, Robinson et al. (2011) categorized two broad dimensions of instructional
leadership: direct instructional leadership (e.g. setting and ensuring goals, leading
teacher teaching and instruction) and indirect instructional leadership (e.g. organizing
instructional program, protecting instructional time) (cited in Louis and
Robinson, 2012). Heck et al. (1990) suggested that instructional leadership consists
of leadership practices addressing high academic expectations, professional Instructional
development, the use of data to monitor academic progress and strong emphasis leadership
on instruction.
Recent OECD research (2009) based on the TALIS data further affirms that practices
instructional leadership is a multidimensional construct. Drawing on data from
23 countries the OECD report (2009) showed that effective instructional leaders tend to
engage actively in three domains: management for school goals, instructional 589
management and direct supervision of instruction. the first domain, management for
school goals, is similar to the first dimension of PIMRS in that it highlights principals’
explicit management of instruction through school goals. The second domain,
instructional management, focusses on developing and improving curriculum,
curriculum knowledge and pedagogy. This domain overlaps with the second
dimension in the PIMRS. The final domain, direct supervision of instruction, refers to
“actions to directly supervise teachers’ instruction and learning outcomes” (OECD,
2009, p. 194) – thus it also partially overlaps with PIRMS’s second dimension
(see Hallinger and Murphy, 1985). In sum, OECD’s conceptualization of instructional
leadership is similar to the PIMRS, the most widely used instrument internationally for
exploring instructional leadership. At the same time, however, the OECD framework
further partitions the second dimension of the PIMRS by proposing instructional
management and direct supervision of instruction.
In summary, there is a now tacit agreement that instructional leadership is a
multidimensional construct. Despite such agreement there is relatively less consensus,
or empirical work, about whether different dimensions of instructional leadership
have different levels of impact on student learning. Little is known about which
dimensions work better, or conversely, which have less impact (or even a negative
effect) on student learning. Further investigation of this may shed greater light on the
intricacies of effective instructional leadership.
In this study we adopted OECD’s instructional leadership framework for a number
of conceptual and practical reasons. First, the framework is conceptually consistent
with previous research. Second, the framework is built on analysis of internationally
validated data. The OECD study is widely regarded as the largest data set available on
instructional leadership. Third, in practical terms, it is conceptually and analytically
compatible with our dataset in terms of the composition of survey items.

Instructional leadership in a high accountability context

The introduction of SBM in Hong Kong in the early 1990s first focussed interest on
the link between principal leadership and school effectiveness (e.g. Chan and Cheng,
1993; Cheng, 1994; Wong, 2003). However, research focussing explicitly on principal
instructional leadership schools remains under researched. Only a handful of relevant
“empirical” studies have been published (e.g. Chan and Cheng, 1993; Cheng, 1994; Lee
and Dimmock, 1999; Lee et al., 2009, 2012a,b; Wong, 2003).
Using OECD’s conceptual framework, we noted two major findings relevant to
instructional leadership research in Hong Kong. In terms of the first dimension,
management for school goals, research has found that leadership practices around
setting goals and building shared vision are significantly associated with school
improvement. For example, Cheng (1994) found that principal leadership was critical
to school performance through providing clear organizational goals which were
used to hold staff accountable. Similarly, Wong (2003) found that leadership practices
focussing on “vision, mission and goals” (p. 243) contributed to effective school
JEA management. Chan and Cheng’s (1993) research found that the instructional leadership
50,5 practices of Hong Kong secondary principals focussed on providing incentives for
learning, maintaining high visibility and enforcing academic standards. These
practices connect with elements of OECD’s second domain – instructional
management.
No research has been conducted explicitly into direct supervision of instruction in
590 Hong Kong. This is a somewhat surprising given the growing recognition of
instructional leadership as a policy lever for accountability (Education Department,
2002)[1]. In other words, although leadership practices related to direct supervision
of instruction appear to be more closely related to external accountability policy
measures than the other two dimensions, there is no empirical research in this area.
As such, it remains unclear how different dimensions of instructional leadership
play out in terms of student learning outcomes in Hong Kong schools, especially when
the broader institutional environments is largely defined by centrally directed
accountability measures.
Given that imposed accountability is one of the most influential policy levers in
Hong Kong, it is important to investigate the impact of direct supervision of instruction
on instructional leadership in schools (e.g. Cheng, 2009; Linn, 2003). Cheng’s (2009)
review of education reform in Hong Kong showed that accountability policies are
embedded in a range of regulatory mechanisms built upon a market orientation (e.g.
School Places Allocation Scheme) and a commitment to decentralization (e.g. SBM). At
the school level, accountability plays out managerially and professionally. These
intersect to re-shape school cultures and operation. Walker and Ko (2011) explain:
[T]he dominant approaches to accountability employed by policy makers fall predominantly
into the managerial and professional categories. For example, in an attempt to more closely
link school development and accountability policy makers introduced the Quality Assurance
Framework in 2003. This policy made it compulsory for schools to prepare annual
development plans, apply performance indicators, such as stakeholders’ surveys,
value-added information and norm-referenced outcome measures and implement internal
and external school reviews. Schools were also placed under increased scrutiny through the
territory-wide assessment system. At the same time the government introduced initiatives
designed to enhance professionalism in schools. For example, they offered territory-wide
information technology training, commissioned the development of a benchmark assessment
instrument for language teachers and introducing formal accreditation and professional
training for aspiring principals. The Quality Education Fund and the Chief Executive’s
Award for Teaching Excellence were launched as ways to encourage professionalization
through funding school-initiated action research and to reward teachers (pp. 12-13).
In brief, demands for externally imposed accountability are seen as forming a key
contextual influence on Hong Kong principals’ work (Walker and Ko, 2011). As such,
we suggest that principals’ instructional leadership practices (e.g. direct supervision of
instruction) which respond to accountability demands may not have a positive impact
on student learning. Although viewing instructional leader practices as
counterproductive runs counter to most research in the area in Hong Kong, it has a
basis in local literature. Lee and Dimmock (1999) found that when principals focussed
too strongly on implementing practices associated with accountability and quality
assurance, this increased pressure on teachers. Almost a decade later Walker and Ko
(2011) found that working in a demanding accountability environment had a negative
impact on school conditions. Drawing on these perspectives we further assume, for the
purposes of this study, that direct supervision of instruction, which fits neatly within
regulatory accountability environments, may generate a negative impact on teachers’ Instructional
work conditions and student learning. leadership
Exploring the effects of instructional leadership on student learning through school practices
attachment
In this paper we hold that student learning is influenced by student perceptions of
school attachment, or how students view their school, and more specifically, the sense 591
of belonging they feel to their school communities ( Johnson et al., 2001). Our
assumption is that students are more likely to engage actively in learning if they feel a
positive attachment to the school (e.g. sense of belonging, valuing teacher instruction,
intriguing class lessons, etc.). Research supports this assumption. For example, studies
have shown that student attachment to their school is associated with positive student
outcomes and academic performance (e.g. Libbey, 2004; Marchant et al., 2001). Blum
(2005) concluded that a high level of school attachment promotes student “motivation,
classroom engagement and improved school attendance” (p. 6), which in turn increases
academic achievement.
School cultures are partly shaped through teachers influencing student perceptions
of school attachment. This produces a social setting which includes regular interaction
between students and teachers. Student learning in this milieu cultivates certain
values, expectations and images about their schools. Given that research suggests that
principal leadership behaviors around accountability generates negative pressure on
teachers (Lee and Dimmock, 1999; Walker and Ko, 2011), it is reasonable to assume that
this pressure will in turn influence student attachment to their schools through daily
classroom interactions. We further hold that principal leadership practices such as
regular inspections of student school work, observation of classroom activities and the
use of student assessment data for monitoring teaching may generate additional
pressure on teachers, and that this will play out in classrooms. We use these as
measures of principal leadership practice related to direct supervision of instruction.
Likewise, we assume that an unhealthy organization climate may result when
instructional leadership behaviors driven by external accountability demands are
limited to classroom inspection and bland judgments of teaching quality only.
Therefore, in this study we investigate the effect of student attachment to their schools
on student learning outcomes and how this is moderated by principal leadership
practices such as direct supervision of instruction. Our interest in this angle was
supported by preliminary hierarchical linear modeling (HLM) analysis that showed the
school attachment slope differed significantly across schools, unlike other student-level
variables. This suggests that the effect of school attachment is more likely to vary
across the sample schools than do other student-level variables.

Methodology
This study employed cross-level interaction analysis of HLM to examine how the two
leadership dimensions (i.e. instructional management and direct supervision of
instruction) contribute to student achievement by decomposing variation in student
achievement into within- and between-schools portions, when other important school
and student characteristics are controlled for.

Data collection
All secondary schools in Hong Kong in 2009-2010 (498 secondary schools excluding
English Schools Foundation and international schools (EDB, 2011)) were invited to
JEA participate in the study. Of the 498 schools, 52 schools agreed to participate[2]. While a
50,5 low participation rate is not unusual in leadership research in Hong Kong, further
disaggregation of the data showed that schools using English as the medium of
instruction (MOI) were overrepresented (see Appendix 1 for more details). Partly
because of this overrepresentation, the average score of Hong Kong Certificate of
Education Examination (HKCEE) of the students sampled from the 42 schools (61.6)
592 was higher than that the estimated average of the entire population (50.0)[3]. In this
regard, caution needs to be exercised in interpreting results.
We gathered survey data from 180 key staff working in the sample schools who
were seen by the principal as playing an important role in schools improvement[4].
Approximately 74 percent held administrative positions such as vice-principals or
department heads, the remainder were senior teachers. All had worked in the same
school for at least three years prior to data collection (see Appendix 1). They were
asked to rate their principals’ leadership practices related to instructional management
and direct supervision of instruction on a six-point Likert scale (see Table I).
Another survey dataset was gathered from 2,032 Secondary 7 students enrolled
in the schools. On average, 48 students from each of the sampled school participated in
the study. They were asked to indicate their perceptions of school attachment, peer
academic orientation and parental involvement, also on a six-point Likert scale

Variables Question items a

Direct Regularly inspect student homework

supervision of Regularly observe classroom activities
instruction After observing classroom activities, work with teachers to improve
their teaching 0.817
Instructional Initiate school-based instructional projects
management Articulate high expectations for student academic achievement
Design ways to improve student learning
Encourage staff to consider new ideas for their teaching 0.870
School I feel that I belong at this school.
attachment Most of what I learn at school is interesting.
I think schoolwork is really important.
It is really important to me that I learn and develop my skills at school.
I really enjoy school most of the time.
I feel proud to be a student at this school.
I like most of the lessons in my school.
This school is a friendly place.
This school has a good reputation.
Students get good results at this school.
Students behave well at this school.
I feel confident that I will be successful in school 0.924
Peer academic Most students at this school are interested in learning.
orientation Most students at this school want to continue their education after
finishing secondary school.
Most students at this school want to do well in tests and exams 0.672
Perceived My parent(s)/guardians feel welcome in school and like to visit it.
parental My parents/guardians come to parents’ evening/events.
involvement I often discuss my schoolwork with my parents/guardians.
Table I. My parents/guardians are always willing to help me with my
Survey question items schoolwork 0.733
(see Table I). Basic demographic information such as gender was collected (see also Instructional
Appendix 1). Importantly, we gathered self-reported HKCEE scores from the leadership
students.
We also collected school demographic data such as MOI and school size from practices
school archival data. School-level performance data were obtained from the EDB.
The EDB’s value-added data collected over a three-year period of time (i.e. 2006-2008)
were released with the permission of the sample schools (see Measures for more 593
details).
Finally, we note that our final analysis included 42 out of the 52 participating
schools only. We excluded ten schools in the final analysis because the key staff survey
had relatively lower rates of non-responses (e.g. 480 percent), and so could not be
adequately handled, even with multiple imputation (MI) methods.

Measures
The study included two broad categories of independent variables: student-level
characteristics and school-level characteristics, as the dataset incorporated a unit of
analysis (i.e. students) that was nested within a larger unit (i.e. schools). The dependent
variable focussed on standardized student achievement scores, we define these as
follows.
Student-level variables (control variables): student-level characteristics were
comprised of student perceptions of school attachment, peer academic orientation,
parental involvement and demographic variables such as gender and number of years
the students had attended the school. All the level-1 variables were used as control
variables (see Appendix 2 for a correlation matrix among the control variables).
. Gender: since studies conducted in Hong Kong have reported mixed findings of
gender differences in terms of educational outcomes (Wang, 2006; Wong et al.,
2002), we included the gender variable in our model to control for any gender
effect on academic achievement. With this in mind, as a key demographic
characteristic, female was coded as 1 and male coded as 0.
. Years of enrollment in the current school: we assumed a possible association
between the number of years students had attended the school and student
achievement. Our rationale was that longer student attendance meant greater
exposure to different school-level factors related to achievement. We also
included this variable because a number of students had transferred to the
sample schools. Research has shown that when students’ change schools for
reasons other than grade promotion (e.g. primary to junior secondary) the effect
is negatively associated with educational outcomes such as low school
performance, higher dropout rates and more frequent absenteeism (e.g. Lee and
Burkam, 1992; Rumberger and Larson, 1998; South et al., 2007). Given this,
length of attendance was used in the model; high values indicate that students
have been enrolled in the school for a longer time.
. School attachment: students’ perceptions of school attachment were included in
the model. This variable was measured with 12 items (a ¼ 0.924) gauging
students’ agreement with items such as “I feel that I belong at this school” and
“I like most of the lessons in my school” (see Table I for the all survey items).
High values (on a six-point scale) indicate that students have a positive
perception of school attachment.
JEA . Peer academic orientation: we included students’ perceptions of their peer
50,5 group’s academic orientation. This is an important factor influencing student
achievement particularly in the Hong Kong context. Students who attend a high-
performing school where average achievement is high are more competitive;
this has been found to negatively influence academic outcomes (Marsh et al.,
2000). In addition, Salili et al.’s (2004) research as cited in Leung and Choi (2010)
594 reported that Hong Kong teachers tended to show greater appreciation and
pay more attention to academically able students. This often resulted in a
negative classroom atmosphere. The peer academic orientation variable was
therefore derived from three items (a ¼ 0.672) measuring student perceptions
of the academic orientation of their peers (e.g. “Most students at this school want
to do well in tests and exams” and “Most students at this school are interested in
learning”). High values (on a six-point scale) indicate that peers are highly
academically oriented.
. Parental involvement: the effect of parental involvement on academic
achievement has been documented internationally. However, findings present
a mixed picture. While a number of studies support the positive impact of
parental involvement on different types of academic outcomes (e.g. Ho and
Willms, 1996; Horvat et al., 2003; Madyun and Lee, 2010; McNeal, 1999), other
research suggests it has either an insignificant or a negative influence on student
achievement. For example, Catsambis’s (2001) study found that indicators of
parental involvement were not associated with achievement growth between
the 8th and 12th grades in US high schools. In the Hong Kong context, Chen
(2008) reported that perceived parental support was negatively linked to
academic achievement for Form 4 students. This suggests that the effect of
parental involvement on academic achievement may either disappear or even
morph into a negative in certain youth developmental contexts. We considered
these contradictory findings when we set up our model by incorporating
a parental involvement variable. The variable was based on four items
(a ¼ 0.733) measuring students’ perception of parent involvement, such as parent
participation in school events and parent help with schoolwork[5]. High values
(on a six-point scale) indicate strong parental involvement (see Table I for
details).

School-level variables (control variables): key school characteristics were employed

as control variables. Specifically, MOI and school size were considered important
school-level conditions associated with student achievement (see Pong, 2009; Pong and
Tsang, 2010). We also added school-level performance data and the value-added data
offered by the EDB as a school-level control variable in order to further isolate the effect
of school leadership on student achievement:
. MOI: the role of English language is crucial in Hong Kong. English competency
is associated with increased career prospects. English medium instruction (EMI)
schools are therefore preferred by most families (Pong, 2009). This trend has
intensified since 1998 when the Education Department introduced a new
language policy to encourage schools to adopt Chinese as the medium (CMI) of
instruction. To use EMI schools had to meet three requirements: student ability,
teacher capability and support measures. Consequently, the majority of
secondary schools became CMI schools (Education Commission, 2005). This has
 highlighted the value of attending an EMI school to improve career prospects. Instructional
In other words, high-achieving students tend to select EMI schools because leadership
“English linguistic capital continues to be linked to cultural and economic
capital and to reproduce the existing stratification of society and schooling” practices
(Morrison and Lui, 2000, p. 482). This is supported by research that shows that
EMI schools tend to perform better than CMI schools on standard achievement
tests, particularly in Chinese, English and Mathematics (Salili and Lai, 2003). 595
Within this context, we included the MOI variable in our model, CMI schools
were coded as 0 and Chinese/English medium schools (mixed mode) 1 and EMI
schools as 2.
. School size: a body of research indicates that school size plays an important role
in improving student school life in general and student achievement in particular
(Leithwood and Jantzi, 2009). School size research has overwhelmingly affirmed
that small schools are more effective than larger schools in terms of educational
outcomes, including academic achievement (e.g. Howley, 1994; Lee and Smith,
1995; Stiefel et al., 2000), academic equity among different racial ethnic groups
(Howley et al., 2000; Lee and Friedrich, 2007; Stiefel et al., 2000) and school safety
(Cotton, 2001). At the same time, however, a number of studies support the
finding of a positive relationship between large size and academic achievement
(e.g. Barnett et al., 2002; Sawkins, 2002, cited in Leithwood and Jantzi, 2009;
Schreiber, 2002). These studies attribute this positive relationship to the fact that
larger schools have a more diverse teacher population and can therefore offer
greater instructional and curriculum specialization (Leithwood and Jantzi, 2009).
Although these findings are inconsistent, the bottom line is that school size
matters. We therefore added the variable of school size in our model. Small
schools (o800 students) were coded as 0, mid-size schools (between 800 and
1,100 students) as 1 and large schools (41,100 students) as 2.
. Student streaming: Hong Kong secondary students are “streamed” into schools
under the Secondary School Place Allocation System in accordance with their
academic achievement. Using the Pre-Secondary 1 Hong Kong Attainment Test
primary school graduates are allocated to one of three equally sized secondary
school bands – Band 1 (highest academic achievement), Band 2 and Band
3 (EDB, 2011). This cannot be ignored, but given that Band classifications
are not publicly available we used the value-added data as a proxy measure.
Value-added data is measured by using Stanines – normalized standard scores
ranging from 1 to 9. Since Stanines have a mean of 5 and a SD of 2, we coded
0 for Stanines from 1 to 3 as low-performing schools (corresponding to Band 3),
1 for Stanines from 4 to 6 as mid-performing schools (corresponding to Band 2) and
2 for Stanines from 7 to 9 as high-performing schools (corresponding to Band 1),
based on the average of value-added scores over three years (2006-2008).
School-level variables (key independent variables of interest): we used the two
instructional leadership dimensions as school-level variables to examine how key staff
saw their association with student achievement:
. Instructional management: measured by four items (a ¼ 0.870) comprising
principals’ practices focussed on teaching and instruction – encouraging
new ideas about teaching, initiating instructional projects, designing measures
for improving student learning and articulating high expectations for student
JEA learning. The average score of the four items was used for the analysis; high
50,5 values (on a six-point scale) indicate that the practices are important in the
schools (see Table I for more details).
. Direct supervision of instruction: derived from three items (a ¼ 0.817) of key staff
perceptions of principal’ focus on direct supervision of instruction aligned to
student learning: regular inspection of student homework; regular classroom
596 observation; and post-observation classroom activities, work with teachers to
improve their teaching. The average score of three items was used for the
analysis; high values (on a six-point scale)[6] indicate that direct supervision of
instruction and learning outcomes is emphasized in the schools (see Table I for
more details).
It should be noted that the two instructional leadership variables had a moderately
high correlation (0.625, po0.001). Conceptually, this interdependency is not surprising
given that both represent the construct of instructional leadership. However,
analytically this might cause collinearity between the two variables. Our preliminary
analysis indicated that the variance inflation factor was not substantially 41 and thus
we concluded that collinearity was not an issue. In addition, given the interdependency
between the two constructs, we also conducted confirmatory factor analysis of the
two constructs. Specifically, we compared a two-factor structure model, separating
instructional management from direct supervision of instruction, with a single factor
model as a competing model which combines items of both instructional management
and direct supervision of instruction. Results indicated that the two-factor model
showed a significantly better model fit (X2 ¼ 23.7, df ¼ 13, CFI ¼ 0.983, TLI ¼ 0.964
and RMSEA ¼ 0.057) than the single factor model (X2 ¼ 76.3, df ¼ 14, CFI ¼ 0.902,
TLI ¼ 0.804, RMSEA ¼ 0.132). A model comparison confirmed that the two-factor
model fit the data significantly better. In fact, the single factor model did not even
meet standard cutoff recommendations for fit indices[7]. This suggests that while the
two constructs are conceptually interdependent under the heading of instructional
leadership they are also conceptually distinguishable in that they reflect different
aspects of instructional leadership.
Dependent variable: mandated standardized test scores (i.e. HKCEE) were used as a
dependent variable, based on students’ self-reports. The original scale of HKCEE
(i.e. 1 to 30) was transformed for easier interpretations of the analysis – i.e. ranging
from 3.33 to 99.9 with a mean of 61.6 and SD of 14.0.

Analytical strategies
Because the dataset had a nested structure in terms of units of analysis (students
within schools) we employed a two-level hierarchical linear model (Raudenbush
and Bryk, 2002). As in many large datasets, there were missing values in both of
the key staff and student surveys. These ranged from 0.4 percent (gender) to
2.4 percent (student perceptions of school attachment). To address missing values
we conducted a single imputation of the school-level data[8]. For student-level data, we
conducted a MI by using a custom imputation model with constraints on the variables
to prevent imputed values from falling outside a reasonable range[9]. Consequently,
five completed datasets representing simulated versions of the sample were
created[10]. These complete datasets were analyzed using HLM 6.8 software. The
estimated parameters for variables in the model from the five datasets were averaged
to yield a single estimate[11].
 By setting up a random effects ANOVA model (i.e. null model), we identified an Instructional
intra-class correlation coefficient (ICC) for the dependent variable. We then built leadership
explanatory models by adding level-1 (student characteristics) and level-2 variables
(school characteristics) in that order. The final HLM model was constructed using practices
an intercepts- and slopes-as-outcomes model (Raudenbush and Bryk, 2002) in order to
examine cross-level interactions.
597
Results
Descriptive results
Figure 1 presents the variation in student achievement across the 42 schools. The
median student achievement score across the schools was 59.9 (see the dotted line in
the figure). The boxplots in the figure provide more detailed information regarding
variation in student achievement by illustrating the distribution within and between
schools. The slightly thicker horizontal lines in the tinted boxes indicate median
student achievement in each school. The tinted boxes show the middle 50 percent of
students’ achievement scores. The distance between the top edges of the tinted boxes
and the upper horizontal lines indicate the top 25 percent of students’ achievement
scores. Likewise, the distance between the bottom edges of the tinted boxes and the
bottom horizontal lines indicate the bottom 25 percent of students’ achievement scores.
The wide range of medians in the boxplots highlights the striking variation in student
achievement scores between and within the schools.

HLM on student achievement

Subsequent HLM analyses confirmed this impression of substantial between-school
variance in student achievement. A random effects ANOVA model (null model) showed
that average student achievement varied significantly across the 42 schools.
Specifically, the associated ICC is 0.349. In other words approximately 35 percent of
the total variance lies between the schools. This again supports the striking variation
in student achievement scores between the schools shown in Figure 1. In all, 35 percent

100

90

80

70

60
HKCEE

50

40

30

20

10

0
Figure 1.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
Within and between
Schools
variation in HKCEE
Notes: The figure is constructed from the dataset before the multiple imputations (N = 1,994). by schools
Outliers are included in the figure, yet they are not visualized
JEA of the total variance in student achievement is related to school-level characteristics,
50,5 including principal leadership practices.
Based on this dependency, we built explanatory models (see Table II) by adding
level-1 student characteristics (Model 1), level-2 school characteristics (Model 2) and
cross-level interactions (Model 3). In the explanatory models, we allowed the school
attachment slope to vary across the schools, whereas we specified the other level-1
598 slopes as fixed. We used this approach for several reasons. First, the deviance statistic
indicated a better model fit when the school attachment slope was allowed to vary
across the schools[12]. Second, based on school attachment literature, we assumed
that student perceptions of school attachment differ between schools in as much
as they mold or sustain the different cultures that influence these perceptions.
Preliminary analysis also indicated that, except for the school attachment slope, other
level-1 slopes did not significantly vary across the schools as we built our explanatory
models. Third, drawing on relevant research our process assumed that principal
leadership practice moderates the effect of school attachment on student achievement.
As such, we examined cross-level interactions through the school attachment slope.
We added two level-2 predictors “instructional management” and “direct supervision
of instruction” in the slope. Other school-level characteristics were not added for the
parsimony of the final model[13]. The final results are presented in Table II.
Effects of individual characteristics: all the student-level characteristics, except the
perception of parental involvement, had significant effects on student achievement.
Notably, gender was a salient factor; males tended to outperform their female peers
(2.86***) when the other predictors were held constant. Consistent with previous
studies (e.g. Marsh et al., 2000), students’ perceptions of their peers’ academic

Fixed effects Model 1 Model 2 Model 3

For adjusted grand mean (b0j )

Intercept g00 60.66 (1.34)*** 54.50 (1.16)*** 54.51 (1.14)***
Medium of instruction g01 5.16 (0.90)*** 5.28 (0.88)***
School size g02 5.57 (1.24)*** 5.48 (1.21)***
School performance g03 4.19 (1.18)** 4.25 (1.16)**
Direct supervision of instruction g04 1.05 (1.66) 0.63 (1.48)
Instructional management g05 2.57 (1.85) 2.14 (1.70)
For gender slope g10 2.70 (0.63)*** 2.81 (0.62)*** 2.86 (0.61)***
For years in school slope g20 0.88 (0.43)* 0.92 (0.43)* 0.93 (0.43)*
For school attachment slope b3j
Intercept g30 1.00 (0.44)* 1.05 (0.44)* 1.12 (0.43)*
Direct supervision of instruction g31 2.53 (0.95)*
Instructional management g32 2.26 (0.92)*
For image of peer academic orientation slope g40 1.25 (0.44)** 1.32 (0.43)** 1.33 (0.43)**
For parental involvement slope g50 0.01 (0.27) 0.04 (0.27) 0.02 (0.27)
Random effects v.c. df p.v. v.c. df p.v. v.c. df p.v.
Level-1 effect gij 130.1 130.0 130.0
Mean u0j 75.5 41 0.000 22.8 36 0.000 21.6 36 0.000
School attachment u3j 3.4 41 0.009 3.4 41 0.009 2.3 38 0.032
Table II.
Intra-class correlation coefficient (ICC) 0.349
Hierarchical linear models
predicting student Notes: v.c., variance component, df, degree of freedom, p.v., p-value; 2,037 students from 42 schools;
achievement *po0.05, **po0.01, ***po0.001
orientation were negatively associated with their academic achievement (1.33**). Instructional
There was also a negative association between the number of years students attended leadership
their current school and student achievement (0.93*). When other student- and
school-level variables were held constant, student attachment to their schools (at the practices
student level) had a positive effect on student achievement (1.12*).
Effects of school contexts: results indicate that school contextual characteristics
make a substantial difference in student achievement. More specifically, students 599
from EMI schools were much more likely than their counterparts from CMI schools to
have higher HKCEE scores – this was expected (5.28***). Students from high-
performing schools were more likely than their counterparts to have higher HKCEE
scores (4.25**). However, somewhat surprisingly, students from larger schools
were more likely than their counterparts to have higher HKCEE scores, when other
predictors were held constant (5.48***).
Effects of principal leadership: student perceptions of school attachment were
positively associated with HKCEE scores. More interestingly, principal leadership
practices which focussed on instructional management positively moderated the
relationship between students’ perceptions of school attachment and their HKCEE
scores (2.26*). In other words, leadership practices centered on managing the
instructional program further elevated the positive effect of school attachment on
student achievement. Unlike the positive moderating effect of instructional leadership,
leadership practices that emphasize direct supervision of instruction negatively
moderated the relationship between students’ perceptions of school attachment and
their HKCEE scores (2.53*). That is, the positive effect of school attachment on
HKCEE scores were exacerbated by principal leadership practices related to direct
supervision of teaching.

Discussion
Principal instructional leadership focussing on instructional management boosts the
positive effect of school attachment on student learning. The moderating effect of
principal instructional leadership suggests a linkage between key staff perceptions of
leadership and students’ perception of school attachment, an area relatively less
charted in empirical research. In essence, this suggests that if key staff hold a positive
view of their principal’s focus on instructional management, students are likely to have
a positive image of their schools. This connection seems reasonable given that key
staff, including teachers, are best positioned through their daily interactions to
influence student perceptions of the values, expectations and the images students
hold about their school (see Stanton-Salazar, 1997; Stanton-Salazar and Spina, 2000).
Lee and Dimmock (1999) found that when teachers and department heads in Hong
Kong focus on improving teaching and learning, students follow, this becomes
“a driving force for promoting academic achievement” (Lee and Dimmock, 1999,
pp. 475-6). The bottom line of this finding is that if key staff have a positive perception
of their principals’ instructional management this seems to influence student
achievement by heightening students’ perception of school attachment.
This also suggests that principal leadership practices which focus on encouraging
teachers to value new ideas and innovative instructional designs are seen as
positive motivators by key staff (Elmore, 2003, 2005). When teachers and key
staff perceive principal instructional leadership practices as promoting
professional growth, they are motivated to reflect on their teaching routines and
seek new pedagogical approaches (Blase and Blase, 1999; Blase and Kirby, 2000;
JEA Robinson et al., 2008). Within this process principal instructional leadership behaviors,
50,5 such as initiating school-based instructional projects (e.g. implementing action
research to inform instructional development), supporting new ideas or redesigning
programs, all play a key role in changing teachers’ behaviors around teaching (Blase
and Blase, 1999). In other words, principals are viewed as facilitators of teacher
professional growth rather than directive supervisors, especially when teachers
600 see their principals as effective instructional leaders (Blase and Kirby, 2000;
Poole, 1995).
In contrast to the positive effect of instructional management on student learning
through school attachment, principals’ practices closely intertwined with direct
supervision of instruction undermined student achievement through school
attachment. This negative moderating effect of direct supervision of instruction on
student achievement can be explained by examining the following survey items:
. regular inspection of student homework;
. regular observation of classroom; and
. working with teacher based on classroom observation.
These practices seem to be perceived negatively by key staff because they create
negative pressure on teachers. Lee and Dimmock (1999) found that “intangible
pressure”(p. 470) was loaded on teachers’ in two Hong Kong secondary schools when
principals exercised curriculum leadership practices aligned to accountability and
quality assurance as a the prime strategy for improving student learning. Similarly,
Walker and Ko (2011) found that working in an accountability environment could
undermine the school conditions supporting student learning.
The key question then is why principal leadership practices focussing on direct
supervision of instruction and learning outcomes generate a negative school
atmosphere for teachers? We propose three possible explanations. First, there seems to
be an intellectual disconnection – i.e. inconsistency in the uniformity of the messages
about a particular type of leadership behavior (Walker, 2006; Walker and Qian,
forthcoming) – between principals and key staff. In practice this might be issues
around how teachers decode the intentions embedded in principals’ direct supervision
of instruction. For example principals may see direct supervision of instruction
as a means of authentic, technocratic control, which is welcomed by parents from a
consumerist stance. Conversely, within a highly regulated accountability context,
principals may be pushed to define instructional leadership simply as direct
supervision. They may also be attracted to direct supervision as an easier, more
efficient pathway to increase standardized test scores. In either case, instructional
leadership practices are geared primarily around inspection and a one-dimensional
judgment of classroom instruction.
Such complex situations facing principals as instructional leaders seem interwoven
with high accountability policy environments. In other words, while instructional
leadership is a critical leadership construct, and related practices contribute
significantly to school improvement, it may not be a given that such practices
automatically have a positive effect. Some instructional leadership domains, such
as direct supervision of instruction, may actually generate a negative impact on school
improvement by weakening teacher empowerment or autonomy (Walker and Qian,
forthcoming). This seems especially so when institutional contexts are largely shaped
by external accountability measures. It is worth noting Elmore’s (2005, p. 135)
suggestion about how accountability ought to be understood: “accountability is defined Instructional
by what individual teachers think students can do, not by their work environment leadership
or by the supervision of school leaders.” In a similar vein, Linn (2003) proposed that
“shared responsibility” (p. 3) should be emphasized in accountability systems. This practices
does not appear to be case in Hong Kong where a substantial portion of responsibility
flows onto teachers’ desks.
Another explanation relates to the socio-cultural context of Hong Kong. Walker and 601
Qian (forthcoming) used the term “cultural disconnection” to refer to “disconnection
between what reforms demand and the cultural realities of teaching in and leading
schools” (pp. 162-77). This poses questions about whether instructional leadership
emphasizing direct supervision and inspection, triggered by the current accountability
framework in Hong Kong, is actually congruent with “the broader culture of Hong
Kong and the deep teaching and leadership structures and values which guide
relationships and behaviors in Hong Kong schools” (not paginated yet). Like many of
the current educational reform measures, instructional leadership is driven by
global educational trends. As such, educational practices borrowed from other
countries can be accompanied by conflicting values and incompatible conditions to
the host society (Phillips and Ochs, 2003) – this raises concerns about cultural
appropriateness (Walker and Dimmock, 2000). While instructional leadership as a
whole is generally understood as an effective leverage for improving schools across
diverse socio-cultural contexts, including in Hong Kong (e.g. Chan and Cheng, 1993),
it is informative to note that some instructional leadership practices may have a
negative impact on school improvement and student learning, especially if it is
contextually inappropriate. Our analysis clearly indicates that principals’ direct
supervision and inspection is a case in point. Interestingly, research has shown that
even in hierarchically structured societies such as Hong Kong, observation or
inspection of teachers’ classroom activities for the purpose of accountability is
interpreted as principal intrusion into teachers’ traditional domains (e.g. Lee, 2005;
Walker and Dimmock, 2000).
Third, the negative moderating effect of principals’ direct supervision suggests
that there is a detrimental linkage between negative perceptions held by about
leadership practices and student attachment to the school. As noted, “what teachers
do in classrooms” is the most influential factor shaping students’ perceptions of their
school (Stanton-Salazar, 1997; Stanton-Salazar and Spina, 2000). It thus appears
reasonable to speculate that teachers working with the pressure generated by
accountability-oriented leadership practices such as direct supervision and inspection
will either intentionally or unintentionally negatively influence students’ attachment
school to their school.

Limitations
There were several limitations of the study. First, because of the perceived sensitivity
of to public exposure of value-added data we were forced to elicit voluntary
participation. Consequently, the overall response rate in terms of the overall sample
was very low. The number of sampled schools was also restricted by the fact that
we only included schools where principals and key staff had worked in the same school
for three consecutive years. We did this in order to get a better picture of the impact of
leadership over time. Although low participation is not unusual in Hong Kong, the
limitation might have generated potential problems of selection bias and certainly
reduced the generalizability of the findings[14].
JEA A second limitation is the cross-sectional dataset used. A longitudinal design with
50,5 the available data would have provided more significant effects of instructional
leadership.
Third, given that HKCEE scores were not available we relied on self-reported
scores for the dependent variable. While this can be considered a limitation in terms
of reliability, there was a significantly positive correlation between the average
602 HKCEE score and the value-added data of school performance in the same schools
(0.358, po0.05), thus suggesting the self-reported data is a fairly reliable measure.
Even though the correlation is not particularly high it should be noted that the
value-added school performance data is an aggregated index incorporating a range of
sub-measures. In other words, the value-added data is not the exactly same as the
aggregate of individuals’ HKCEE scores. In this regard, the moderate correlation is
understandable.
Finally, because of data inaccessibility and sensitivity, some important student
characteristics (e.g. family SES), which are predictive of student achievement in
similar studies, were not included in the level-1 equation in the final model. This absence
of adequate control variables in the level-1 model is a weakness of the present study.

Implications for policy, practice, and research

Several implications for policy, practice and research emerge from our analysis.
This result of the negative influence of principals’ direct supervision of instruction
supports previous findings that teachers in East Asian societies value a reasonable
level of autonomy, especially in terms of curriculum development (Lee, 2005; Lee
and Dimmock, 1999). In these societies teachers seem to interpret a lack of overt
engagement by administrators in curriculum as a signal that the principal trusts their
curricular and instructional expertise, and that management of these areas is teacher
business (Lee, 2005; Walker and Dimmock, 2000). The negative impact of direct
supervision of instruction on student learning identified in the study suggests that the
enactment of instructional leadership practices makes a difference. As Lee and
Dimmock (1999) pointed out, if principals focus too much on practices associated
with accountability and quality assurance, negative pressure on teachers increases.
Whereas some pressure through quality assurance and accountability is necessary,
it is useful to consider the extent to which such pressure is applied and how it is
communicated to shape teachers’ mindsets. Teacher instructional behaviors appear
influenced by whether they see principal practices as stemming from student good or
policy mandate.
Some “instructional management” practices identified in this study may inform
principal practices. For example, strengthening practice around the following areas
may guide instructional action: stimulating innovative, school-based, and
contextualized instructional designs and projects; consciously articulating links
between teaching and learning, and encouraging teachers to navigate effective
instructional approaches to improve student academic achievement.
The study also shed light on how instructional leadership as a multidimensional
construct plays out differently, especially within regulatory accountability policy
environments. Principals’ direct supervision of instruction and learning outcomes
emerged as a key dimension within this environment. It appears that principal
instructional leadership practices “encouraged” by centralized policy interfere with the
values driving school life and pedagogies through somehow giving teachers the
impression that their autonomy is being challenged in their traditional spheres of
control – curriculum and pedagogy. As Reynolds et al. (2002) pointed out, assertive Instructional
principal instructional leadership, which is often supported by US-based research, leadership
may not work in differing socio-cultural contexts[15]. Based on our analysis, this seems
to be true in Hong Kong. practices
This does not imply that direct supervision of instruction should be removed
from the arsenal of principal instructional leadership. Rather, the point is that in some
educational systems, in this case Hong Kong, the way principals’ enact this supervision 603
should be sensitive school context and teaching cultures, not just policy mandates
and/or decontexualized research. For example, principals may provide more reflective
and formative, rather of judgmental, comments when monitoring the know-how.
How supervisory practices are enacted appears closely linked to the intentional
basis of principal actions. Principal instructional leadership practices are more likely
to be effective if congruent with explicitly understood vision and within an
understanding of collective formal and informal responsibilities, rather than simply
molded in reaction to centralized accountability measures[16]. Teachers are more
likely to respond to principals’ direct supervision of instruction when they believe
that their leader’s intentions are underpinned by student good and teacher professional
growth.
Notes
1. For example, reflecting the critical role of instructional leadership in implementing
accountability policy measures in Hong Kong, over the last decade instructional leadership
has been integrated into school leadership preparation and certification as a central
component (Education Department, 2002). In other words, instructional leadership has
underpinned leader development programs for aspiring, beginning and experienced
principals in Hong Kong for almost a decade. Therefore, we speculate that instructional
leadership plays an important role in shaping leadership practices in Hong Kong schools.
2. The agreement rate was low for two reasons. First, the EDB could not release school-specific
value-added data without the agreement of individual principals. Therefore, we were forced
to approach each principal. Concerns over loss of control over test results (not currently
public information) became a significant obstacle to obtaining school participation. Second,
the school level survey data contained questions about principal leadership which would
be answered by other staff. In this high accountability context concerns for public school
test results and perceptions of their own leadership led principals to decline participation
in this study.
3. Notably, the original scale of HKCEE ranges 1-30. This was rescaled for easier
interpretations of our analysis, ranging 3.33-99.9. This transformed scale works
mathematically identical with the original scale in our statistical modeling.
4. All of the principals in our sample schools had been working as principal in the same school
for three consecutive years.
5. Note that parental involvement in this study was measured by the perception of each
student. This suggests that actual parental involvement might be different from students’
perceptions.
6. Note that all the variables using a six-point Likert scale in this study have the following
response categories: not at all, very little, little, partially, a lot, very significantly.
7. We used w2 test statistic, root-mean-square-error of approximation (RMSEA), comparative fit
index (CFI), and Tucker-Lewis index (TLI). In particular, we relied more on standard cutoff
recommendations for the RMSEA, CFI and TLI (Hu and Bentler, 1999; Fan and Sivo, 2007).
For the RMSEA, values o0.05 and 0.08 suggest a good model fit and an acceptable model fit,
JEA respectively. For the TLI and CFI, values 40.95 and 0.90 indicate goodness of fit and
acceptable fit, respectively.
50,5
8. Since HLM 6.8, a multilevel analysis program, accommodates multiply imputed datasets at
the level-1 only, for the level-2, we used a single imputation data.
9. MI techniques have been reported as significantly more effective than traditional
techniques (e.g. listwise deletion, stochastic regression imputation) in addressing missing
604 data (Schafer and Graham, 2002).
10. The imputation model is compatible with the analytical model used in this study (see
Allison, 2002 for more details).
11. Standard errors were calculated by considering the within- and between-imputation
variation in the parameter estimates.
12. We calculated the deviance statistics from each of five imputed datasets, respectively. The
average deviance statistics was, then, calculated from using a SAS macro that combines w2
statistics from the five separate HLM analyses.
13. While adding a common set of level-2 predictors (or the same level-2 predictors) in
level-1 slopes of interest is more common in the analysis of cross-level interactions, in this
study we selectively added level-2 predictors in the school attachment slope (see
Raudenbush and Bryk, 2002, p. 151) based on the aforementioned conceptual and analytical
reasons.
14. Despite the low participation rate the study is the largest scale investigation into the effect of
instructional leadership practices on student learning conducted in Hong Kong.
15. Reynolds et al. (2002) noted that assertive principal instructional leadership is not a
significant predictor that determines effective school status in the Netherlands.
16. We note that Louis and Robinson (2012) article in this special issue found that
external accountability policies can have a “positive” impact on principals’ instructional
leadership practices in the US schooling context. This suggests a twofold meaning related
to our paper. First, the finding suggests that external accountability policy is a key driving
force that shapes instructional leadership behaviors in the US as well as Hong Kong. In other
words, principals’ attitudes toward accountability policies appear to be an important
predictor of principals’ instructional leadership practices in the US school context (Louis
and Robinson, 2012), which we think seems to be equally true for Hong Kong. Second,
US principals tend to make sense of external accountability policy in a “positive” way,
especially when they internalize accountability policy measures as “aligned with their own
values and preferences” and when they view “district administrators as supportive of
school-driven accountability initiatives” (Louis and Robinson, 2012,p. 1). However, unlike US
counterparts, probably Hong Kong principals might have a difficulty in aligning their
leadership values around certain positive goals embedded in external accountability policies
due to their different socio-cultural and organizational contexts. While this statement should
be empirically investigated by following studies, this perspective offers an implication for
educational policy makers in Hong Kong by making them think of how current external
accountability approaches should resonate with principals’ own leadership values with more
supportive manners.

References
Allison, P.D. (2002), Missing Data, Sage, Thousand Oaks, CA.
Barber, M., Whelan, F. and Clark, M. (2010), Capturing the Leadership Premium: How the World’s
Top School Systems are Building Leadership Capacity for the Future, Mckinsey &
Company, available at: www.mckinsey.com/clientservice/Social_Sector/our_practices/
 Education/Knowledge_Highlights/B/media/Reports/SSO/schoolleadership_final.ashx Instructional
(accessed February 15, 2011).
leadership
Barnett, R.R., Glass, C., Snowdon, R.I. and Stringer, K.S. (2002), “Size, performance and
effectiveness: cost-constrained measures of best-practice performance and secondary- practices
school size”, Education Economics, Vol. 10 No. 3, pp. 291-310.
Blase, J. and Blase, J. (1999), “Principals’ instructional leadership and teacher development:
teachers’ perspectives”, Educational Administration Quarterly, Vol. 35 No. 3, 605
pp. 349-78.
Blase, J. and Kirby, P. (2000), Bringing out the Best in Teachers: What Effective Principals Do, 2nd ed.,
Corwin Press, Thousand Oaks, CA.
Blum, R.W. (2005), “A case for school connectedness”, The Adolescent Learner, Vol. 62 No. 7,
pp. 16-20.
Bossert, S., Dwyer, D., Rowan, B. and Lee, G. (1982), “The instructional management role of the
principal”, Educational Administration Quarterly, Vol. 18 No. 3, pp. 34-64.
Bryk, A.S., Sebring, P.B., Allensworth, E., Luppescu, S. and Easton, J.Q. (2010), Organizing
Schools for Improvement: Lessons from Chicago, University of Chicago Press, Chicago, IL.
Carnoy, M. and Loeb, S. (2002), “Does external accountability affect student outcomes?
A cross- state analysis”, Educational Evaluation and Policy Analysis, Vol. 24 No. 4,
pp. 305-31.
Catsambis, S. (2001), “Expanding knowledge of parental involvement in children’s secondary
education: connections with high school seniors’ academic success”, Social Psychology of
Education, Vol. 5 No. 2, pp. 149-77.
Chan, Y.-C. and Cheng, Y.C. (1993), “A study of principals’ instructional leadership in Hong Kong
secondary schools”, Educational Research Journal, Vol. 8, pp. 56-67.
Chen, J.J. (2008), “Grade-level differences: relations of parental, teacher and peer support to
academic engagement and achievement among Hong Kong students”, School Psychology
International, Vol. 29 No. 2, pp. 183-98.
Cheng, Y.C. (1994), “Principal’s leadership as a critical factor for school performance: evidence
from multi-levels of primary schools”, School Effectiveness and School Improvement., Vol. 5
No. 3, pp. 299-317.
Cheng, Y.C. (2009), “Hong Kong educational reforms in the last decade: reform syndrome and
new developments”, International Journal of Educational Management, Vol. 23 No. 1,
pp. 65-86.
Cooley, V.E. and Shen, J. (2003), “School accountability and professional job responsibility: a
perspective from secondary principals”, NASSP Bulletin, Vol. 87 No. 634, pp. 10-25.
Cotton, K. (2001), “New small learning communities: findings from recent literature?”, available
at: www.nwrel.org/scpd/sirs/nslc.pdf (accessed May 4, 2006).
Education Bureau (EDB) (2008), “Performance indicator for Hong Kong schools 2008”, available
at: www.edb.gov.hk/FileManager/EN/Content_6456/pi2008%20eng%205_5.pdf (accessed
April 14, 2011).
Education Bureau (EDB) (2011), “Press releases and publications: secondary education”,
available at: www.edb.gov.hk/index.aspx?langno¼1&nodeid¼1039 (accessed April 14,
2011).
Education Commission (2005), Report on Review of Medium of Instruction for Secondary Schools
and Secondary School Places Allocation, Education Commission, Hong Kong.
Education Department (2002), “Guidelines for principals’ continuing professional development”,
available at: www.edb.gov.hk/filemanager/en/content_2129/Guideline1809.pdf (accessed
April 13, 2011).
JEA Elmore, R.F. (2003), Knowing the Right Thing To Do: School Improvement and Performance-
Based Accountability, NGA Center for Best Practices, Washington, DC, available at: www.
50,5 nga.org/center/divisions/l.1188.T_CEN_EDS#C_ISSUE_BRIEF#D_584300.html (accessed
May 2, 2011).
Elmore, R.F. (2005), “Accountable leadership”, The Educational Forum, Vol. 69 No. 2, pp. 134-42.
Fan, X. and Sivo, S.A. (2007), “Sensitivity of fit indices to model misspecification and model
606 types”, Multivariate Behavioral Research, Vol. 42 No. 3, pp. 509-29.
Hallinger, P. (2003), “Leading educational change: reflections on the practice of instructional and
transformational leadership”, Cambridge Journal of Education, Vol. 33 No. 3, pp. 329-51.
Hallinger, P. (2005), “Instructional leadership and the school principal: a passing fancy that
refuses to fade away”, Leadership and Policy in Schools, Vol. 4 No. 3, pp. 221-39.
Hallinger, P. and Heck, R. (1996), “Reassessing the principal’s role in school effectiveness: a
review of empirical research, 1980-1995”, Educational Administration Quarterly, Vol. 32
No. 1, pp. 5-44.
Hallinger, P. and Murphy, J. (1985), “Assessing the instructional leadership behavior of
principals”, Elementary School Journal, Vol. 86 No. 2, pp. 217-48.
Heck, R., Larson, T. and Marcoulides, G. (1990), “Principal instructional leadership and school
achievement: validation of a causal model”, Educational Administration Quarterly, Vol. 26
No. 2, pp. 94-125.
Heck, R.H. and Hallinger, P. (2009), “Assessing the contribution of distributed leadership to
school improvement and growth in math achievement”, American Educational Research
Journal, Vol. 46 No. 3, pp. 626-58.
Ho, S.-C.E. (2005), “Effect of school decentralization and school climate on student mathematics
performance: the case of Hong Kong”, Educational Research for Policy and Practice, Vol. 4
No. 1, pp. 47-64.
Ho, S.-C.E. and Willms, D.J. (1996), “Effects of parental involvement on eighth-grade
achievement”, Sociology of Education, Vol. 69 No. 2, pp. 126-41.
Horvat, E.M., Weininger, E.B. and Lareau, A. (2003), “From social ties to social capital: class
differences in the relations between schools and parent networks”, American Educational
Research Journal, Vol. 40 No. 2, pp. 319-51.
Howley, C. (1994), “The academic effectiveness of small-scale schooling”, (ED 372897), available
at: www.eric.ed.gov (accessed April 22, 2006).
Howley, C., Strange, M. and Bickel, R. (2000), “Research about school size and school
performance in impoverished communities”, (ED 448968) available at: www.eric.ed.gov
(accessed April 22, 2006).
Hu, L. and Bentler, P.M. (1999), “Cutoff criterion for fit indexes in covariance structure analysis:
conventional criteria versus new alternatives”, Structural Equation Modeling, Vol. 6 No. 1,
pp. 1-55.
Ingram, D., Louis, K.S. and Schroeder, R.G. (2004), “Accountability policies and teacher decision
making: barriers to the use of data to improve practice”, Teachers College Record, Vol. 106
No. 6, pp. 1258-87.
Johnson, M.K., Crosnoe, R. and Elder, G.H. (2001), “Students’ attachment and academic
engagement: the role of race and ethnicity”, Sociology of Education, Vol. 74 No. 4,
pp. 318-40.
Lee, J.C. and Dimmock, C. (1999), “Curriculum leadership and management in secondary schools:
a Hong Kong case”, School Leadership & Management, Vol. 19 No. 4, pp. 455-81.
Lee, J.C., Dimmock, C. and Au-Yeung, T.Y. (2009), “Who really leads and manages the curriculum
in primary schools? A Hong Kong case-study”, Curriculum Journal, Vol. 20 No. 1, pp. 3-26.
Lee, M. and Friedrich, T. (2007), “The ‘smaller’ the school, the better?: the smaller learning Instructional
communities program in US high schools”, Improving Schools, Vol. 10 No. 3, pp. 261-82.
leadership
Lee, M., Hallinger, P. and Walker, A. (2012a), “A distributed perspective on instructional
leadership in International Baccalaureate (IB) schools”, Educational Administration practices
Quarterly, in press.
Lee, M., Louis, K.S. and Anderson, S. (2012b), “Local education authorities and student
learning: the effects of policies and practices”, School Effectiveness and School 607
Improvement, Vol. 23 No. 2, pp. 133-58.
Lee, S.-Y. (2005), “Effects of school organizational culture and moral leadership on teachers’
professional zone of acceptance”, The Korean Journal of Educational Administration,
Vol. 23 No. 1, pp. 71-92.
Lee, V.E. and Burkam, D.T. (1992), “Transferring high schools: an alternative to dropping out?”,
American Journal of Education, Vol. 100 No. 4, pp. 420-53.
Lee, V.E. and Smith, J.B. (1995), “Effects of high school restructuring and size on early gains in
achievement and engagement”, Sociology of Education, Vol. 68 No. 4, pp. 241-70.
Leithwood, K. and Jantzi, D. (2009), “A review of empirical evidence about school size effects:
a policy perspective”, Review of Educational Research, Vol. 79 No. 1, pp. 464-90.
Leithwood, K., Louis, K.S., Anderson, S. and Wahlsttom, K. (2004), “Review of research: how
leadership influences student learning”, available at: www.wallacefoundation.org/
KnowledgeCenter/KnowledgeTopics/CurrentAreasofocus/EducationLeadership/Pages/
HowLeadershipInfluencesStudentLearning.aspx (accessed May 5, 2010).
Leung, C.H. and Choi, E. (2010), “A qualitative study of self-esteem, peer affiliation, and academic
outcome among low achieving students in Hong Kong”, New Horizons in Education,
Vol. 58 No. 1, pp. 22-42.
Libbey, H.P. (2004), “Measuring student relationships to school: attachment, bonding,
connectedness, and engagement”, Journal of School Health, Vol. 74 No. 7, pp. 274-83.
Linn, R.L. (2003), “Accountability: responsibility and reasonable expectations”, Educational
Researcher, Vol. 32 No. 7, pp. 3-13.
Louis, K.S. and Robinson, V.M.J. (2012), “External mandates and instructional leadership:
principals as meditating agents”, Journal of Educational Administration, Vol. 50 No. 5.
Louis, K.S., Leithwood, K., Wahlstrom, K.L. and Anderson, S.E. (2010), Investigating the Links to
Improved Student Learning: Final Report of Research Findings, University of Minnesota,
Minneapolis, MN, available at: www.cehd.umn.edu/carei/Leadership/Learning-from-
Leadership_Final-Research-Report_July-2010.pdf (accessed August 14, 2011).
McNeal, R.B. Jr (1999), “Parental involvement as social capital: differential effectiveness on
science achievement, truancy, and dropping out”, Social Forces, Vol. 78 No. 1, pp. 117-44.
Madyun, N. and Lee, M. (2010), “Effects of religious involvement on parent-child communication
regarding schooling: a study of black youth in the United States”, Journal of Negro
Education, Vol. 79 No. 3, pp. 295-307.
Marchant, G.J., Paulson, S. and Rothlisberg, B.A. (2001), “Relations of middle school students
perceptions of family and school context with academic achievement”, Psychology in the
Schools, Vol. 38 No. 6, pp. 505-19.
Marks, H. and Printy, S. (2004), “Principal leadership and school performance: an integration of
transformation and instructional leadership”, Educational Administration Quarterly,
Vol. 39 No. 3, pp. 370-97.
Marsh, H.W., Kong, C.K. and Hau, K.T. (2000), “Longitudinal multilevel models of the big fish
little pond effect on academic self-concept: counterbalancing contrast and reflected glory
effects in Hong Kong schools”, Journal of Personality and Social Psychology, Vol. 78 No. 2,
pp. 337-49.
JEA Morrison, K. and Lui, I. (2000), “Ideology, linguistic capital and the medium of instruction
in Hong Kong”, Journal of Multilingual and Multicultural Development, Vol. 21 No. 6,
50,5 pp. 471-86.
Ng, P.T. and Chan, D. (2008), “A comparative study of Singapore’s school excellence model with
Hong Kong’s school-based management”, International Journal of Educational
Management, Vol. 22 No. 6, pp. 488-505.
608 O’Day, J.A. (2002), “Complexity, accountability and school improvement”, Harvard Educational
Review, Vol. 72 No. 3, pp. 293-329.
O’Donnell, R. and White, G. (2005), “Within the accountability era: principals’ instructional
leadership behaviors and student achievement”, NASSP Bulletin, Vol. 89 No. 645, pp. 56-71.
Organization for Economic Co-operation and Development (OECD) (2009), Creating Effective
Teaching and Learning Environments: First Results from TALIS, OECD, Paris.
Phillips, D. and Ochs, K. (2003), “Processes of policy borrowing in education: some explanatory
and analytical devices”, Comparative Education, Vol. 39 No. 4, pp. 451-61.
Pong, S.L. (2009), “Grade level and achievement of immigrants’ children: academic redshirting in
Hong Kong”, Educational Research and Evaluation, Vol. 15 No. 4, pp. 405-25.
Pong, S.-L. and Tsang, W.K. (2010), “The educational progress of Mainland Chinese immigrant
students in Hong Kong”, in Hannum, E., Park, H. and Butler, Y.B. (Eds) Globalization,
Changing Demographics, and Educational Challengesin East Asia, (Research in Sociology
of Education), Vol. 17, Emerald Group Publishing Limited, Bingley, pp. 201-30.
Poole, W. (1995), “Reconstructing the teacher-administrator relationship to achieve systemic
change”, Journal of School Leadership, Vol. 5 No. 6, pp. 565-96.
Raudenbush, S.W. and Bryk, A.S. (2002), Hierarchical Linear Models: Applications and Data
Analysis Methods, Sage, Thousand Oaks, CA.
Reynolds, D., Stringfield, S., Teddlie, C. and Creemers, B. (2002), “The intellectual and policy context”,
in Reynolds, D., Teddlie, C., Stringfield, S. and Creemers, B. (Eds), World Class Schools:
International Perspectives on School Effectiveness, Routledge, New York, NY, pp. 3-14.
Robinson, V.M.J., Bendikson, L. and Hattie, J. (2011), “Leadership and student outcomes: are
secondary schools unique?”, in Robertson, J. and Timperley, H. (Eds), Leadership and
Learning, Sage, London, pp. 131-42.
Robinson, V.M.J., Lloyd, C.A. and Rowe, K.J. (2008), “The impact of leadership on student
outcomes: an analysis of the differential effects of leadership types”, Educational
Administration Quarterly, Vol. 44 No. 5, pp. 635-74.
Rumberger, R.W. and Larson, K.A. (1998), “Student mobility and the increased risk of high school
dropout”, American Journal of Education, Vol. 107 No. 1, pp. 1-35.
Salili, F. and Lai, M.K. (2003), “Learning and motivation of Chinese students in Hong Kong:
a longitudinal study of contextual influences on students’ achievement orientation and
performance”, Psychology in the Schools, Vol. 40 No. 1, pp. 51-70.
Salili, F., Lai, M.K. and Leung, S.S.K. (2004), “The consequences of pressure on adolescent
students to perform well in school”, Hong Kong Journal of Pediatrics, Vol. 9 No. 4, pp. 329-36.
Sawkins, J.W. (2002), “Examination performance in Scottish secondary schools: an ordered logic
approach”, Applied Economics, Vol. 34 No. 16, pp. 2031-41.
Schafer, J.L. and Graham, J.W. (2002), “Missing data: our view of the state of the art”,
Psychological Methods, Vol. 7 No. 2, pp. 147-77.
Schreiber, J.B. (2002), “Institutional and student factors and their influence on advanced
mathematics achievement”, Journal of Educational Research, Vol. 95 No. 5, pp. 274-86.
South, S.J., Haynie, D. and Bose, S. (2007), “Student mobility and school dropout”, Social Science
Research, Vol. 36 No. 1, pp. 68-94.
Stanton-Salazar, R.D. (1997), “A social capital framework for understanding the socialization Instructional
of racial minority children and youth”, Harvard Educational Review, Vol. 67 No. 1,
pp. 1-40. leadership
Stanton-Salazar, R.D. and Spina, S. (2000), “The network orientations of highly resilient urban practices
minority youth: a network-analytic account of minority socialization and its educational
implications”, The Urban Review, Vol. 32 No. 3, pp. 227-61.
Stiefel, L., Berne, R., Iatarola, P. and Fruchter, N. (2000), “High school size: effects on budgets and 609
performance in New York City”, Educational Evaluation and Policy Analysis, Vol. 22 No. 1,
pp. 27-39.
Vanderhaar, J.E., Muñoz, M.A. and Rodosky, R.J. (2006), “Leadership as accountability for
learning: the effects of school poverty, teacher experience, previous achievement, and
principal preparation programs on student achievement”, Journal of Personnel Evaluation
in Education, Vol. 19 Nos 1-2, pp. 17-33.
Walker, A. (2004), “Constitution and culture: exploring the deep leadership structures of
Hong Kong schools”, Discourse: Studies in the Cultural Politics of Education, Vol. 25 No. 1,
pp. 75-94.
Walker, A. (2006), “Divided they stand, united they fall: reform disconnection in Hong Kong”,
unpublished manuscript.
Walker, A. and Dimmock, C. (2000), “One size fits all? Teacher appraisal in a Chinese culture”,
Journal of Personnel Evaluation in Education, Vol. 14 No. 2, pp. 155-78.
Walker, A. and Ko, J. (2011), “Principal leadership in an era of accountability: a perspective
from the Hong Kong context”, School Leadership and Management, Vol. 31 No. 4,
pp. 369-92.
Walker, A. and Qian, H. (forthcoming), “Reform disconnection in China”, Peabody Journal of
Education, Vol. 87 No. 2, pp. 162-77.
Wang, J. (2006), “An empirical study of gender difference in the relationship between self-concept
and mathematics achievement in a cross-cultural context”, Educational Psychology, Vol. 26
No. 5, pp. 689-706.
Wiseman, A.W. (2004), Principals Under Pressure: The Growing Crisis, Scarecrow Press,
Lanham, MD.
Wong, E.O.W. (2003), “Leadership style for school-based management in Hong Kong”,
International Journal of Educational Management, Vol. 17 No. 6, pp. 243-7.
Wong, K.C., Lam, Y.R. and Ho, L.M. (2002), “The effects of schooling on gender differences”,
British Educational Research Journal, Vol. 28 No. 6, pp. 827-43.

Further reading
Getzels, J., Lipham, J. and Campbell, R. (1968), Educational Administration as a Social Process,
Harper & Row, New York, NY.
Hofstede, G. (2001), Culture’s Consequences: International Differences in Work-Related Values,
Sage, Thousand Oaks, CA.
Hoyle, E. and Wallace, M. (2007), “Educational reform: an ironic perspective”, Educational
Management Administration & Leadership, Vol. 35 No. 1, pp. 9-25.
Lee, M. and Hallinger, P. (2012), “National contexts influencing principals’ time use: economic
development, societal culture, and educational system”, School Effectiveness and School
Improvement.
Purkey, S. and Smith, M. (1983), “Effective schools: a review”, Elementary School Journal, Vol. 83
No. 4, pp. 427-52.
JEA Appendix 1
50,5
Schools
School type Government/aided 36 (85.7%)
Direct subsidy scheme 6 (14.3%)
Medium of instruction Chinese 12 (28.6%)
Chinese and English 10 (23.8%)
610 English 20 (47.6%)
School size Small 7 (16.7%)
Mid 17 (40.5%)
Large 18 (42.9%)
School performance Low 11 (26.2%)
Mid 23 (54.8%)
High 8 (19.0%)
Key staff
Gender Male 107 (59.8%)
Female 70 (39.1%)
Role Vice-principals 56 (31.3%)
Panel chairs 96 (53.6%)
Senior teachers 46 (25.7%)
Years of teaching in the present schools 0-3 years: 0 (0%)
4-7 years 15 (8.4%)
8-11 years 18 (10.1%)
12 years or above 132 (73.7%)
Students
Gender Male 964 (47.3%)
Female 1,065 (52.3%)
School attachment Mean 4.27
SD 0.75
Peer academic orientation Mean 4.56
SD 0.73
Perceived parental involvement Mean 2.96
SD 0.98
Years of attending the present school Mean 6.69
SD 1.25
HKCEEb Mean 61.8
SD 14.03
Table AI.
Characteristics of the Notes:aN ¼ 42 schools, 180 staff, and 2,037 students. However, figures in the table are based on the
sample schools, key staff, original data with missing values; bthe original scale of HKCEE (i.e. 1-30) was transformed to the scale,
and studentsa ranging from 3.33 to 99.9 for easier interpretations

Appendix 2

Years of enrollment School Peer academic Perceived parental

in the school attachment orientation involvement Gender

Years of enrollment
in the school 1 0.084** 0.054* 0.04 0.021
Table AII. School attachment 0.084** 1 0.46** 0.378** 0.039
Correlation matrix among
level-1 control variables (continued)
 Years of enrollment School Peer academic Perceived parental
Instructional
in the school attachment orientation involvement Gender leadership
practices
Peer academic
orientation 0.054* 0.46** 1 0.228* 0.061**
Perceived parental 611
involvement 0.04 0.378** 0.228** 1 0.06**
Gender 0.021 0.039 0.061** 0.060** 1 Table AII.

About the authors

Moosung Lee is Associate Professor of Educational Administration at the University of Hong
Kong and was previously Assistant Professor of Education Policy and Leadership at the Hong
Kong Institute of Education. His research focuses on social capital, social networks, international
schools and school improvement. Moosung Lee is the corresponding author and can be contacted
at: [email protected]
Allan Walker is the Joseph Lau Chair Professor of International Educational Leadership,
Head of the Department of Education Policy and Leadership and Director of the Asia Pacific
Center for Leadership and Change, Hong Kong Institute of Education. His research centers on
leadership in Chinese societies, leadership in international schools, leader development, and the
impact of leadership on student outcomes.
Yuk Ling Chui is a Research Assistant at the Asia Pacific Center for Leadership and Change,
Hong Kong Institute of Education. Her research interests are educational policy, education
reforms, and school leadership.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints
 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0957-8234.htm

JEA
50,5
The strength of accountability
and teachers’ organisational
citizenship behaviour
612 Eyvind Elstad
Department of Teacher Education and School Research, University of Oslo,
Received 29 November 2010
Revised 21 January 2011 Oslo, Norway
14 February 2011 Knut-Andreas Christophersen
Accepted 6 March 2011
Department of Political Science, University of Oslo, Oslo, Norway, and
Are Turmo
Department of Teacher Education and School Research, University of Oslo,
Oslo, Norway

Abstract
Purpose – Organisational citizenship behaviour (OCB) involves discretionary behaviour
advantageous to the organisation that goes beyond existing role expectations. The purpose of this
paper is to explore the link between the strength of accountability and teachers’ OCB within three
different management systems in which teachers are working: a system of assessment-based
accountability; a system of the gradual introduction of accountability devices; and a system with no
tests or examinations.
Design/methodology/approach – Structural equation modelling of cross-sectional surveys from
the three different management systems was used to estimate the path coefficients and to compare the
strength of relationships between concepts in the models.
Findings – The analysis shows that the factors that influence OCB in an accountability regime are
clearly different from those in a regime with weak or no accountability devices.
Research limitations/implications – A cross-sectional study does not allow us to test causal
relationships among antecedents of organisational citizenship behaviour. The use of self-reported
questionnaire data is another shortcoming. Furthermore, the response rates leave uncertainty about
whether the samples are representative.
Practical implications – The strength of accountability in education governance might influence
OCB among teachers. Educational administrators could benefit from exploring this issue to help the
establishment of institutional arrangements.
Social implications – The paper shows that OCB amongst teachers is essential for the smooth
functioning of schools for several reasons.
Originality/value – The study integrates three strands of theories that have their focal points in
employees’ perceptions of exchange: Shore’s theory on employee-organisation relationships; Bryk and
Schneider’s theory on trust in schools; and theories on accountability.
Keywords Accountability, Teachers, Educational administration, Governance,
Organizational citizenship behaviour, Social exchange theory, Teachers’ work, Leadership
Paper type Research paper

Introduction
Schools originally arose as a result of local initiatives, with local stakeholders driving
Journal of Educational
Administration
the progress (Cuban, 1993, 2009). Over a period, however, comprehensive institutional
Vol. 50 No. 5, 2012 hierarchies have been constructed in connection with the operation of schools.
pp. 612-628
r Emerald Group Publishing Limited
0957-8234
DOI 10.1108/09578231211249844 All three authors have contributed equally to this paper.
For instance, in Norway there are now over 20 hierarchical stages between the teacher Accountability
and the minister of education (Christophersen et al., 2010). During the past decades, and teachers’
schools have increasingly become part of management systems that include targets
and controls. A central purpose has been to contribute to increasing the quality OCB
of one of the school’s primary purposes: influencing pupils’ learning processes in a
positive way.
Elements of the management systems have included setting clear targets and 613
keeping check on the use of resources, the learning processes, and learning outcomes
(Elstad et al., 2009). Accomplishing this has involved the decentralisation of
responsibility (White Paper, 2003-2004). Local accountability systems have been
developed to make the schools responsible for the results they achieve. Incentives have
been introduced to motivate the head teachers, and negative repercussions exist if
results are not satisfactory. Within individual schools, the head teachers use new
tools that, to some extent, can make teachers responsible: measurements of pupil
satisfaction with the teaching; and results measurements by means of tests and exams
(Elstad, 2009). The vision has been to give “the teacher clearly-defined responsibility
for what the pupil learns”[1].
Such measurements can give the head teacher certain indications of the quality of
the work carried out by the teacher (naturally taken together with other impressions
that can form the basis of judgements). If measurement tools are used for assessments,
they will have consequences for the teachers; however, the reliability and validity of the
measurements must be subject to psychometric quality requirements. For instance, we
lack measurements of knowledge levels at both the beginning and end of the school
year in Norway. It is possible that such measurements would also improve the ability to
assess a teacher’s contribution to value added. Still, it would be very demanding in
terms of resources to carry out the comprehensive year-beginning and year-end
assessments that would satisfy all the quality requirements (Koretz, 2008).
In addition, there are a number of validity problems related to measurements. We
must acknowledge that the measurements are not good enough to establish a teacher’s
contribution to pupil learning during the course of a school year (Koretz, 2005).
In contrast to a strawberry picker’s productivity, for instance, the primary productivity
of a teacher is not easily measureable, and thereby not readily controlled by the
employer. Indicators such as days presence, minimum performance requirements, and
pupil satisfaction can give the school authorities some information, but form an
imperfect measurement of teacher job performance (Bevan and Hood, 2006). We should
draw the conclusion that there is a limit to the effectiveness of management through
targets and controls in our efforts to improve schools. Whether or not we believe in new
public management (NPM) (Christensen and Laegreid, 2001) in the education system,
we have to conclude that the school system is a difficult place to apply NPM. One
important element that is difficult to capture with available tools is organisational
citizenship behaviour (OCB), which this paper will address. In teaching, OCB involves
discretionary behaviour that is advantageous to the school and goes beyond existing
in-role expectations (Oplatka, 2006, 2009). OCB is “behaviour that is discretionary, not
directly or explicitly recognized by the formal reward system, and that in aggregate
promotes the effective functioning of the organisation. [y] the behaviour is not an
enforceable requirement of the role or the job description. [y] the behaviour is a matter
of personal choice” (Organ, 1988, p. 4).
School accountability and teacher accountability have increasingly become features
of the education systems of a number of countries. We know from research that the
JEA introduction of accountability devices can lead to both desirable and undesirable
50,5 effects (Hanushek and Raymond, 2005; de Wolf and Janssens, 2007). In educational
management these devices should ensure that the behaviour of individuals serves the
interests of the principals (national authorities and school owners), but there are a
number of challenges. Jacob and Levitt (2003) claim that “high-powered incentive
schemes are designed to align the behaviour of agents with the interest of the principal
614 implementing the system” (p. 843). However, whilst schools might respond to
incentives by working harder, incentive systems could also backfire towards unwanted
distortions. Accountability systems alter the incentives faced by schools (Besley and
Ghatak, 2003). Both experience and research show that it is demanding to design a
management system that is fair, and at the same time does not provide incentives to
game the system, such as artificially inflating grades, targeting instruction to near-
failing pupils (Reback, 2008), classifying more pupils as special needs ( Jacob, 2005) or
as disabled (Heilig and Darling-Hammond, 2008), shifting the amount of time devoted
to test subjects (Sturman, 2003), giving teachers a reason to cheat on standardised tests
( Jacob and Levitt, 2003), or altering the test-taking pool by strategically assigning
suspensions to low-performing pupils close to the test-taking period.
Podsakoff et al. (2000) identified several contextual categories of OCB, among which
were organisational and management behaviour. The present study explores how
accountability strength influences OCB. This is important because “cultural context
may affect the forms of citizenship behaviour observed in organisations” (Podsakoff
et al., 2000, p. 556). The context addressed in this paper is the strength of accountability
devices. Principals are accountable to school governors, and teachers are accountable
to principals. Accountability strength is the external pressure on schools to improve
pupil achievement (Carnoy and Loeb, 2003), such as requirements for goal attainment
and repercussions of results for schools. Average pupil scores on tests and
examinations are the main gauges of performance.
It is important that teachers are motivated to make an effort, above and beyond the
minimum requirements, during the course of their professional practice. Therefore, the
present study will explore the strength of accountability, and how accountability
influences the nature of exchanges in the school organisation and the impact on OCB.
Although OCB has generated much scholarly attention over the last three decades,
research on OCB has mainly been neglected in educational research. However, some
researchers use this approach in organisational studies of schools and have
investigated several aspects of factors influencing OCB (Oplatka, 2007; Tschannen-
Moran and Hoy, 1998; Goddard et al., 2001; Christ et al., 2003; Dick et al., 2006; Somech
and Bogler, 2002; Bragger et al., 2005; Somech and Drach-Zahavy, 2004; Somech, 2007;
Nguni et al., 2006; Yilmaz and Tasdan, 2009; Garg and Rastogi, 2006). However, none of
these studies have investigated the relationship between the strength of accountability
and OCB.
The purpose of this paper is to explore the link between the strength of
accountability and teachers’ OCB within three different management systems:

(1) teachers working under assessment-based accountability;

(2) adult teachers who are experiencing the gradual introduction of accountability
devices; and
(3) Nordic folk high-school (FHS) teachers who work without any tests or
examinations.
Three different management systems Accountability
Teachers working under assessment-based accountability with hard stakes: the and teachers’
accountability system for schools in Oslo
The local schools governing body of the City of Oslo has established a result-oriented OCB
external accountability system that makes the head teacher responsible for attaining
statistical targets in terms of the school’s activity[2]. Quality-assurance systems have
been established that make the head responsible for results achieved and allocate the 615
teacher four annual grades and result-linked bonus. The governing body for the city
municipality schools has given head teachers the responsibility of attaining clear,
result-based goals. The achievement of these goals would result in local salary bonuses
and performance grades given by local authority leaders. The intention is that this
component of quality-assurance systems should increase the quality of teachers and
heads. Quality teachers need to know the subject matter they are teaching and be
effective pedagogues. Head teachers need to know how to manage schools effectively.
The work of teachers in Oslo has been heavily affected by accountability devices,
such as tests, checks on disparities between teacher-allocated grades, and exam
results.
Interviews with school head teachers (Elstad, 2009) revealed an image of people
who pass onto the teachers the governors’ expectations of result improvements for
which they themselves have been given responsibility and use accountability tools
(e.g. average class results in tests, exams, etc.) as leadership devices: “I am able to
monitor learning in the various grades in a completely new way” (Elstad, 2009, p. 180).
Up to a certain point, we also observed heads teachers that attempted to absorb some
of the pressure put onto teachers.

Teachers working under the gradual introduction of accountability devices: adult

education in Norway
Adult immigrant schools in Norwegian municipalities supply host-language and
citizenship education of adult immigrants. About 23,000 persons were registered for
immigrant language training in October 2007, of which 60 per cent were persons with
the rights and obligations to access host-language training. Measurements of the
results and effectiveness of the training are based on the number of passed
examinations. Approximately 89 per cent pass the oral part of the exam, but only
around 59 per cent pass the written exam (2009)[3].
The media has over time reported hard political criticism of the immigrant schools.
For instance, the contents of a report from the Office of the Auditor General of
Norway[4] (OAGN, 2006) criticised the ministries and their external directorate (Vox),
as well as the county and local councils. Administrative accountability devices are
being rushed into a sector that is under strong political pressure because of the
relatively low proportion of adult learners who take and pass the written exams.
The directorate of adult education uses metrics such as average test results, pass rates,
and preparation time before, but the accountability strength is supposed to be only
weak to moderate.

Teachers working under a regime with no tests or examinations: FHSs

The FHSs do not grant academic degrees, but the schooling emphasises mutual
teaching and conversation and concrete formation of student participation. The core
values of the FHS are in opposition to the current emphasis on tests and clearly defined
teacher roles that have otherwise been prevalent in education governance in recent
JEA years. FHSs are characterised by the notion of humanistic-oriented self-formation of
50,5 the individual (Solhaug and Knutas, 2010). Since there are no exams in FHSs, it is quite
simply not possible to control learning outcomes in the same way as result-based
control is carried out in other types of educational establishments, without changing
the whole nature of this kind of school. The situation of the FHS stands as a clear
contrast to the way in which Oslo schools and adult education are run.
616 At a time when typical result-based management by educational authorities is a
feature of most aspects of education in Norway, FHSs still stand out as a free zone in an
educational sector otherwise infected by the new forms of public management.
External accountability devices and visions of commands via clear leadership are at
odds with the core values inherent in the FHS.

Theoretical framework
Many aspects of life can be conceived in terms of exchange (Blau, 1964). Shore et al.
(2006) made a fruitful distinction between economic exchange and social exchange.
Social exchange operates on the norm of reciprocity: “social exchanges require a long-
term orientation, since the exchange is ongoing and based on feelings of obligation”
(Shore et al., 2006, p. 839). Economic exchange focuses on bounded obligations that
reflect only basic expectations for the relationship: “Economic exchange does not
imply long-term or open-ended and diffuse obligations, but rather emphasis on
economic agreement such as pay for performance” (Shore et al., 2006, p. 839). In this
paper we suppose that both types of exchange perceptions exist concurrently in the
school, but only perceptions of social exchange induce OCB (Shore et al., 2006; Kuvaas
and Dysvik, 2009).
Social exchange theory is a possible theoretical explanation for OCB (Shore et al.,
2009). Teacher perception of exchange lies at the heart of our model. The study
examines further a concept of accountability strength transformed via the head
teacher’s clear leadership, and relational building between employer and employee as
antecedents and OCB as a consequence of employee-employer exchange (i.e. the head
teacher’s “coffee-cup diplomacy” during break-time to build relationships and underpin
the feeling of “all being in the same boat”, etc.). We focus on exchange concepts as
mediating variables between leader influences and teachers’ engagement in activities
directed towards helping their pupils. In doing so, we integrate two strands of research
on employee-organisation relationships (EOR) that have their focal point on employees’
perceptions of exchange: literature on EOR (Shore et al., 2006, 2009) and Bryk and
Schneider’s theory on trust in schools. Both theoretical strands are rooted in Blau’s
(1964) theoretical framework, and both suggest that perceptions of social exchange
could be an important determinant of employee behaviour (Organ et al., 2006).
Social exchange implies that teachers perceive that they are treated favourably
by their leader and feel a commitment to return the positive behaviour in their
teaching. Our study is also designed to address whether or not Bryk and Schneider’s
concept of leader-teacher relationships and EOR concepts of exchanges are distinct
constructs.
It is seen as useful for society that teachers develop goodwill and loyalty towards
the school in which they teach. Therefore, OCB is a central factor in school
improvement. A premise in our theoretical approach is that quality development
depends on school employees being able to identify with, involve themselves in, and
engage themselves on behalf of the school where they work, as well as demonstrate
effective administrative management. When employees do this, they work harder,
more responsibly, and smarter (Pfeffer and Veiga, 1999), possibly leading to higher task Accountability
performance. and teachers’
A foundation of the EOR literature is that “employees develop exchanges for
socio-emotional and economic reasons, and that the type of exchange relationship can OCB
predict employee motivation, attitudes and behaviour in relation to the employer”
(Kuvaas and Dysvik, 2009, p. 3500). Social exchange theory emphasises personal
associations on the part of teachers and of students. Shore et al. (2006) found support 617
for the idea that social exchange and economic exchange “can operate relatively
independently” (p. 858). The significance of the two forms of motivation is an empirical
question explored in this study. A high degree of social exchange and a low degree of
economic exchange for teachers suggest that we must hypothesise social exchange as
an important factor for OCB. As previously mentioned, the primary processes[5] in
the school are hard to pin down, despite attempts to measure the central aspects of the
processes (student surveys, etc.) and the output measures (test results, exams, value-
added indicators, etc.). For this very reason, it is important that teachers are motivated
during the course of their professional practice to make an effort that is above and
beyond the minimum requirements on behalf of the school. This is seen as an
important prerequisite for high learning outcomes in pupils. OCB is the dependent
variable in our hypothesised theoretical model, and social exchange is regarded as a
significant prerequisite for OCB (Organ et al., 2006).
Teachers’ work situations are quite different when compared with careers that
demand a high degree of immediate and direct interaction with colleagues during the
course of work. Therefore, we argue that teachers are of particular interest for EOR
research.

Methods
Sample and procedures
Three separate electronic questionnaire surveys were carried out in 2009 and 2010. The
surveys were distributed via e-mail and non-responding teachers were sent reminders
at least once. The surveys included the following groups of teachers:
. A total of 18 schools participating in a school development project in the Oslo
municipality were invited to take part in a teacher survey. In all 11 schools
responded positively to this invitation. These schools are located in areas of
differing socio-economic composition within Oslo.
. A survey was distributed to all teachers working in the schools for adult
immigrants in Norway (Vox).
. A survey was distributed to all teachers working in Norwegian FHSs.
Table I shows the number of participating teachers and the response rates for the
three surveys. The response rates were within the range typically found in surveys

Survey N Response rate (%) Table I.

The number of
Secondary teachers 236 54 participants and the
Folk high-school teachers 366 56 response rates for the
Adult teachers 764 44 three surveys
JEA of teachers. Significantly higher responses that those reported here are not usually
50,5 found in the literature.

Background characteristics of teachers in the three samples

Table II shows significant variation in the background characteristics of the teachers
in the three samples. The majority of the teachers in the Oslo and Vox samples were
618 female, whereas three out of five teachers were male in the FHS sample. Furthermore,
the FHS teachers were clearly younger than the teachers in the two other samples. Only
5 per cent of the FHS teachers were above 60 years old, while 20 per cent of the Vox
teachers were in this age category. Finally, the Oslo sample had, by far, the largest
proportion of teacher with a master’s degree, while the FHS sample had the smallest
percentage. In all, one in four FHS teachers had a higher education with duration of
between two and three years. These results reflect systematic differences among the
school types. Therefore, it was not tenable to assume that the teachers have been
randomly assigned to the three school types.
If the attributes of teachers in the two selections had been roughly similar,
we could have regarded the study as a quasi-experimental analysis; that is to say, we
could have assumed that the selections had arisen as though they had been taken from
a common pool of teachers. Possible differences in the measurements could, in such a
case, have been attributed to teachers being subjected to differing treatment. Such an
assumption is not tenable in this study because the differences between the two
selections (gender, age, and educational level) can be regarded as a consequence of a
systematic selection effect in how teachers are distributed among the Oslo schools,
Vox, and the FHS.

Measurement instruments
Measurement instruments previously reported in the literature were adapted and
translated into the Norwegian language. In the surveys, teachers responded to items on
a seven-point Likert scale ranging from “strongly disagree” to “strongly agree”, where
alternative 4 represented a neutral mid-point.

Oslo Vox FHS

N ¼ 243 N ¼ 764 N ¼ 366

Gender
Female 64 77 41
Male 36 23 59
Age (years)
o25 0 0 2
25-29 14 3 15
30-39 31 15 31
40-49 15 24 27
50-59 27 38 21
460 14 20 5
Table II. Level of education
Background Two to three years higher education 3 7 27
characteristics of the Four to five years higher education, bachelor level 60 72 48
teachers in the three Master’s degree 31 16 14
samples Others/not specified 6 5 11
Dependent variable. OCB was the dependent variable in our study. There exist various Accountability
measures of OCB in the literature. To avoid confounding conceptions of the variables, we and teachers’
followed Van Dyne and Lepine’s (1998) measure of OCB. This measure was translated
and adjusted. Three items were used. One sample item was: “Teachers in this school OCB
share their knowledge and discuss pupil-delivered work with other teachers”.
Mediating variables. We used social exchange and economic exchange measures
developed by Shore et al. (2006) and selected three items from each. Sample items were: 619
“My relationship with my organisation is strictly an economic one: I work and they pay
me” (economic exchange) and “I try to look out for the best interests of the organisation
because I can rely on my organisation to take care of me” (social exchange).
Independent variables. Bryk and Schneider’s (2002) “teacher-principal relationship”
construct was used. Sample items were: “In this school it is OK to discuss feelings,
concerns and frustrations with the school’s leadership” and “The head expresses a
personal interest in teacher professional development”. Furthermore, we developed a
construct called clear leadership. One sample item was: “Communication with the
management helps me to understand what is expected of me in order that the school
can achieve its goals”.
Analysis. Repeated confirmatory factor analysis was used to identify the best
indicators for the different constructs. Based on the factor analysis, items that did not
meet our criteria were removed. Three items for each construct were used to put the
latent variable into effect. Structural equation modelling (Blunch, 2008; Kline, 2005)
was of central importance as a method for the analysis of survey data. We used SPSS
18 and AMOS 18 as our analytical tools.

Results
Before testing the hypothesised model, we first investigated the measurement model.
The results showed that Cronbach’s a was above 0.70 for all the scales used. The
internal consistency was considered acceptable.
The main focus was to analyse how the independent and mediating variables were
related to OCB. The assessments were based on the p-value for the w2 statistic, root
mean square error of approximation (RMSEA), normed fit index (NFI), goodness-of-fit
index (GFI), and comparative fit index (CFI). The standard criteria p40.05, NFI,
GFI and CFI40.95 and RMSEA o0.05 were used for a good fit and criteria p40.05,
GFI and CFI40.90; RMSEAo0.08 indicated an acceptable fit between the model and
the data (Kline, 2005; Blunch, 2008). According to these criteria and the values of
RMSEA, p, GFI, and CFI (Figures 1-4), both the measurement and structure models
provided acceptable fit to the data.
Each figure combined five measurement models and one structure model. There
was one measurement model for each construct (latent variable). Large ellipses
identified the constructs, small ellipses (circles) the error terms (latent variables), and
rectangles the indicators (measured variables). The structure model consisted of the
constructs and the paths (arrows) between them. The paths indicated theoretically
assumed causal relations and the path coefficients (standardised regression
coefficients) the strength of the relation. Numbers linked to ellipses and to rectangles
represented variances and the double arrows with numbers indicated correlations.
The results from the structure model can be summarised as follows. All the fit
indices for RMSEA, CFI, and GFI indicated acceptable fit. The w2 statistics did not
quite support this conclusion. However, w2 statistics are heavily influenced by sample
size, and should count less in the conclusion than the other fit indices.
JEA e12 e11 e10

50,5
0.60 se05 0.70 se04 0.45 se03

e17 cl01 0. 0.7

88 7 0.84

67
0.
0.79
620 e18 cl02
0.87
cl
0.01 0.81
se

0.76
83 e_oc
0.

–0
e20 cl04 0.05 0.29

.06
0.86 ocb01 e1
54
e_se 0. 0.29
0.78
–0.19 ocb ocb02 e2
0.81
e_ee 0.66
0.20
–0.24 0.
82
ocb03 e3
e13 pt01 9
0.67
0.8

0.47 0.68

e14 pt03 pt 0.63 ee

0.86
–0.12
0.74
.87 0.
9
e15 pt04 0 81
0.2

0.32
0.75

0.08 ee05 0.10 ee04 0.66 ee03

e6 e5 e4

Standardised estimates
Figure 1.
A structure model of the RMSEA = 0.049; p-KJI = 0.000; GFI = 0.967; CFI = 0.974
three samples spliced
together Notes: cl, clear leadership; ee, economic exchange; pt, principal-teacher relation;
se, social exchange; and ocb, organizational citizenship behaviour

In all three sub-models, it was the path PT-SE-OCB that was the most significant
in explaining antecedents of OCB. This was not surprising for the FHS, but was
somewhat surprising in the case of the Oslo schools and Vox. A possible interpretation
is that the quality of relationships amongst school professionals is the most central
prerequisite for typical school quality-ensuring processes. If this is the case, it indicates
a limitation of NPM techniques, or at least a complementary quality for human
relationships that a management system should take into account.
Furthermore, the results showed that the relationship between clear leadership,
social exchange, and OCB was much stronger in the Oslo sample than in FHS and Vox
samples. The results further indicate that clear leadership had positive effects on both
social exchange and economic exchange in the Oslo sample, while no corresponding
positive effects were established in the FHS sample. The latter is not surprising
because clear leadership has no clear tradition within the FHS; there are no external
management systems to back up the communication of goals for statistical targets
within the FHS institution. No other results-based statistics than simple completion of
the courses exist within the FHS. The influence of the leadership occurs principally
through relationship building within the FHS.
It is surprising, in the emerging accountability regime of adult education in Norway,
that clear leadership has a small negative effect on social exchange perceptions in the
 e12 e11 e10 Accountability
and teachers’
OCB
0.51 se05 0.71 se04 0.57 se03

e17 cl01 0. 0.7

0.84

76
95 1

0.
e18
0.90

cl02
0.89
cl
0.35 0.73
se
621
0.79
e_oc
80
0.2

e20 cl04 0. 0.09 0.88

2

ocb01 e1
0.79
62
e_se 0. 0.38
0.80
–0.39 ocb ocb02 e2
0.65
e_ee 0.86 0.42
0.
0.07 53
ocb03 e3
e13 pt01
5

0.28
0.5

0.42 0.65

e14 pt03 pt 0.43 ee

0.89 –0.51
0.79
8
0.8
5

0.

e15 pt04
0.3

0.57
84

0.78

0.12 ee05 0.33 ee04 0.70 ee03

e6 e5 e4

Standardised estimates

RMSEA = 0.062; p-KJI = 0.000; GFI = 0.919; CFI = 0.963 Figure 2.

A structure model of the
Notes: cl, clear leadership; ee, economic exchange; pt, principal-teacher relation; Oslo sample
se, social exchange; and ocb, organizational citizenship behaviour

Vox sample. However, it is not surprising that CL-SE-OCB had a fairly prominent
path in the Oslo school. In contrast, CL-EE-OCB was much weaker. Furthermore,
strong positive relations between clear leadership and positive head-teacher
relationships, and between head-teacher relationships and social exchange, were
established in all three samples. Head-teacher relationships were negatively related to
economic exchange in both samples: the poorer the relationship, the higher the teacher
emphasis on economic exchange. Finally, economic exchange was negatively related to
OCB in both FHS and Vox samples.

Discussion
The purpose of this paper was to explore the link between the strength of
accountability and teachers’ OCB within three different management systems:
(1) teachers who work under assessment-based accountability;
(2) adult teachers who experience the gradual introduction of accountability
devices; and
(3) FHS teachers who work without tests or examinations.
The management systems for Oslo schools and for FHS represented the extreme ends
of a scale running from moderate accountability strength to no accountability strength.
JEA e12 e11 e10

50,5
0.59 se05 0.75 se04 0.40 se03

e17 cl01 0. 0.87

0.
89

63
77

0.
622 e18
0.79

cl02
0.88
cl
-0.05 0.88
se

0.78 e_oc

–0
84
e20 cl04 0. 0.07 0.20

.09
ocb01 e1
0.88
56
e_se 0. 0.31
0.78
–0.19 ocb ocb02 e2
0.89
e_ee 0.14 0.79
0.
– 0.24 90
ocb03 e3
e13 pt01
8

0.81
0.9

0.51 0.72

e14 pt03 pt 0.01 ee

0.88 –0.02
0.77
8
0.8

0.
8

e15 pt04
0.2

82
0.30
0.78

0.07 ee05 0.09 ee04 0.67 ee03

e6 e5 e4

Standardised estimates

Figure 3. RMSEA = 0.055; p-KJI = 0.000; GFI = 0.954; CFI = 0.971

A structure model of the
Vox sample Notes: cl, clear leadership; ee, economic exchange; pt, principal-teacher relation;
se, social exchange; and ocb, organizational citizenship behaviour

We regarded adult education as occupying a middle position. Path coefficients

measured the degree of effect induced by one variable in the arrow-pointed variable.
There was barely any trace of substantive influence of clear leadership on OCB within
adult education. As such, adult education lies far closer to FHS than to the Oslo
schools. This finding was surprising. We observe that for FHS and Vox, PT featured
strongly as a factor in the path from PT-SE-OCB, whilst CL had no loading from
CL-SE. For the Oslo schools, the former path had a somewhat lower loading than for
FHS and adult education, whilst the path CL-SE-OCB had a relatively clear loading
(though still weaker than the path PT-SE). This could be interpreted that the clear
leadership was the motivating energy for OCB in the Oslo schools. This interpretation
might be related to the Oslo teachers’ understanding of their roles and their acceptance
of the management system with its clear goals for the schools. This understanding of
roles could also involve the Oslo teachers perceiving the means that are used as
legitimate. The extent to which the latter interpretation is valid should be the subject of
more research.

Limitations
As with all similar studies, this study had certain limitations from a methodological, as
well as a conceptual perspective. We acknowledge these limitations and argue that
 e12 e11 e10 Accountability
and teachers’
OCB
0.62 se05 0.51 se04 0.42 se03

e17 cl01 0. 0.7 0.72

65
83 9

0.
e18
0.70

cl02
0.85
cl
0.02 0.68
se
623
4
0.71 0.8
e_oc
–0

e20 cl04 –0.05 0.30

.16

ocb01 e1
0.88
55
e_se 0. 0.30
0.80
–0.02 ocb ocb02 e2
0.80
e_ee 0.64
0.15
0.
–0.24 88
ocb03 e3
e13 pt01
1

0.78
0.8

0.33 0.57

e14 pt03 pt ee
0.76 –0.06 0.04
0.57
0 .83
9

e15
0.

pt04
0.1

97

0.23
0.68

0.03 ee05 0.05 ee04 0.94 ee03

e6 e5 e4

Standardised estimates
RMSEA = 0.051; p-KJI = 0.000; GFI = 0.946; CFI = 0.965 Figure 4.
A structure model of the
Notes: cl, clear leadership; ee, economic exchange; pt, principal-teacher relation; FHS sample
se, social exchange; and ocb, organizational citizenship behaviour

they contributed to a foundation for future studies. First, cross-sectional studies

represent only instant images of these three organisations and do not allow us to test
causal relationships among antecedents of OCB or inducement of exchanges amongst
employer and employees. More experimental and longitudinal research is needed to
address the complexity of interaction dynamics between head teachers and teachers,
and the associated inducements on teachers’ motivation to put pressure on pupils’
learning processes. We cannot defend the assumption of selections as though they were
taken from a common pool of teachers. However, more research along this line
(homogenous samples of teachers) could help in understanding the assumed causal
relationships. A second shortcoming was the use of self-reported questionnaire data.
The subjective component of such data is undeniable. Independent judgements can
provide interesting data about an employee’s performance.
A third shortcoming is that we did not have the opportunity to couple teacher’s
self-reporting with objective goals in terms of their task performance, because it was
not possible to examine the associations between OCB in teaching and pupil learning
outcomes. We found that this was extremely difficult in this instance because FHS,
Vox, and the Oslo schools all have a pupil base that does not allow comparisons in
terms of ages. For future research it would be good to develop different accountability
samples that are all based on teaching the same student age groups in mainstream
settings; there are possibilities here of cross-national research.
JEA Only a limited number of antecedents of teachers’ OCB have been examined. One
50,5 challenge in relation to measuring such factors is that measurement becomes
increasingly difficult when the remoteness of the factor increases, separated from the
school and teacher by the tiers of hierarchical organisation in the education sector.
A possible improvement is to examine some cases in-depth in order to attain a better
understanding of antecedents of OCB. It would, thus, be possible to research these
624 phenomena in depth.
The relation between the strength of external pressure via accountability
systems-attitudes-behaviour might not be linear. The functioning of external
pressure via accountability systems could be lower than an optimal value of pressure.
If this were the case, one would see a positive linear relationship between pressure and
performance. If, on the other hand, the pressure is higher than the optimal value, one
could possibly identify a negative relationship between pressure and performance.
A final limitation is that the response rate leaves uncertainty about whether the
selection is representative. However, this response rate is not unusual for social science
studies.

Implications for leadership practice

Despite its shortcomings, this study contributes to an understanding of how
perceptions of exchange mediate antecedents and consequences. If the statistical
associations between the variables represent causal relationships, our findings could
have implications for practice.
This survey joins the ranks of studies highlighting that the quality of human
relations among school professionals makes a difference in how OCB unfolds in spite of
accountability strength. A significant category of quality-enhancing mechanisms in
our survey was the pathway principal-teacher relation-social exchange-OCB.
Uncertainty is the teacher’s lot in life (Lortie, 1976, p. 133); good relationships can
contribute to reducing teacher uncertainty and vulnerability, which in turn can be a
catalyst for energy that will be put to service on behalf of the school. When
relationships with the head teacher are good, collective processes amongst the teaching
staff can also be strengthened at the same time as the unfortunate individualistic
characteristics of some teaching cultures (Lortie, 1976) are reduced. A further
mechanism inherent in the mentioned path is that the head teacher can be regarded as
a role model and, thereby, a moral calling for school improvement.
The challenges for school leaders are situated in employee-organisation
relationships. Even if western countries use substantial resources on education,
there are substantial challenges in relation to school quality. There are, therefore,
grounds for emphasising the factors that contribute to quality promotion of primary
processes in the school system and how management affects these primary processes.
In practice, this involves concentrating on how school leadership, management, and
administration affect the quality of human resources in the school, which are seen as
important for promoting student learning outcomes.

Conclusion
The purpose of this study was to explore the link between the strength of
accountability and teachers’ OCB within three different management systems. The
analysis shows that the factors that influence OCB in an accountability regime
are clearly different from those in a regime with weak or no accountability devices. The
strength of accountability in education governance may influence OCB among
teachers. The main implication of the study is, therefore, that educational administrators Accountability
could benefit from exploring this issue to help the establishment of institutional and teachers’
arrangements. OCB among teachers is essential for the smooth functioning of schools.
The results could be seen to support the main tenets of social exchange theory: good OCB
relations between educators and leaders are important also for OCB, and educators
respond to social exchanges by way of positive behaviours. Our findings show that the
strength of the path from leadership to economic exchange is weak in all three samples. 625
Conversely, the path from relational trust between leaders and teachers to social
exchange is rather strong in all three samples. This result emphasised the importance
of strengthening human relationships between leaders and teachers. We found some
moderate support for the importance of clear leadership for OCBs, but only in the
regime of assessment-based accountability. Furthermore, our pattern of results in
general suggested that social exchange theory is appropriate for explaining OCB in
the three management systems. There were, however, grounds for emphasising the
complexity of factors. This paper serves as a starting point that will stimulate further
research.
Notes
1. www.regjeringen.no/nb/dep/smk/aktuelt/taler_og_artikler/statsministeren/statsminister_jens_
stoltenberg/2008/nyttarstale-2008.html?id¼495221
2. www.utdanningsetaten.oslo.kommune.no/kvalitetsportalen_oslo/
3. www.folkeuniversitetet.info/avd_filer/ls/spraaktest/Npr_2_og_3_des.2005-d.d.pdf
4. Document No. 3:14 (2007-2008). The National Auditor’s survey of the provision for adults of
elementary/secondary-school education and of teaching at sixth-form level was presented to
the Norwegian Parliament on 11 September 2008.
5. Primary processes in schools are all the instructional methods, curriculum choices, and
organisational preconditions that make it possible for pupils to acquire knowledge.

References
Besley, T. and Ghatak, M. (2003), “Incentives, choice, and accountability in the provision of public
services”, Oxford Review of Economic Policy, Vol. 19 No. 2, pp. 235-49.
Bevan, G. and Hood, C. (2006), “What’s measured is what matters: targets and gaming in the
English public health care system”, Public Administration, Vol. 84 No. 3, pp. 517-38.
Blau, P. (1964), Exchange and Power in Social Life, John Wiley and Sons, New York, NY.
Blunch, N.J. (2008), Introduction to Structural Equation Modelling Using SPSS and AMOS, Sage
Publications, London.
Bragger, J.D., Rodriguez-Srednicki, O., Kutcher, E.J., Indovino, L. and Rosner, E. (2005),
“Work-family conflict, work-family culture, and organizational citizenship behaviour
among teachers”, Journal of Business and Psychology, Vol. 20 No. 2, pp. 303-24.
Bryk, A.S. and Schneider, B. (2002), Trust in Schools: A Core Resource for Improvement, Russell
Sage Foundation, New York, NY.
Carnoy, M. and Loeb, S. (2003), “Does external accountability affect student outcomes?
A cross-state analysis”, Education Evaluation and Policy Analysis, Vol. 24 No. 4, pp. 305-31.
Christ, O., Dick, R.V., Wagner, U. and Stellmacher, J. (2003), “When teachers go the extra mile: foci
of organizational identification as determinants of different forms of organizational
citizenship behaviour among schoolteachers”, British Journal of Educational Psychology,
Vol. 73 No. 3, pp. 329-41.
JEA Christensen, T. and Laegreid, P. (2001), New Public Management: The Transformation of Ideas
and Practice, Ashgate, Aldershot.
50,5
Christophersen, K.A., Elstad, E. and Turmo, A. (2010), “Is teacher accountability possible? What
influence does the teacher have over the pupil’s learning?”, Scandinavian Journal of
Educational Research, Vol. 54 No. 5, pp. 413-29.
Cuban, L. (1993), How Teachers Taught: Constancy and Change in American Classrooms
626 1890-1990, Teachers College Press, New York, NY.
Cuban, L. (2009), Hugging the Middle: How Teachers Teach in an Era of Testing and
Accountability, Teachers College Press, New York, NY.
de Wolf, I.F. and Janssens, F.J.G. (2007), “Effects and side effects of inspections and accountability
in education: an overview of empirical studies”, Oxford Review of Education, Vol. 33 No. 3,
pp. 379-96.
Dick, R.V., Grojean, M., Christ, O. and Wieseke, J. (2006), “Identity and the extra mile:
relationships between organizational identification and organizational citizenship
behaviour”, British Journal of Management, Vol. 17 No. 4, pp. 283-301.
Elstad, E. (2009), “Schools which are named, shamed and blamed by the media: school
accountability in Norway”, Educational Assessment, Evaluation and Accountability, Vol. 21
No. 2, pp. 173-89.
Elstad, E., Nordtvedt, G.A. and Turmo, A. (2009), “The Norwegian assessment system: an
accountability perspective”, CADMO, Vol. 17 No. 2, pp. 89-103.
Garg, P. and Rastogi, R. (2006), “Climate profile and OCBs of teachers in public and private
schools of India”, International Journal of International Management, Vol. 20 No. 6,
pp. 529-41.
Goddard, R.D., Tschannen-Moran, M. and Hoy, W.K. (2001), “A multilevel examination of the
distribution and effects of teacher trust in students and parents in urban elementary
schools”, Elementary School Journal, Vol. 102 No. 1, pp. 3-17.
Hanushek, E.A. and Raymond, M.E. (2005), “Does school accountability lead to improved
performance?”, Journal of Policy Analysis and Management, Vol. 24 No. 2, pp. 297-327.
Heilig, J.V. and Darling-Hammond, L. (2008), “Accountability Texas-style: the progress and
learning of urban minority students in a high-stakes testing context”, Educational
Evaluation and Policy Analysis, Vol. 30 No. 2, pp. 75-110.
Jacob, B. (2005), “Accountability, incentives and behaviour: evidence from school reform in
Chicago”, Journal of Public Economics, Vol. 89 Nos 5-6, pp. 761-96.
Jacob, B. and Levitt, S. (2003), “Rotten apples: an investigation of the prevalence and predictors of
teacher cheating”, Quarterly Journal of Economics, Vol. 117 No. 3, pp. 843-78.
Kline, R.B. (2005), Principle and Practice of Structural Equation Modelling, The Guildford Press,
New York, NY.
Koretz, D. (2005), “Alignment, high stakes, and the inflation of test scores”, in Herman, J. and
Haertel, E. (Eds), Uses and Misuses of Data in Accountability Testing. Yearbook of the
National Society for the Study of Education, Blackwell Publishing, Malden, MA, pp. 99-118.
Koretz, D. (2008), Measuring Up: What Educational Testing Really Tells Us, Harvard University
Press, Cambridge, MA.
Kuvaas, B. and Dysvik, A. (2009), “Perceived investment in permanent employee development
and social and economic exchange perceptions among temporary employees”, Journal of
Applied Social Psychology, Vol. 39 No. 10, pp. 2499-524.
Lortie, D. (1976), School Teacher, University of Chicago Press, Chicago.
Nguni, S., Sleegers, P. and Denessen, E. (2006), “Transformational and transactional leadership
effects on teachers’ job satisfaction, organizational commitment, and organizational
 citizenship behaviour in primary schools: the Tanzanian case”, School Effectiveness and Accountability
School Improvement, Vol. 17 No. 2, pp. 145-77.
and teachers’
OAGN (2006), “Riksrevisjonens undersøkelse av opplæringen i grunnskolen”, [The office of the
auditor general of Norway’s investigation of primary and lower secondary school; in OCB
Norwegian], Dokument 3:10 (2005-2006), The Office of the Auditor General of Norway, Oslo.
Oplatka, I. (2006), “Going beyond role expectations: towards an understanding of the
determinants and components of teacher organizational citizenship behaviour”, 627
Educational Administration Quarterly, Vol. 42 No. 3, pp. 385-423.
Oplatka, I. (2007), “Managing emotions in teaching: towards an understanding of emotion
displays and caring as nonprescribed role elements”, Teachers College Record, Vol. 109
No. 6, pp. 1374-400.
Oplatka, I. (2009), “Organizational citizenship behaviour in teaching: the consequences
for teachers, pupils, and the school”, International Journal of Educational Management,
Vol. 23 No. 5, pp. 375-89.
Organ, D.W. (1988), Organizational Citizenship Behaviour: The Good Soldier Syndrome,
Lexington Books, Lexington, MA.
Organ, D.W., Podsakoff, P.M. and MacKenzie, S.B. (2006), Organizational Citizenship Behaviour:
Its Nature, Antecedents and Consequences, Sage, Beverly Hills, CA.
Pfeffer, J. and Veiga, J.F. (1999), “Putting people first for organizational success”, Academy of
Management Executive, Vol. 13 No. 2, pp. 37-48.
Podsakoff, P.M., MacKenzie, S.B., Paine, J.B. and Bacharach, D.G. (2000), “Organizational
citizenship behaviours: a critical review of the theoretical and empirical literature and
suggestions for future research”, Journal of Management, Vol. 26 No. 3, pp. 513-63.
Reback, R. (2008), “Teaching to the rating: school accountability and the distribution of student
achievement”, Journal of Public Economics, Vol. 92 Nos 5-6, pp. 1394-415.
Shore, L.M., Bommer, W.H., Rao, A.N. and Seo, J. (2009), “Social and economic exchange in the
employee-organization relationship: the moderating role of reciprocation wariness”,
Journal of Managerial Psychology, Vol. 24 No. 8, pp. 701-21.
Shore, L.M., Tetrick, L.E., Lynch, P. and Barksdale, K. (2006), “Social and economic exchange:
construct development and validation”, Journal of Applied Social Psychology, Vol. 36 No. 4,
pp. 837-67.
Solhaug, T. and Knutas, A. (2010), Som en sang i sinnet – som et eneste sollyst minne. Elevers
utbytte av folkehøgskolen, Folkehøgskolerådet, Oslo.
Somech, A. (2007), “Promoting organizational citizenship behaviour in schools: the impact of
individual and organizational characteristics”, Educational Administration Quarterly,
Vol. 43 No. 1, pp. 38-66.
Somech, A. and Bogler, R. (2002), “Antecedents and consequences of teacher organizational and
professional commitment”, Educational Administration Quarterly, Vol. 38 No. 4, pp. 555-77.
Somech, A. and Drach-Zahavy, A. (2004), “Exploring organizational citizenship behaviour from
an organizational perspective: the relationship between organizational learning and
organizational citizenship behaviour”, Journal of Occupational and Organizational
Psychology, Vol. 77 No. 3, pp. 281-98.
Sturman, L. (2003), “Teaching to the test: science or intuition?”, Educational Research, Vol. 45
No. 3, pp. 261-73.
Tschannen-Moran, M. and Hoy, W.K. (1998), “Trust in schools: a conceptual and empirical
analysis”, Journal of Educational Administration, Vol. 36 No. 4, pp. 334-52.
Van Dyne, L. and Lepine, J.A. (1998), “Helping and voice extra-role behaviours: evidence
of construct and predictive validity”, Academy of Management Journal, Vol. 41 No. 1,
pp. 108-19.
JEA White Paper (2003-2004), Culture for Learning, Ministry of Education, Oslo.
50,5 Yilmaz, K. and Tasdan, M. (2009), “Organizational citizenship and organizational justice in
Turkish primary schools”, Journal of Educational Administration, Vol. 47 No. 1, pp. 108-26.

Further reading
Bogler, R. and Somech, A. (2004), “Influence of teacher empowerment on teachers’ organizational
628 commitment, professional commitment and organizational citizenship behaviour in
schools”, Teaching and Teacher Education, Vol. 20 No. 3, pp. 277-89.
Culver, S.M., Wolfe, L.M. and Cross, L.H. (1990), “Testing a model of teacher satisfaction for
blacks and whites”, American Education Research Journal, Vol. 27 No. 2, pp. 323-49.
DiPaola, M.F. and Hoy, W.K. (2005), “Organizational citizenship of faculty and achievement of
high school students”, The High School Journal, Vol. 88 No. 3, pp. 35-44.
Figlio, D.N. and Getzler, L.S. (2002), “Accountability, ability, and disability: gaming the system”,
Working Paper No. 9307, National Bureau of Economic Research, Cambridge, MA.
Fresko, B., Kfir, D. and Nasser, F. (1997), “Predicting teacher commitment”, Teaching and Teacher
Education, Vol. 13 No. 4, pp. 429-38.

About the authors

Eyvind Elstad is a Professor in the Department of Teacher Education and School Research,
University of Oslo. Eyvind Elstad is the corresponding author and can be contacted at:
[email protected]
Knut-Andreas Christophersen is an Assistant Professor in the Department of Political
Science, University of Oslo.
Are Turmo is an Associate Professor in the Department of Teacher Education and School
Research, University of Oslo.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints
 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0957-8234.htm

School leaders
External mandates and as mediating
instructional leadership: school agents
leaders as mediating agents
Karen Seashore Louis 629
University of Minnesota, Minneapolis, Minnesota, USA, and
Received 25 October 2011
Viviane M. Robinson Revised 28 March 2012
University of Auckland, Auckland, New Zealand 16 May 2012
Accepted 16 May 2012

Abstract
Purpose – The purpose of this paper is to examine how US school leaders make sense of external
mandates, and the way in which their understanding of state and district accountability policies affects
their work. It is posited that school leaders’ responses to external accountability are likely to reflect a
complex interaction between their perception of the accountability policies, the state and district
contexts in which those policies are situated and their own leadership beliefs and practices.
Design/methodology/approach – The authors use both principal and teacher survey data to
explore the question of how perceptions of external policy are associated with instructional
leadership behaviors. Cases of seven principals are employed to flesh out the findings from the survey
analysis.
Findings – It is concluded that external accountability policy may have a positive impact on
instructional leadership – where they see those policies as aligned with their own values and
preferences, and where they see their district leaders as supportive of school-driven accountability
initiatives. In these cases, school leaders internalize the external accountability policies and shape
them to the particular needs that they see as priorities in their own school. Where one or the other
of these factors is weak or missing, on the other hand, leaders demonstrate more negative attitudes
to external accountability and weaker instructional leadership.
Originality/value – This analysis draws on a unique, large-scale data base and uses a mixed
methods approach to answer the question.
Keywords United States of America, Schools, Leaders, Leadership, Educational policy,
Implementation, Improvement, Legislation, Organizational behaviour, Principals
Paper type Research paper

Introduction
Educators around the world are living in a period of almost unprecedented policy
activism. State pressures for schools to be more publicly accountable for their results
were observed in the USA by the end of the 1980s (Wills and Peterson, 1992;
Wohlstetter, 1991), and similar initiatives were felt in other countries (Gordon, 1995;
Louis and van Velzen, in press). Whether the pressures for improvement generated
by elected officials are due to the increased availability of comparative data (such as
PISA or TIMSS internationally, and NAEP within the USA) or the more general
circulation of theories about how to improve the management of public services (new
public management), one of the consequences has been a steady stream of research on
the nature and impact of accountability policies. Journal of Educational
Through the mid-1990s, policy analysts began to examine the logic underlying state Administration
Vol. 50 No. 5, 2012
accountability and testing systems, most of which operated under the assumption that pp. 629-665
public test results would motivate school-based educators to work harder (because r Emerald Group Publishing Limited
0957-8234
their performance would be made public) and smarter (because they would have clear DOI 10.1108/09578231211249853
JEA objectives for improving student learning, McDonnell, 1994; O’Day, 2002). Initially,
50,5 the responsibility for creating accountability systems rested with the states, and most
responded to the expectation, albeit in different ways (Louis et al., 2005, 2010b).
In the USA, the passage of the federal No Child Left Behind (NCLB) Act, which
took effect in 2003, added another layer to the school accountability hierarchy. This
legislation tied federal funding to state testing, public reporting and a set of
630 increasingly harsh consequences for schools that failed to improve. Like most federal
education mandates, NCLB provided an umbrella framework which states adapt to
suit their existing political, and accountability contexts (Louis et al., 2010b). It also
provided additional guidance and support, and, in particular, created the expectation
that states would assist their districts in becoming better at supporting schools. The
additional expectations for states that were embedded in NCLB have garnered less
research attention than the law’s effects on schools. This is, perhaps, surprising
because the law was clear in its expectations that districts were seen as key
institutional actors in creating results (Rorrer et al., 2008). Over the last decade,
increasing research attention to districts suggests that, while some may be up to the
task of setting strong improvement agendas and providing support for schools
(Honig, 2006; Kerr et al., 2006; Togneri and Anderson, 2003), others are not (Coburn
et al., 2009; Spillane, 1998). The NCLB requirements are, thus, contextualized within
state and school district policies related not only to accountability but to any related
reform initiatives.
Given this layered accountability context, school leaders responses to federal
accountability mandates are likely to reflect a complex interaction between their
perception of state policies and support, the specific district contexts in which
those policies are situated, including ongoing district reform initiatives and their
own leadership beliefs and practices. In this paper, we investigate the relationship
between these three forces, as we address a number of questions:
. Do school leaders’ perceptions of state or district school improvement policies
and procedures influence how they lead their schools?
. Are those school leaders who perceive their accountability context (state or
district), as supportive more likely to behave as instructional leaders?
. How do school leaders integrate their own leadership beliefs and agendas
with the external mandates to which they are subject? To what extent do they
perceive conflict between the two?
. To what extent do school leaders’ relationships with the district office enable
them to craft coherence between the external accountability policies and their
own agendas?

The focus of NCLB policy is on student outcomes, not on leadership. Although in most
states, NCLB has led to specification of the achievement targets to be met by each
school, the federal legislation is silent about the role of leadership in achieving them.
Yet the implications of NCLB for school leaders are profound. If targets are to be
achieved, at least in the population of schools relevant to this paper, then instructional
leadership that is skilled in monitoring student achievement data, in using that data to
identify student needs and in building collective teacher capability to address those
needs, are just a few of the leadership skills and responsibilities that are required.
In schools without high levels of capability in these areas, skilled leadership of
teacher change will also be required. NCLB not only sets highly ambitious student School leaders
achievement targets, but by implication, also sets a very particular leadership agenda. as mediating
The reaction of leaders to this agenda is, we predict, partly determined by the extent to
which it is consistent with their current leadership beliefs and practice, and partly agents
determined by their perception of district capacity to support them in achieving it.

Related literature 631

While our research questions belong within a long tradition of research on educational
policy implementation, we draw, in particular, on two more recent theoretical trends
within this literature. The first is that of sensemaking – the central insight of which
is that implementing agents interpret policies through relevant prior beliefs and
understandings (Spillane et al., 2002b). It is their interpretations of policy, rather than
an invariant and objective policy idea, that is implemented. The explanation of policy
implementation, therefore, requires investigation of the nature and determinants of
actors’ policy interpretations.
The second theoretical idea is that of crafting coherence (Honig and Hatch, 2004).
How NCLB policies are enacted by leaders is not just a function of their independent
sensemaking processes. The quality of their instructional leadership is also likely to be
influenced by the degree of coherence achieved between their own leadership agendas
and the policy agenda set by the district office (Youngs et al., 2011). We use the term
“coherence achieved” deliberately, because we see coherence as the result of sustained
interactions between policy makers and implementing agents – one in which there is
mutual influence and adjustment so that there is a principled integration of internal
and external policy agendas.
The third conceptual resource we draw on is that of instructional leadership. We
explain the particular aspect of this concept that informed the items we included in our
principal and teacher surveys.

Sensemaking and policy implementation

How people and organizations act is determined, in part, by the way in which they
interpret and make sense of ambiguous event and environments (Gioia et al., 1994;
Thomas et al., 2001; Walsh, 1995). A sensemaking framework is, thus, particularly
important for understanding how ambitious and comprehensive external policies,
like NCLB, may affect leaders within organizations. As Weick et al. (2005) point out,
“sensemaking involves turning circumstances into a situation that is comprehended
explicitly in words and that serves as a springboard into action” (p. 526).
A sensemaking perspective thus allows us to investigate why accountability policies
have engendered such disparate responses among implementing agents (Coburn, 2006;
Mintrop and Sunderman, 2009; Spillane et al., 2002a; Louis et al., 2005). Drawing on
research on situated cognition, researchers have shown that how reform policies are
understood and implemented is “a function of the interaction of (a) the policy signal;
(b) the implementing agents’ knowledge, beliefs, and experience; and (c) the
circumstances in which the local actor attempts to make sense of policy” (Spillane et al.,
2002b, p. 420).
Policies themselves comprise complex ideas that are often ambiguous and open to
multiple interpretations. How educators understand a given policy will depend on the
constructs, schema, values and emotional reactions they bring to the policy content
(O’Day, 2002). These prior understandings act as lenses or filters through which
the policy is understood, and those understandings, in turn, shape decisions about the
JEA changes that the policy requires (Spillane et al., 2002b). One key determinant of the
50,5 response of implementing agents to accountability policies is the degree of consistency
they perceive between the policy content and their own values, goals and strategies for
achieving them. For example, school leaders who believe that fair treatment of special
needs students is incompatible with assessing them against age-related benchmarks
will have a different response to AYP targets for this categorical group than school
632 leaders who do not share this belief.
When information, such as that communicated through accountability policies, is
judged to be consistent with prior understandings, it is less likely to generate defensive
reactions than if it is seen as challenging those understandings (Hart et al., 2009).
On the other hand, perceived consistency may lead to over-assimilation – that is to
the mistaken interpretation that the policy requires no change in the school or its
leadership (Spillane et al., 2002 b). Given the prevalence of defensive motivation – that
is the propensity to select and interpret information in ways that confirm rather
than challenge prior beliefs (Hart et al., 2009), unmediated policy interventions will
have less chance of motivating change than those that are mediated in skillful
interactions designed to engage the views of implementing agents.
When a large gap is perceived between the policy content and their current
understandings, considerable cognitive effort may be required to interpret the policy
accurately, let alone change practice accordingly. How a leader understands a policy, in
other words, is partly a function of what they already understand and value. That is
why we report in some detail about how our sample of principals and assistant
principals understood their leadership at the level of both values and practices.
These descriptions provide the backdrop against which we can better understand their
responses to the external accountability policies.
The interaction of policy content and prior beliefs and experience is evident in a
study of how three Chicago elementary schools responded to accountability policies
(Spillane et al., 2002b). Consistent with the policy intent, all three responded to the
policy by focussing on test results, increasing their instructional leadership and
concentrating on math and literacy. Despite these commonalities, there were important
differences in their sensemaking. The principal with a deep commitment to school and
teacher autonomy and high-level skill in data analysis, used external data to mobilize
staff to understand and resolve uneven math results. A principal newly appointed to a
school in probation mediated the policy as an opportunity to “get off probation” and
focussed teachers on the test content and test taking skills to achieve that goal. In the
third school, where student performance was high, the principal struggled to convince
her teachers that the policy had relevance to their context. This study points to
the importance of the principal’s ability to help teachers “make sense” of accountability
by connecting policy to goals and values that are already important to their staff.
Features of the situation in which sensemaking occurs also shape leaders’ cognitive
and behavioral responses to accountability policy. A critical part of the situation is the
district office and we know there is wide variation in how districts support schools
in their pursuit of improvement. Some school-district relationships reflect a more
bureaucratic and others a more professional type of accountability (O’Day, 2002). In
bureaucratic accountability the district’s role is largely confined to monitoring school’s
performance to ensure that personnel are focussed on externally set targets and on
employing resources and tools for achieving them (Daly and Finnigan, 2011; O’Day,
2002). In professional accountability relationships, the district role goes beyond
monitoring, to the development of common frameworks for instruction and the
provision of support focussed on the improvement of teaching and learning (Youngs School leaders
et al., 2011). Such districts share the challenge with school leaders of raising the as mediating
capability of teachers to improve performance.
agents
Crafting coherence between external and internal improvement agendas
Researchers have repeatedly pointed out the negative consequences of policy
incoherence on school improvement (Hess, 1999; Newmann et al., 2002). Multiple 633
simultaneous school reform initiatives, each with their own accountabilities, can
fragment the efforts of leaders and teachers, create stress, and increase cynicism
without achieving the intended improvement (Hess, 1999). While there have been
multiple calls for increased policy coherence, neither of the two dominant strategies
have proved effective.
The first, an “outside-in” strategy, treats incoherence as a problem of policy design
(Honig and Hatch, 2004). Increased coherence is sought by “creating conglomerate
policies that subsume the different strands of reform activity into one carefully
orchestrated whole” (Knapp et al., 1998). Increased policy coherence, in other words,
will increase coordination and unity of teacher practice at the school level by making
fewer and better aligned reform demands on schools (Coburn, 2003). There are a
number of reasons why reliance on outside-in approaches has not proved sufficiently
effective. First, national and global political forces that are likely to produce multiple,
discrete and short-lived reform responses in many countries (Louis and van Velzen,
in press). Second, our previous discussion of sensemaking showed that even when
a policy is technically coherent it may not be experienced as such by implementing
agents. “Coherence depends on how implementers make sense of policy demands
and on the extent to which external demands fit a particular school’s culture,
political interests, aspirations, conceptions of professionalism and on-going
operations” (Honig and Hatch, 2004, p. 18). The inevitable variety of context and
cognition means that the coherence of a given policy will vary across implementing
agents.
A common alternative, an inside-out approach to creating coherence, sees schools
as the source of the solution. Rather than mandate a specific policy, a policy framework
or menu is promulgated that provides leaders with considerable discretion about how
they will meet broad policy goals (Youngs et al., 2011). For example, rather than
requiring implementation of a specific reform approach, the US comprehensive school
reform initiative provided criteria against which schools’ own reform proposals would
be evaluated and funded (Borman et al., 2003). It was up to school leaders to shape their
proposals in ways that integrated what they saw as school priorities with federal
requirements. In practice, these inside-out, or bottom up approaches to coherence may
generate local ownership, but they may also fail to create reasonable fidelity to the
intent of reform policies (Coburn, 2003). When leaders cannot establish a clear internal
agenda, a discretionary external policy will not, in itself, help them find a more
coherent focus. In addition, even those leaders who have a clear reform focus do not
necessarily have the strategic skills required to integrate it with the external policy
framework. For example, leaders may choose school reform initiatives for reasons that
have little to do with their alignment to existing school reform purpose and strategies
(Datnow, 2000; Datnow, 2002). Inside-out strategies may also fail when district leaders
do not make the shift from a top down authority relationship to a more collegial one in
which school leaders and policy-makers work together to integrate their internal and
external agendas (Datnow and Stringfield, 2000; Coburn, 2003).
JEA Both the outside-in and inside-out approaches treat coherence as an objective
50,5 condition that is predominantly the responsibility of either the policy-maker or
school-based educators. Instead, Honig and Hatch (2004) argue that a third approach is
needed – one that treats coherence as a process that “requires school and school
district central office leaders to work in partnership to continually ‘craft’ or ‘negotiate’
the fit between external demands and schools’ own goals and strategies” (p. 17).
634 Honig and Hatch (2004) have proposed three broad conditions as conducive
to this on-going interaction. First leaders need to establish improvement goals that are
flexible enough to be integrated with external agendas but specific enough to enable
monitoring of goal progress. Second, they need to use these goals to make decisions
about how to respond to demands by either “bridging” and increasing interaction with
external actors or “buffering” teachers from outside influences that they see as
distractions from school priorities. Third, if coherence is crafted through on-going
interaction between policy makers and implementing agents, district officials need to
define their interactions with schools as a partnership rather than an authority
relationship. Reducing leader regulatory roles also limits the degrees of freedom that
are needed to bring differing internal and external agendas into alignment, and may
result in re-introducing a more authoritative approach (Honig, 2006). If district officials
see their job as monitoring results, administering sanctions and providing centrally
determined resources, then the system is unlikely to engage with schools about the
support they require from the district to achieve improvement goals. Building leaders
may be left on the periphery, without the networks and access to the expertise they
need to meet accountability goals (Daly and Finnigan, 2011).
Under a professional accountability relationship, the school-district relationship is
one of reciprocal rather than top down influence – external policies are adjusted or
interpreted in ways that serve local goals and local efforts are shaped by overall policy
purposes (Coburn, 2003; Louis et al., 2005). It is the on-going interaction around
the pursuit of particular school goals that produces a coherent approach to school
improvement. This interactive conception of coherence suggests that neither district
officials nor school leaders acting alone will be able to craft coherence in school
improvement efforts – it requires both to be working in partnership to learn about,
refine and support the school’s own improvement goals.
In the context of our study, the interacting agents are principals and district leaders.
In our qualitative cases we investigate the extent to which these two parties were
able to craft coherence between school goals and strategies and those of the district
office – within the larger context of state accountability policies.

Instructional leadership
External accountability policies treat the school as the unit of analysis and intervention
and yet it is the practice of teachers within those schools that determines improved
performance (O’Day, 2002). The causal logic of accountability driven improvement
requires school leaders to intervene to improve teacher practice so that student
performance comes closer to accountability targets. Leadership work that is focussed
on the improvement of teaching and learning is generally known as instructional
leadership (Hallinger, 2005). A distinction is sometimes made between those leader
behaviors that involve direct involvement with teachers, as in observation of
classrooms, feedback to teachers, discussion of results and leadership of teacher
learning, and more indirect instructional leadership practices involving the
organization of curriculum and instruction and the creation of classroom and school
routines which protect time for student and teacher learning (Robinson et al., 2011). School leaders
What makes all these diverse behaviors illustrative of instructional leadership is that as mediating
they are deliberately directed by leadership to the improvement of teaching and
learning. agents
While evidence about the impact of instructional leadership shows that, all else
being equal, students achieve more in schools with strong instructional leadership
(Marzano et al., 2005; Robinson et al., 2008), our particular concern here is its role in 635
contexts of district-driven accountability. Formal leaders must motivate teacher
change by setting goals and expectations, resourcing those goals and providing
high-quality opportunities to learn the knowledge and skills to achieve the goals
(Finnigan, 2010; Leithwood et al., 2002). Teacher commitment to the pursuit of
ambitious goals requires teachers to believe that the goals are important and that they
have the capacity to achieve them (Latham and Locke, 2006; Locke and Latham, 1990).
Building that capacity is a key responsibility of school and district leaders.
The relationship between leadership and teachers’ sense of efficacy was tested in
a recent study of Chicago schools under accountability sanctions (Finnigan, 2010).
Teacher efficacy was assessed as the strength of their belief that they could make a
significant difference to their students’ achievement, including that of their most
difficult students. Of four leadership constructs, two (instructional leadership and
support for change) were significantly related to teachers’ sense of efficacy. The
instructional leadership measure included items assessing leaders’ goal setting (vision,
setting high standards for both staff and students), the application of professional
learning in classrooms and leaders’ understanding and monitoring of student learning.
These findings support the conclusion of an earlier study which also showed the
substantial impact on school performance of the combination of instructional and
change oriented leadership (Marks and Printy, 2003).
There is ample evidence that the desire of policy makers for strong instructional
leadership falls short of the reality (Cooley and Shen, 2003; Horng et al., 2009). Part of
the difficulty is that increased instructional leadership requires leaders to spend
relatively more time on the educational and less on the management aspects of their
role, or at least to integrate instructional concerns into all aspects of their managerial
decision making. Making this shift poses considerable professional and organizational
challenges. The professional challenges include developing the capabilities required to
engage in the practices described as instructional leadership (Nelson and Sassi, 2005;
Robinson, 2011; Stein and Nelson, 2003). Broadly specified, those capabilities involve
deep knowledge of teaching and learning, the ability to bring that knowledge to bear in
context-specific management and instructional decision making and to build relational
trust in the process. It is the integration of the knowledge, context-specific problem
solving and trust building that characterizes the work of instructional leadership
(Robinson, 2011). The organizational challenges include aligning the organizational
and systemic conditions that shape school leaders’ work to the goal of stronger
instructional leadership.
Since states provide the regulatory framework within which school leaders’
work, we have assessed school leaders’ perceptions of the degree to which they
see state policies as broadly supportive of their instructional leadership. We also
assessed leaders “perceptions of the capacity of their district office to support
their school’s improvement by providing reliable information on school and student
performance and using it to offer teacher learning opportunities that are aligned to
student learning needs. We predict a relationship between school leaders” perceptions
JEA of district capacity and their teachers’ reports of the strength of their instructional
50,5 leadership.

Data collection methods

This paper is based on a secondary analysis of an existing data base. The data sources
for this paper are drawn from a larger multi-method study that included nine randomly
636 sampled states[1], 45 districts and approximately 175 randomly sampled schools
within these states[2].

Sampling
The random sample of districts was stratified to ensure that schools from larger
and smaller districts would be selected, as well as district with higher and lower
poverty and racial/ethnic minority enrollments. Size definitions followed the categories
established by the National Center for Educational Statistics. As our measure of
poverty, we used the percentage of students in the district who were eligible for free
or reduced price lunch. Ethnic/minority enrollment was measured by the percentage of
non-white students in the district. Schools were selected within each district to include
one high school, one middle school and two elementary schools. Two districts per
state (one larger, one smaller) were selected for site visits, and within these we
conducted site visits in one secondary and one elementary school. When districts
declined to participate, we replaced them with another district or school that had
similar characteristics. The larger project and the complex sampling strategy for
schools, and teachers are described in detail elsewhere (Louis et al., 2010a). Teachers
and principals filled out surveys that tapped assessments of building, district and state
leadership, as well as school culture, in 2005 and 2008 (the second and fourth year of
the five-year study).

Data collection
The teacher and principal surveys were mailed to individual schools and were
typically completed by all teachers during a school staff meeting. Each survey form
was accompanied by a blank envelope that could be sealed to ensure confidentiality
so that none of the principals had access to their teachers’ responses. The teacher
response rate was slightly over 65 percent in 2005, and 55 percent in 2008[3]. Principals
filled out their surveys at the same time, with at least one principal or assistant
principal responding from most of the sampled schools in 2008, with responses
received from 138 principals and 119 assistant principals, representing 37 districts
and 146 schools. The principal response rate was 78 percent in 2005 and slightly over
60 percent in 2008. Data on school background characteristics (student socio-economic
status, urbanicity, school enrollment, etc.) and student achievement were gathered from
publicly available information. This paper draws from the 2008 teacher and principal
surveys.
We collected three rounds of site-visit data from the 36 schools and 18 districts.
These occurred in years 2, 3 and 5 of the study. As noted above, two districts in each
of the nine states, one larger and one smaller, had agreed to be site visit districts.
Typically we visited the two sampled buildings, but in two of the small districts we
visited three buildings each, which were all the regular buildings in those two districts.
Interviews were held with the principal (and assistant where they were present) during
all three site visits, as were interviews with district leaders[4]. Interview protocols for
varied between site visits, but focussed on their leadership values, actual leadership
practices, improvement strategies, responses to state accountability mandates and School leaders
relationships with the district, community and state. as mediating
Limitations agents
Overall, our analysis and questions were motivated by some of the survey findings
that we will present below. However, the paper’s main value, we believe, lies in the more
in-depth examination of the principal interviews. This paper, like most secondary 637
analyses, has both strengths and limitations. Because the study from which these data
were drawn was large in scale, we have access to a data base and a far more diverse
population of schools, districts and states than one typically finds in qualitative
investigations of how leaders adapt to the new pressures of accountability. On the other
hand, because the study was not designed with our questions in mind, we are forced to
rely on the survey and interview data that exists. To give just one example,
superintendents and other district office staff were not asked specific questions about
the principals included in this analysis, so we are only able to use their interviews as a
generic “validity check” for consistency with the principal’s responses. Some survey
data that would have been useful, such as the Annual Yearly Progress status for the
survey schools, is not available.
Finally, although the study is relatively comprehensive in its coverage of the USA,
the findings are clearly embedded in a specific national context. However, because
many countries are experiencing similar accountability policies, we believe that there
may be some value in the results for non-US readers.

Data analysis: survey data

We first developed a merged data base that allowed us to match principal and teacher
survey responses in 2008[5]. Due to non-response, either from teachers or any
principal/assistant principal, on key items, our useable school sample size was reduced
from 157 to 147, which includes 201 principal and assistant principal respondents[6].
We used the 2008 principal survey to develop measures of our two key predictor
variables. A positive state policy index was measured by principal’s degree of
agreement with four positively worded items about state policy. Principals’ perceptions
of district accountability policies were assessed by their degree of agreement that the
district was focussed on and could support the school’s improvement efforts, which we
called the positive district accountability index.
An index of instructional leadership was developed from the 2008 teacher survey. It
comprised seven items about school leader involvement in such things as setting
standards and providing instructional advice (Louis et al., 2010a). Since the design of
the wider study involved analyses of the relationships between selected school
variables and student achievement, the instructional leadership measure was based
on published evidence about the relative impact of various types of practice on
student outcomes (Robinson et al., 2008; Marzano et al., 2005). Most of the items
included in the survey were about direct instructional leadership, including
frequency of classroom visits; provision of teaching advice and involvement in
teacher planning. We judged that while it might be difficult for high school
leaders to engage in these more direct forms of instructional leadership, they
provided a realistic assessment of the focus of our elementary and middle school
leaders.
The description of each of these variables, along with associated survey items and
scale reliability is reported in Appendix.
JEA In addition, we assumed that a number of measures of school characteristics might
50,5 mediate the effects of accountability policies on principal behavior. In particular,
secondary schools are typically under considerably more pressure on accountability
measure, since they must not only perform on state tests, but also on graduation rates
and increasing expectations about students being “college ready.” We therefore
included school level (elementary/secondary) in our analysis. Student poverty levels
638 (percent eligible for free or reduced price meals) are known to be associated with lower
student test scores, which may put schools in jeopardy of not passing muster on state
indicators.
For this paper we examined the distribution of leaders’ responses on key variables.
Multiple regression analysis was used to examine predictors of administrative
instructional leadership. The useable sample size for this analysis includes all
respondents who identified themselves as an individual with the title of principal. This
decision was based on the assumption that both principals and assistant were in
a position to make assessments of state and district policies, that including both
increased the sample size, and that teachers were asked to include all leaders in their
rating of instructional leadership (see Appendix)[7].

Case study sampling and data analysis

As already indicated, site visits were conducted in 18 districts and 36 schools in nine
states. For this paper, an outlier sampling strategy was used to further reduce the
number of case study schools. All site visit schools were ranked by the instructional
leadership score described above. While the relationship between this score and school
type (elementary or high school) was 0.33 in the survey sample, the relationship was
considerably stronger in this sub-group, with none of the high schools ranked above
the median. As a consequence, we decided to eliminate high schools from the case
analysis and focus on elementary and middle schools serving students in grades
eight or lower. We sampled the four top and four lowest scoring schools for analysis.
As we proceeded, we discovered that one of the lowest scoring sites had incomplete
data so we limited our analysis to the four highest and three lowest instructional
leadership schools.
The purpose of the qualitative analyses is to deepen our understanding of the inter-
relationships between principals’ attitudes to state and district policies and teachers’
ratings of their leadership practices. We begin with a brief description of the context of
each the seven cases and then provide a thematic analysis organized as a comparison
between three principals categorized as low instructional leaders (LIL) and four
categorized as high instructional leaders (HIL). This comparison is organized by
four themes: principals’ espoused leadership theory, instructional leadership emphasis
and practices; response to state and district accountability policies, and their
relationship with district office. Under each heading, we note the contrasts between
the two principal groups while also being careful to note the variation within each
group and the overlaps between them.
The seven schools, which were located in six different states, varied widely in size,
demographic characteristics of the student body and the administrative experience of
their principal (Table I). Both LIL and HIL schools came from larger and smaller
districts; both groups had relatively new and relatively experienced principals; both
included schools with students from varying backgrounds (although the LIL schools
had somewhat fewer minority students). The only obvious imbalance across the two
groups is that all of the LIL principals were male, while all of the HIL principals were
 School characteristics Principal characteristics
Principal years Principal total
Approximate Student SES Principal as administrator years in
School School location enrollment background % minority AYP sanction sex in the school administration

Low-instructional leadership schools

Pinewood Medium-sized town (NJ) 850 Higher Low No (2005, 2008) M 9 9
Elementary
Chillwater Medium-sized city (NM) 500 Mid-level Mid-level No, 2005; Yes, M 6 30
Elementary 2008
East Starr Middle Small town (NY) 500 Higher Low No (2005, 2008) M 2 7
School
High-instructional leadership schools
Antica Elementary Medium-sized town 420 Low High Yes, 2005; No, F 4 17
(NM) 2008
“Maple Island Medium-sized town (NC) 500 Low Mid-level No (2005, 2008) F 7 14
Elementary”
Walker Elementaryb Large suburb (OR) 450 Mid-level Mid-level No (2005, 2008) F 2 2
Overton Elementary Small suburb (MS) 250 Mid-level Mid-level No (2005, 2008) F 4 8

Notes: aWe do not include specific demographic data in order to protect site and individual anonymity. For both poverty and percent minority, low signifies
o20%, while higher more than 60%. bThere was principal turnover in this school during the study. Principal characteristics refer to the principal in place
when the 2008 survey was carried out
agents
as mediating
School leaders

their principalsa
study schools and
Characteristics of case
639

Table I.
JEA female. However, Walker Elementary, an HIL school, was led by a male principal
50,5 during the first two and a half years of data collection. Only Maple Island had an
assistant principal, and it was therefore decided to confine our analysis to the seven
who served in similar positions.
Given our focus on sensemaking and coherence we chose a largely inductive
approach to interview analysis. Each author independently read interviews from
640 both higher and lower instructional leadership and proposed themes that might be
relevant to the research questions. After discussion we settled on seven coding
categories that were discussed by all principals and that were relevant to the focus
of our inquiry:
(1) features of school and community context that they believed shaped their
leadership;
(2) descriptions of their personal leadership theories, including leadership vision
and values;
(3) descriptions of their leadership priorities, particularly with respect to
instructional leadership;
(4) connections between federal and state policies and their own leadership
priorities and actions;
(5) connections between district policies and their own leadership priorities and
actions;
(6) discussion of resources and their effects on their leadership; and
(7) principals’ sense of “ownership” of external policies from federal, state or
district sources and their integration of external and internal initiatives.
Responsibility for the initial analysis of each principal’s interviews was divided
equally between the two authors. Quotes and summaries of interview data were
entered in an excel spreadsheet under each of the seven headings and then a draft case
was written for each school. Before the cross-case analysis was completed the seven
themes were reduced to the four which connected most closely to our research
questions and captured the contrasts between HIL and LIL: leadership vision; reported
instructional leadership practices; response to external accountability policies and
relationship with district office. Once each case was written, a cross-case analysis
was conducted in which the leadership of higher and lower scoring schools was
systematically compared on each of the four themes. Claims made in the final cross-
case analysis were checked against the entries in the excel spreadsheet and, in some
cases, against the original interview transcripts. Finally, the first author read the
district interviews with superintendents and other key leaders to cross-check
principals’ claims about both the district and the district’s response to state and federal
policies.

Survey results: instructional leadership and responses to accountability

policies
Our analysis of the survey data was focussed on two questions: first, how did the
principals describe the policy context? Second, did their views of their policy context
help to explain their instructional leadership behaviors? Both of these questions
address, albeit indirectly, the context in which they “make sense” of the policy
environment in which they are located, and whether they view them as coherent School leaders
and incoherent. as mediating
School leader ratings of their policy context agents
In general, school leaders had a relatively positive view of their district’s
policies supporting accountability for student learning. On a six point scale,
40 percent of the leaders marked “strongly agree” with the statement that “Our district 641
has the capacity for reliable student performance” while almost as many agreed
that “Our district incorporates student performance data in district level decisions”
and “Our district assists schools in the use of school/student performance data.”
The only area where districts were viewed somewhat less positively was on the item
“The district uses student achievement data to determine teacher professional
development needs and resources,” where o30 percent indicated that they strongly
agreed.
The principal survey also reveals a positive assessment of the effects of state
policy. For example, the mean for ratings on the item “State standards stimulate
additional professional learning in our school” was 4.39 on a six-point scale, with more
than 60 percent of the respondents giving the item a rating between somewhat
agree and strongly agree. Although fewer gave the items “State policies help us to
accomplish our school’s learning objectives” and “The state communicates clearly
with our district about educational priorities” the highest rating of “strongly agree,”
both items suggest that most have positive views of the state’s role in these areas. Only
one of the items, “The state gives schools the flexibility and freedom to do their work,”
garnered a response suggesting that most respondents disagree[8].
An additional question is whether perceptions of state policy vary by the state in
which they are located. While it is beyond the scope of this paper to analyze the
different policy environments in which the schools were located, we know from a
variety of media sources and other analyses that, for example, North Carolina and
Texas were early adopters of state-wide tests and accountability, while Oregon and
Missouri made limited use of tests for accountability until NCLB, and Nebraska did not
institute state-wide tests until four years after NCLB had passed[9]. We might
reasonably expect this history to make a difference, since research on teachers’
perceptions of accountability policy suggests that familiarity breeds acceptance
(Ingram et al., 2004). We therefore conducted an ANOVA and also examined boxplots to
look for patterns. The ANOVA was significant, but as both the post hoc Schefes test
(not tabled) and the boxplots (Figure 1) indicate, this is largely due to a difference
between Texas (which was highly rated, although it places high accountability
demands on) and Oregon (which was rated lower, although it has a tradition of giving
more leeway to local districts). Furthermore, Texas and North Carolina, with the
longest consistent history of using tests for accountability, are perceived more
positively than those that are later into the testing game. No other paired comparison is
significant.
This finding hints that, as we might expect, school leaders’ world views are not
conditioned by deep knowledge of the larger national policy context in which they find
themselves. Rather, they arrive at assessments of their state’s policies by comparing
today with what they may have experienced a few years ago or on the basis of what
local conversations suggest they should expect. As one pre-NCLB study noted, leader’s
views of state accountability policies tend to be deeply affected by the amount of time
they have had to adjust to them: new policies are regarded as onerous; time allows
JEA 2.00
50,5 44
147
1.00

0.00

ZSTATE
642
–1.00 62
49

–2.00 96 129

–3.00
Indiana

Missouri

North Carolina

Nebraska

New Jersey

New Mexico

New York

Oregon

Texas
Figure 1.
Positive perceptions of
state policy (ZSTATE)
by state

people to “make sense” of the requirements so that older policies are accepted as
a reasonable status quo (Louis et al., 2005).

The relationship between school leader ratings of policy context and instructional
leadership
Two regression analyses were carried out to examine the relationship between school
leaders’ assessments of district and state policy and their leadership behavior. In the
first we looked only at the association of the positive state policy index and teachers’
ratings of their leader(s) instructional leadership, controlling for two key school
characteristics (building level, coded as elementary or secondary; and the percentage
of students in poverty, or eligible for free and reduced-price lunch), as well as the
respondent’s position as a principal or assistant principal. We then examined
the positive district accountability index in an alternative model in order to determine
whether district policy and accountability priorities were associated with school leader
instructional leadership behaviors. The results of these parallel regressions, presented
in Table II, reveal three key findings:
(1) The first regression shows that school leaders’ positive perceptions of
state policy are significantly associated with teachers’ ratings of their school
leaders’ instructional leadership behavior. The relationship could signal either
that state policy shapes school level leadership, or that school leaders’ current
level of instructional leadership shapes their attitudes to an instructionally
focussed policy or that there is a reciprocal relationship been the two.
(2) The second regression suggests that district policies are equally important.
(3) In both cases, school level and the demographic composition of the school are
stronger predictors than the associations with policy.
 Predictorsa b coefficients t Significance R2 Adjusted R2
School leaders
as mediating
1 (Constant) 29.43 agents
State policy index 0.125 2.099 0.037
Building level 0.326 5.36 0.000
Poverty 0.326 5.24 0.000
Principal/AP (dummy) 0.007 0.113 0.91 643
F ¼ 21.714 o0.001 0.30 0.29

2 (Constant) 29.143
District accountability index 0.135 2.26 0.025
Building level 0.309 4.94 0.000
Poverty 0.347 5.54 0.000
Table II.
Principal/AP (dummy) 0.008 0.12 0.899
Positive state policy index,
F ¼ 21.208 o0.001 0.30 0.29
district accountability
Notes: aBuilding level, elementary or secondary dummy coded; poverty, percent free and/or reduced- index and the leaders
price lunch; your title, principal (0) and assistant principal (1). See Appendix for items in state policy as instructional
index and district accountability index leaders (N ¼ 201)

Overall, these findings suggest that that the school leaders’ perceptions of both the
district and the state’s role are associated with their instructional leadership behaviors.
They do not, however, answer the question of how these two policy levels may augment or
interact with each other. We therefore decided to examine both variables simultaneously
using structural equation modeling. The initial model made the following assumptions:
first, school level and student characteristics may have a direct effect on the school leaders’
perception of district and state accountability policies as well as on the leader(s)
instructional leadership; second, leader perceptions of state policy are not independent
of district policy. Rather, they are likely to see state policies through the accountability
lens that is articulated to them by district personnel and through district policies.
An initial analysis suggested that the school’s level was a weak predictor of the
leaders’ attitudes toward both state and district policy, and that student poverty
was a weak predictor of leader attitudes toward district policy. When these insignificant
paths were eliminated, the model was robust (see Figure 2). The results suggest that:
. leader’s perceptions of a positive district accountability policy environment are a
significant predictor of their instructional behaviors (as perceived by their
teachers);
. leader’s perceptions of a positive district accountability policy environment are a
significant predictor of their positive attitudes toward state policies;
. secondary schools have significantly less instructional leadership, but school
level is not a significant predictor of attitudes toward either district or state
policy;
. schools with higher poverty levels are significantly more likely to have positive
attitudes toward state policy. Although they also have more positive attitudes
toward district policy, the path is not significant; and
. schools with higher levels of poverty are more likely to have teachers who
experience strong instructional leadership.
JEA
f
50,5

Building
level
644 –0.32
0.00 g
District 0.12
accountability 0.30

–0.27 36
0.16 Pinstfoc
Positive 0.08
state policy
index
0.17
0.32

% Free or
Figure 2. reduced
Predicting instructional lunch
leadership with state and
district policy and f
accountability (SEM)
(N ¼ 211)
Notes: RMSEA = 0.00; Fmin = 0.06; Cmin = 0.00

Overall, these findings suggest that that the districts may play a moderating role
with respect to attitudes to state policy. They also raise a tentative proposition that
will be explored in more detail as we examine our case data – namely, that unless the
district is able to build on state policy to augment the local agenda, the effects of
state policies at the school level will be minimal. The finding that leaders in higher
poverty buildings are more likely to have positive attitudes toward state policy
and are also more likely to be viewed as instructional leaders by their teachers
addresses a “holistic fallacy” that is often portrayed in the media, which assumes
that high poverty schools are also “stuck schools” (Rosenholtz, 1991) with lackluster
staff.

Responses to accountability policies: a comparison of low and high

instructional leadership
In the remainder of our paper, we deepen our understanding of the impact of external
accountability policies by looking at the relationships between school leaders’ attitudes
to and experience of state and district policies and their instructional leadership
practices. Our findings are organized as a comparison between three school leaders
who teachers rated as among the lowest on instructional leadership (LIL) and four
who were assessed as among the highest (HIL). Table III presents an overview of
this comparison organized under four themes relevant to our research questions.
 Low-instructional leaders High-instructional leaders

Leadership vision Describe their leadership as involving implementation of Describe an ambitious inclusive vision based on social
state and district policies justice or high-performance values
create a positive family atmosphere prior to tackling big Develop trusting respectful relationships which serve these
educational issues ambitious purposes
provide support so teachers can do their job are clear about the need for teacher change
Reported instructional Manage the instructional program by: Manage the instructional program by:
leadership practices organizing student placements monitoring test results
Instructional monitoring test results allocating resources to support educational goals
management ensuring teacher familiarity with state standards
Teacher professional Organize the teacher learning by: Lead or participate in the teacher learning by:
learning communicating the expectation that teachers learn developing school-based teacher learning communities
providing access to a rich repertoire of teacher learning using student learning needs to determine the focus of
opportunities teacher learning
providing considerable teacher discretion over the focus observing teaching and giving feedback based on
of teacher learning common instructional framework
Response to external Are positive about the principle of increased accountability Are positive about the principle of increased accountability
accountability policies but and
are critical of aspects of implementation, e.g. misalignment discuss how state and district policies provide leverage for
of state and district standards pursuit of their educational vision and instructional
are negative towards the impact on subgroups e.g., high leadership
poverty and special education students see district policies as aligned with school needs
largely focussed on compliance with state and district view themselves as active agents in shaping policy
requirements
Relationship with district Hold mostly positive attitudes to districts which are Hold mostly positive attitudes to districts which are seen as
office reported to exercise benevolent control through: a partner in the pursuit of educational goals through:
personal support and mentoring of a team approach
generous resourcing of principal development collaborative capacity building
increased control of instructional decisions which relieves responsive school – specific assistance
the principal of some instructional leadership responsibility freedom to pursue own vision while strongly accountable
for results
agents
as mediating
School leaders

leaders
(n ¼ 4) instructional
Thematic comparison of

low (n ¼ 3) and high

principals classified as
Table III.
645
JEA In discussing these findings, we will attend to both similarities and differences between
50,5 the two groups of cases and to the differences within each group.

Thematic analysis: espoused leadership theories[10]

There were important differences between the stories HIL and LIL told about their
leadership vision and their instructional leadership practices.
646 LIL. LIL spoke predominantly about their relationships with staff and students.
For the principal of Chillwater Elementary, creating a school which was warm and
friendly was a priority. He believed he could get things done by explaining the direction
he wanted to go and working with those staff who wanted to go with him rather than
assigning or directing staff accordingly. For example, he explained to his staff how he
agreed with the district goal for greater inclusion of special education students and
told them:
I am not going to assign people to do that. You need to come to me and tell me if you’re
interested [y] they all came in and volunteered, we gave them some in-service and we’ve
been doing it, this is our second year and it’s great. But again, it was not, to me, if I require
and I mandate and I say “you must”, it’s not there [y]. they have to want to do it, they have to
[y] understand what it’s about and be ready to go forward. That’s how I operate.
Maintaining a positive school climate was also stressed by the principal of Cedar Grove
Elementary who described himself as a “laid back leader” who strives to maintain
a family atmosphere by focussing on the positive despite the challenge of some
negative teachers and difficult parents and students.
The principal of East Starr Middle School articulated very deliberate strategies for
building staff trust and mutual respect. He believed that trust developed over time as
conflicts were resolved and teachers learned that their principal did not make decisions
to appease particular groups. He also believed that trust was a pre-requisite to tackling
the bigger educational issues and defended the delay this created in making important
changes by arguing that the end result was likely to be a more sustainable change. While
still focussed on relationships, the East Starr principal articulated a more assertive and, at
times, directive leadership style than the other three LIL principals. His goal was less
ensuring a family atmosphere than providing his teachers with an environment where
they could be productive. His job was not to “make them happy,” but to give them
direction, provide them with support and protection from criticism.
Beyond relationships, the only other theme that emerged from the LIL principals’
description of their leadership vision was the implementation of state and district
policies. The Pinewood principal, for example, described his leadership vision as “to
bring state policies back to the school.”
HIL. While relationships were also important to HIL principals, their accounts of
their leadership transcended the personal and interpersonal because relationships
were embedded in a wider vision of what they wanted to achieve in their schools. For two
of the four HIL principals, that wider purpose was informed by a deep commitment to
social justice. The principal of Maple Island emphasized that creativity in addressing
the needs of their very low-income pupils had to come first in all decisions. Similarly, the
principal of Antica Elementary school brought her respect for and experience with Native
American communities to her leadership of a school that had a reputation as the worst in
the district. She saw her role as demonstrating that her pupils would succeed:
The role of my school in this community, my own vision of it, would be that my school is
the only school that is predominantly Native American. I think the role of Antica is to show
 the community that Native American kids can and will be successful. I carry that vision School leaders
because of my previous life [living on an Indian reservation] of knowing that. I think that’s my
mission in life, to show that to this community. as mediating
Although the leadership vision pursued by the remaining two HIL principals was less
agents
clearly based on a personal vision for social justice, they placed student achievement
and well-being at the center. At Overton Elementary, the principal saw her leadership
role as growing teacher leaders and supporting teacher teams that were focussed on 647
instructional issues; at Walker, relationships of trust between parents and teachers
were essential to building a more inclusive school in which parents and teachers could
work together to help students break out of the cycle of poverty that many families
experienced.
In short, the vision articulated by HIL principals was educationally richer and more
ambitious than that of their LIL counterparts. They articulated educational purposes
that were broader than relationships, broader than accountability imperatives and yet
inclusive of both. While relationships were critical to achieving these broader
educational and social justice purposes, they were not treated as ends in themselves.
Respectful relationships with teachers were pursued through the work of developing
strong teacher leadership and professional learning communities. Trust was a social
resource for achieving shared educational and community vision (Bryk and Schneider,
2002). In the next section, we see how these differences in leadership vision translate
into reported instructional leadership practice.

Thematic analysis: reported instructional leadership

Leaders descriptions of their instructional leadership exemplified the wide range
of practices that are embraced by this concept (Hallinger, 2005). We have made
a distinction in Table III between instructional leadership that involves instructional
management and that which involves leadership of, or participation in, teacher
learning. While both principal groups reported very similar engagement in the former,
the HIL principals were far more closely involved in their teachers’ professional
learning. This is a particularly interesting difference given the research showing
that the effects of this type of leadership on student outcomes are, on average, stronger
than the effects of other types of leadership practice (Robinson et al., 2008).
LIL principals. The three LIL leaders managed instruction by organizing student
placements, monitoring test results and ensuring teacher familiarity with, and adherence
to, state standards. The Chillwater principal, for example, gave a detailed account of how
he made student placement decisions after seeking parental opinion and using test results
to determine the best match between teacher, student and grade level.
Like his counterpart at Chillwater, the Pinewood Elementary principal acted as
instructional leader through his organization of the instructional program and focus
on student placements. He believed the job of teachers was to “deliver the curriculum,”
which he monitored by classroom observations and close review of test results.
He ensured that there was both vertical and horizontal alignment of the curriculum to
state standards and that teachers at any given grade level understood what students
who were entering and leaving their grade should have achieved. Official test results
were used for placement purposes by identifying students at the boundaries of
achievement standards and ensuring they were placed with teachers who could
provide the required assistance.
JEA When it came to their leadership of teacher professional learning, the LIL principals
50,5 relied heavily on communicating their expectations that all teachers should learn.
The Chillwater principal described it this way:
My attitude toward the staff is that I expect everybody to grow every year, including myself.
I said, “I want to see, I want you to identify what areas you want to grow in and what
things you’re interested in. I want to see that growth every year. So keep moving forward.”
648 Some grow a little bit, some grow immense amounts. But again, it’s, that’s a challenge to
them all the time to see where we’re growing.
The principal did not, however, set a common direction for teacher “growth.” Teachers
exercised considerable discretion over how they set their own learning agenda.
There was no explicit connection made by the principal between student and teacher
learning needs. Topic selection and teacher choice seemed to be the paramount
considerations. The same was true at Pinewood. The teachers’ professional learning
agenda was mainly set by the teachers themselves, because the district required grade
level teams to do so. When asked how he would know if teachers’ professional learning
had influenced their classroom practice, the Pinewood principal replied that he would
only really know if they told him.
The East Starr principal also described how he communicated his expectations
for teacher change. He saw his main challenge as ensuring teacher familiarity with and
responsibility for achieving the state standards. As he saw it, this change required
convincing teachers that it was their responsibility not just to teach the material but to
ensure that it was learned. His strategies included talking to teachers about their test
results and expecting them to produce ideas for improvement:
A bunch of [kids] didn’t do well on the social studies assessments so the social studies teacher
and I had a little conversation. “These are the results. What does this mean to you?” The poor
lady doesn’t know, she is like in tears. “I’m not going to fire you, but we need to look at this.
We need to take a serious look and say Okay, our reading kids didn’t do well [y] This social
studies is a reading test. How do we bridge the gap?” [y] I said “Don’t even start with
‘I taught the material. Tell me what we are going to do differently.’” Solve the problem. It’s not
a personal attack.
In addition to communicating their expectation of change, the LIL principals supported
teacher change by monitoring what was happening in classrooms through brief visits
and informal feedback, reviewing plan books and resourcing a range of teacher
professional learning opportunities, including bringing in expert speakers. None of
the LIL principals made explicit the links between the topics chosen and the learning
needs of particular groups of students and their teachers.
HIL. Like the LIL principals, the HIL principals played an active role in the
management of the instructional program, and stated that they spent a lot of time
on it. HIL principals, however, combined detailed knowledge of student learning needs
and a vision of how to meet them. As a consequence, they played a stronger role in
steering and coordinating improvement activities and teacher choice played a far less
important role in determining the professional development agenda.
The extensive classroom teaching and literacy coaching experience of the Antica
and Walker elementary provided the platform from which they evaluated teachers,
engaged in data discussions and led professional learning. We will highlight Walker,
because there was principal turnover during the study, but no gap in the emphasis
on instructional leadership because of the sustained focus on the role of principal as
coordinating teachers’ instructional learning.
 The first principal, interviewed in 2006, described how district-wide consultation School leaders
had led to a common literacy curriculum supported by literacy coaches. He pointed as mediating
out that this research-based, consensus-driven process allowed him to provide
direction to the teachers in his school as they worked through their own plans. One of agents
the strategies that he pointed to was filming in a teacher’s class and then giving the
film to the teacher to review in private before discussing it together. While this process
made teaching visible to the two persons involved in the review, the principal was now 649
searching for ways to have more teachers involved in peer observation and coaching.
The new principal, whose background was in literacy, continued the momentum by
focussing on developing common instructional practice. She described precisely what
she was looking for when she visited classrooms:
I’m looking for some commonality in the way we practice here [y] Are they giving kids time
for guided practice before they ask them to do independent work and that kind of thing? I’m
looking for differentiation all the time. I’m also looking for classroom management. I don’t
think learning happens if there is not good classroom management. So those are some of
the things that I look for. Then what happens is I get to know the kids because I am in the
classrooms all the time. The teachers aren’t so uncomfortable with somebody coming in for
more formal evaluation because I have been in on a daily basis.
Bi-weekly staff meetings were reconfigured as professional learning opportunities,
internal expertise was identified and data teams met regularly to identify and support
students with particular learning needs. The close involvement of this principal in the
data teams meant she had a detailed knowledge of what teachers were up against and
how she could best support them in achieving change:
Data teams [y]. I sit in on them and I think it’s really important that I’m there because then
I hear what the concerns are and I can help address them [y] Our literature coach is the
facilitator. What happens as teachers look at the data base, is they say “There is a group of
kids really struggling with this. What should we do next?” So there is our literacy coach who
has ideas and resources and I actually sometimes chime in. They have said they want me to
[y] because that was my role in my previous life.
In contrast to Walker and Antica, the instructional leadership of the remaining two HIL
principals was somewhat more indirect, involving careful organization and allocation
of resources to support the achievement of school-wide educational objectives. In the
case of Maple Island, the whole school was involved in the development of an
International Baccalaureate curriculum that would prepare students of poverty for a
globalized, high-technology world (and that requires external certification). The
principal emphasized that it was important that any of her teachers be as capable as
she was of explaining the way in which the curriculum would operate. In Overton, the
principal also focussed on the creation of a vibrant, inclusive learning community, and
noted that it was important to keep the vision very simple, and to allow many people to
be teachers-of-teachers:
Our vision is to be an excellent learning community, period [y]. I think that’s important [for
me to be up on the research] and I think I need to give those strategies and things to teachers,
but who do you learn best from? Your colleagues. So my role in instructional leadership is
providing time and providing opportunities for teachers to [y] learn from each other. Job
embed it within the regular day, providing them professional development and then under
that umbrella meaning providing time for them to study groups. Providing opportunities
for them to collaborate. Bringing in a consultant, that is not me, that is an unbiased person
that can sit down in a neutral setting and walk them through what good writing looks like
JEA and anchor papers and assessments. Have them start learning to coach each other. That’s
my role.
50,5
In both cases the work was intimately connected to student learning needs and the
principals positioned themselves as people who learned from their students and with
their teachers about how to bring the vision to a reality.
In summary, both LIL and HIL principals saw themselves as instructional leaders,
650 but they articulated different theories of change. The LIL group acted as if they
believed they could improve teaching and learning by laying a foundation of positive
or trusting relationships, communicating high expectations for improved teaching and
teacher learning and giving teachers access to a rich repertoire of professional learning
opportunities. Teachers were then given discretion over which opportunities they
engaged in – a leadership strategy that precluded a closely coordinated collective
improvement effort directed to specific student learning needs. This strategic
orientation was largely disconnected from state and district policies and frameworks
for school accountability, and few connections were drawn between what they could
do to make their school a better place and external expectations. In contrast, the HIL
principals were closer to the action – “hip deep in curriculum and instruction”
(Hallinger, 2007). They either led the teacher learning themselves or accessed experts
who could support a school-wide, tightly coordinated approach to instructional
improvement that was guided by their vision, which was informed by the expectations
of the district and state. If they themselves lacked the requisite knowledge of how to
achieve their vision, they led as lead learners, researching good practice and joining
with their staff in the sustained learning needed to achieve it. They brought about
improvement by participating in the learning with their teachers rather than by telling
teachers to learn or solve problems. It is this stance that best distinguishes the
instructional leadership of the LIL and HIL principals.

Thematic analysis: leaders responses to external accountability policies

LIL. It would be wrong to characterize the three LIL attitudes to external
accountability policies as wholly positive or negative, for their attitudes varied
depending on whether they were describing their attitudes to the principle of increased
external accountability (about which they were positive), their views about how
the policies had been implemented (about which they were much more critical), or their
attitudes to particular targets for their schools (about which they were mostly
negative).
The Chillwater principal saw the increased accountability as mostly having a
positive impact because it gave direction. He explained, “I have always felt, ‘just give
me the plan’. Give me the parameters and let’s get on with it and we’ll do the best job
we can [y].” He was, however, scathing in his criticism of the (state-determined)
targets that had led his Title 1 building to fail AYP in the third year of our study – a result
which the principal attributed to the inclusion of free lunch students as a categorical
group and unrealistic targets for the special educational students in his school.
We were one of the few Title 1 schools in the state that are still meeting AYP. When [the state]
changed [the comparison groups], we won’t meet it again. Okay? They changed it last year to
include all kids that were on free lunch. Well, that was 70% of the school. Okay? So, all of
a sudden you had all these kids lumped as one big lump. And we have increased our special
ed. numbers. [y] The parents want their kids here, so the numbers increase. But those kids
aren’t going to be on level so we’re not going to make it. We know it [y] Parents accept it [y]
 What idiot created that plan? [y] I tell parents in that manner. I say, “You know, realistically School leaders
[y] Every school is going to be in need of improvement by 2014 if nothing changes”.
as mediating
While implementation problems also contributed to this principal’s critical attitudes, agents
the most important source of his negativity was the fundamental incompatibility he
perceived between his own and the state’s approach to assessing student progress. For
this principal, success is judged by demonstrating that students are making progress,
not by reaching proficiency targets. His response to the AYP challenge was to change 651
the assessment schedule and the tests used, so that teachers could provide more
concrete and more regular representations of student “growth,” but he was unwilling to
accept predetermined benchmarks of adequate progress. While the external policy had
pushed the principal to be more specific and transparent about the “growth” in his
students’ learning, his demonizing of the target setting aspects of the policy, made it
difficult for him to see how the pursuit of “growth” could be integrated with those
external policy requirements. In addition, reconciliation of the two agendas was
inhibited by high turnover at the district office, which disrupted relationships and led
to inconsistent district reactions to state policies.
When asked how external policies affected his leadership, the principal of Pinewood
Elementary replied that “It’s probably close to a hundred percent of every decision
or every program that’s brought into the school.” State standards, tests and
programs shape practically everything the school does, and there is little evidence of
an independent instructional or educational agenda through which those
requirements are mediated. The transmission of the state agenda does not imply
complete agreement with it. The principal believes both the state standards and the
targets for special education students are unrealistic, and this self-described “laid
back” principal fears the consequences of NCLB for both his school and his own
career:
My biggest fear would be when you really look at the sanctions-punishments-consequences,
whatever you want to call them, [that] you are never going to get to the point where 100% of
your kids are proficient [y] When you take a look at your special ed. population, there is a
reason they are in there. They have a learning disability [y] My philosophy has always been
“Let’s hold on as long as we can.” You don’t want to be the first who ends up in school
choice or any other type of sanction. We’ve done very well and it’s great being on the top, but
[y] But in the back of my mind, I’m very concerned, because obviously I have a lot of
years to go.
In short the response of the Pinewood principal to external accountability requirements
is wholesale adoption. While he fears that the challenge is too tough and somewhat
unfair, he brings no alternative to the table.
The principal of East Starr Middle School gave a considerably more positive
account of state policies than the other two LIL leaders. At the time of the interviews,
he was not experiencing AYP pressure and believed his school would exceed the state
targets. He saw his job as ensuring that his teachers teach to the standards:
[Our] professional responsibility is to make sure those lessons are aligned to the standards. So
teaching to the test here is not a bad concept because it is standards-based. Years ago, if you
said “teaching to the test”, it was taboo [y] But in this case, you should be teaching to the test
because those tests are a reflection of the standards that you should be teaching all year long
[y] They take the guesswork out [y] It’s very clear. We have access to the web site. We have
resources now with canned lessons in how to teach to the standard. So in essence, we’ve made
it very easy for teachers [y].
JEA This principal shares the theory of improvement that is at the heart of NCLB itself. His
50,5 theory of change is that the combination of clear expectations, aligned resources and
accountability is sufficient. His role is to implement what the state wants and ensure
his teachers do the same. It would be a mistake to judge this as an unprincipled stance,
for he genuinely believes in what his district and state are doing.
HIL principals. The HIL had largely positive views about state policies. They were,
652 in particular, positive about their increased accountability for the learning of poor and
minority students even when they had reservations about some aspects of the
implementation. The first Walker Elementary principal strongly approved of state
standards and the “transformative” focus on high need student groups. Unlike the LIL
principals, the generally positive attitudes of the HIL principals were not explained by
the achievement levels of their students. The Antica principal (located in the same state
as the Chillwater principal) had taken over a school which had been in restructuring
“since the dinosaurs,” and for which meeting AYP targets was a considerable
challenge. She mediated external accountability requirements, not through
communicating her personal views, nor by requiring and expecting change, but by
focussing on developing supportive and capacity building internal accountability
processes (Elmore, 2004). This is how she described her own leadership response to the
external accountabilities:

I have concentrated my efforts on data, on having realistic, real time data to give back
to the staff to influence how they’re doing their instruction [y] We do the MAP testing
four times a year, so we have updated data all the time [y] [teachers] have access to that
[information] anytime they want it. And that way we can see each individual
student, whether they’re progressing or not. And then we can get the help for those
who are not [y] And for the accountability part of it, I think one of the things I’ve learned is
that you have to look at student data in groups first, you’ve got to look at groups. So having
the tiers made a good group to have. And I know how many Tier 2 readers there are,
and how many Tier 3 readers there are, and how many Tier 1. And that can show us our
progress.

The principal reported proudly that her emphasis on the two mantras of data and
student success allowed her to out-perform schools on state tests that were assumed to
be better than her own. She understood the external requirements as supportive of
her own vision of demonstrating that Native American students could be as, or more
successful, than other students in the district.
In sum, the voices of these seven enrich the survey finding that principal attitudes
toward state policy were associated with their teachers’ assessments of their
instructional leadership behaviors. Overall, it is reasonable to say that the HIL
principals had a “story” of external accountability as allowing them to be internally
coherent, while the LIL understood external demands, but had not incorporated
them in a coherent way into their own agenda. The HIL principals were remarkably
similar in the degree to which they interpreted both state and federal external
accountability as a lever for gaining an increased focus – on the part of the public
and their teachers – on improving the performance of students in their schools. They
viewed it as a tool to create a “no excuses” culture that focussed on learning and did
not blame external accountability for the increasing likelihood that their schools would
face some consequences.
Two of the LIL principals had incorporated state standards as the focus of their job,
but neither of them connected state standards with their own role as instructional
leaders, presenting a classic case of limited efforts at active sensemaking may lead to School leaders
incoherence. The third LIL principal had the most ambivalent perspective on state as mediating
and federal requirements, focussing mainly on how targets would cast his school in a
negative light. This principal’s impassioned defense of “growth” as a measure of agents
success, combined with his personal criticism of policy makers, meant he was very
unlikely to find a principled way through the impasse. He had made sense of external
demands, and had rejected them. 653
Thematic analysis: leaders relationships with their districts
According to Honig and Hatch (2004) the quality of relationships between policy
makers and implementing agents affects the extent to which coherence is achieved
between the goals and agendas of each group. Relationships that are stable and involve
rich opportunities for interaction are more likely to foster the understanding and
mutual adjustment required to craft coherence, than those which are less so. In this
fourth theme, we compare the relationships LIL and HIL principals report having
with the district offices, as a way of explaining their enactment of accountability
policies through their instructional leadership.
LIL. Among the LIL principals, the Chillwater Elementary principal was the only
one who expressed a predominantly negative attitude about the district’s role. While
most of his concerns were directed at the state, he was also critical of district
leadership, which, he believed, had let the teacher unions get too strong, provided
weak coordination for some improvement efforts, introduced initiatives that were
not well linked to student learning (a “wellness” initiative) and placed increasing
restrictions on school-centered development. In other words, he saw the district as
an actor that made state policies (with which he disagreed) worse, largely because they
responded to the state’s agenda by trying to standardize education in a very diverse
district. By the third interview, he was concerned about increasing constraints
on schools’ choice of curriculum materials and about how union restrictions and
a shortage of substitute teachers restricted his teachers’ access to the district’s
professional development opportunities. Despite these criticisms, he was grateful for
the district’s provision of an instructional coach whom he described as both supporting
teachers and as relieving him of some of his responsibilities as an instructional
leader.
At Pinewood Elementary, the district took a strong central role in determining
curriculum and instruction. It set standards for classroom observation of teachers,
and requirements for principal professional development that were higher than those
of the state. In general, the principal felt supported by the district’s central control.
He was pleased about a district-wide math program which had raised test scores
considerably, and wanted teachers to keep teaching to the tests (as recommended
by the district) even though they reported that students were tired of them. In this
regard, he saw the district’s steady policies as buffering his school from state policies
that he perceived as often incomprehensible. Even onerous district requirements were
interpreted as protecting the principals. For example, district revisions to a form
for reporting on required classroom observations were sufficiently consultative to
please him:
So again the superintendent’s office, the curriculum department really was working with a
group of teachers and leaders to come up with a new form that would make it easier for you to
observe forty teachers but really pinpoint some areas that we wanted to work on. And it even
gave you a spot where you could write comments.
JEA He reserved his greatest enthusiasm for the support that he received from his district
50,5 supervisor, who appeared to be highly available for mentoring and consultation.
However, the support that he received, while valued, rarely focussed on instructional
leadership or meeting district or state standards.
In East Starr, the third LIL school, the principal was constantly in the district
office, which was located in his building. He spoke of the district in largely positive
654 terms, citing generous conference provision and access to resources.
This district is great and the resources are great. If I say I want to go to a conference,
obviously I don’t have carte blanch to everything but if it’s within reason they will send me to
any reasonable conference. I’m assigned a mentor [y] The superintendent pretty much has
an open door policy.
However, despite the proximity, the district staff seldom visited the school and its
classrooms, although the principal noted that he would have welcomed it if they had
taken responsibility for communicating accountability messages to his teachers. The
few leadership development opportunities the district had organized, such as
a seminar on “Good to Great” did not appear to be closely aligned to teacher and
student learning needs, and he did not see any connection to the state’s accountability
focus. The principal explained that he and the district staff did not get together to talk
about leadership, but rather the “minutiae of the job.” He saw the district as risk
averse, and failing to provide the support of principal initiative and risk taking
that was needed to counter the cautiousness engendered by a high stakes
accountability environment. As he saw it, “Because the accountability has gone up,
we’ve all retracted into our little safe zones. Everybody wants that safe harbor. We are
never going to go anywhere. You can’t steal second base without taking your foot off
of first.”
For the three LIL principals, relationships with the district office are best
characterized as that of benevolent control, with districts providing personal support
while at the same time directing curriculum and instructional choices. There was
surprisingly little negative reaction to this control, possibly because these LIL did not
bring a strong personal educational agenda, to the table. After all, wholesale adoption
of state and district policies was the leadership vision articulated by two of these four
principals.
HIL principals. The HIL viewed their districts as partners rather than supervisors in
the process of school improvement. The partnership was evident in the formulation of
district policies and initiatives, and in district support for school-initiated strategies
for achieving broad district goals for improved achievement. When district policies
and programs are formulated in consultation with building leaders, the boundary
between internal and external agendas blurs as leaders and district staff learn together
about how to address context-specific needs and priorities while still satisfying policy
constraints. The first Walker principal described such a partnership in which 30
elementary school principals collaborated with district staff to hash out a framework
for literacy instruction and assessment, while the second indicated that most of
the school’s reform work was focussed around district priorities that were
collaboratively set. Forging a district-wide collaborative approach was also fostered
by a staffing policy which enabled building leaders to move in and out of central office
positions. At Walker, the first principal left the school to go to a district-level
coordinating job; his replacement came to the school from a district level coordinator’s
position. The movement back and forth solidified personal and professional
relationships in ways that translated into a sense of being part of a collaborative School leaders
district-wide improvement effort. as mediating
The development of principals’ instructional leadership was supported by regular
meetings to discuss professional readings and district polices as well as by more agents
personal mentoring. The district also organized school clusters which engaged in
structured and facilitated learning walks at each member’s school.
While the Walker example illustrated partnership through the co-construction of 655
district policies, a partnership rather than supervisory approach was also
evident in district support for school-based initiatives that while aligned with
broad district imperatives had their origins in the educational vision and passions
of building leaders. The Maple Island principal was explicit about how her
district had supported her efforts to develop an International Baccalaureate
curriculum:

What is so wonderful is I empower my teachers and at the district level, they empower me.
They valued not only just my opinion but the amount of research and work we had done
as a school. It wasn’t just that I showed up and said “How about this magnet school thing?
Why don’t we do it?” They respected that this is an initiative that I’ve worked on for almost
five years [y]. They have supported my vision all along, whether it was [the former
Superintendent] and now [the current superintendent] [y]. Our [district] is so forward
thinking.

What she emphasized on a number of occasions was the degree to which the
district office staff, from superintendent on down, understood her school and its needs.
They did so because they were available and spent time in the school: “You call them
and they are the most accessible people. Seeing them walk in nobody (gasps) the
district office is here. No one does that. It’s just a natural thing to see them visit your
school.” Rather than being a source of standardized assistance and prescriptive
interpretations of state policy, district staff provided assistance that was specific to the
school.
The principals at Overton and Antica voiced similar perspectives on the role that
their district’s played in developing policies and programs that would support
improvement in their schools. In Overton, where there were only six elementary
schools, the principal emphasized the team work involved in a district-wide approach
to improvement. She would not take a job closer to her home because of the value
that she placed on her own learning within the team, and the commitment she felt to
their collective decisions:
[y] there are not many principals that can say that, at least once or twice a month, they sit
down with their superintendent and make decisions for the district. For the most part, those
of us that are elementary are usually low man on the totem pole and we usually have to get
our information from the superintendent rather than make the decisions with him and that’s
what I love about him [y] Some of it is what I call administrivia [y] Other times, if
questions are brought up we say, “OK, we’ve got this facing us. How do we want to go about
it? Let’s brainstorm it.”

The Antica principal, from a slightly larger district, did not have the same intense
collaborative experience with the district, but emphasized that they worked as a
team. She, like the other HIL principals, described herself as accountable for results
rather than for the means for achieving them. This results focus gave her a sense of
JEA autonomy within a tight yet supportive external accountability environment. As she
50,5 explained:
I hate to say that they’ve given me carte blanche, but I feel a great deal of support from my
bosses. And if I can justify whatever it is that we’re doing, I have been allowed to do that. So I
don’t feel a lot of constraints [y]. But they are interested in Antica getting some academic results
and have pretty much given me free reign to get those results in a way that I think we need to.
656 In summary, one of the conditions for crafting coherence between external and internal
improvement and accountability agendas is opportunities for regular and respectful
interaction between those responsible for each agenda (Honig and Hatch, 2004). This
implies a relationship between and district officials where the latter are responsive and
consultative rather than hierarchical and authoritative. Both LIL and HIL principals,
with the exception of Chillwater, had forged what they considered to be “good”
relationships with those in district office, which they saw as providing support and
resources which would help them to meet external policy imperatives. The nature of
the relationship was, however, distinctly different.
The relationship described by HIL principals was largely responsive and
consultative. In three of the four HIL settings, the district was seen as a strong
partner who gave them the opportunity to pursue educational agendas about which
they were passionate. In other words, the HIL principals tended to see district policy as
a work-in-progress, to which they could make contributions.
Although the relationships described by the LIL principal group were more
authoritative, they did not necessarily engender a negative reaction. In the three LIL
schools, responses to questions about the district’s role suggested a sense of subservience,
even when the relationship was deemed positive. The mental image of the principals was
that the districts were in charge – for better or worse – and that the schools were obliged
to implement policy choices on which they had limited input. Given that two of the
schools were in small- or medium-sized districts, it is notable that their sense of being
relatively powerless stands in contrast to the explicit stories of the HIL principals, which
emphasized that they were at the table when decisions were made and were part of a
network of peers and district staff whose job was to make sense of how state and district
priorities could serve and be served by the needs and agendas of building leaders.
Summary and discussion
The evidence from our analysis of the survey data suggests that most principals in this
study had positive perceptions of state policies, a finding that is corroborated from
our qualitative data, where five of the seven had largely positive comments, while a
sixth had mixed perceptions that included both positive and some negative
comments. Given the public controversies about NCLB, school leaders’ predominantly
positive perceptions of state policy were something of a surprise. Overall, many
journalistic reports about NCLB’s effects on schools have tended to emphasize the
problems that have been caused by particular stipulations of the law, such as tutoring
requirements, or inflexibilities related to testing. At least one characterizes NCLB as
heading toward “predictable failure” (Mintrop and Sunderman, 2009). This study suggests
a need to reframe the discussion of accountability, at least in the context of the HIL.

Accountability and instructional leadership

The first two questions that we posed for our analysis were: Do school leaders’
perceptions of state or district school improvement policies and procedures influence
how they lead their schools? Are those who perceive their accountability context (state School leaders
or district), as supportive more likely to behave as instructional leaders? Our survey as mediating
data suggest that the attitudes held by school leaders toward accountability policies
may be particularly important as a predictor of their behaviors. In particular, agents
principals and assistant principals who have positive attitudes about state and district
policies are more likely to be judged as stronger instructional leaders by their teachers.
Our findings suggest, furthermore, that positive perceptions of state policy are 657
important, but that perceptions of support from the district moderate this positive
relationship. These findings are important because they come from nine randomly
sampled states operating different accountability policies under the broad NCLB
federal legislation. The analysis of interviews with principals in seven schools
confirmed these overall positive attitudes and also shed light on the underlying issues
for those who were less positive. The interviews suggest that more negative views
about accountability policies may have resulted from more ambitious categorical
targets, especially those set for special education students.
The cases revealed subtleties in leaders’ attitudes to policy that are easily
overlooked, but that affect the degree to which they engage in active efforts to direct
improvement in classrooms. It is important to note that there was no overall
relationship between the attitudes expressed by leaders and the level of sanctions, if
any, being applied to the school. The principal of Antica Elementary (HIL), which had
been in corrective action “since the dinosaurs walked,” was positive because it helped
her pursue her strong equity agenda. The principal of Pinewood Elementary (LIL),
which had never failed to achieve its testing targets, was, on the other hand, upset
about the use of targets. The cases tell us more about how principal sensemaking
and crafting coherence – the frameworks that were set out at the beginning of the
paper – help to explain the quantitative findings. They also parallel an earlier
examination of how superintendents’ interpreted accountability requirements, which
also ranged from a focus on implementing state expectations to one of exceeding and
re-defining accountability goals (Louis et al., 2010b).

Principal’s sensemaking
The third question that we posed was: How do leaders integrate their own leadership
beliefs and agendas with the external mandates to which they are subject? To what
extent do they perceive conflict between the two? In our framework and analysis,
we drew on a sensemaking perspective to address this topic. Sensemaking is the result
of an interaction between policy content, the relevant prior understanding and
experience of implementing agents and the context in which they are working (Spillane
et al., 2002b). Our seven cases have focussed on both the state and the district as
relevant policy contexts and have shown the importance of taking each of these into
account. With one exception, principals reported supportive relationships with their
district offices and these relationships enabled them to either adopt the state’s external
improvement agenda or to integrate it with their pre-existing educational vision. The
principals, again with one exception, viewed their relevant policy context as the state
and the district, and made only limited references to NCLB.
School leaders differed considerably, however, in the educational and instructional
knowledge that they brought to the enactment of the external agenda. The HIL
principals brought more such knowledge to this task than did their LIL counterparts,
and this difference substantially shaped how they led the work. The LIL principals
led by building inclusive relationships, managing placements, monitoring data,
JEA making quick visits to classrooms and urging their teaches to “grow,” “show progress”
50,5 and “solve” their teaching problems. In addition to the above relational and indirect
instructional leadership roles, the HIL principal led the improvement of instruction
by participating as either leader of or learner in teacher professional development
that was aligned to student learning needs. While the LIL principals managed the
structures and processes around instruction, such as placement, reporting and record
658 keeping, the HIL principals were more directly involved, in addition, with the core
business of teaching and teacher learning (Elmore, 2004).
In only two cases were these differences mitigated by district offices because of
professional development experiences sponsored by the district. In these cases (Walker
and Overton, both HIL), learning related to instructional leadership was an informal
byproduct of collaborative administrative work that allowed school leaders to
participate in developing instructional policies. For the remaining four HIL and LIL
principals who had a positive view of the district, their appreciation was framed in
terms of the district’s providing of a supportive and consultative environment, which
included the exchange of expertise, rather than because of a specific focus on the
instructional learning needs of their principals.

Crafting coherence
Our final question was: To what extent do leaders’ relationships with the district office
enable them to craft coherence between the external accountability policies and their
own agendas? In our introductory discussion, we described three conditions that are
required for crafting coherence between the school improvement agendas of school
leaders and policy makers (leader; Honig and Hatch, 2004): Specific yet flexible
improvement goals, appropriate “bridging” and “buffering” to link schools to external
actors, and a partnership orientation. These three conditions were very apparent in our
HIL cases. All four principals saw district policies as enabling the achievement of
educational goals to which they were already strongly committed. While there were
occasional tensions around timing or implementation, the policies provided an enabling
structure that sharpened goal focus and brought additional resources and expertise. The
personal relationships they developed with the district officials made the power they
wielded less threatening, and provided expertise and feedback. Sustained interaction
about how district mandates could serve school needs and their educational vision built
up a level of trust between them and district officials – a basis of trust from which
together, they could figure out how to make the policy work for their school.
The evidence from two of our three LIL principals (East Starr and Pinewood)
provides a more complex story. These two described a leadership vision that was
predominantly about relationships between staff, students and community. Instructional
and educational goals were given far less emphasis, and, in one case (Pinewood), were
not articulated. Coherence was achieved, not by integrating two agendas, but simply by
adopting that of the state and district office. The vision of the Pinewood principal was “to
bring state policies back to the school.” The East Starr principal described his role as
ensuring his teachers taught to the standards, which in conjunction with the “canned
lessons” would make it “very easy for teachers.” For these two, coherence was achieved
by policy adoption rather than integration. They were grateful for the greater clarity the
policy and associated standards provided and their positive relationships with the
district office meant they were happy to be directed.
One could argue that this wholesale policy adoption was desirable because it
strengthened principals’ instructional focus within these two schools. Close
examination of these two cases, however, suggests that this new focus is unlikely to School leaders
deliver improved teaching and learning because the district offices are not developing as mediating
the instructional leadership capability of these two. In East Starr there is little focus
on leadership development and what there is, is not instructionally focussed. At agents
Pinewood, the principal was appreciative of district professional development
opportunities, but once again they were not tailored to his instructional leadership
needs. The coherence that has been achieved may be at the level of words rather than 659
actions, because, despite their genuine support of the policies, these two may lack the
instructional knowledge they need to make the envisaged educational changes in
their schools. Coherence in the end, lies not in the integration of words but in the
integration of policy intent and school practice (Honig and Hatch, 2004). An obvious
limitation of this study is that we do not have data on these practices.

Implications for policy and practice

These findings have important implications for principal preparation and support.
They have confirmed the importance of developing school leaders with a clear
educational vision that is grounded in understanding of their teachers and families,
and the knowledge and skills to partner with their own teachers and their district office
in its pursuit. They develop the capacity to bridge the external accountability
demands and specific localized needs. Principals who lack a rich educational vision are
overly reliant on relationships and externally oriented compliance to motivate teacher
change. Principals’ ability to think strategically about how external accountabilities
provide important opportunities to advance their educational agendas is also critical to
their ability to craft coherence between increasingly tight accountability requirements
and their own agendas. Oppositional thinking, that sees the conflict rather than
the common ground between school-based improvement strategies and those of the
state and district, will inhibit formulating a goal that is inclusive of both.
While we lacked systematic data on the background training and experience of the
seven cases, there was some evidence that their curriculum and pedagogical
knowledge was different. HIL principals were confident providers of advice, leaders
of data teams, and partners in professional learning. Unlike the LIL principals, they
went beyond the monitoring of results and appeals for improvement, to collaborative
planning and implementation of reform strategies. To put it in its simplest terms, they
seemed to know more about teaching their students. While these findings are
suggestive of the importance of deep knowledge of teaching and learning for
instructional leadership, it should be remembered that the instructional leadership
of HIL principals also included a considerable emphasis on instructional organization.
It was their leadership of teacher professional learning, however, which distinguished
the instructional leadership of the HIL group.
There are obvious implications of this study for district offices. While all but one
principal described supportive and cordial relationships with their district office, the
focus of those relationships was very different. Strong instructional leadership from
principals seems to be related to a strong and collaborative instructional focus from
district offices. Collaboration is important because it seems to ensure that experience
support as appropriate to their school. Focus, rather than a menu of opportunities is
important, so sufficient expertise can be developed within the district to achieve
improvement goals. Of the four HIL principals, only two had participated in formal
professional development opportunities provided through their districts. None of the
LIL principals reported district-sponsored professional development. This suggests
JEA that, while the emphasis on instructional leadership was already prominent at the time
50,5 when this study was conducted, there is likely to be considerable room for districts to
develop more collaborative opportunities for principals to learn to link educational
visions to practical behaviors that might help them to enact them within their schools.
Other survey analyses from the survey data base used here suggest that district-wide
principal networks, such as those reported by the Walker and Overton are particularly
660 critical to developing stronger school leadership (Lee et al., 2012).

Concluding remarks
At the beginning of this paper we posited that “school leaders’ responses to NCLB
are likely to reflect a complex interaction between their perception of the accountability
policies, the state and district contexts in which those policies are situated and their
own leadership beliefs and practices.” Other analyses, from the larger study that
provided the data for this paper, also established the empirical connections,
demonstrating that district practices are important in determining school leaders’
sense of efficacy and their behavior (Leithwood and Louis, 2011) and that the way in
which superintendents interpret state policy affects the way in which they design
responses and interpret it to the schools (Louis et al., 2010b). This in-depth analysis
adds to the previous work in several ways.
First, it suggests that external accountability policy can have a positive impact
on the behaviors of school leaders where principals see those policies as aligned with
their own values and preferences, and where they see their district leaders as
supportive of school-driven accountability initiatives. In these cases, principals
internalize the external accountability policies and shape them to the particular needs
that they see in their staff and among their students. Where one or the other of these
factors is weak or missing, external accountability policies will not develop the
instructional leadership that is needed to bridge state and district policy intentions
with improved school performance.
Second, while the survey data suggest that both state and district policies
are positively associated with instructional leadership, the case study analysis
indicates that the principals who are assessed as effective instructional leaders by their
teachers have a nuanced and well-articulated perspective about their district’s policies,
but see the state largely through the lens provided by the district. Their relationship
with their districts suggests that they value the district’s bridging role because
district policies and practices support the alignment of accountability demands
and the individual school’s student development aspirations. They also value the
buffering role and trust that the district will support their efforts to provide
the right kind of educational experiences for their students, even when they may be a
little risky.
Third, our analysis shows that effective instructional leaders internalize the
external accountability policies articulated by both their state and district, and shape
them to the particular priorities in their own school. Where cannot create a personal
story about how external accountability demands support their aspirations for their
school or where they see their districts as unsupportive, on the other hand,
demonstrate more negative attitudes to external accountability and are assessed
by their teachers as having weaker instructional leadership.
We cannot, of course, make causal attributions based on this research, which covers
only a few years in any of the participating school leaders’ careers. However, the use of
in-depth interview data, collected over time, enriches our understanding of how they
use their larger narrative of the purposes of state and district accountability initiatives School leaders
to reinforce the leadership strategies and values that they have developed. School as mediating
leaders who appear to have subscribed to a “big narrative” of state and district
accountability – that the purpose of leadership in schools is to create settings in which agents
students from diverse backgrounds can succeed – work diligently to shape the
environments in which this will be achieved, even when the “small narrative” of test
scores suggest, in the short run, that the school is still imperfect. 661
Notes
1. The nine states were: New York, New Jersey, North Carolina, Texas, Missouri, Indiana,
Nebraska, New Mexico and Oregon.
2. The term approximate is used because, over the course of a five-year study some schools
dropped out entirely, some had only partial data, and some were added. In a few small
districts, middle and secondary schools were combined, while others used a K-8
configuration, with no middle school. The actual N of schools varies across components of
the very large data base.
3. In both administrations, a few schools failed to return surveys; whole school non-response
was greater in 2008. Other response rates were limited because teachers were absent or
(infrequently) chose to turn in unusable questionnaires. In 2008, we mailed the surveys to 177
schools with a total teacher population of 7,075. Teachers returned 3,900 surveys from
134 schools. A total of 211 principals responded, including 116 principals and 90 assistant
principals. When teacher and principal responses were merged, the total number of schools
in the 2008 sample was 146. We cannot calculate within-school non-response, but in general
it is clear that the lower response rate was due to the absence of entire schools rather than
spotty response within schools.
4. Interviews and observations with teachers, and interviews with other members of the school
community, also occurred but are not part of this analysis because they were not asked about
the questions addressed in this study. In this paper we draw primarily on 20 principal
interviews (one principal was only interviewed twice), with validity check from 15
superintendent interviews.
5. The 2005 and 2008 surveys had some overlapping items for both principals and teachers, but
were different. In addition, the same questions were not asked of principals or district leaders
during the three site visits.
6. The minimum number of teachers responding in any school was seven. One school was
eliminated due to low teacher response. Another factors leading to a reduction in the N of
schools was due to the fact that a number of schools “dropped out” of the 2008 survey
administration and that there were no principal or assistant principal respondents in a few
schools.
7. To ensure that this decision did not have a significant effect on the results, we conducted
ANOVAs to determine whether there were any significant differences between principals
and assistant principals on any of the key variables used in our analysis. The F-statistics for
the ANOVAs have significance levels of between 0.22 and 0.88. We also included principal/
AP status in some analyses.
8. Are these assessments of state policy, obtained in 2008, different from those collected at the
beginning of the project, when leaders had less experience with the effects of state responses
to NCLB? The answer is, not surprisingly, that they are. While there are some differences
between states, respondents in all states rated state policies less positively in 2008 than in
2005. Note that the respondents in 2005 and 2008 were not matched due to principal turnover
and non-response in some schools.
JEA 9. Although we turn to the interview data in detail later in this paper, the principals in our case
studies tended to view the state as the primary source for their public accountability and
50,5 rarely mentioned the federal role.
10. Argyris and Schön (1974) distinguish between espoused theories of action and theories-in-use.
Espoused theories are derived from reports of actions or intended actions and theories-in-use
from behavioral evidence (observations or incident reports). Since the following accounts of
662 principals’ leadership are derived from interview data, they are, by definition, espoused theory.

References
Argyris, C. and Schön, D. (1974), Theory in Practice: Increasing Professional Effectiveness,
Jossey-Bass, San Francisco, CA.
Borman, G.D., Hewes, G.M., Overman, L.T. and Brown, S. (2003), “Comprehensive school reform and
achievement: a meta-analysis”, Review of Educational Research, Vol. 73 No. 2, pp. 125-230.
Bryk, A.S. and Schneider, B. (2002), Trust in Schools: A Core Resource for Improvement, Russell
Sage Foundation, New York, NY.
Coburn, C.E. (2003), “Rethinking scale: moving beyond numbers to deep and lasting change”,
Educational Researcher, Vol. 32 No. 5, pp. 3-12.
Coburn, C.E. (2006), “Framing the problem of reading instruction: using frame analysis to
uncover the microprocesses of policy implementation”, American Educational Research
Journal, Vol. 43 No. 3, pp. 343-79.
Coburn, C.E., Honig, M.I. and Stein, M.K. (2009), “What’s the evidence on districts’ use of
evidence?”, in Bransford, J., Gomez, L., Lam, D. and Vye, N. (Eds), Research and Practice:
Towards a Reconciliation, Harvard Education Press, Cambridge, MA, pp. 67-87.
Cooley, V. and Shen, J. (2003), “School accountability and professional job responsibilities: a
perspective from secondary principals”, NASSP Bulletin, Vol. 87 No. 634, pp. 10-25.
Daly, A.J. and Finnigan, K.S. (2011), “The ebb and flow of social network ties between district
leaders under high-stakes accountability”, American Educational Research Journal, Vol. 48
No. 1, pp. 39-79.
Datnow, A. (2000), “Power and politics in the adoption of school reform models”, Educational
Evaluation and Policy Analysis, Vol. 22 No. 4, pp. 357-74.
Datnow, A. (2002), “Can we transplant educational reform and does it last?”, Journal of School
Change, Vol. 3 Nos. 3-4, pp. 215-39.
Datnow, A. and Stringfield, S. (2000), “Working together for reliable school reform”, Journal of
Education for Students Placed at Risk, Vol. 5 Nos 1-2, pp. 183-204.
Elmore, R.F. (2004), School Reform from the Inside Out: Policy, Practice, and Performance,
Harvard Education Press, Cambridge, MA.
Finnigan, K.S. (2010), “Principal leadership and teacher motivation under high-stakes
accountability policies”, Leadership & Policy in Schools, Vol. 9 No. 2, pp. 161-89.
Gioia, D.A., Thomas, J.B., Clark, S.M. and Chittipeddi, K. (1994), “Symbolism and strategic
change in academia: the dynamics of sensemaking and influence”, Organization Science,
Vol. 5 No. 3, pp. 363-83.
Gordon, L. (1995), “Controlling education: agency theory and the reformation of New Zealand
schools”, Educational Policy, Vol. 9 No. 1, pp. 54-74.
Hallinger, P. (2005), “Instructional leadership and the school principal: a passing fancy that
refuses to fade away”, Leadership and Policy in Schools, Vol. 4 No. 3, pp. 221-39.
Hallinger, P. (2007), “Leadership for learning: reflections on the practice of transformational and
instructional leadership [Online]”, Australian Council for Educational Research,
Melbourne, available at: www.acer.edu.au/documents/RC2007_Hallinger-Presentation.pdf
(accessed July 10, 2011).
Hart, W., Albarracin, D., Eagly, A., Brechan, I., Lindberg, M.J. and Merrill, L. (2009), “Feeling School leaders
validated versus being correct: a meta-analysis of selective exposure of information”,
Psychological Bulletin, Vol. 135 No. 4, pp. 555-88. as mediating
Hess, F. (1999), Spinning Wheels: The Politics of Urban School Reform, The Brookings Institution, agents
Washington, DC.
Honig, M. and Hatch, T. (2004), “Crafting coherence: how schools strategically manage multiple,
conflicting demands”, Educational Researcher, Vol. 33 No. 8, pp. 16-30. 663
Honig, M.I. (2006), “Street-level bureaucracy revisited: frontline district central-office
administrators as boundary spanners in education policy implementation”, Educational
Evaluation and Policy Analysis, Vol. 28 No. 4, pp. 357-83.
Horng, E., Klasik, D. and Loeb, S. (2009), Principal Time-Use and School Effectiveness: School
Leadership Research Report 9-03, Stanford University, Palo Alto, CA.
Ingram, D., Louis, K.S. and Schroeder, R. (2004), “Accountability policies and teacher decision
making: barriers to the use of data to improve practice”, Teachers College Record, Vol. 106
No. 6, pp. 1258-87.
Kerr, K.A., Marsh, J., Ikemoto, G., Darilek, H. and Barney, H. (2006), “Strategies to promote data
use for instructional improvement: actions, outcomes, and lessons from three urban
districts”, American Journal of Education, Vol. 112 No. 4, pp. 496-520.
Knapp, M.S., Bamburg, J.D., Ferguson, M.C. and Hil, P.T. (1998), “Converging reforms and the
working lives of frontline professionals in schools”, Educational Policy, Vol. 12 No. 4, pp. 397-418.
Latham, G.P. and Locke, E.A. (2006), “Enhancing the benefits and overcoming the pitfalls of goal
setting”, Organizational Dynamics, Vol. 35 No. 4, pp. 332-40.
Lee, M., Louis, K.S. and Anderson, S. (2012), “Local education authorities and student learning:
the effects of policies and practices”, School Effectiveness and School Improvement, Vol. 23
No. 2, pp. 133-59.
Leithwood, K. and Louis, K.S. (2011), Linking Leadership to Student Learning, Jossey-Bass,
San Francisco, CA.
Leithwood, K., Steinbach, R. and Jantzi, D. (2002), “School leadership and teachers’ motivation to
implement accountability policies”, Educational Administration Quarterly, Vol. 38 No. 1,
pp. 94-119.
Locke, E.A. and Latham, G.P. (1990), A Theory of Goal Setting and Task Performance, Prentice
Hall, Englewood Cliffs, NJ.
Louis, K.S. and Van Velzen, B. (Eds) (in press), Educational Policy in an International Context:
Political Culture and its Effects, Palgrave/Macmillan, New York, NY.
Louis, K.S., Dretzke, B. and Wahlstrom, K. (2010a), “How does leadership affect student
achievement? Results from a national US survey”, School Effectiveness and School
Improvement, Vol. 21 No. 3, pp. 315-36.
Louis, K.S., Febey, K. and Schroeder, R. (2005), “State-mandated accountability in high schools:
teachers’ interpretations of a new era”, Educational Evaluation and Policy Analysis, Vol. 27
No. 2, pp. 177-204.
Louis, K.S., Thomas, E. and Anderson, S.E. (2010b), “How do states influence district leadership”,
Educational Leadership and Policy, Vol. 9 No. 3, pp. 328-66.
McDonnell, L.M. (1994), “Assessment policy as persuasion and regulation”, American Journal of
Education, Vol. 102 No. 4, pp. 394-420.
Marks, H.M. and Printy, S.M. (2003), “Principal leadership and school performance: an
integration of transformational and instructional leadership”, Educational Administration
Quarterly, Vol. 39 No. 3, pp. 370-97.
Marzano, R., Waters, T. and McNulty, B. (2005), School Leadership that Works: From Research to
Results, ASCD, Alexandria, VA.
JEA Mintrop, H. and Sunderman, G. (2009), “Predictable failure of federal sanctions-driven accountability
for school improvement – and why we may retain it anyway”, Educational Researcher, Vol. 38
50,5 No. 5, pp. 353-64.
Nelson, B.S. and Sassi, A. (2005), The Effective Principal: Instructional Leadership for High
Quality Learning, Teachers College Press, New York, NY.
Newmann, F., Smith, B., Allensworth, E. and Bryk, A.S. (2002), “Improving Chicago’s schools:
664 school instructional program coherence”, ERS Spectrum, Vol. 20 No. 2, pp. 38-46.
O’Day, J. (2002), “Complexity, accountability, and school improvement”, Harvard Educational
Review, Vol. 72 No. 3, pp. 293-329.
Robinson, V. (2011), Student-Centered Leadership, Jossey-Bass, San Francisco, CA.
Robinson, V.M.J., Bendikson, L. and Hattie, J. (2011), “Leadership and student outcomes: are
secondary schools unique?”, in Robertson, J. and Timperley, H. (Eds), Leadership and
Learning, Sage, London, pp. 131-42.
Robinson, V.M.J., Lloyd, C. and Rowe, K. (2008), “The impact of leadership on student outcomes,
An analysis of the differential effects of leadership types”, Educational Administration
Quarterly, Vol. 44 No. 5, pp. 635-74.
Rorrer, A.K., Skrly, L. and Scheurich, J.J. (2008), “Districts as institutional actors in educational
reform”, Educational Administration Quarterly, Vol. 44 No. 3, pp. 307-57.
Rosenholtz, S.J. (1991), Teachers’ Workplace: The Social Organization of Schools, Teachers
College Press, New York, NY.
Spillane, J., Diamond, J.B. and Burch, P. (2002a), “Managing in the middle: school leaders and the
enactment of accountability policy”, Educational Policy, Vol. 16 No. 5, pp. 731-62.
Spillane, J., Reiser, B.J. and Reimer, T. (2002b), “Policy implementation and cognition: reframing
and refocusing implementation research”, Review of Educational Research, Vol. 72 No. 3, pp. 387-431.
Spillane, J.P. (1998), “State policy and the non-monolithic nature of the local school district: organizational
and professional considerations”, American Educational Research Journal, Vol. 35 No. 1, pp. 33-63.
Stein, M.K. and Nelson, B.S. (2003), “Leadership content knowledge”, Educational Evaluation and
Policy Analysis, Vol. 25 No. 4, pp. 423-48.
Thomas, J.B., Sussman, S.W. and Henderson, J.C. (2001), “Understanding ‘strategic learning’:
linking organizational learning, knowledge management, and sensemaking”, Organization
Science, Vol. 12 No. 3, pp. 331-45.
Togneri, W. and Anderson, S.E. (2003), “How high poverty districts improve”, Leadership, Vol. 33
No. 1, pp. 12-6.
Walsh, J.P. (1995), “Managerial and organizational cognition: notes from a trip down memory
lane”, Organization Science, Vol. 6 No. 3, pp. 280-321.
Weick, K.E., Sutcliffe, K.M. and Obstfeld, D. (2005), “Organizing and the process of sensemaking”,
Organization Science, Vol. 16 No. 4, pp. 409-21.
Wills, F.G. and Peterson, K.D. (1992), “External pressures for reform and strategy formation at
the district level: superintendents’ interpretations of state demands”, Educational
Evaluation and Policy Analysis, Vol. 14 No. 3, pp. 241-60.
Wohlstetter, P. (1991), “Accountability mechanisms for state education reform: some organizational
alternatives”, Educational Evaluation and Policy Analysis, Vol. 13 No. 1, pp. 31-48.
Youngs, P., Holdgreve-Resendez, R. and Qian, H. (2011), “The role of instructional program
coherence in beginning elementary teachers’ induction experiences”, The Elementary
School Journal, Vol. 111 No. 3, pp. 455-76.

Further reading
Grissom, J.A. and Loeb, S. (2011), “Triangulating principal effectiveness”, American Educational
Research Journal, Vol. 48 No. 5, pp. 1091-123.
Appendix School leaders
as mediating
Minimum- agents
Scale maximum/
Measure Stem and items reliability mean/SD

Positive state policy To what extent do you agree that: 0.83 (n ¼ 4) Minimum/ 665
index: degree of state standards stimulate additional maximum:
agreement with professional learning in our school; state 5-30
positively worded policies help us accomplish our school’s Mean: 22.4
statements about the learning objectives; the state gives schools SD: 4.08
instructional focus freedom and flexibility to do their work;
and responsiveness the state communicates clearly with our
of state policy (from district about educational policies
principal survey)
District To what extent do you agree that: 0.87 (n ¼ 5) Minimum/
accountability: our district has the capacity for reliable maximum:
degree of agreement assessment of student and school 1-6
that the district was performance; our district incorporates Mean: 5.32
focussed on and student and school performance data in SD: 0.65
could support the district-level decision making; our district
school’s assists schools with the use of student and
improvement efforts school performance data for school
(from principal improvement planning; the district uses
survey) student performance data to determine
teacher professional development needs
and resources
Principal To what extent do you agree or disagree 0.82 (n ¼ 7) Minimum/
instructional with the following statements about your maximum on
leadership: teacher school leader(s): factor score:
ratings of specific my school leader clearly defines standards 2.71 to 1.96
types of principal for instructional practices; Mean: 0
instructional how often in this school year has your SD: 1
leadership – school leader(s) [y]
frequency of or observed your classroom instruction;
agreement about its attended teacher planning meetings; made
occurrence (from suggestions to improve classroom
2008 teacher survey) behaviour or classroom management;
given you specific ideas for how to
improve your instruction;
buffered teachers from distractions to Table AI.
their instruction? Survey measures

About the authors

Karen Seashore Louis is a Regents Professor at the University of Minnesota, USA. Karen
Seashore Louis is the corresponding author and can be contacted at: [email protected]
Viviane M. Robinson is a Distinguished Professor at the University of Auckland,
New Zealand.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints
 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0957-8234.htm

JEA
50,5
Managing the intersection
of internal and external
accountability
666 Challenge for urban school leadership
Received 27 October 2011 in the United States
Revised 2 March 2012
Accepted 18 April 2012 Michael S. Knapp
Education Leadership and Policy Studies, University of Washington, Seattle,
Washington, USA, and
Susan B. Feldman
Research Center for Learning and Leadership, Vancouver, Washington, USA

Abstract
Purpose – The purpose of this paper is to direct attention to the intersection of external and internal
accountability systems within urban schools, and the role of school leadership, especially that of the
principal, in managing this intersection. In particular, the paper explores how school leaders are able to
strengthen and sustain the school’s internal accountability system, in pursuit of school-defined
learning improvement agenda, and at the same time respond productively to external accountability
demands. The paper also seeks to identify consequences of these leaders’ efforts to navigate an often
problematic set of converging demands.
Design/methodology/approach – This paper draws on findings from a larger multi-case study of
learning-focused leadership in 15 schools in four urban school districts in the USA. Schools were
chosen to represent those that were “making progress” (by local measures). Data were collected over
18 months, spanning two school years, from Spring 2007 to Fall 2008. Data collection included multiple
site visits, semi-structured interviews and observations of leadership activity across school and
district settings, and a variety of documentary evidence.
Findings – Though working in substantially different contexts, these leaders found remarkably
similar ways of crafting tools and creating occasions, from the array of external accountability
demands and resources, to serve internal accountability purposes. They did so by internalizing
external expectations and developing accountable practice within the school, leading through data,
and modelling what it meant to learn to lead in a fully accountable way. As they did so, they reshaped
the scope of instruction and the instructional improvement conversation, and also made teaching and
leadership practice more public.
Originality/value – This paper extends discussions of school-level accountability in two ways. First,
it updates scholarship on accountability by examining school-level responses at a time five years into
the new accountability context in the USA defined by strict system-wide expectations and
mechanisms. Second, the paper demonstrates ways in which the often onerous demands of external
accountability systems can be treated as a resource by school leaders and used in ways that bolster the
school’s capacity for accountable professional practice.
Keywords United States of America, Schools, Urban areas, Leadership, Educational administration,
Policy, Accountability
Paper type Research paper

Journal of Educational
Administration In current policy discourse across national contexts, the term “accountability” is likely
Vol. 50 No. 5, 2012
pp. 666-694 to conjure up images of system-wide arrangements for ensuring the proper
r Emerald Group Publishing Limited
0957-8234
expenditure of public funds and for encouraging or even compelling educators to
DOI 10.1108/09578231211249862 improve their performance to acceptable levels. The term almost always concerns
systems of expectations, rewards, and sanctions that surround the school and originate Internal and
outside of it. But in this discourse, we easily and often forget that the ultimate goal external
is to get accountable practice within the school and doing so inevitably means
that school-based educators must themselves enact and adhere to a system of accountability
accountability internal to the school. In some fashion, they must hold themselves and
each other to account for their contributions to students’ learning and for their
collective performance. Absent that, no external accountability arrangements can have 667
any useful effect. Put another way, as recent scholarship has recognized (Abelmann
et al., 1999; Carnoy et al., 2003), externally defined accountability arrangements can
only work through the school’s internal accountability system.
Currently in the USA, nearly a decade into an era of intensified, system-wide
accountability pressures under the No Child Left Behind law, is an important time to
consider the interaction of external and internal accountability systems in schools. The
conditions set up by this nation-wide policy add to the external pressures on schools, and
increase the temptation to see the problem of accountability in externalized terms.
This paper redirects attention to the intersection of external and internal accountability
systems within urban schools, and the role of school leadership, especially that of
the principal, in managing this intersection. To pursue this matter, we explore issues
concerning the way system-wide, high-stakes accountability systems can contribute to
the goal of developing internally accountable practice for high performance. In this regard,
we are especially interested in the school leaders’ response to external accountability
expectations, directives, resources, and constraints, and how these responses work inside
the school, especially inside the school-specific culture of accountability and professional
work, which the leaders may have responsibility for forging. Of particular concern to us
is the response of the school principal, though other school leaders are implicated – people
in this position are likely to be held responsible for meeting external demands, are in
position to translate this responsibility into expectations for their own staffs, not to
mention themselves; they must also consider how external expectations can be reconciled
with internal school priorities. We address these matters by reviewing findings from
a recent study we have undertaken of urban school leadership and its relationship to
learning improvement. Our work is located in lines of research that accumulate insights
about external accountability systems, internal school accountability, and the relation
between the two.

Accountability systems that surround the school

Strict, high-stakes accountability systems are a central feature of standards-based
reform at local, state, and federal levels these days, and are sometimes thought of as
a recent phenomenon. Yet schools have always been accountable to external actors and
interests (Cuban, 2004; Beadie, 2004). External accountability systems can be understood
as a complex arrangement of policies, created by actors and interests outside of schools,
who are in position to reward and punish schools, aimed at impacting practices inside
schools, and requiring reporting to diverse external audiences.
From the perspective of school leaders, the external accountability system is an array
of political, bureaucratic, and market-driven supports and constraints aimed at defining
what educators in the school should be doing and producing. Because multiple interests
outside of the school have expectations of it, these supports and constraints may not be
educationally coherent and may require activity that conflicts with the educational
interests of the school (e.g. O’Day, 2002; Skrla and Scheurich, 2004; Firestone and Shipps,
2005). Orchestrating the ongoing instructional practices generated in this contentious
JEA zone between external actors and interests, and internal school actors and interests is
50,5 thus a central concern of school leadership (Leithwood, 2005).
Over time, the array of external expectations and mechanisms for encouraging or
compelling educators to meet these expectations has shifted. Captured in the mid-1990’s
as “the new accountability” (e.g. Elmore et al., 1996; Carnoy et al., 2003; Elmore, 2004;
Cohen et al., 2007), these shifts include at least the following four developments. First,
668 accountability systems paid more explicit attention and placed far greater emphasis on
demonstrated results, generally through the vehicle of student test scores. In other words,
rather than concentrating on whether the inputs to education were in place (Noguera,
2004), the system asked for evidence of outputs. Second, the system developed and put in
place clear, and highly specified common reference points for performance (e.g. state
learning standards). Third, the units of accountability broadened to include the whole
school, the individual teacher (in any given school year), and school leader. While testing
is not new to school leadership, testing individual students for the purpose of measuring
the success of the whole school and particularly the success of the principal is a new
arrangement for school leadership (e.g. Leithwood et al., 1999; Elmore, 2000; O’Day, 2002).
Fourth, the system increased the stakes in accountability, by attaching more stringent
consequences to school, educator, or student performance.
Because external accountability systems make demands on school staff,
individually and collectively, they necessarily, but not deterministically, intersect and
interact with internal accountability systems. Leadership practice uses tools and
resources available in the external environment to generate the instructional practices,
assessments, and incentives for the work of school staff to align and exceed external
expectations. While principals have always worked in this contentious zone, and they
have always taken advantage of the external environment to conduct their work, the
tools and resources now available through system-wide accountability arrangements
(e.g. data warehouses and associated data displays, online testing, leadership coaching)
shape their options and their actions in particular ways.
A growing body of research documents how educators and students experience
external accountability systems. The research makes clear that these system demands can
be experienced as onerous, punitive, intrusive, and de-skilling, and they have the potential
to work these effects in inequitable ways that disproportionately punish schools serving
historically underserved children (e.g. Sunderman et al., 2005; Sirotnik, 2004; Diamond and
Spillane, 2002; Darling-Hammond, 2007). While these concerns have prompted a major
national policy debate, commitment to some version of strict accountability for producing
results continues. There is no sign in the process of reauthorizing No Child Left Behind, for
example, that authorities or the general public wish to back off of a strict and demanding
accountability stance vis-a`-vis the public schools, though specific requirements (such as the
“adequate yearly progress” provision) may be loosened.
However, they are experienced in a high-stakes world, external accountability demands
are hard for school-based educators to ignore. In other words, they are consequential.

Accountability systems within the school

Some researchers point to the school’s own internal accountability system as the key
factor that explains a school’s capacity to meet external expectations. Described here
by researchers focussed on high school reform:
[y] Schools actually have conceptions of accountability embedded in the patterns of their
day-to-day operations, and [y] a school’s conception of accountability significantly
 influences how it delivers education [y] Schools must solve the problem of accountability in Internal and
some way in order to function and [y] the way they solve this problem is reflected in the way
teachers, administrators, students and parents talk about the fundamental issues of schooling external
(Carnoy et al., 2003, pp. 3-4). accountability
This explanation of internal accountability is so broad it could be replaced with
the term “school culture.” However, the authors add some specificity to this
explanation, which helps to pinpoint features of the school that constitute its 669
accountability system:
Internal accountability has three tiers: the individual’s sense of responsibility; parents’, teachers’,
administrators’, and students’ collective sense of expectations; and the organizational rules,
incentives, and implementation mechanisms that constitute the formal accountability system in
schools (Carnoy et al., 2003, p. 4).
What we take from these authors is the idea that an internal set of expectations
permeate schools. Individuals in the school have expectations for themselves
and often for others, and those expectations are demonstrated and discerned
in the instructional decisions educators make, their descriptions of their work,
and their interactions with each other. In varying degrees, the actors in a given
school also allocate responsibility for meeting these expectations, and for
demonstrating that they have been met. Nothing ensures that expectations are
coherent, shared, or acted upon – the default, in fact, is for schools to rely on
“atomistic” accountability arrangements, in which individual teachers are left
to hold themselves to account for successfully educating the students in their
charge (Elmore, 2006).
In this way, schools, as a professional collective, always have some kind of
accountability system in place – generally more implicit than explicit, and generally
(by default) placing most or all of the responsibility for accountable practice in the
hands of individual teachers. Conducting the continuous transformation from
atomistic and personally derived professional expectations to professional
expectations defined collectively by the school as a whole implies a major step for
school staff and leaders. The further step of anchoring collective internal expectations
to those defined externally (and enacted internally) is likely to entail a redefinition of
classroom teaching and school leadership.

The challenge for school leadership at the intersection of external and

internal accountability
Accountable practice within the school includes that of school leaders, as well as those
whose daily work is situated in classrooms. With standards, assessments, and public
reporting required under current standards-based reform policies, school leaders –
especially principals – face a practical shift from evaluating individual teachers
as the measure of the school’s productivity, to themselves being evaluated based on
the assessed academic productivity of their students and, by extension, their teachers
as a whole. School leaders able to quickly align their teachers’ instruction to the
annual tests are likely to be evaluated as successful, thus contributing to a new
expression of school leadership. Under these conditions of highly specified, high-stakes
expectations, school leadership requires tightly tying classroom activity to state
standards and assessments rather than to teachers’ personal standards,
repositioning historically personal and individualistic classroom activity as
essentially interdependent.
JEA Under conditions that put increasing pressure on the school as a whole, the natural
50,5 tendency of schools to rely heavily on individualized atomistic accountability
arrangements sets the stage for a major learning challenge for school staff – those
exercising leadership as well as everyone else. In other words, school staff have a lot to
learn together concerning how to be responsible to each other and to high expectations
(their own and those of the larger system), and to meet new expectations for their work.
670 Education researchers studying the early efforts of school leaders to adjust their
schools’ practices to align to external expectations have found resistance (Mintrop,
2004b), confusion (Carnoy et al., 2003), and denial (Ingram et al., 2004). In these early
years of standards and assessment expectations (beginning in 1989 with an education
summit of state governors – see Goertz et al., 2001; Phelps and Addonizio, 2006),
leaders had few tools to use for internalizing these expectations.
Once standards and the associated assessments were established, these tools were
certainly going to be used to evaluate their success, and school leaders could use these
tools to reposition instructional practice in their schools to meet or exceed external
expectations. Some school leaders were apparently able to align the work inside their
school to meet the expectations of external audiences while other school leaders were
not (Diamond and Spillane, 2002). Some school leaders who had been leading popular
public schools suddenly found themselves leading “failing schools,” in which teachers
who had seen themselves as successful suddenly found themselves with a new identity
defined by external expectations.
So, on top of the enduring challenge to a school principal of ensuring and
supporting productive work across the school while doing what needs to be done to
manage the building, school leaders find themselves with new responsibilities – and
potentially new tools – at the intersection of internal and external accountability. Under
the current high-stakes environment in which they work, their schools’ continuation in
its current configuration, not to mention the current staff and their own jobs, depend in
a basic way on their ability to manage a productive merger of what the outside world
asks of the school and what the school staff ask of themselves and their students.

The puzzle for research on school leadership

Given their position as a central figure in the affairs of the school and simultaneously
its chief point of connection with the larger educational system, school principals are
positioned to notice, experience, and respond to the dynamic interactions between
external and internal accountability systems. On the one hand, system leaders and
community constituents for public education are likely to expect them to “make
good things happen” in the school; and at the same time, they are best able to act as
a chief architect and central participant in the school’s own accountability system.
The puzzle has to do with the management of multiple demands, internal and external,
the potential for expectations to work at cross purposes, and the sheer
multidimensionality of the accountability challenge. Looking ahead to the past
decade of leadership research, a prominent review anticipated the need to study this
point of intersection more closely:
[y] Internal-professional mechanisms [of accountability] substantially govern practice, and
are more easily aligned with teaching and learning, and thus are more likely to improve
student performance. The management question regarding school-based professional
communities becomes how to support school-based modes of accountability while aligning
them with demands from schools’ bureaucratic or political superordinates (Adams and
Kirst, 1999, p. 470).
This puzzle is made more salient in the current climate in the USA of federally driven Internal and
standards-based reform, which creates for schools and their leaders ambitious and external
highly specified improvement targets, short timelines, and severe consequences for
failing to meet them. Relatively little research has investigated the matter empirically accountability
five or more years into the No Child Left Behind era, in other words, at a point in time
where an intensified multi-level external accountability system has been in place
for enough years to alter the way school staffs configure and pursue their work. In this 671
context, it is important to ask several questions, concerning schools that are making
progress on learning improvement goals:
(1) How do school leaders negotiate and navigate the intersection of internal and
external accountability systems?
(2) What consequences do their actions at this intersection have for efforts to
improve teaching practice and student learning outcomes?
These questions matter because of the enormous investment in external
accountability-driven reforms, and the hopes pinned to these strategies. In parallel,
reformers hope for a professionalized and competent teaching workforce, and a quality
of instructional practice that leads the full range of students served by public schools
to new levels of performance and preparation for their future lives. Yet, the early signs
are clear: standards-based accountability systems may be having counterproductive
effects on practice, as much as productive impacts. The possibility remains that
these effects may be distinctly inequitable. We need to understand where and how
school leaders can make best use of the opportunities and constraints present in
accountability dynamics both outside and inside their schools.

Framing ideas
A closer look at ideas related to accountability and leadership helps us to pinpoint
what may be taking place at the intersection of internal and external accountability
systems. These ideas emerge from research over the last decade focussed on the reform
of elementary and secondary schools, often in the settings and circumstances that pose
significant challenges to reformers.

Forms and logics of accountability

Researchers have distinguished different forms of accountability related to the work of
educators in schools. Their logics differ in fundamental ways from each other, and each
presumes a different way of diagnosing high or low school performance and locating
the path to educational improvement. Each offers a different set of answers to the basic
questions: accountable to whom? For what, in relation to what standards? On whose
authority? Through what practices? With what metrics or ways of demonstrating that
standards have been met? With what consequences, for whom?
The following typology (adapted from Adams and Kirst, 1999; Ranson, 2003;
Leithwood, 2005; Firestone and Shipps, 2005) locates four main forms of accountability in
different configurations of authority and approaches to ensuring high performance. The
first positions accountability in decisions made by the managers of the educational
system, through one of two arrangements (or combinations of the two):
. “management” (Leithwood, 2005), “bureaucratic” (Adams and Kirst, 1999;
Firestone and Shipps, 2005), or “corporate” (Ranson, 2003) accountability:
those in charge of the education system (e.g. district, state leaders) develop
JEA mechanisms and measures for assuring that schools perform effectively,
50,5 and through strategic planning and regular monitoring of performance, and
rewards and sanctions based on performance, strive to motivate and achieve
high performance; and
. “decentralized” accountability (Leithwood, 2005): those in charge of the
system may also devolve authority for making decisions about program
672 (e.g. curriculum, budget, hiring) and for ensuring high performance to the
school level, thereby “empowering” school leaders and often school-based
decision-making bodies that may include other stakeholders (e.g. community
members, sometimes teaching staff).
The second form locates accountability in professional consensus about good practice:
. “professional” accountability (Adams and Kirst, 1999; Ranson, 2003; Leithwood,
2005): here, the ultimate touchstone for good practice and high performance is
consensus among professional teaching staff in a school – and more broadly, in
the teaching profession. In this view, through norms and other more explicit
mechanisms involving peer scrutiny, members of the profession hold each other
accountable to standards for good practice that they have set for themselves.
The third source anchors accountability to non-professional interests and
preferences, either as expressed by parents or within the political system in
which public education sits:
. “market” (Adams and Kirst, 1999; Leithwood, 2005) or “consumer” (Ranson,
2003) accountability: here, the ultimate touchstones for good practice are
the preferences and decisions of parents, conceived of as the “clients” or
“consumers” of education, whose expressed desires for schooling and
ultimate decisions about where their children are schooled motivate and guide
educators’ work; and
. “political” accountability (Adams and Kirst, 1999; Firestone and Shipps, 2005):
instead of parents’ preferences and decisions to enroll students in this or that
school, the ultimate touchstone shifts to the established political process,
whereby broader community interests are expressed through elected
representatives, who make decisions about educational funding, programs,
even educators’ job parameters and specifications. School performance resides in
the eye of the electorate, and is judged as much in terms of constituents’
perceptions as any objective measures.

One other form anchors accountability to beliefs about the “right thing to do” for school
children:
. “Moral accountability” (Adams and Kirst, 1999; Firestone and Shipps, 2005):
identifies a different ultimate touchstone for good practice in the school. Here,
compelling conceptions of what is right to do, as in many current notions of
“social justice,” become the guiding light for educators and others to judge their
work and what it produces.
These forms of accountability are not inherently incompatible, and the types can
exist in combination, as in decentralized arrangements that offer school leaders
considerable autonomy (a form of decentralized accountability) while insisting that
the school adhere to system-wide learning standards, submit to regular monitoring, Internal and
and demonstrate performance on system-wide measures (a form of management external
accountability). At the same time, they can easily conflict (Firestone and Shipps, 2005),
as when the political accountability system finds schools wanting, which professional accountability
educators see as solid and sound, or vice versa. That said, in a given school or district
context and larger policy environment, one form of accountability is likely to dominate
the others, as in present standards-based reform policies, in which a management 673
accountability system overshadows the efforts of most school-based educators. And
over time, one dominant mode of accountability may give way to another, dramatically
demonstrated in the transition from a largely professional form of accountability
during the 1970s in the UK to accountability forms that prioritize more corporate and
consumer interests in the 1980s and beyond (Ranson, 2003).
Each form of accountability arises and operates in ways that are external to the
school, though they may also manifest themselves internally. But the basic logic of
external and internal accountability are likely to differ in predictable ways, with
important ramifications for leaders’ practice and for efforts within the school to
improve teaching and learning.
The logic of external accountabilities lies in the notion that professional work
is, or needs to be, extrinsically motivated, guided by a larger set of interests residing
in the community served by public education, and compelled or enforced by the
system-level leaders (located outside individual schools) who serve these
interests. External accountabilities thus rely on actors, positioned at some
distance from the actual interface between teachers and learners, to specify what
needs to happen in that interaction, how the quality of that interaction will be
known, and what consequences to attach to that interaction. From this relatively
distant vantage point, management, market, and political accountabilities are likely
to predominate.
Accountability located and exercised in schools sits within a set of ongoing
relationships among professional people who work alongside each other and in ways
that are, by degrees, interdependent with one another. The logic of internal school
accountability typically assumes that those who work closest to teaching and learning
interactions are in the best position to judge each others’ work. Internal accountability
is thus more likely to favor intrinsic motivation, as it often presumes a sense of mutual
responsibility for the quality of work. Here, decentralized, professional, and moral
accountabilities are likely to predominate, all other things being equal. That said, there
is nothing to prevent a school’s internal accountability system from being largely
management-driven or “political,” as an autocratic principal tries to “make things
happen” to satisfy external constituencies.
Presiding over the intersection of external and internal accountabilities are school
leaders, especially the principal, who is the official link between the school and the
larger system in which the school sits, yet at the same time the person who takes
responsibility for managing the affairs of the school. But to understand this
individual’s work at the intersection presumes a picture of leadership and its relation to
professional and system learning, which we turn to next.

Learning-focussed school leadership

While there are many ways to construe what school leadership is all about (Portin et al.,
2003), a set of ideas we and others have been developing over the last decade directs
attention toward its connections to learning, on the premise that learning and learning
JEA improvement is ultimately the greatest concern of school leaders and, indeed, of all
50,5 staff members in a school. Using the term “learning-focussed leadership,” this view
relates school leaders’ work to student, professional, and system learning, and
highlights the ways that particular kinds of actions forge strong connections among
them (Knapp and Copland, 2006; Knapp et al., 2003). This view of leadership further
presumes that all three arenas of learning operate simultaneously and
674 interdependently, and that to maximize the performance of the school means to
maximize the learning of all three. These ideas build on others’ work using similar
terms, for example, writings that have directed attention to “learning-centered
leadership” (Murphy et al., 2006), and “leadership for learning” (e.g. Resnick and
Glennan, 2002; Stoll et al., 2003; Swaffield and Macbeath, 2009). (Still others use
the same or related terms, though not necessarily in the ways that we or these
authors do[1].)
Notions of learning-focussed and learning-centered leadership direct attention to
certain kinds of processes at work in the school among the various actors who do or
can exercise leadership aimed at instructional improvement. With roots in theory and
empirical findings concerning distributed leadership, instructional leadership, the
urban principalship, and organizational learning, this lens puts a great deal of
emphasis on the collective leadership work of the school, among which are steps
leaders take to move the school beyond an atomistic accountability culture. Also
central to this way of framing what goes on in the school is the development of a
school-wide “learning improvement agenda” (Knapp and Copland, 2006; Portin et al.,
2009), which becomes a natural reference point for accountable practice. The agenda
can help to anchor a set of relationships within teams and other collaborative
structures that bring matters of instructional practice regularly into the view of
peers or others who might help individuals feel responsibility for improving practice,
and act accordingly.
This framing of leadership work in the school is particularly appropriate for
examining questions about the way internal and external accountabilities intersect. For
one thing, the framework highlights the focus of accountability expectations (the
improvement of students’ learning) and also the means for meeting these expectations
(new learning about how to meet these expectations as a professional collective).
Furthermore, embedded within the idea set concerning learning-focussed leadership
is the notion of shared accountability (Macbeath, 2008). Because this approach
to school leadership assumes an active distribution of effort to guide and support
practice aimed at the improvement of teaching and learning, it is only natural that
the responsibility for improving learning (and for failures to do so) resides within
the collective, and that schools will develop practices that make this the result.
Furthermore, shared accountability can subsume the sharing of expectations by
internal and external groups, if school staffs are able to internalize and/or adapt
external expectations for their own purposes. Absent that, the work of leaders within
the school takes on the burden of external accountability in such a way that they
become “more concerned with accounting than learning, with control than with
teaching, with compliance than risk-taking, and with public relations than student
experiences” (Sackney and Mitchell, 2010, as cited in Macbeath, 2008, p. 139).
Finally, learning-focussed leadership frameworks emphasize the leaders’ treatment
of the external environment as a resource (Knapp and Copland, 2006; Knapp et al.,
2013). As learning-focussed school leaders look inward to develop their professional
communities and bring school staff’s best efforts to bear on instructional improvement,
they simultaneously “interact with the local community and other external environments Internal and
in ways that define and create opportunities for learning improvement. [y]” while at the external
same time being “on the lookout for challenges such as external pressures, demands,
crises, or other events that may preempt or constrain attempts to advance a learning accountability
improvement agenda” (Knapp and Copland, 2006, p. 56). In other words, leadership
practice aimed at learning improvement is simultaneously internal and external work,
with school leaders approaching the environment as a source of opportunities and 675
resources as well as constraints.
In sum, understanding changing expectations and emerging practices requires
learning for all school staff, and especially for school leaders. Principals are naturally
positioned as a “lead learner,” in the sense that they need to know what their teachers
and students will need to be learning, at the same time they may wish to cultivate a
learning stance among their staff. As they do so, they are in a good position to engage
the external environment and all its demands as a resource for the learning
improvement work of the school (Knapp et al., 2013).

Framework for our research

Our study draws together ideas from these bodies of work in ways that focus attention
on leadership work at the intersection of external and internal accountability. In broad
strokes, we assume, in line with research described earlier, that the practice of internal
school accountability reflects three overlapping spheres – the individual staff members’
sense of responsibility for their own performance as professionals (and the performance
of the students they teach); a collective set of expectations held by parents, teachers,
administrators, and students about performance; and a set of rules, mechanisms, and
incentives for improving, and ultimately attaining high levels of, performance. These
three many not be coherently aligned with each other, or even well established. By
default, schools are likely to rely most heavily on individuals’ sense of responsibility,
thereby relegating the school to a fairly atomistic accountability pattern.
The challenge to a school staff, then, and hence to its leaders, is to develop practices
that maximize the overlap between the three spheres of activity. There, at this
intersection, collective understandings of the work of the school increasingly shape,
and are shaped by, individuals’ sense of responsibility for high-quality teaching and
learning; and both are reinforced by various mechanisms and incentives which
encourage and reward the collective work and what results from it. What happens here
is a central feature of learning-focussed leadership work, as suggested schematically in
Figure 1. Maximizing the interconnections among school accountability elements is
unlikely to come about by accident. Rather, it calls on what leadership is always about –
the nurturing of collective purpose in the organization and the mobilization of effort in
pursuit of that purpose. And because the dominant purpose of concern is the
improvement of teaching and learning, this task calls for leadership work that is
persistently and publicly focussed on learning and learning improvement.
The school’s internal accountability patterns and routines – and hence, school
leadership work – occur within a larger system of accountability, in fact, within an
environment of multiple accountabilities (e.g. bureaucratic, political, market, and moral
accountabilities, as noted earlier). This fact highlights a second intersection for school
leaders, especially the principal. As signaled by Figure 2, the principal manages not
only the intersection among the three spheres of activity that comprise the school’s
internal system of accountability, but also the intersection between internal and
external accountability systems. Each requires artful navigation and negotiation; each
JEA School context
50,5
Learning-focussed
Leadership work

676 Collective
Individuals’ expectations for
sense of performance
responsibility for (parents’, teachers’,
performance administrators’,
students’)

Organizational
rules, incentives,
Figure 1. implementation
Leadership and the mechanisms
internal school
accountability system

Larger External accountability system:

environment expectations, pressures, tools, resources

School context
Learning-focussed
Leadership work

Individuals’ Collecitve
sense of expectations for
responsibility performance
for performance

Figure 2. Organizational rules,

incentives,
The dual accountability
implementation
challenge for school mechanisms
leaders

offers opportunities as well as constraints. Taken together, the two pose a many-
faceted accountability challenge for the principal and other school leaders.
Navigating these intersections can take many forms. Obvious default pathways can
be taken at either intersection. Within the school, for example, school leaders can
follow lines of least resistance by allowing the school staff to slip into an atomistic
accountability pattern, while insisting that staff simply comply straightforwardly with
external dictates about test score improvement. But doing so is likely to be
counterproductive, and school leaders who care about serving students well, and are
willing to be held accountable for this, are likely to explore various other avenues that: Internal and
expand and strengthen collective expectations for performance in the school, heighten external
individuals’ sense of personal responsibility for outcomes, and establish reinforcing
mechanisms and incentives for doing so. Their efforts are likely to be both informed accountability
and shaped by what the external accountability system asks for. Our research was
intended to capture in some detail what was taking place at these intersections, and
especially how school principals viewed and managed the possibilities and pressures 677
that existed in this crucial aspect of their roles.

Design and methods

To pursue these and other facets of learning-focussed leadership, we undertook in
2006-2008, with support from The Wallace Foundation, a multiple-site case study
investigation of four moderate-to-large-sized urban school districts undertaking
district-wide learning-improvement initiatives that involved active, though different
attempts to realize substantial improvements in teaching and learning within the
four districts (Atlanta public schools, Atlanta, GA; New York City Department of
Education/Empowerment Schools Organization; public schools, Springfield, MA,; and
Unified School District, Norwalk-La Mirada, CA). We used a purposive sampling
strategy (Patton, 2002) to identify 15 schools in which to study learning-focussed
leadership. Schools were selected that were:
(1) making progress (by whatever local metrics school staff used to demonstrate
progress) on improving student learning across the full range of diverse students;
(2) sharing leadership work in ways that maximized leaders’ attention to learning;
and
(3) experimenting with the allocation of staffing resources, to facilitate more
equitable student learning.
Within this boundary, we used a “maximum variation” strategy, to include large and
small schools, non-traditional and traditional school structures, and schools at each
level of the K-12 system (elementary, middle, and high schools). We also chose schools
to reflect a range of experience among the school leaders, with some having long-
standing principals (in one case more than 15 years) and others with relatively new
leaders (e.g. having come to the principalship the year before our study began).
The schools we studied were accomplishing ambitious learning goals, under
demanding conditions. At the time of this study, under the definition of district
achievement in the federal No Child Left Behind policy, all four urban districts
were emerging as “failing” at the district level – a new classification for school districts.
The schools in our study, however, were all outperforming the district average for
school achievement. Table I provides state report card data for the schools. Keeping in
mind that each state and district organizes their achievement data and goals
differently, these schools’ achievement cannot be compared to each other.

Data collection
Each school was visited at least four times during the study by one or two researchers,
across a year and a half, including the spring of the 2006-2007 school year, the
2007-2008 school year, and the first three months of the following year. Across our
visits, we interviewed repeatedly all staff within each school who were identified as
exercising leadership, and selected others (teachers, support staff) who could reflect on
JEA Predominant
50,5 demographic
School levels School size groups School achievement

NYC Department of Education/Empowerment schools organization

Elementary 650-1700 82-96% Latino Met AYP in all subject
schools (3) 85-98% FRL areas in all grades
678 Middle school (1) 512 students 59% Latino Made AYP in math, language
grades 6-8 40% African- arts, and science
American
81% FRL
Middle-High school 386 100% Latino Made AYP in math and
(gr 6-12) (1) 86% FRL science, not in English
language arts

Atlanta public schools

Elementary school 475 students 90% African- Met AYP in all years
(1) American
81% FRL
Middle school (1) 950 students 97% African- Met AYP in all years
American
78% FRL
High school (2) 155 students 94-97% African- Both high schools made AYP
950 students American
81-82% FRL

Norwalk-La Mirada unified school district

Elementary school Approximately 550 81% Latino Made AYP in 17 of 17 categories
(1) students 71%
FRL34% ELL
Middle school (1) Approximately 86% Latino Made AYP in all areas
1,050 students 73% FRL except ELL and special
education students
High school (1) Approximately 52% Latino Made AYP in the all categories,
2,300 students 32% white except in Latino, white,
25% gifted and or low socioeconomic advantage
talented

Springfield (MA) Public schools

Elementary school 650 students 48% Latino School made AYP in all areas,
(1) 21% African- except in English on aggregate in 2006
American
14% white
Middle school (1) 1,200 students 84% FRL 2006 made AYP in language arts
77.6% Latino in all areas; 2007, whites made
AYP in language arts
In math, African-Americans
made AYP in 2007
High school (1) 2,000 students 53% FRL School made AYP in all areas except
36% Latino Special Education students
29% African-
American
Table I.
School environments, by Notes: FRL, participates in the Free-or-Reduced Price Lunch program; ELL, English language
district, 2006-2008 learners; AYP, Adequate Yearly Progress
the nature and consequences of this leadership work. The purpose of these semi- Internal and
structured interviews was to gather detailed descriptions of how different people in the external
school: brought attention to particular learning problems; generated and over time
reconsidered particular definitions of the problems to be solved; identified courses of accountability
action that would guide and support the improvement of teaching and learning; carried
out these courses of action; and learned from and about the results of these
improvement efforts. 679
We coupled interview data with evidence from actual interactions between leaders
and others. Specifically, researchers shadowed school leaders doing their work and
observed instruction in order to learn more about leadership practice in relation to the
context of the school, the learning climate, and the community the school served.

Data analysis
This nested, multi-case study of 15 schools and their school districts (Merriam, 2009;
Yin, 2005) utilized a grounded theory analysis approach (Strauss and Corbin, 1997),
married to a two-stage, within-case and cross-case analytic process (Miles and
Huberman, 1994). Debriefing reports were written on each school case from two
researchers’ perspectives throughout site visits and revised at the end of each site visit.
A formal, iterative analysis of interviews and observations was on-going during data
collection. Interviews and observations were transcribed and uploaded in qualitative
research software (Nvivo). Code sets were developed before coding began, and free
codes were added in some cases, through a modified “open” coding process, as
productive themes emerged from the coding process. All documents were base-coded
for state, school district, school type, and participant. An additional set of codes
were developed to reflect overall research questions and analytic categories that were
built into an emerging conceptual framework guiding the research. Axial coding
(Strauss and Corbin, 1997) was used to explore relationships between codes.
As explained in greater detail in Portin et al. (2009), the coding phase set in motion
an analytic process resulting in lengthy (60-80 pages) analytic memos concerning
each school. These memos offer a detailed description and analysis of each school case
with special emphasis on the nature of leadership work in the school along with
the main conditions and events associated with it. Site visitor pairs who had been
assigned to the schools in question created these accounts to reflect that state of
the school as accurately as possible. Following this within-site analytic work, we
undertook a cross-site analysis (as described in Miles and Huberman, 1994) by reading
all the school site memos by analytic category; this analysis produced emerging
patterns, possible hunches, and new categories or relationships that needed deeper
exploration.
Accountability was one such category. We did not go into this study with a
primary focus on accountability dynamics in or around the schools under study.
But accountability emerged as a central and unmistakable theme across study sites
that signaled a central pre-occupation of school leaders. Although these schools
were situated in different districts, working at different levels of the K-12 system,
the external accountability environment was remarkably similar across districts.
The attention to accountability prompted us to look more close at internal
accountability arrangements. Likewise, through a wide variety of learning
agendas, leadership configurations, and school and district arrangements of
discretion and support, we found strikingly similar internal accountability themes
that linked external accountability expectations with internal accountability
JEA practices, often quite tightly. In the following two sections, we detail the findings
50,5 that emerged from these accountability-focussed analyses, first, to create a portrait
of the convergence of external and internal accountability in leaders’ practice, and
second, to identify key consequences of this convergence for instruction and
instructional improvement.
Before laying out our findings, several limitations of this study design for the
680 analytic purposes undertaken here deserve mention. First, because we did not enter
the study with a well-developed framing of accountability dynamics, the framework
in the preceding pages and related findings emerged more inductively, and through
continued exploration of related literature as the study proceeded. Second, given the
study’s primary focus on leadership practice, we spent relatively little time in
classrooms, and cannot offer a definitive rendering of actual effects on classroom
instruction, though our findings are suggestive of impacts on the ways teachers and
administrators approached their instructional improvement work.

How external and internal accountabilities converged in leadership

practice within the study schools
Our analytic work revealed a common theme across a disparate set of elementary,
middle, and high schools and in districts that were strikingly different from one
another: external and internal accountability systems and logics converged in
remarkable ways. This convergence was managed and even prompted by school
leaders, especially the principal. Internally, the schools largely exhibited collective
forms of internal accountability, anchored to a school-wide learning improvement
agenda that school staff had for the most part bought into (and often helped to
develop), or were in the process of accepting and owning, though sometimes with
significant resistance. At the same time, the internal accountability system was
aligned in numerous ways with the expectations, procedures, and tools of the external
system, and in certain respects made good use of what the external system provided
and required.
We unfold this theme in several stages, showing, first, what school principals and
other leaders did to internalize external expectations while they were developing
practices that were more internally accountable within the school. Second, we
show how they made extensive use of the information and tools that the external
accountability system provided, to further the school’s own learning improvement
agenda. In so doing, they were leading through and with data. Finally, we underscore
the extensive and difficult learning that school leaders themselves needed to
undertake, often publicly, as they guided their schools and their colleagues toward
a more powerful outcome.

Internalizing accountability and developing accountable practice within the school

The school principals and some other key leaders in these schools saw in the external
systems ways to realize goals they had already committed to, or were developing. Both
rhetorically and in their own thinking about their work, school leaders took up the
main principles and expectations that external accountability systems asked of them.
Leaders and school staff have to internalize external expectations for performance –
and the notion that their practice is accountable – in order to produce the results
that are expected by their environment. This meant two things at least: first, the
school principal needed to own the expectations and make them his or her own
(with or without modifications). And then, these expectations needed to become
expectations that guide the school as a whole so that, as a collective, the school staff Internal and
internalized both the expectations and a sense of responsibility for meeting them in external
a demonstrable way.
Owning external expectations. Our study has relatively little light to shed on accountability
how school principals came to see accountability expectations as reasonable and
compatible with their own deepest commitments in education – or whether the
external expectations were instrumental in shaping those convictions. But some kind 681
of matching process appeared to be at work. School leaders who believed that students
should be able to perform well on annual assessments of math and literacy and who
saw the paramount importance of these academic skills to the students’ long-term
prospects were willing to work in the environments we were studying. Over time, those
who were not so willing were likely to leave or be removed.
We commonly heard from principals that, while this work was hard, it was the right
work to be doing, an expression of a moral accountability anchor. Some of the
principals had been waiting a long time for an official mandate that all students must
learn. One principal in New York City, who had worked her way up through the school
system from a lunch room para-professional to a classroom para-professional, then to a
teacher, a coach, assistant principal, and now a principal, deeply believed in the
capacity of every child to learn and to lead. One of her assistant principals had been a
student in her school and the other had followed a similar trajectory as she had
working her way up through the system. Her parent coordinator had been a parent in
the school and had begun learning to work in the school as a volunteer, eventually
working into a paid position on the school leadership team, with significant influence
on the school community. This sort of direct experience had taught this principal that
every student had the potential to be the next school leader, community leader,
community doctor, or community lawyer, and that the community needed every
student to be successful if it was going to realize its potential. For this principal,
internalizing high expectations had come long before the external accountability
system promoted them. As she had no doubt that all her students could meet high
expectations, her focus was on making sure her teachers were not creating obstacles
for students by expecting anything less than she expected. In this sense, she and other
principals in our data set were less driven by external demands, but rather married
their own convictions (and those of many of their staff ) to the external demands.
Spreading the word and spreading responsibility. Spreading this marriage of external
and internal expectations and the sense of responsibility for meeting them among a
school staff was done in various ways. Beyond the bully pulpit – simply using
positional authority to declare and reinforce the importance of meeting accountability
targets – the following three were common avenues for principals in the schools we
studied to communicate expectations and encourage staff to own them:
(1) creating incentives for assuming and demonstrating desirable practice;
(2) using professional development events as a mechanism for internalizing
expectations, among other agendas; and
(3) communicating clear and ambitious performance expectations through
supervision and other one-on-one interactions with teaching staff.
First, the school leaders created tangible and intangible incentives for assuming and
demonstrating desirable practice. School principals had various resources to offer their
school staffs, as an enticement or reward for improving practice in ways that conform
JEA to accountability expectations. These resources were varied, and were likely to have
50,5 local meaning and potency in relation to the particular staff members’ needs: time,
teaching assignments, funds, access to desired equipment, trips to other sites engaged
in related work, scheduling adjustments, and team configurations were among the
tools that the school principals in our study were using to encourage their staffs to
orient their work toward accountability expectations. But beyond these fairly
682 straightforward steps, the leaders generated more intangible incentives through peer
pressure, by creating numerous occasions in which staff talked about and shared their
work in ways that others could see and react to. Over time, these less “private”
dimensions of practice could have the effect of motivating staff to reconsider their
practice and display their progress toward improvement goals.
Second, school leaders used professional development as a mechanism for
internalizing the idea of being accountable. Recognizing that being accountable
for progress toward improvement targets set by the larger system – or being
accountable for realizing the school’s own, internalized learning improvement agenda –
generally means learning better ways of teaching (and leading), school leaders were
naturally likely to use various forms of school-based professional development to
communicate the relevant reform messages. All of the principals we studied, for
example, seemed to believe that their teachers needed to learn new ways to meet
higher instructional expectations. These leaders’ operative theory of action was: teach
the staff what you want them to know and do, and then expect them to demonstrate
what they have learned. Whether the principals encouraged staff to participate in
district-wide professional development (where this function was centralized), created
school-specific professional development events (in more decentralized settings), or
took on the professional development role themselves, their intention was similar:
engage their staffs in a continuing conversation about instructional practice that was
more likely to produce the desired results, along with developing more specific pictures
of what those results looked like in practice. One Atlanta principal did it through one-
on-one coaching, using video technology:
The most important thing that I do is I go inside the classrooms constantly, every day – every
day – and assess what’s going on. And what I find is that some teachers don’t really realize
that they’re not doing what they need to do to support children and to affect student success
in terms of learning. So for the last couple of years I went in with a video camera. Of course
I was courteous. And then we have a conversation – one-on-one conversation about what
I saw. We talk about the best practices in terms of what I should see. And first I give the
teacher an opportunity to talk about how do you think the lesson went? What do you think
you did to support children learning? Did you bring closure? Review what you taught them?
Did you have a smart objective on the board? Did the children really know what they were
supposed to do and learn before they left that particular lesson? And a lot of teachers say
“yes, they knew,” but then when I show them the video [y] It takes a lot of time, but I tell you
it is so worth it. Because teachers don’t realize that they’re not doing what they need to do to
make children learn and to have children learn. And so when they see that, it’s like a rude
awakening.
Finally, school leaders communicated clear and ambitious performance expectations
through one-on-one supervision and other interactions. This leadership approach to
establishing internal accountability was seen in all the schools we studied to some
extent, and in some schools this was a major leadership activity. In one such school, the
year started off with a display board in the teachers’ lounge, on which the previous
spring’s test scores were displayed, disaggregated by classroom. The principal had
built a culture in the school in which teachers welcomed the data and the clear message Internal and
it conveyed about the work to be done. The principal noted: “[y] they look forward to external
this data. It’s like even though your name is going to be up on the board [y]. you’re not
going to look at your data in isolation. And I think that is the biggest thing to help [y] accountability
Everybody says, ‘Oh wow – I can’t hide.’” The ensuing year’s supervisory work took
the board display as a reference point.
Observing, coaching, documenting, and evaluating teacher performance is time 683
consuming work for principals that take this approach to helping the school internalize
the aspiration for accountable practice, but it was often done through other members
of the school’s instructional leadership team. However, a “bottom line” accountability
rested with the principal and other staff who had supervisory authority; they needed
to make a judgment regarding whether staff were meeting expectations sufficiently
well to remain at the school. One principal in New York City described her work
“evaluating-out” ineffective teachers as a distraction from the work she wished she
could be doing supporting excellent teachers, yet the fact that she did so sent a broader
message to all staff that accountability was serious business and the consequences
were real. Principals who took this approach were able to improve their staff
effectiveness especially when they had discretion over the new hires.
In summary, an important step in meeting external expectations, and at the same
time developing school-wide collective expectations, is finding a way to internalize
those expectations among the staff. While principals in our study were primarily
responsible for meeting accountability expectations, all three of the approaches above,
in effect, spread responsibility for accountability across all the staff. There is not one
way to accomplish this internalization process, and whichever approach a principal
takes will consume much time and attention, but not necessarily new or additional
resources.

Leading through data and accountability tools

Data of various kinds became a central part of the way the school leaders we
studied merged external and internal accountability. In short, data were a constant
accompaniment to leadership work, and a medium for conversation about instructional
improvement, if not for thinking about the improvement work to begin with. To this end,
the leaders themselves paid regular attention to data of all kinds, helped establish formal
or informal “data systems” within the school, and made regular use of externally
provided data, as well as external assistance for data use, where available.
The variety of data within play in the schools we studied was striking. In addition to
the ever-present student test score data (e.g. from last year’s annual assessment, from
periodic formative assessments undertaken across the year), the school leaders were
considering environmental survey results, observational data generated by themselves
(e.g. through walk-throughs) or other staff (e.g. from external visitations), students’
work of various kinds, and the results of local school-based inquiry into particular
problems of practice (e.g. from interviewing teachers and students about what they
were doing in this or that subject, looking at grade trends or patterns of attendance and
other school-based measures).
All of these practices for generating and interpreting data were mechanisms for
staff to internalize both school-wide and external expectations for their work, and
provided a way to make these expectations concrete. On the less formal end of
the spectrum of data use, feedback of various kinds pinpointed specific aspects
of practice to work on, while reinforcing a sense of responsibility for addressing
JEA these improvement targets, as in the video feedback example noted above. On the more
50,5 formal end, fairly elaborate accountability systems, generally set up outside the school,
became useful tools for leadership work within the school. Consider, for example, the
set of accountability tools developed in New York City and required of all schools:
annual assessments, periodic assessments (testing every six weeks to see how
and whether students were making progress), an annual environmental survey
684 (of parents, teachers, students), a school quality review by an external visiting team,
and the findings of a school-based “inquiry-team” process. All of these data were
simultaneously available to the school for its own improvement work, and fed into
the school’s annual progress report, a public document that characterized the quality of
the school’s work and gave it a grade (from A to F). A high school principal in this
system described how the school quality review process fed her own sense of progress:
[The School Quality Review] does a lot. I mean it’s a one and a half day visit. So again, it’s as
bad as a test in the sense that it’s a real snapshot. But I like the process of it because really
they’re judging you against your own evaluation of the school. So I mean what a quality
review judges in my opinion is, how well does the school know itself and how well does
a school know what it needs to do next, and is the school reflective of that? And so, what the
process did for me the last 2 years – it was very affirming in the sense of a lot of the stuff that
I put down the reviewer saw [y] the review itself is thorough enough that you can’t fake it
totally and I’m proud that the last year’s, both the reviewers, the feedback I got has been like,
“You really know the school well, you’ve analyzed the problems,” and a lot of the next steps
that they’re giving me are next steps that I’ve already thought of or already articulated. And
so they’re just – it gives me faith that I’m going in the right direction as a school leader and
that I’m making the right decisions and that we are growing in the right way.
This principal and others confirmed that the thorough, on-site process, while nerve-
racking, got at what was really happening at the school, and both validated practice as
well as highlighting specific aspects that could be improved. As another principal put
it, “You cannot put on a show for it.”
Data have the potential to significantly enhance accountability of both teaching
and leadership practice. Implied or explicitly required by most school district’s
accountability systems is regular use of data in the school by teachers and others who
are helping them improve their practice. To this end, school leaders found they needed
to lead through data – that is, in ways that made data a medium of the daily
transactions between them and other staff. This meant finding new ways to generate
data, collecting various forms of data and evidence more regularly, analyzing data
periodically and referencing the analysis in interactions with teachers and other
staff, and using data routinely to make a wide variety of decisions (e.g. regarding
resource allocation, student grouping, staff configurations, instructional strategies, and
strategic goal setting). In effect, data and its regular use by teachers and administrators
provided a basis for professional forms of accountability within the school, at the same
time that some of the data (e.g. annual test scores) figured prominently in the more
bureaucratic forms of accountability set up by the district and state.
The districts in which the school leaders worked varied in the comprehensiveness,
timeliness, accessibility, and “user friendliness” of data available for school use, and the
district central office was more or less helpful in building school capacity to make use
of the data. Accordingly, whatever their initial levels of comfort and expertise with
data, school leaders operating in these demanding accountability environments often
sought out assistance in whatever way they could find it – among them, through
coursework, strategic hiring of staff who were data-savvy, designating more expert
staff as “data specialists” for the school, or regular interaction with available experts Internal and
from the central office or in outside organizations. external
Across the variation in data practices one thing was clear: these principals
appreciated having data to point objectively to the gaps in achievement between accountability
different groups within their schools, or between their student populations and other
more advantaged ones in the larger community. Even if the data reflected poorly on
their own practice, it was important to have it. Data were the tools that allowed 685
principals to say Yes, there is a problem and, Yes, we can and will solve this problem, if
we pay attention to it.

Learning to lead for learning improvement in a fully accountable way

It is a challenge for a principal to both lead – and at the same time learn to lead – in
accountable ways, and yet that is what we found happening in the schools we studied.
The demands of the external accountability environment, coupled with the internal
impulse to engage in improvement conversations, had direct implications for these
principals’ learning – in particular, their capacity to lead in ways that model and
embody accountable practice.
In some districts we found principals learning fairly simple things, such as how to
use Excel, how to file new reports, how to engage with a new school review process, or
how to work with a new central office network design. In other cases, we found
principals:
. learning complicated new instructional strategies alongside their teachers, and
more sophisticated ways of viewing instruction, guided by instructional
frameworks and other tools;
. learning to work more collaboratively where they had worked in isolation, and
especially learning to coach instructional teams; and
. learning to interact with new data systems, and to some extent to construct data
systems of their own that would serve the purposes of learning improvement in
the school.

Across these instances, they were learning in greater depth what the school was being
held accountable for, how they and the school staff would be held accountable (to each
other or to outside audiences), and how they might seek to promote more accountable
practice (see Portin et al., 2009, for a more extended discussion of the new learning
these leaders were doing). As they did so, they were internalizing a way of working
that embedded a more explicit form of accountability into their daily work, and by
extension into the practice of others around them. And they did so publicly, in full view
of their staffs, often intentionally modeling their learning to help others embrace a
focus on learning improvement.
The process hinged, in part, on the facts of employment for these principals:
their jobs often depended on their performance. In several systems, this matter was
enshrined in their contracts. Absent evidence in two to three years’ time of the school’s
progress, they were at serious risk of losing their jobs. School staff were aware of this,
and often assumed a kind of mutual responsibility for the leader’s and their collective
performance, as one technology coach in a New York City elementary school put it:
[Becoming an Empowerment School] puts more pressure on [the principal] because if we
mess up, he messed up and then he’s fired. So that’s really the empowerment [system]. When
JEA we turned empowerment, we realized that the risk is all him actually – or us. If we don’t work
hard enough, we’re going to lose him, but it’s more pressure on him than it is on us.
50,5
But it would be a mistake to assume that principals’ learning to lead in accountable
ways was mainly a matter of extrinsic motivation. They generally welcomed
the circumstances they found themselves in, and assumed they would and should be
able to show results, as a matter of good practice.
686 And by making their learning public among their staffs, the principals we studied
were modeling what it meant to strive for a more professionally accountable practice.
In New York City, for example, the district-wide reform required that principals
develop inquiry teams in their schools to continuously examine the instructional
practices that were creating obstacles to student learning for groups of struggling
students that the school staff identified as targets for improvement. Learning to lead
this work and, through it, attempting to transform their schools into inquiring
communities meant that the teacher leaders and administrators who were part of the
inquiry team had to open up their own practice and the instructional practices in the
classroom to inquiry. In Norwalk-La Mirada, we found principals collaboratively
learning to observe instruction and give specific and constructive feedback to
a teacher. This was highly revealing of both the current instructional practices in the
schools and their own expertise as an instructional leader. Principals worked with
skilled coaches who helped them improve their instructional leadership skills, while
collectively learning about their weaknesses in this area and how to address them.
The principals we studied were not isolated. We found them in classrooms, in
various staff gatherings leading teams of other school leaders, and in other sessions
working with principals from other schools. The majority of this work was public
learning. In Norwalk-La Mirada, principals met in job-alike cadres facilitated by
a leadership coach, to experience facilitation strategies they were then expected to
introduce to their own leadership teams in their schools. The coach and often other
cadre members observed as they attempted to use the new strategies, making
their learning public to their staff, other principals in the districts, and the leadership
coach. A principal in New York explained that, while she could have participated in
a professional development session specially designed for busy principals to quickly
become familiar with what their teachers were learning, she choose instead to participate
alongside her teachers:
When I took the summer institutes, I took them as a teacher, not as a principal, because what
they give you as a principal is not as useful for me [y]. I think that’s more helpful for me,
when I go and observe someone, if I haven’t done it, I’m not really going to know what I’m
looking for. I’m also not giving the teacher the opportunity to say, “You didn’t do it, so how do
you know that I’m not doing it right?” They know I know how to teach the writing and the
reading [y], so that’s not a problem because I’ve done it in the classroom before. That’s
another thing that you have to show teachers that you can do – teach the same lessons.
This approach highlights multiple uses of learning as a source of accountability:
both principal and teachers learned together new pedagogical practices; teachers
learned that their principal knew what they had learned and what changes to expect
in their practice; and together they were learning what instruction should be able
to accomplish with students. In these ways, this leadership work was strengthening
the school’s internal accountability system – by bolstering collective expectations
and installing a mechanism for exposing practice to scrutiny – while at the same time
aligning practice with external expectations.
 While it could be argued that principals and other school leaders have Internal and
always learned on the job, the difference in the case of learning-focussed external
leadership within strict accountability environments such as those we studied is
that the work is often uncharted and unfamiliar. In this context, significant stakes accountability
are attached to one’s capacity to grasp and actualize this work, and the effort to
lead in an explicitly accountable way can be turned into a teachable moment,
an opportunity to model accountable (leadership) practice in ways that help 687
solidify the school’s internal accountability system, if not the larger system
as well.

Consequences of leadership in the “convergence zone”

School leaders’ efforts to facilitate and shape the convergence of external and internal
accountability in the schools we studied had identifiable consequences for teaching
practice and for efforts to improve it. Two stand out – first, the reshaping of the scope
of instruction itself and the conversation about improving it, and, second, the idea that
both teaching and leadership practices are, and need to be, public.

Reshaping the scope of instruction and instructional improvement conversation

By positioning leadership work within the intersection of external and internal
accountability, school leaders were participating in and contributing to a different
conception of teaching itself and, therefore, of the discourse within the school about
improving practice. We have discussed these changes more extensively elsewhere (see
Knapp et al., 2013; Portin et al., 2009), but, in brief, they concerned the language for
talking about student learning and student progress; the scope of the curriculum; and
the ways in which instruction could be productively differentiated to meet the needs of
a diverse student population. Each of these changes was anchored simultaneously to
external accountability pressures (as interpreted and internalized by school leaders),
and to the school’s internal accountability system.
Across the sites in this study there was a striking similarity in the way
principals talked about the teaching and learning work in their schools. They
were “finding and filling gaps,” “making gains,” “hitting or missing targets,” or
“monitoring progress to the benchmarks,” and pegging students (e.g. as “level ones”
and “level twos”) to the levels of proficiency set up by state and local assessments.
The deep concern was to “move” students, as a grade-level team leader in Atlanta
commented:
The main focus, I think, is definitely just to move students. I mean even if they’re in third
grade, if they came to you reading at a second-grade level, by the end of the year, of course,
you’re trying to get them up to the level because they still have to take the third-grade test.
But just being able to have data to document the fact that they are moving, whether or not it’s
at the pace that the next student is moving, just so long as you’re able to show growth in the
student.
A thousand miles away, in New York City, the director of a district-wide school
leadership preparation program had similar words for talking about the essential work
facing school principals in that system:

We have a very specific vision [of good school leadership.] At its essence, it’s the focus on
moving student learning. And so if the outcome is moving student learning, then what [y]
does a leader of the school need to be able to understand, [to know] how to assess a student
and where their learning gaps are, in order to help teachers do that work? So [our leadership
JEA standards] map back from the student learning needs to the adult learning needs, and then
they’re organized for those adult learning needs [y]:
50,5
This language was not common ten years ago, and it seems to reflect a commitment to
assessment as a source for understanding student progress and school success, and
even as a redefinition of student learning goals. In obvious ways, this commitment was
formalized in the design of external accountability systems to which these schools
688 were responding, but in less obvious ways, the idea of helping students progress
through defined stages of mastery in essential subjects was a deeply held value among
the professional staff of the schools, and one that they felt some responsibility for
mastering.
In their single-minded attempt to “move” student learning, the combined efforts of
principals and other members of each school’s instructional leadership cadre were
simultaneously focussing and narrowing the curriculum that students were taught. In
some instances, if it was not on the test, it was not taught. In other instances, learning
to take the tests became almost a content area itself: various forms of preparation for
tests appeared in all the schools we studied, in before- and after-school programs, on
Saturdays, and in a variety of ways integrated into instruction itself. To be sure,
this meant de-emphasizing important things in the curriculum, or leaving them
out altogether, as many prominent critics of high-stakes accountability have asserted
(e.g. McNeil, 2000; Kohn, 2000; Amrein and Berliner, 2002; Sirotnik, 2004), but there
was little question across the schools we studied about the paramount importance of
helping students master those subjects that were tested. If there were doubts, they lay
in the worry, articulated by one principal, that the tested curriculum would somehow
not capture the “thinking curriculum” that his students desperately needed to further
their lives, educational options, and careers.
Additionally, a common interest in “differentiated instruction” across the schools
was a natural part of efforts to realize a more accountable practice, in which the goal
was to move students – all students – as fast and as far as they could go in pursuit of
ambitious learning standards ensconced in the external accountability system as well
as the school’s own learning improvement agenda. What this meant varied across the
schools, but the common idea was that in order to get all students to the same place by
the end-of-the-year test, teachers would need to do different things for different
students. In one school district, differentiation went as far as not promoting a student to
the next grade to give them more time to meet the grade-level expectations in their
current grade. In another school, differentiation was created by tracking students into
different classes so the students who were ahead of grade-level expectations could
accelerate their instruction and the students who were below grade level could have
intensified instruction. Some schools offered “double doses” of reading or math to
struggling students, arranged either to pre-teach them the material that was coming up
in the following week or to review with them what had been taught in the previous
week. While schools varied in their approach to differentiation, the variation seemed to
revolve around a limited set of possibilities, such as homogeneous v. heterogeneous
grouping of students; pacing of instruction; and intensification of instruction.

Making teaching and leadership practice more public

Inevitably, leadership practices that merged external and internal accountability
systems brought the daily practice of teachers and school leaders more squarely into
view of others in the school and even beyond it. We have already noted the leadership
strategy of assuming a public learning stance, and various ways in which leaders’ Internal and
efforts to guide, direct, and support school staff were inherently more open to scrutiny external
in a high-stakes accountability environment.
But in a more basic way, internalizing external accountability expectations, in accountability
alignment with an increased and more collective internal accountability, meant that
daily practice was not left up to individual professionals behind closed doors (e.g. of the
classroom, school office, or school building). A veteran high school teacher in Norwalk- 689
La Mirada spoke for many as he described how things were different at present:
[y] Now times have changed. Anybody can come in at any time. Your classroom should be
open – you should be in a fishbowl, basically, and showing what you’re doing at all times.
And kids should be monitored for whether or not they’re actually getting the information. So
I think it’s just a different time. And I think accountability is – can be – a negative word. But
we are accountable, just like somebody in an office [y].
He and a number of his colleagues had been somewhat resistant to the changes
introduced by the current principal and the district-wide literacy improvement
initiative, which touched all classes and subject areas in the school, but he had come to
a new view of his teaching as accountable practice. A similar sentiment was voiced by
school principals, as they embraced what it meant to be regularly scrutinized through
various forms of data about the school, as in the School Quality Review case noted
above. Their practice had changed, and they were embracing the changes.

Conclusions and interpretations

Stepping back from our findings, several observations and continuing questions are in
order. First of all, in these schools, internal accountability systems were in place that
asserted a strong and reinforced sense of collective responsibility for the school’s
performance. Second, school leaders, especially the principal, were actively internalizing
(or had already internalized) external expectations for the school and were spreading the
word among school staff, though often with adjustments to the accountability messages
from the larger system, to make them fit with the school’s own learning improvement
agenda. The net effect was a merger of external and internal accountability systems, in
ways that have often been thought to be unlikely (Adams and Kirst, 1999).
It is instructive to return to the findings and framing of the seminal research on
internal school accountability, which drew on data that preceded the No Child Left Behind
era and in contexts with relatively weak external accountability controls (Abelmann et al.,
1999; Elmore, 2006). Across this sample of schools, the dominant pattern was that:
Teachers and principals often dealt with the demands of formal external accountability
structures (curriculum guidance, testing, and the like) either by incorporating them in
superficial ways – claiming, for example, that they were consistent with existing practice
when they clearly were not – or by rejecting them as unrealistic for the type of students they
served [y] (Elmore, 2006, p. 196).
While noting that some schools which had a more collective approach to internal
accountability had deliberated about how they might align with external accountability
systems, or had even done so:

[y] in most cases teachers and principals viewed external accountability systems like
the weather – something that might affect their daily lives in some way, something they
could protect themselves against, but not something they could or should do much about
(Elmore, 2006, p. 196).
JEA Our research paints a different picture, and the picture sits in a backdrop that has
50,5 substantially changed the conditions of schooling and is exerting more consequential
demands on the schools. Under such conditions, the schools we studied appear to
demonstrate in various ways a merger of different logics of accountability, as well
as the specific routines, tools, and expectations that are associated with each. In
specific terms, they appear to have found a way to merge an externally driven logic,
690 reflecting management, bureaucratic, and political accountabilities, with one that is
more professionally driven, and anchored to patterns of professional and moral
accountability. Doing so means embracing somewhat contradictory forces and
conditions, and it is noteworthy, even remarkable, that this appears to be happening
in a productive fashion.
Our findings in no way suggest that the convergence of external and internal
accountability is happening in large numbers of schools, or is even possible in all. The
schools we studied, after all, were not typical of urban schools in the USA, nor were
they struggling in ways that so many of their counterparts in urban systems in this
country or elsewhere are, under what is experienced as a “yoke” of high-stakes
accountability. They were not “schools on probation” (Mintrop, 2004a, b) or otherwise
designated as cases of “persistent failure.” But as schools that were making progress in
difficult circumstances, they do offer images of possibility – ways to visualize what
it might mean to reconcile what the larger educational system is asking for and
what school-based professionals want and need, to make their work fulfilling and
productive. These images have special relevance given the fact that most of the schools
in our sample had been, in recent years (e.g. from two to six years back), a “struggling
school,” some on the usual lists of needing improvement or at risk of more serious
interventions. Therefore, the way they appeared at the time of our study was a stage in
an evolutionary process that placed then squarely within the mainstream of conditions,
challenges, and constraints that most urban schools face.
What we have described and concluded about the leadership work in these schools
and its consequences leave open various questions which deserve mention and
further exploration. First, in these schools, progress was already being demonstrated,
by the time we arrived; in other words, they were already meeting accountability
expectations, and would therefore be receiving the approval of system leaders. What if
they were not? How would schools start to make progress in situations of chronic low
performance combined with the pervasive demoralization of staff in many urban
schools? (Payne, 2008). Second, at what cost do leaders in the kinds of schools we
studied embrace the accountability processes we are describing? Are the consequences
for instructional practice encouraged by these approaches to leadership the right or
best consequences for young people’s learning? Put another way, do the benefits of
increased focus outweigh the narrowing of the curriculum on which so many critics
have concentrated? Third, what does it mean to model accountable leadership practice,
and what images can we develop of how this can be done successfully, across a range of
school settings beyond what we were able to study? Fourth, what sorts of school
leadership preparation programs and experiences are most likely to push aspiring
leaders to engage accountability environments productively and proactively, rather
then reactively and defensively? Fifth, what kinds of ongoing leadership support
systems will enhance the accountability of leadership practice, while at the same time
enhancing the learning that must accompany such practice? Finally, are there
alternative accountability designs at the system level that would strike a better balance
between pressure for performance and support for improvement?
 These questions deserve continuing inquiry of various kinds, as the field tries to Internal and
maintain the benefits of increasing the accountability of educators’ work, while external
minimizing the counterproductive or dysfunctional aspects of demanding accountability
systems. These questions would also benefit from cross-national evidence and accountability
theorizing, drawing on the now substantial array of cases across the world in which
national policies address the persistent failure of a subset of the nation’s schools through
high-stakes accountability measures. The ultimate goal is more accountable practice at 691
the level of student learning, and we have much still to learn about how to accomplish
this goal, across a wide range of educational settings.

Acknowledgments
This article is adapted from a longer report: “Leadership for Learning Improvement in
Urban Schools” (Portin, B.S., Knapp, M.S., Dareff, S., Feldman, S., Russell, F., Samuelson, C.
and Yeh, T.L. (2009), Center for the Study of Teaching & Policy, University of
Washington, Seattle, WA) carried out with support from The Wallace Foundation. The
authors acknowledge the contributions of other research team members to the
analyses and conclusions reported here, which reflect those of the authors and not
necessarily The Foundation. See Knapp, M.S., Copland, M.A., Honig, M.H., Plecki, M.L.
and Portin, B.S., “Learning-focused Leadership and Leadership Support: Meaning and
Practice In Urban Systems” (Center of the Study of Teaching & Policy, University of
Washington, Seattle WA: 2010) for a brief summary of all strands of the larger
investigation of which this study was a part. All reports from this study appear on the
CTP web site (www.ctpweb.org) and the Wallace Foundation’s Knowledge Center web
site (www.wallacefoundation.org).
Note
1. For example, see Glickman’s Leadership for Learning (2002), that focuses primarily on the
direct guidance that school principals (or others) offer their teaching staff; Schlechty’s
Leading for Learning (2009), which concentrates instead on how schools can be transformed
into learning organizations; and Learner-centered Leadership (A.B. Danzig, K.M. Bormann,
B.A. Jones, and W.F. Wright, Mahwah NJ: Lawrence Erlbaum, 2009), which emphasizes
leadership training approaches that foster learning communities. While these latter works
do share some resemblances with our own, they were not central to the development of our
thinking.

References
Abelmann, C., Elmore, R., Even, J., Kenyon, S. and Marshall, J. (1999), When Accountability
Knocks, Will Anyone Answer? CPRE Research Series, University of Pennsylvania Graduate
School of Education, Philadelphia, PA.
Adams, J.E. and Kirst, M.W. (1999), “New demands and concepts for educational accountability:
striving for results in an era of excellence”, in Murphy, J. and Seashore, K. L. (Eds), New
Demands and Concepts for Educational Accountability: Striving for Results in an Era of
Excellence, Jossey Bass, San Francisco, CA, pp. 463-90.
Amrein, A. and Berliner, D. (2002), “High-stakes testing, uncertainty and student learning”,
Education Policy Analysis Archives, Vol. 10 No. 1, pp. 950-95.
Beadie, N. (2004), “Moral errors and strategic mistakes”, in Sirotnik, K. (Ed.), Holding
Accountability Accountable, Teachers College Press, New York, NY, pp. 35-50.
Carnoy, M., Elmore, R. and Sisken, L. (2003), The New Accountability: High Schools and High
Stakes Testing, RoutledgeFalmer, New York, NY.
JEA Cohen, D., Moffet, S. and Goldin, S. (2007), “Policy and practice”, in Fuhrman, S., Cohen, D. and
Mosher, F. (Eds), The State of Education Policy Research, Lawrence Erlbaum, Mahwah, NJ,
50,5 pp. 63-86.
Cuban, L. (2004), “Looking through the rearview mirror”, in Sirotnik, K. (Ed.), Holding
Accountability Accountable, Teachers College Press, New York, NY, pp. 18-34.
Danzig, A.B., Bormann, K.M., Jones, B.A. and Wright, W.F. (2009), Learner-centered Leadership,
Lawrence Erlbaum, Mahwah NJ.
692
Darling-Hammond, L. (2007), “Race, inequality and educational accountability: the irony of
‘No child left behind’ ”, Race, Ethnicity and Education, Vol. 10 No. 3, pp. 245-60.
Diamond, J. and Spillane, J. (2002), High Stakes Accountability in Urban Elementary Schools:
Challenging or Reproducing Inequality? Institute for Policy Research, Evanston, IL.
Elmore, R. (2000), Building a New Structure for School Leadership, Shanker Institute,
New York, NY, pp. 274-96.
Elmore, R. (2004), “Conclusion: the problem of stakes in performance-based accountability
systems”, in Fuhrman, S. and Elmore, R. (Eds), Redesigning Accountability Systems for
Education, Teachers College Press, New York, NY.
Elmore, R. (2006), School Reform from the Inside Out, Harvard Education Press, Cambridge, MA.
Elmore, R., Abelmann, C. and Fuhrman, S. (1996), “The new accountability in state education
reform: from process to performance”, in Ladd, H. (Ed.), Holding Schools Accountable,
Brookings Institution, Washington, DC, pp. 65-98.
Firestone, W. and Shipps, D. (2005), “How do leaders interpret conflicting accountabilities to
improve student learning?”, in Firestone, W. and Riehl, C. (Eds), New Agenda for Research
on Educational Leadership, Teachers College Press, New York, NY, pp. 81-100.
Glickman, C. (2002), Leadership for Learning, Association for Supervision and Curriculum
Development, Alexandria, VA.
Goertz, M., Duffy, M. and Le Floch, K. (2001), Assessment and Accountability Systems in the
50 States 1999-2000, CPRE Research Reports, Consortium for Policy Research in
Education, Philadelphia, PA.
Ingram, D., Seashore, K.L. and Schroeder, R. (2004), “Accountability policies and teacher decision
making: barriers to the use of data to improve practice”, Teachers College Record, Vol. 106
No. 6, pp. 1258-87.
Knapp, M.S. and Copland, M. (2006), Connecting Leadership with Learning: A Framework for
Reflection, Planning, and Action, Association for Supervision & Curriculum Development,
Alexandria, VA.
Knapp, M.S., Copland, M. and Talbert, J. (2003), Leading for Learning: Reflective Tools for School
and District Leaders, Center for the Study of Teaching and Policy, Seattle, WA.
Knapp, M.S., Feldman, S. and Yeh, T.L. (2013), “Learning-focused leadership in urban high
schools: response to demanding environments”, Journal of School Leadership, Vol. 23,
forthcoming.
Kohn, A. (2000), The Case Against Standardized Tests: Raising the Scores, Ruining Schools,
Heinemann, Portsmouth.
Leithwood, K. (2005), “Accountable schools and the leadership they need”, International
Handbook of Educational Policy, Springer, Dordrecht, pp. 439-56.
Leithwood, K., Edge, K. and Jantzi, D. (1999), Education Accountability: The State of the Art,
Berelsmann Foundation, Gutersloh.
McNeil, L. (2000), Contradictions of School Reform: Educational Costs of Standardized Testing,
Routledge, New York, NY.
Macbeath, J. (2008), “Shared accountability”, in Macbeath, J. and Dempster, N. (Eds), Connecting
Leadership and Learning: Principles for Practice, Routledge, London and New York, NY,
pp. 137-56.
Merriam, S. (2009), Quantitive Research: A Guide to Design and Implementation, Jossey-Bass, Internal and
San Francisco, CA.
external
Miles, M. and Huberman, M. (1994), Qualitative Data Analysis, SAGE Publications,
Beverly Hills, CA. accountability
Mintrop, H. (2004a), Schools on Probation: How Accountability Works (and Doesn’t), Teachers
College Press, New York, NY.
Mintrop, H. (Ed.) (2004b), “Pressure of sanctions”, Schools on Probation: How Accountability 693
Works (and Doesn’t), Teachers College Press, New York, NY, pp. 19-35.
Murphy, J., Elliott, S.N., Goldring, E. and Porter, A.C. (2006), Learning-Centered Leadership:
A Conceptual Foundation, The Wallace Foundation, New York, NY.
Noguera, P. (2004), “Standards for what? Accountability for whom? Rethinking standards-based
reform in public education”, in Sirotnik, K. (Ed.), Holding Accountability Accountable,
Teachers College Press, New York, NY, pp. 66-81.
O’Day, J. (2002), “Complexity, accountability and school improvement”, Harvard Educational
Review, Vol. 72 No. 3, pp. 293-329.
Patton, M.Q. (2002), Qualitative Research and Evaluation Methods, SAGE Publications,
Thousand Oaks, CA.
Payne, C. (2008), So Much Reform So Little Change: The Persistence of Failure in Urban Schools,
Harvard Education Press, Cambridge, MA.
Phelps, J. and Addonizio, M. (2006), “How much do schools matter? A production
function approach to school accountability”, Educational Considerations, Vol. 33 No. 2,
pp. 51-62.
Portin, B., DeArmond, M., Gundlach, L. and Schneider, P. (2003), Making Sense of Leading
Schools: A National Study of the Principalship, Center on Reinventing Public Schools,
University of Washington, Seattle, WA.
Portin, B., Knapp, M.S., Dareff, S., Feldman, S., Russell, F., Samuelson, C. and Yeh, T.L. (2009),
Leadership for Learning Improvement in Urban Schools, Center for the Study of Teaching
and Policy, University of Washington, Seattle, WA.
Ranson, S. (2003), “Public accountability in the age of neo-liberal governance”, Journal of
Education Policy, Vol. 18 No. 5, pp. 459-80.
Resnick, L. and Glennan, T. (2002), “Leadership for learning: a theory of action for urban
school districts”, in Hightower, A., Knapp, M.S., Marsh, J. and McLaughlin, M.W. (Eds),
School Districts and Instructional Renewal, Teachers College Press, New York, NY,
pp. 160-72.
Schlechty, P. (2009), Leading for Learning, Jossey-Bass, San Francisco, CA.
Sirotnik, K. (2004), Holding Accountability Accountable, Teachers College Press, New York, NY.
Skrla, L. and Scheurich, J. (2004), Educational Equity and Accountability, RoutledgeFalmer,
London and New York, NY.
Stoll, L., Fink, D. and Earl, L. (2003), It’s About Learning (And it’s About Time): What’s in it for
Schools? RoutledgeFalmer, London and New York, NY.
Strauss, A. and Corbin, J. (1997), Grounded Theory in Practice, Corwin Press, Thousand Oaks, CA.
Sunderman, G., Kim, J. and Orfield, G. (2005), NCLB Meets School Realities, Corwin Press,
Thousand Oaks, CA.
Swaffield, S. and Macbeath, J. (2009), “Leadership for learning”, in Macbeath, J. and Dempster, N.
(Eds), Connecting Leadership and Learning: Principles for Practice, Routledge, London and
New York, NY, pp. 32-52.
Yin, R. (2005), Case Study Research: Design and Methods, Corwin Press, Thousand Oaks, CA.
JEA Further reading
50,5 Cibulka, J. (1999), “Moving toward an accountability system of K-12 education: alternative
approaches and challenges”, in Cizek, G. (Ed.), Handbook of Educational Policy, Academic
Press, New York, pp. 184-210.
Cobb, C. and Rallis, S. (2008), “District response to NCLB: where is the justice?”, Leadership and
Policy in Schools, Vol. 7 No. 2, pp. 178-201.
694 Elmore, R. and Fuhrman, S. (2001), “Research finds the false assumption of accountability”,
Education Digest, Vol. 67, pp. 9-14.
Elmore, R. and Fuhrman, S. (2004), Redesigning Accountability Systems for Education, Teachers
College Press, New York, NY.
Macbeath, J. and Dempster, N. (2008), Connecting Leadership and Learning: Principles for
Practice, Routledge, London and New York, NY.
Mintrop, H. (2010), “Bridging accountability obligations, professional values and students needs
with integrity and succeeding”, paper presented at the Annual Meeting of the American
Education Research Association, New Orleans, LA.
Mintrop, H., Curtis, K. and Plut-Pregelj, L. (2004c), “Schools moving toward improvement”,
in Mintrop, H. (Ed.), Schools on Probation: How Accountability Works (and Doesn’t),
Teachers College Press, New York, NY, pp. 69-88.

About the authors

Michael S. Knapp, a Professor of Educational Leadership and Policy Studies and Director of the
Center for the Study of Teaching and Policy in the University of Washington College of
Education, focuses his teaching and research on educational leadership and policymaking,
school and school system reform, the professional learning of teachers and administrators, and
methods of inquiry and policy analysis. In recent years, his work has probed the meaning
and forms of “learning-focused leadership” in schools, districts, and state systems of education.
While considering various settings and applications, his work pays special attention to the
education of disenfranchised populations, mathematics and science education, and the
professional development of educators. He has written extensively about his research, including
eight books, among them, School Districts and Instructional Renewal (2002), Self-Reflective
Renewal in Schools (2003) and Connecting Leadership with Learning: A Framework for Reflection,
Planning, and Action (2006). Michael S. Knapp is the corresponding author and can be contacted
at: [email protected]
Susan B. Feldman, an education researcher and the Director of the Research Center for
Learning and Leadership at Education Service District 112, in Vancouver, Washington, works on
pressing and persistent problems or practice for school leaders. Recently her work has focused
on how leaders and teachers learn to do the work that is expected of them when that work is new
to the system. A new teacher and principal evaluation systems is a case in point. Feldman is
a recent PhD graduate from the University of Washington, Education Leadership and Policy
Studied program. Her dissertation, “Inquiry-focused reform: how teachers learn new practices
from their current practice”, was completed in 2010.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints
 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0957-8234.htm

Accountability
Bridging accountability obligations
obligations, professional
values and (perceived)
student needs with integrity 695
Received 29 October 2011
Heinrich Mintrop Revised 22 March 2012
University of California at Berkeley, Berkeley, California, USA 8 May 2012
Accepted 13 May 2012
Abstract
Purpose – The purpose of this paper is to explore the tensions between external accountability
obligations, educator’s professional values, and student needs. Strategic, cognitive, and moral
dimensions of this tension are captured with the central category of integrity.
Design/methodology/approach – This is a mixed methods study that compares five exceptionally
high performing middle schools with four exceptionally low performing middle schools in the state
of California (USA), controlling for demographics, school context factors, and below average
performance range.
Findings – It is found that schools under similar circumstances differ on the degree of integrity.
Schools with high integrity have a good balance between values and reality, are more cohesive and
more open to dissent. In each case, integrity was associated with an expansion of agency that
combined moral earnestness with prudent strategizing and actively constructing interpretive frames
that maintained a school’s sense of self-worth. Integrity develops or survives with a good dose of
educational leaders’ personal strength, but also depends on leaders’ insistence to fully exhaust
the moral horizon of an institution which obligates educators to balance equity, system efficiency,
child-centeredness and professionalism with prudence.
Research limitations/implications – This is a case study of nine schools in one state. Explanatory
relationships can be explored, but not generalized.
Practical implications – The research has implications for leadership. It demonstrates the power of
integrity as a key virtue of leadership under accountability pressures. It shows the different ways integrity
can be forged in schools and the different ways it can be missed with consequences for school life.
Social implications – The paper stresses the point that it is quite conceivable that ideological zeal,
Machiavellian strategizing, or eager system conformism may produce more forceful agency than integrity.
But as everyday responses they are not as realistic, ethical or productive as the striving for integrity.
Originality/value – The practitioner literature often points to integrity as a desirable quality when
dealing with tensions of the sort addressed in this paper, but little systematic theoretical thinking and
empirical exploration of this concept exists. The paper makes an advance in both areas.
Keywords United States of America, Schools, Management accountability, Ethics, Integration,
Leadership, Organizational culture
Paper type Research paper

Integrity is an emotionally charged word. [y] Nevertheless, a breach of the system is precisely
the terminology that applies to those situations in which the practice of a system so profoundly
contradicts its values. [y] The pursuit of integrity requires a comparison of our present activities
to our goals, the welcoming of dissatisfaction, and the painful removal of layers of obsolete and
potentially harmful practice before new layers of successful pedagogy can take hold. [y]
Journal of Educational
Integrity must also be at the heart of the accountability system (Reeves, 2000, pp. 61-2). Administration
Vol. 50 No. 5, 2012
pp. 695-726
r Emerald Group Publishing Limited
The author wishes to thank Tina Trujillo for her earlier contribution to this research, as well as 0957-8234
the anonymous reviewers whose comments were very helpful in improving the manuscript. DOI 10.1108/09578231211249871
JEA Tensions among demands issuing from government and administration, professional
50,5 values, and student needs are probably typical in public school systems all over the
world. For the USA, Reeves, an outspoken and well-known school reform advocate,
characterizes the present experience of this tension in vivid terms: as a breach between
a system’s practices and values in the wake of powerful external accountability
demands, a dissonance between educators’ goals and practices, a quest to benefit, and
696 not harm, the recipients of teachers’ services, accompanied by feelings of “welcoming
dissatisfaction” and painful learning. For Reeves, writing for an audience of
USA-American school practitioners, the intensity of the situation requires “integrity,”
an emotionally charged quality or state of being that makes the dissonances productive
for good educational practice. While the shape of these dissonances may depend
on uniquely national constellations, the striving for integrity may be a more widely
shared quest.
For the USA, the introduction of powerful high-stakes test-driven accountability
systems into the arena of American public schools has activated (or reactivated, as it
may be) a uniquely American force field. Fault lines with a long historical tradition
around first principles and values and a central conflict between policy intent and
educational realities on the ground make bridging accountability demands with
professional values and student needs a challenging undertaking for educators. In this
paper, I describe what forces may constitute this constellation and how educators
might productively deal with it. I argue that the category of “integrity” is a key concept
that aptly captures what might constitute productive agency under these conditions.
My arguments draw from history, philosophy, the sociology of school reform, and an
empirical study of nine urban middle schools that found themselves on opposite
ends of the performance spectrum by the criteria of their state accountability system.
I advance my argument in several steps: first, I ask what specific conditions might
produce the need for bridging; second, develop the idea of “integrity;” third, report on
an empirical study that illustrates the role integrity may play in leadership for school
improvement; and finally, discuss the idea of “integrity” in the context of today’s
reform challenges.

What needs bridging: a USA-American constellation

Public school educators in the USA commonly see their work as moral and
professional (Johnson, 1990). They readily share control over standards, materials, and
assessments with external agents (Ingersoll, 2003), but have traditionally reserved for
themselves, and given space for, some independent judgment about appropriate
pedagogy and the students’ well-being. Society charges public school educators to do
what’s best for children, but within the relatively narrow confines of governmental
control and normative expectations that “real school” be conducted (Metz, 1978).
Educators presumably have first-hand knowledge and experience of what students
need and fancy. But they are expected to steer youthful energy toward official school
knowledge and inculcate norms and values that lie within the band of societal
consensus. Unlike universities that have license, if not charter, to challenge knowledge
and disrupt traditions, primary and secondary schools, across many countries, are
rarely permitted to veer far off their socializing mission (Gardner, 1999; Graham, 1993).
Yet, in an open and pluralist society, educators are free to adhere to competing
educational philosophies, scientific theories, and pedagogical preferences (Kliebard,
1987). Traditionally, these pluralist impulses have been accommodated within the
institutional order of the democratic state (Fuller, 2003) with more or less ease.
 Test-based accountability systems, designed by governments with the expressed Accountability
purpose of equalizing outcomes and standardizing educators’ work, presumably obligations
condense agency, as they force or encourage schools to focus on a narrowed scope of
authoritative performance goals (Fuhrman and Elmore, 2004). But even when guidance
and pressure – the combination of standards, assessments, performance targets, and
sanctions for non-performance – make accountability systems irrefutable for schools
and school districts, competing beliefs, convictions, and professional traditions, the 697
stress of contending environmental influences and demands, and the encompassing
struggle to engage students in learning, regardless of official goals and preferences,
simply do not go away. Quite the opposite, resistant realities of children’s needs,
enduring value traditions, and claims to (at least marginal) professional autonomy and
agency may make bridging a strenuous undertaking for educators.
Pluralist values: historians of the American educational system have shown how a
host of grand ideas, foundational values, myths, and interests have traditionally
inspired and motivated actors in their struggle for influence. Four of these defining
ideas are especially pertinent to the introduction of accountability systems. Already at
the founding of the public school system in this country, the idea of education as the
“great equalizer” (Cremin, 1961) of the human condition took hold and has remained
anchored in the public imagination (Meyer, 2006). Child-centeredness (Proefriedt, 2008),
the idea that educational offerings ought to serve the interests and curiosities of the
individual child rather than flow from societal prerogatives, has since the times of
Dewey been a core concern and value for educators. Social efficiency (Tyack, 1974) has
been described as a theme so powerful within the American tradition that some have
called it a “cult” (Callahan, 1970). The idea that educators ought to be professionals
who derive their occupational status from technical expertise and knowledge of
children’s learning has strongly guided the strivings of the occupation (Labaree, 2006;
Popkewitz, 1994; Joseph and Burnaford, 2001). It is the subject of historical analysis
to describe how these four strands, equalization, child-centeredness, social efficiency,
and professionalism, among others, have enlivened history as myths, justifications
for underlying interests, or inspirations for true strivings, and how they have been in
conflict with each other or synergistically fed on each other. For our purposes, it
suffices to recognize that a political project of such magnitude as high-stakes
accountability may powerfully activate these foundational values and in redistributing
their relative weights may require actors at the receiving end, most notably teachers
and principals in schools, to reformulate some workable balance[1].
Differentiated needs: in the case of the USA, as elsewhere, working out a balance is
made harder as educators face two social-structural trends that run counter to each
other. On one hand, accountability systems increase the momentum to standardize
educational offerings via curricular alignment, prescriptive programs, pacing guides
and the like in order to fulfill accountability system expectations and insure a baseline
proficiency for all children. On the other hand, students’ learning needs have become
increasingly differentiated socially and individually. Immigration, cultural diversity, as
well as discrepancies between the poor and the middle class regarding opportunities
and lifestyles have increased social differentiation (Buchmann, 1989; Gilbert, 2008;
Lareau, 2003; Neckerman, 2004; Beck et al., 1994). Individualization is indicated
by rising special learning needs (Banks and Banks, 2009), for example, by special
education assignments. Clamor for personal regard and emotional support has
intensified among the student population while teachers’ authority and students’
norms of group cooperation and comportment have become more precarious (Pace and
JEA Hemmings, 2007). As the press for standardization rubs against a resistant reality of
50,5 social and individual differentiation, inconsistencies between the needs of children and
the demands of the system may arise that cannot be bridged with facility, particularly
in schools that serve disadvantaged students. An example may be the requirement
to give a test in English to large numbers of limited English speakers. These stresses,
if they are indeed perceived by educators, in turn may galvanize traditional value
698 conflicts within the American teaching profession (e.g. between child-centeredness and
system efficiency).
Accountability systems, via standards, assessments, data and the like, can be
powerful technical drivers of school change, but by their very nature, such systems
pivot on judgment (Hargreaves, 2004; Mintrop, 2004). How high should equality
expectations be, how narrow or wide, complex or simple should content be, what
should be the role of the child in the educational endeavor, and what ought to count as
high quality professional work or human service delivery, all these are matters of
potential “harm” and “pain,” to speak with Reeves again, that is, they may entail
value conflict, moral dilemma, and emotional intensity. Thus when educators craft
local coherence (Honig and Hatch, 2004) or make sense of cognitive dissonances
(Coburn, 2006), moral concerns may be accentuated.
Responses to accountability: based on the literature on educators’ responses to
accountability, one may analytically distinguish three patterns: resistance, alignment,
and coherence (Mintrop and Sunderman, 2009; Supovitz, 2009; Leithwood et al., 2002;
Louis et al., 2005). In real life, however, schools may craft responses that contain
elements of all three. Thus patterns will overlap.
Educators may develop resistance (Hursh, 2003; McNeil, 2000; Mintrop, 2004;
Valenzuela, 2005; Achinstein and Ogawa, 2006 ) as a result of their evaluation of
system pressures in light of their own standards of good teaching and care for
students. Overemphasis on control, narrow learning goals, and disregard for students’
social needs may generate attempts to devise subversive strategies (McNeil, 2000) or
may result in exit (Achinstein and Ogawa, 2006). Resistance to accountability demands
is probably not widespread (Hursh, 2003) as it may imperil one’s job and organizational
survival.
Alignment seems to be a wide-spread response pattern (Jacob, 2005; Hamilton et al.,
2005). When schools align to the accountability system, they internally reorder goals,
programs, and data with system elements. They focus on the system goals,
deemphasize non-tested subjects, carefully orient instruction to test items that recur on
state tests, select standards-based materials, use the system’s performance information
to monitor teachers, focus on remedial learning needs, and so on (Herman, 2004;
Koretz, 2008). Schools responding with alignment ration time and energy to optimize
measured results (Booher-Jennings, 2005). Alignment is a technical or strategic
response that excludes moral concerns from the conversation (Anagnostopoulos, 2006;
Anagnostopoulos and Rutledge, 2007).
When schools, in response to system demands, pull together around goals, develop
a sense of shared responsibility for high performance, and establish consistency
between external accountability and a school’s internal accountability culture, they
have established coherence (Elmore, 2004; Carnoy et al., 2003). Coherence may be
programmtic or normative. Coherence comes about more easily in organizations that
have sufficient internal capacity to answer to external pressures. Lower capacity
schools under performance pressures situated in lower SES environments may find it
harder to develop coherence and tend to opt for alignment strategies (Diamond and
Spillane, 2004). The concept of integrity, developed in this paper, is meant to be an Accountability
elaboration of the coherence pattern and may be especially salient for schools obligations
situated in disadvantaged communites at the lower end of the performance
spectrum. As schools claim their space for internal action, interrogate accountability
demands within the existing spectrum of pluralist values, and maintain sensitivity
for countervailing student needs, while perhaps using the impetus of accountability
systems to leverage desirable improvements, they strive toward integrity. 699
The concept of integrity
Integrity is a moral quality of social life that inheres cognitive effort, but also binds and
mobilizes positive emotions (Turiel, 2005). Common sense notions of integrity conjure
honesty, sticking to one’s principles, courage in the face of challenges, and wholeness
in the face of fragmentation, conflict, or fragility. Integrity is at the core of relational
trust (Louis 2007; Bryk and Schneider, 2002). Integrity in schools, according to Bryk
and Schneider, hinges upon a reliable consistency between word and deed around core
educational values.
In a nutshell, for the purpose of this study we say that educators as individuals or
collectives have integrity when they strive for agency in pursuit of valued internal
purposes; establish coherence or consistency among values, word, and deed;
acknowledge compromise, rupture, and conflict with honesty and truthfulness;
evaluate action in light of perceived client needs; and address institutional role
obligations in the face of multiple values and moral demands of the institution.
The philosophical and psychological literature on (personal) integrity seems to
agree that integrity has a moral and integrative dimension. A person with integrity
affirms one’s core values and commitments within a normative frame while
integrating, giving unity, coherence, or identity, to the manifold and conflicting
demands placed on the self (McFall, 1987; Ramsay, 1997). But there is disagreement as
to the “moral texture” of integrity. Carter (1996) gives a “tightly” textured version.
Persons have integrity when they have developed a sense of right and wrong,
discerned a course of action, and avowed to stay true to principles even when the
environment does not reward the conduct. Integrity always involves risk. An
individual in complete harmony with self and their social environment cannot develop
integrity. In the words of McFall (1987), “Where there is no possibility of its loss,
integrity cannot exist” (p. 9). For Carter, there are standards of the morally right. These
standards may not always be apparent, but one may gain access to them through
reflecting with sincerity, earnestness, and commitment.
Opposite to Carter’s morally stringent concept of integrity is Steele’s social
psychological conceptualization. Here integrity is about a sense of self-worth and
psychic integration, rather than external normative standards. Steele (1988) defines
integrity as a sense of moral or adaptive adequacy in the face of environmental
forces that threaten individuals’ sense of self-worth, for example, as a result of
negative judgments, sanctions, and the like. Integrity is restored with images that
affirm “the larger self” (Steele, 1988). These images may not necessarily address
the specific situation, nor may they be able to actually resolve the material threat.
In fact, individuals may “tolerate specific inconsistencies with no attempt at resolution,”
(p. 268) as long as a broad balance, a workable whole is maintained.
Benjamin’s (1990) conception of integrity is at once looser than Carter’s, but ethically
more substantive than Steele’s. For him commitments to a set of values are essential,
and these commitments, as for Carter, are coupled with self-reflection, willingness to be
JEA self-critical, and avoidance of self-deception. But his view of modern society as highly
50,5 differentiated precludes the existence of overarching world views that could support firm
and discernible moral standards of the kind Carter has in mind. Given multiple roles we
play, situations we negotiate, and diverse people with unique sets of commitments
we encounter, value conflicts are rampant and not easily solvable. For Benjamin, it is
tolerance of dissent, flexibility, dialogue, and compromise with others that lead to an
700 integrated life and morally superior outcomes, achieved collectively. Though individuals
strive for coherence, values and core commitments may change as long as individuals
(or groups) can account for the reasons in an intelligent and honest way. Integrity is
in opposition to ideological rigidity. This more loosely textured conception of integrity
appears to be more applicable to the public school setting where people of many different
persuasions and experiences typically interact with each other.
Conceptions differ on the strength of coherence or unity that integrity requires.
McFall (1987) describes as integrated those individuals who have developed stable and
well-ordered value hierarchies and consistencies among their various unconditional
and conditional commitments. Critics of this tight definition point out that the need for
stability and order should not be overplayed since a person with integrity always needs
to engage ambivalences and inconsistencies in their lives that persistently challenge
coherence (Cox et al., 2003). But even with definitions of integration as more tightly
textured, integrity is never a coherent unity. Rather it is formed out of component parts
that are still recognized as unique qualities in tension with others. One does not have
integrity, but strives for it.
We have so far looked at conceptions of integrity at the individual level. But as
Montefiori (Montefiore and Vines, 1999, Chapter 1) has pointed out, individuals act in
concrete and social situations. They carry out certain tasks in specific social settings
that are more or less structurally unified or rife with conflict or contradictions. They
occupy certain social roles in specific institutions that provide the normative frame, to
them and others, for evaluative standards of integrity. Because educators in their work
settings act out a public role that is defined by institutional task structures and values,
we are not only concerned with personal integrity, but also with public integrity.
Moreover, we are concerned with the behavior of organizations that have specific
cultures and adhere to shared core philosophies and programmatic commitments that
may be in tension with their environment.
When one moves from the personal to the public realm, conceptions of integrity
become more loosely textured for two reasons: first, the media through which much of
public life is lived, such as money, power, or bureaucracy, operate to a large degree in
impersonal or non-moral ways with which individuals and groups need to compromise
in order to function (Paine, 1994; Selznick, 1994); and second, congruence cannot be
assumed between personal commitments and the kinds of commitments institutions
permit and oblige public officials to express, what Dobel (1999) calls an institution’s
“moral horizon.” If institutions “represent inheritances of valued purpose with
attendant rules and moral obligations” (Heclo, 2008, p. 38) that result in given, often
unequal, distributions of power, wealth, and prestige, then moral horizons of
institutions derive from the institution’s multiple core values and the consistency of
those values with its practices. The military’s moral horizon does not include pacifism,
nor is it restricted to the obligation of killing one’s enemy in times of war. The moral
horizon of the educational system as it relates to the specific bridging challenge
discussed in this paper is circumscribed by multiple and inconsistent core values
activated by the structural disparities that I have discussed above.
 Dobel (1999) has developed a model of public integrity for higher echelon officials Accountability
that nevertheless is applicable to street-level officials, such as school administrators obligations
and teachers. It consists of three overlapping domains that together set up a “triangle
of judgment” through which officials may move when they make decisions: obligations
of office, personal integrity, and prudence. Personal integrity is at the heart of public
integrity in liberally constituted societies because institutional structures never abdicate
personal responsibility for one’s actions. However, in the public realm, personal integrity 701
is hemmed in by institutional structures and the need to be efficacious in one’s actions to
accomplish results.
Offices are grants of power over citizens and come with duties to abide by the rules,
norms, and aspirational values of the institution. At the very basic level, this means
obeying laws and regulations and striving to act according to foundational and
constitutional principles. Institutions prevent people from being harmful to others and
encourage them to improve, for example, when new legal norms protect minorities
from discrimination or educational authorities press for more equitable educational
outcomes. But sometimes it is the other way around, and it is up to individual citizens
or office holders to prevent institutions from doing harm and press them to live up to
higher standards. Thus obligation of office is not blind obedience, compliance, or
abdication of personal responsibility, rather it requires careful weighing of the
purposes and consequences of an institutional structure, program, or policy in light of
one’s own commitments and the broadest ethical standards. But it is the obligation to
take external normative guidance seriously, whether it comes from governmental or
professional sources, and not reject it on the grounds that it constrains one’s individual
commitments.
Officials are in their offices, principals in their schools, and teachers in their
classrooms, in order to achieve results. Given the increasingly impersonal, disintegrative,
and amoral functioning of much of institutional life, we should not be surprised, as the
philosopher Oksenberg Rorty polemically remarks, that it is “the willful, boastful
or sniveling egotists, people of low integration and even lower integrity” (Rorty, 1999,
p. 117) who get things done, and not necessarily those who pay homage to integrity.
Integrity in the public realm needs to be augmented by prudence, the practical wisdom,
skill, and forethought to marshal the forces and means needed to accomplish valued
outcomes. Prudence is distinguished from expediency. For the latter, any means are
welcome that produce effects in the moment; for the former, the choice of means, with
sometimes needed ethical compromises, is oriented toward the longer term and with
an outlook on a broader picture. Thus, for the street-level office, integrity is shot
through with pragmatic, but prudent, politicking, managing, and strategizing (Honig
and Hatch, 2004).
Organizational integrity is a matter of both culture and structure. It is about crafting
a coherent whole, both technically and normatively (Elmore, 2004; Newmann et al.,
2001), and a conscious ethical response in reference to the institution’s moral horizon,
internal core commitments of personnel, and the need to produce results (Fullan, 2003).
With Schein (1985), we can presume that organizational integrity centrally hinges on
the agency of leaders whose role it is to embody organizational and personal
commitments and integrate the organization’s core values with its needs to solve
problems of adaptation to its environment. Such leadership, we can presume is at once
moral and strategic-managerial.
We are now in a better position to analytically distinguish integrity from other
response patterns to accountability discussed above. (In reality, there will probably
JEA always be overlap.) Technical alignment privileges the demands of the system over
50,5 internal values and student needs as they present themselves to teachers. Coupled
with authoritatively established consensus, alignment may preclude moral-ethical
considerations and suppress internal values and student needs that run counter to
“what works” in terms of measured performance. Integrity moves beyond coherence,
though the two of them are related. Coherence by itself implies unity, an active forging
702 of productive congruence and consensus between external demands and internal
programs and orientations. Integrity stresses the precariousness of this congruence
and the ethical struggle to bridge what may be in conflict with each other. After
careful weighing of external obligations in light of internal values and observed
needs, integrity may entail elements of resistance, but resistance within the bounds of
ensuring the viability and survival of the organization. For, the organization is the
means for actors’ collective agency.

Operationalizing integrity
Integrity in schools is at base about finding a good balance between external demands
that emanate from district and state administrations, educators’ pluralist educational
values, and differentiated student needs. Good balance does not mean equal weights,
but assigning weights according to a rank order of normative importance. Schools with
high integrity have found a place in the system, but have retained something of their
own. They have given external demands their rightful place, as weighted against
internal values and perceived student needs. They have done so without (self-)
deception, but with sincerity, honesty, dialogue and tolerance for dissent. They act out
of respect for the institution and for self. While the concrete bridging “solutions” may
differ and may privilege some values over others, schools with high integrity may
craft these solutions in reference to the moral horizon of the American educational
system with its spectrum of relevant values. This means that even when their
solution veers toward “efficiency,” administrative rationality, and standardization, for
example, they keep child-centered concerns and uniquely professional responsibilities
in the conversation, or vice versa. Being open to dissent makes it more likely that
these concerns are at the table and inform the school’s collective tinkering with
solutions. Integrity is a pragmatic aspiration that de-emphasizes ideological wars, but
encourages ethical and complex problem solving (Campbell, 2008).
Integrity is reflected in concrete beliefs, norms, and practices at the organizational
level. For this study, integrity is indicated by the shared belief that a good balance is
struck at a given school among external demands, professional values, and student
needs. Good balance in substance is cultivated in organizations with certain, more
formal, characteristics: weighing tensions is reflected in open communication,
toleration of dissent, and learning; external obligations are reflected in the school
having raised expectations in response to the demands of the accountability system;
coherence is indicated by norms of shared responsibility and pulling together around
common goals. Principals play an essential role in bridging the organization to the
external environment (Goldring and Rallis, 1993), particularly in response to
accountability (Rutledge, 2010). Leadership that furthers integrity presumably creates
a sense of normative and programmatic coherence in conjunction with toleration of
dissent. Thus, leadership may range from managerial or instructional to moral
emphases (Hodgkinson, 1991; Fullan, 2003; Louis et al., 2010; Sergiovanni, 1992;
Goldring and Rallis, 1993). Lastly, schools that exhibit this pattern of integrity may
relate more positively or negatively to the accountability system (Mintrop and Trujillo,
2007a). A more positive response would be indicated by schools’ perceptions Accountability
of meaningfulness of the system, the latter by perceptions of pressure (Mintrop, 2004). obligations
I explore this basic pattern of organizational integrity with quantitative survey
(and test) data.
Integrity necessitates the perspective of actors who craft coherence, make sense of
dissonances, and struggle through accountability judgments, value tensions, and
structural disparities with moral effort. This actor perspective is captured in 703
qualitative interviews primarily with school leaders (principals, assistant principals,
and teacher leaders) and those receiving the leaders’ messages. Questions and
interpretive codes revolve around the triangle of judgment: what personal core
commitments do leaders have, how do they interpret accountability demands, and how
do they strategize so that valued outcomes are achieved and the organization survives.
I ask how leaders frame accountability demands in light of the moral horizon of the
institution, how they remain sensitive to student needs that run counter to external
demands, and how they make sure voices are heard that can articulate varied values
and perceptions. I query how leaders make the organization “whole” (i.e. maintain an
internal locus of control and sense of worth while avoiding self-deception), and what
strategies they use to make it through the system in one piece (i.e. avoid sanctions and
corrective actions). I examine how they strive for unifying goals and aspirations while
accommodating dissent and self-examination; and lastly how they instill courage in the
face of risk.

The California accountability context

Across the USA, accountability systems differ widely. Some are bare bones, others are
fairly elaborate (Mintrop and Trujillo, 2005). Measuring student performance through
regular testing, providing performance information, setting simple quantitative goals
that force schools on a continuous improvement path, and the threat of sanctions for
low performers, from corrective action all the way to the dissolution of the school, are
the main drivers in the bare bones version. More elaborate systems add to this a
catalogue of state learning standards, cognitively ambitious state assessments,
instructional materials aligned with state assessments, and resources for professional
development and school intervention around mandated instructional programs. In the
bare bones version, schools are left alone in figuring out how to meet their performance
targets. In the more elaborate systems, schools are potentially more regulated; and
what teachers do in their classrooms may be subject to central instructional
management. The bare bones version may move schools through the sheer might of
goals and sanctions; the more elaborate version may work through a more subtle multi-
pronged systemic approach that combines incentives with controls.
The California accountability system, my place of study, is of the elaborate kind. At
the time of this study, the state had a fairly well-aligned system in place. State
standards were aligned with the California Standards Tests which were the weightiest
components in the formula with which the state computed each school’s academic
performance index (API) annually (CDE, 2006). Each year, schools received a new API
growth target that was calculated as 5 percent of the difference between a school’s
present API and the state goal of 800. (In addition to the state index API, the federal
Adequate Yearly Progress (AYP) performance target, not particularly well aligned with
API, was making inroads into the schools when we collected data.) Textbook
publishers, ever mindful of the state’s market power, adapted their materials explicitly
to state standards. The state adopted a small number of English language, arts, math,
JEA and remedial literacy programs that schools were required to use. Monies were
50,5 allocated for professional development explicitly tied to state-adopted curricula.
Additional funds, quite generous compared to other states (Mintrop and Trujillo, 2005),
were given to schools in the lowest performing brackets through, so called, High
Priority Schools Grants. As to sanctions, California followed the procedures laid out
in the federal No Child Left Behind legislation that stipulates five program
704 improvement stages at which end a persistently low-performing school ceases to exist.
Thus, in California, many schools, especially in poor urban and rural districts
(Kim and Sunderman, 2005), have experienced strong accountability pressures, have
wrestled with labels of low performance and failure, and become subject to the
regulatory reach of districts all the way into the classroom.

The study
The data for the study were collected from nine middle schools, urban in character, that
found themselves in the bottom half of the state’s performance distribution. Within this
band, five schools were rather high performing and four rather low performing. But all
nine were as similar as possible with respect to social background and internal
capacity so that the relationship among salient variables surrounding the challenge to
integrity could be studied “controlling” for extraneous factors. The study employs
a structured multiple-cases design that allows for quantitative and qualitative
cross-case comparisons (Yin, 2003; Miles and Huberman, 1994). The bulk of the data
were collected in the 2004-2005 school year.

Cases
Table I shows characteristics of the nine schools on which this paper is based. The
schools were overall similar demographically. Two schools (I and C) had relatively
lower proportions of African-American and Hispanic students, but a high proportion
of Hmong students. To explore school context conditions with finer grain size, we went
beyond state-reported data and inquired about student and teacher perceptions of
family background and support for education[2]. Some schools, it appears from this
data, are more challenged in the area of parental support and poverty, while others
more in the area of language and possession of cultural goods, but for the analysis
conducted in this paper, schools overall are reasonably well matched. Similarly, higher
capacity schools might be able to exert more forceful agency in bridging tensions than

A B C D E F G H I

2005 API 653 683 604 573 670 573 653 642 598
2003-2005 API growth 37 78 56 4 36 36 4 47 65
Enrollment 1,628 868 991 1,100 780 866 705 1,818 1,031
African-American (%) 5 1 12 4 6 3 1 0 9
Hispanic (%) 75 93 59 84 81 88 59 97 56
English learners (%) 43 28 26 22 18 29 31 44 39
Free/reduced lunch (%) 83 78 100 59 69 97 85 77 100
Parent educationa 2.09 2.03 2.25 2.13 2.18 1.81 2.02 1.81 2.09
Table I.
Demographic Notes: a1 ¼ not a high school graduate, 5 ¼ graduate school; parental education is subject to the
characteristics of the nine inaccuracies of self-reported data
selected cases, 2004-2005 Source: California Department of Education (2006)
lower capacity schools. Survey data that capture internal teacher capacity as indicated Accountability
by school-wide percentages and averages, respectively, of total years of teaching, obligations
degree completed, full certification, and subjective sense of preparedness, show that
the schools were well matched with regard to internal teacher capacity, though some
differences exist on degrees completed[3]. In sum, the cases exhibit fairly similar
conditions across the nine schools with respect to individual teacher capacity and
socioeconomic environment. 705
Instruments and data
A number of robust research instruments were developed for the quantitative
component of this study. All instruments were repeatedly field tested. Factor and scale
reliabilities were in most instances high and in a few instances acceptable. Some survey
items and scales were validated in previous studies, conducted by the authors
and other researchers in the field; some were specifically developed for this study.
Here I briefly describe the properties of the instruments and the ways they were
administered. For a more in-depth discussion and detailed data collection and analysis
procedures, refer to the CRESST technical report (Mintrop and Trujillo, 2007b). The
teacher questionnaire consisted of over 180 individual response items designed to
collect information on teachers’ perceptions of accountability, school goals, leadership,
organizational strength, motivation, efficacy, school program, and change strategy as
well as teacher background data. Items and scales come from a variety of sources
(CCSR, 2003; Mintrop, 2004; McLaughlin and Talbert, 1993; SRI International,
Policy Studies Associates and CPRE, 2003). The questionnaire was administered to
all teachers in the nine schools; 317 teachers responded in total. Overall response
rate was 83 percent, ranging from 67 percent for School I up to 94 percent for School E.
The analysis in this paper concentrates on a set of relevant variables related to the
organizational-cultural base pattern of integrity, leadership emphases, and responses
to the accountability system.
For the qualitative component, a total of 157 interviews were conducted, between
17 and 20 interviews per school. We interviewed school leaders and classroom teachers,
at least one counselor per school, and the person responsible for administering special
programs for disadvantaged children. Interviews were conducted in two phases.
Interspersed between the two phases were classroom observations which allowed us
a view of school practices independent from teacher testimony. In the first interview
phase (up to ten interviews per school) we asked about goals, values, organizational
culture, leadership, and accountability. In the second interview phase, we focussed on the
instructional program, perceived student needs, and teachers’ own learning, often in
reference to observed lessons. The principals were interviewed in both phases.
The purpose of the interviews was to increase data richness around the quantitative
survey items and scales. Thus, the coding of the 157 interviews followed the variables
of the quantitative component of the study. Broad descriptors were coded, such as
principal leadership, instructional program, response to accountability, school change
process, performance management, professional development, and the role of district.
Data analysis for this paper focusses on material related to the first four. More fine-
grained codes were developed that homed in on matters of integrity. Material coded
with constructive/defensive approach to accountability, fairness, goal setting, goal
integrity, meaningfulness of accountability, morale/commitment, pulling together,
pressure, realism, responsibility/ obligation, reputation/ image help illuminate
educators’ interpretations of system demands. Material coded with academic press,
JEA care, engagement, expectations, student discipline, connections to student lives,
50,5 curriculum differentiation, English language learners (in the California context), “bubble
kids”/expediency, teaching to the test, adaptation/fidelity of instructional program,
extra-curricular activities help illuminate educators’ educational philosophies,
interpretations of student needs, and strategies to answer to these needs. Educators’
own values in light of external demands and student needs, their propensity to adopt
706 expedient, prudent, or value-concordant or discordant strategies are illuminated by
both sets of codes.
Initially a small set of interviews were double coded by two coders. Discrepancies
between the two coders were discussed until agreements could be established. The
bulk of the interviews was coded by one coder. For the data analysis for this
paper, interviews with principals, other administrators, and teachers with special
leadership functions were reread in their entirety (about 20) by the second coder, and
the consistency of codes was re-examined. Based on these 20 or so interviews,
the initial coding turned out to cover the key material needed for this analysis.
Rereading of all other interviews was restricted to the material coded with the above
listed codes.
The interview protocols did contain prompts for “balance,” but did not contain
explicit prompts for the “moral horizon of the institution”, nor was the material
explicitly coded in this regard. However, respondents, particularly principals, tended to
frame the rendering of good balance or imbalance among accountability demands,
internal values, and perceived student needs in terms of broader institutional values
(most notably equalization, professionalism, child-centeredness, and effectiveness – not
necessarily using the exact terms). Following Miles and Huberman (1994), data were
grouped into “case dynamics matrices” (p. 148ff) with categories derived from the
conceptualization of the study and the quantitative patterns. For this paper, the material
is used to compose short vignettes that illustrate the quantitatively established patterns.
Classroom observation data consisting of 270 lesson segments in English Language
Arts serve a very limited purpose in this paper. With a focus on three dimensions listed
in Table II, they enter the case narrative to countercheck claims that respondents make
regarding their schools’ concern for instructional quality.
Thus, this mixed-methods case study design uses quantitative data for descriptive
and correlational analyses to understand the nine cases in comparison. Utilizing
a sequential explanatory strategy (Creswell, 2003), I build on quantitative analyses
with qualitative inquiries. The quantitative data reveal associations among factors, but
we cannot infer directionality without qualitative data that illuminate what makes
these associations come to life (Miles and Huberman, 1994; Green et al., 2006).

Organizational integrity Good balance of external demands, teachers’ values, and student needs
Open communication and toleration of dissent; learning orientation
Shared responsibility for performance; collegiality; pulling together
around performance goals
Leadership strength Moral, instructional, supportive, managerial
Perceptions of Guidance, validity, fairness, pressure
accountability system
Measured performance API scores
Table II. Observed instructional Positive tone, pro-active instructional formats, cognitive complexity
Variables quality
Findings Accountability
Findings are presented in two steps. I first examine and display descriptive and obligations
correlational data to explore integrity-related organizational characteristics and identify
schools with different patterns that I compare in more detail through qualitative data in
case vignettes that, for this paper, interpret the actions of school leaders.

Quantitative data 707

When teachers across the nine schools were asked how system demands, student
needs, and teacher values ought to be balanced normatively, they gave the highest
importance ratings to student needs. Teacher values, according to mean responses
across all nine schools, should count for less than student needs, but more than system
demands. When asked about the reality of these weights in their schools, system
demands were factually perceived as dominant. Thus there is an imbalance. Factually,
teacher values were seen as less important than student needs, roughly as it should be
normatively. Two schools, B and C, stand out with particularly high (factual)
importance ratings for student needs. That is, in these two schools teachers on the
whole perceive their schools factually to attach more importance to student needs (and
to lesser degree to teacher values) than in the other seven schools. One school (E)
is especially conspicuous, according to teachers’ perceptions, in unduly disregarding
student needs, unduly in light of normative expectations (“should” ratings). In two
schools (D and H), the discrepancy between the system’s legitimacy (“should”)
and the reality of the system’s perceived dominance is particularly conspicuous
(Table III)[4].
Among the nine schools, the two schools (B and C) with the highest perceived
regard for student needs also stand out with an overall organizational-cultural pattern
that points in the direction of relatively strong integrity compared to two other schools,
notably schools E and H that point in the opposite direction. The two schools in which
the faculty perceived a better balance among demands, values, and needs were also
relatively stronger in accommodating open communication and tolerating dissent
while at the same time raising expectations according to accountability demands and
forging internal coherence.
Thus, these two schools presumably achieved both moral integrity, here indicated
as a balancing of core tensions, as well as more formal integration, here indicated by

A B C D E F G H I
School n ¼ 44 n ¼ 31 n ¼ 39 n ¼ 42 n ¼ 29 n ¼ 26 n ¼ 28 n ¼ 49 n ¼ 29

System demands
Normative 3.00 3.30 3.21 2.62 3.18 3.12 3.11 2.79 2.97
Factual 4.53 4.55 4.23 4.40 4.82 4.81 4.89 4.85 4.79
Student needs
Normative 4.91 5.00 4.74 4.88 4.82 4.81 4.89 4.85 4.79
Factual 3.56 4.23 4.05 3.07 2.57 3.38 3.46 3.04 3.17
Teachers’ values
Table III.
Normative 3.98 3.81 4.03 3.93 3.96 3.96 4.46 4.02 3.48
Teachers’ perceptions of
Factual 2.74 3.35 3.36 2.69 2.46 2.62 3.00 2.73 2.83
balance of system
Notes: Five-point Likert scale: How important should these forces be?/How important are these forces demands, values, and
in the reality of your school?: “very important” to “not at all important” needs (means)
JEA the poise with which they pull together and at the same time accommodate diverse
50,5 views or judgments[5]. Across the nine schools, associations among the various
organizational-cultural indicators of integrity are strong, allowing us to speak of a
consistent pattern[6]. Figure 1 displays the association between good balance as the
substantive core of integrity in this study and other integrity-related characteristics
that confirm for the nine schools the interweaving of moral and integrative dimensions
708 of integrity postulated by the literature.
Leadership, as predicted, was associated with a more strongly developed pattern
of integrity across the nine schools. Figure 2 displays the relationship between
good balance and various emphases of leadership. It was not moral leadership
alone that seems to have played a role. In the eyes of survey respondents, moral,
technical-instructional, supportive, and managerial emphases, in combination, seem
to have contributed to a shared sense of good balance, reinforcing the sense that
integrity is a multi-dimensional quality that comes about through weighing and
clarifying, but also managing, strategizing, technical support, and personal regard. In
short, integrity may be facilitated by multi-faceted leadership strength.
Integrity, as the literature states, develops in the tension between an internal state
and external challenges. Integrity involves potential loss, risk, and disharmony while
striving toward effectiveness, coherence, sense of worth, and moral core commitments.
The focal external challenge, relevant for our investigation, is the looming presence
of high-stakes accountability systems that in their elaborate forms exert unprecedented
control over schools and classrooms while demanding “gap closing” performance

0.5
Faculty characteristics

0
–1 –0.5 0 0.5 1

–0.5

–1
Figure 1. Good balance of demands, values and needs
Relationships between
“good balance” and Notes: Trend lines based on nine cases. Scatter plots not displayed to
faculty culture facilitate readability. Trend lines for open communication, learning
orientation, pulling together, raised expectations, shared responsibility
 1 Accountability
obligations

0.5
709
Leadership strength

0
–1 –0.5 0 0.5 1

–0.5

–1
Good balance of demands, values and needs Figure 2.
Relationships between
Notes: Trend lines based on nine cases. Scatter plots not displayed to good balance and
facilitate readability. Trend lines for open communication, learning leadership strength
orientation, pulling together, raised expectations, shared responsibility

improvements in short order. How do schools with a strongly developed integrity

pattern bridge to the system and how do they fare in the system?
Table V groups the nine schools according to integrity strength displayed in
Table IV. In that table, Schools B and C were identified with the strongest patterns
while Schools H and E with the weakest ones. As to test performance in the system, no
causal connections can be construed from this data. But Table V does illuminate that a
strong integrity pattern was not a detriment to test performance. If at all, the two
schools with a strong pattern grew relatively well within the two years prior to the time
of data collection, compared to the nine schools. (School B, the school with the strongest
integrity pattern, was the strongest performer in the nine-school sample.) Table V also
shows that schools with a relatively stronger integrity pattern, compared to those with
a weaker pattern, had a more positive outlook on the system’s guidance function,
fairness, and validity. They found the system more meaningful for their practice while
not feeling more pressure. A different way of displaying this relationship is Figure 3,
the “bundled” trend lines based on scatter plots of the nine schools that demonstrate
the relationship between accountability perceptions and “good balance” perceptions.
The quantitative data alone, as stated above, cannot positively establish integrity
because they lack the actor perspective, but the data can indicate a presumed integrity
pattern. In sum, the data reveal that integrity strength is indicated by good balance
among external demands, teachers’ values, and perceived student needs, embedded
in a culture that takes external performance obligations seriously, coheres around
JEA Strong Mixed Weak
50,5 B C I F A G D H E

Good balance of demands, values, and needs þþ þþ     

Open communication and dissent þ þþ þ   þ  
Learning orientation þþ þ  þ þ þ  
710 Raised expectations þþ þ þ þ     
Pulling together þþ þþ þ þ    
Shared responsibility þþ þþ  þ  þ   
Cohesion þþ þþ  þ þ   
Observed instructional quality** þ  þ
Notes: *Table III displays a matrix of the nine schools’ mean perceptions of faculty culture. The
matrix displays scale means by assigning a zero (suppressed) to school means that fell within 0.1 point
Table IV. of the nine-school mean; one plus ( þ ) or minus () to means that fell o1 SD above or below the mean;
Integrity-related school two pluses ( þ þ ) or minuses () to means that fell 41 SD above or below the mean; **see Table in
culture characteristics Appendix 1 classroom observations (in percent of observed snapshots). “ þ ” and “–”, exceptionally
(matrix of school means*) high quality in three dimensions combined

Strong Mixed Weak

B C I F A G D H E

API growth over last two years 78 56 65 36 37 4 4 47 36

Pressure   þ  þ þ þ
Table V. Guidance/focus þþ þ þ  þ    
Response to accountability Validity þþ þ þþ  þ   þ 
by strength of integrity Fairness þþ þ þ  þþ  
pattern (matrix of
school means)* Note: *See Table III footnotes

1
Accountability perceptions

0.5

0
–1 –0.5 0 0.5 1

–0.5
Meaning: guidance, validity, fairness
Pressure
–1
Good balance
Figure 3.
Relationships between Notes: Trend lines based on nine cases. Scatter plots not displayed
good balance and
accountability perceptions to facilitate readability. Trend lines for open communication, learning
orientation, pulling together, raised expectations, shared responsibility
common goals, but at the same time leaves sufficient openness for dissent and learning. Accountability
This pattern may be facilitated by multi-faceted leadership strength at given schools. obligations
Across the nine cases, schools with a relatively stronger integrity pattern tend to
connect to the accountability system in a more meaningful way. The integrity
pattern is also associated with relatively strong growth on the state’s API within a
two-year period of data collection, though other schools with a weaker integrity
pattern posted similar growth. Thus, overall the quantitative data would suggest 711
integrity as a rather desirable property of schools. Teachers at the schools seem to
think so as well. Satisfaction ratings (not displayed here) are highest in the two
high-integrity schools and lowest in one of the low-integrity schools (E). How this
pattern actually plays out in schools is illuminated by qualitative data.

Qualitative case vignettes

My main purpose in this section is to show what concrete beliefs, attitudes, and
practices attach themselves to the previously established patterns. It makes sense to
begin the analysis with the two schools that, relative to the other seven, exhibited the
strongest integrity according to the quantitative indicators. These are Schools B and C.
For a school with relatively lower integrity characteristics, clearly School E qualifies as
Table IV shows. School H, in the same category as School E, could have been chosen,
but School D, a borderline case with a somewhat different organizational culture
pattern that points to integrity challenges of a different kind than School E (or H for
that matter) makes for a more poignant vignette.
School B. On our first visit to School B, we were introduced to one of the senior
teachers who had taken on the responsibility for the school’s remedial literacy
programs. With tears in his eyes, he recounted how the school’s principal over many
years had turned the school around and made it one of the highest performing middle
schools in the state for its demographic profile. This deep connection between faculty
and leadership resurfaced again and again. For the principal, himself part of the
community from which the school drew its students, the accountability system was an
extension of his moral mission: opening up opportunities for mainly poor Latino
immigrants. In his mind, he was following a moral compass that he himself had
experienced in his own middle class upbringing. When legislation passed and the
accountability system came into being, his agenda was already well underway; he
merely seized on new opportunities. For him, the system finally ended all distractions
from achievement and strengthened the hands of those who believed in the possibility
of dramatic learning gains for poor children. Over time, he separated from teachers
who resisted his vision. He came to rely on a cadre of committed senior teachers and
attracted larger numbers of young energetic teachers whose tenure at the school tended
to be relatively short. Accountability to the state was framed as a mere extension
of strong internal commitments. The school lived and breathed a morality of
effectiveness, at the core of which was a leader who had convinced the faculty that the
accountability system was a moral imperative, designed to benefit life chances of
students of color and immigrants. Thus, little tensions between personal commitments
and external obligations surfaced in the interviews with faculty.
Guidance from the system was seen as very helpful in focussing the school’s
resources and energy and making instruction evidence based. Directed by its
administration, the faculty strove to design an instructional program that would
optimize outcomes. But raising test scores was not seen as an end in itself, rather as
a sign of a job well done. The school opted for detailed curriculum alignment and
JEA prescribed instruction. The majority of below-grade-level students were taught for the
50,5 majority of their learning time in remedial literacy programs. Social studies and science
had been de-departmentalized in this middle school and folded into the teaching of the
literacy programs. This was justified on the grounds that arrangements more typical
for elementary schools would work better for the kinds of students and learning needs
that the school had to address.
712 But there was unease. Doubts about the adequacy of science and social studies
instruction were aired openly in this faculty, but for the time being, a joint commitment
had been made to focus on remedial literacy as a deliberate experiment. A number of
teachers and the assistant principal, who functioned as the main instructional leader,
acknowledged that the remedial instructional programs might fall short on being
engaging for students, but this did not prevent them from embarking on an active
quest to improve teaching within these programs. Indeed, data from classroom
observations demonstrate higher instructional quality in this school than in most of the
other schools observed. Moreover, teachers were encouraged to identify gaps in the
school’s program and offer solutions. For example, English teachers, complaining
about a lack of writing skills, identified, procured, and were in the process of trying out
a new writing program. Thus, the needs of children were a constant point of reference,
but they were reflected within the fairly narrow confines of the standards-aligned
programmatic structure. The school was very successful at raising standardized test
scores and points on the state’s API.
How do actions at School B convey a striving for integrity? The school makes a
conscious moral choice in bridging external state and district demands, internal
values, and student needs as perceived by the school, and it pursues its decisions with
zeal and shared faculty commitment. It strongly values efficiency, utility, and
remediation, and interprets those pursuits in the frame of equalizing life chances for
poor students and assuming professional responsibility for expanding those chances.
Even though, by the criteria of Deweyan child-centered philosophy, the school would
surely fall short, child-centeredness as more diffuse institutional inheritance makes
itself felt in the seriousness with which faculty and leadership observe student learning
needs and teaching quality and the openness with which they voice their doubts about
the narrow curriculum and the prescriptive nature of many of their programmatic
offerings. The boundaries around this inquiry are circumscribed by the aligned
technical structure and the moral and positional authority of the principal, but the
potential is there to revisit the present consensus and build new bridges when student
needs demand.
School C. Integrity expressed itself differently in School C. Whereas School B could
look back to a continuous period of success in the accountability system, School C had
been battered by the system as one of the lowest performing schools in the district for
repeatedly missing its growth targets. School C’s principal was fairly new in
administration, but had a solid background in classroom teaching and staff
development. When she began her tenure, she realized that most of all, her staff needed
a break from being preoccupied with the sanctions-oriented nature of accountability
which for the faculty represented mainly negativity and a de-moralizing streak of
failure. Good balance of school internal concerns with external demands in this case
was established through the frame of professionalism. It meant a deliberate shielding
or buffering and rebuilding of faculty coherence through confidence-building and an
emphasis on internal professional commitments. The principal made it clear to her
staff that accountability should not be taken too seriously; that the system would take
care of itself if teachers would focus on good instruction; and that when respected Accountability
professionals would voluntarily participate in instructional improvements, obligations
improvement in test scores would ensue. The principal confided that she herself
was skeptical about the fairness of the accountability system and about the kind of
pedagogy the system rewarded, but she valued the equity goals of the system. She
made the strategic choice not to disclose her skepticism in public because she was
afraid that oppositional dispositions in her faculty would detract from improvement 713
efforts. Instead she maintained with her teachers that the system provided helpful
goals and guidance and that looking at performance data was a useful undertakings.
Her leadership shored up confidence, hope, and renewed effort. She invited teachers to
participate in an open learning community that valued internal preferences and varied
opinions. Her faculty followed her lead. At the time of collecting data, the principals’
agenda was in the initial stages. Apprehension was still present, but the “no fear”
message had given the school breathing space. The profile of system demands was
lowered and internal values and student needs elevated, though the concrete shape of
internal commitments were not (yet) clearly spelled out. Apart from implementing a set
of instructional strategies that the faculty had collectively adopted, teachers were
entrusted, and given the space, to respond individually and flexibly to student needs.
The view was widespread that students’ varied needs strained teachers’ capacities, but
teachers tried their best to provide good service. What that service might actually look
like was not yet a centerpiece of the school’s collective conversation. Lesson
observation data suggest that School C had a bigger problem with instructional quality
than most of the other nine schools. Still, the principal’s tenure at the school coincided
with an upswing on standardized test scores and the API.
How do actions at School C convey a striving for integrity? School C faculty, mainly
through its leader, struggles to re-establish a sense of self-worth by referring to a larger
collective professional self. External accountability obligations are not disregarded or
ignored, but relegated to second place in the school’s value hierarchy. They are present,
but no longer overarching. The principal, prudently, strategizes, buffers, and frames
the problem of good balance in a way that kindles internal commitments without
endangering the school’s external standing. The school pulls together around the
principal’s invitation to participate in an open community of professional learners. But
this openness leaves much unsaid. Professional values of good teaching and care are
invoked, but whether teaching practices actually meet articulated or felt student needs
is largely left unexamined in this school.
School E. The contrasting case of a school suffering from a lack of integrity was
School E. With two-thirds of this middle school faculty holding Master’s degrees,
teachers in School E saw themselves and their faculty as strong and hard working in
the interest of students. They were particularly proud of the many enrichment
activities (e.g. school band) that they organized for their students, often times well
beyond the call of duty. But the school as a whole was fragmented. Observed
instructional quality was about average within the nine-school sample, but the school
ranged at the bottom of both coherence and openness. The faculty as a whole was
skeptical about the meaningfulness of the accountability system and anxious about
pressure although the school was one of the higher performing schools in the sample
and had managed for the most part to meet its state performance targets.
The principal was a young man on a fast career track, as he pointed out in the
interviews, who wanted to switch into district administration sooner rather than later.
He was determined to make his school as high performing as he could. He insisted on
JEA the implementation of specific standards-based activities (e.g. posting state standards,
50,5 implementing specific “research-based “ strategies) and was particularly enamored
with the use of new media. He organized his staff in mandatory workshops that he
himself conducted. Once he taught a strategy, he wanted to see it in action when he
made his frequent rounds through classrooms. He invoked accountability to legitimize
his expectations and vigilant monitoring. But his pro-active stance of control and, what
714 he considered, instructional leadership could not rally the faculty around the goals he
espoused. Instead teachers responded with muted, but tangible disdain and discredit of
both the accountability system and the principal. Teachers voiced that school
improvement efforts were centered on the principal and that he staged himself to
further his own career, rather than the common good of the school. He tolerated few
disagreements with his agenda. But his heavy handedness and confrontation, many
teachers felt, served no purpose in improving their teaching or their interaction with
students. He, on the other hand, saw himself as a forceful leader and skillful instructional
coach who acted in the best interest of children and had the test scores to show for it.
At the end of the school year, he did succeed in moving on to a district position.
The role of a researcher is not to take sides or cast judgments on the integrity of an
individual or a group of people. But research can compare actual behaviors with
defined characteristics associated with a quality such as integrity. In reading through
the interviews from School E, one is struck by a remarkable inconsistency between the
teachers’ and the principal’s descriptions of the situation. When the leader’s belief in his
leadership strength is not in the least confirmed by his followers, the leader has either
isolated himself from his followers and squelched open communication or willfully
misleads himself in an act of self-deception. Either condition is detrimental to integrity.
Career motives by themselves need not detract from integrity. But when they
perceptibly overshadow a leader’s moral commitments, potential connection to the
commitments of his faculty is lost. No overarching idea, no moral impetus embodied in
the leader’s personality, and no process of open and honest conversation about inherent
tensions and inconsistencies bridges external obligations with internal values. Rather,
as in this case, the school breaks up into two cultural layers, one official, the other
unofficial. In the official layer, publicly sanctioned goals and expectations are served
and the aligned technical structure operates, in the unofficial one, largely silent on the
public stage, teachers preserve personal commitments, sense of self-worth, and their
individual notions of what it means to be receptive to student needs.
School D. The fourth school, School D, that I describe now is different from the
other three schools in that here the friction between accountability-related values
(most notably efficiency) and child-centeredness came across as clear philosophical
opposition. School D’s faculty had a dim view of the accountability system, a view that
preceded the principal’s tenure and was unaffected by her. The principal, a former
union leader, had recently been appointed to the school. She considered it her main task
to mollify her very outspoken faculty. The faculty saw her leadership as open and
supportive, but also less well organized and less involved in instructional affairs. In
contrast to the other three cases, the principal at School D did not formulate a narrative
that interpreted accountability demands for her faculty. She herself was not sure what
to make of the system. Teachers were left to their own sense making. Accountability
was seen by many of the staff as incompatible with the school’s philosophy of student-
centeredness and professional criteria of good teaching. Indeed, classroom
observations reveal a higher quality of instruction particularly in the dimensions of
task complexity and active teaching formats, relative to most of the other nine schools.
 There were few requirements to work collectively, nor were teachers expected to Accountability
follow closely any of the state-adopted textbook series or remedial literacy programs. obligations
Alignment activities, prevalent in almost all the other schools were largely absent here.
For example, running remedial prescriptive programs for below-grade level students
and curtailing electives, customary in other schools, were specifically rejected by
faculty members. There was ample room for faculty to learn and disagree, if they so
chose. Faculty members cooperated with each other informally, but no systematic 715
structures were in place that involved the faculty collectively. Due to declining test
scores, the faculty had recently become aware of the pressures and potential threats
that could result from the school’s unwillingness to focus on state assessments and its
inability to raise the API sufficiently to meet state targets. Apparently, district
administrators had sent the message that the school’s abysmal performance on state
assessments could not be tolerated much longer. The faculty swayed between the
desire to maintain its open curricula and individualistic collegial culture and a sense of
doom and surrender. But the school was not organized enough to take a principled
collective stance, either in opposition to the accountability system or in bridging
accountability obligations with student-centered philosophies and what they perceived
to be their students’ needs. As a result the school was unbalanced and beginning to
waver and skid. Across the nine schools, the school posted the lowest gains on the state
performance index over several years and was one of the lowest performing middle
schools in the state for its demographic profile.
For the School D faculty, integrity means to maintain one’s personal and
professional commitments in the face of inimical pressures and controls coming from
the system that are perceived as a disservice to student needs. It is the only school in
the sample that exhibits signs of resistance. Observed instructional quality would
seem to give the school a professionally legitimate base from which to justify this
resistance, but lack of leadership and collective purpose makes the striving for moral
or philosophical integrity an exercise in individualistic defensiveness. In the parlance
of integrity theorists, the school lacks integration. Lack of integration deters the
school from exercising prudence to insure its survival in the sanctions-driven
accountability environment, nor does the school benefit from investigating possible
equity deficiencies that a serious consideration of its external obligations, i.e. raising
test scores, might have surfaced.

Conclusion
It is quite conceivable that ideological zeal, Machiavellian strategizing, or eager system
conformism may produce more forceful agency than integrity, either in resisting
external demands that have been found wanting or embracing them for desired
optimal effect. But as everyday responses they are not as realistic, ethical, or
productive as the striving for integrity. Ideological zeal and resistance, as a matter
of course, are discouraged and negatively sanctioned in an institution whose function
is to socialize children into the established ways of a society. Eager conformism, on
the other hand, though perhaps being rewarded by the system, uneasily rubs against
the institution’s moral horizon which summons a spectrum of human values that
supersede the authority of any one policy, administrative decision, or adopted program.
In the American tradition, as in many other liberal and pluralist societies, educators are
called upon to reflect on their personal responsibilities and the needs of children as felt
by teachers and articulated to them in day-to-day interactions. All four, equalization,
child-centeredness, system efficiency, and professionalism are part of the institutional
JEA inheritance from which schools can collectively draw, and are called upon to draw,
50,5 when they make sense of authoritative system demands, craft coherence with internal
goals and operations, and exert the effort to provide morally acceptable service to
students. This does not mean that a given school always will.
The case vignettes in conjunction with the quantitative data show schools that
strive in different ways to develop or maintain their integrity in the face of
716 incontrovertible accountability demands. For one school, accountability demands, internal
commitments, and student needs are largely interpreted through an overarching morality
of efficiency and equalization. But an uneasy concern for children’s multiple (subject
matter) interests and curiosities, not well served in the aligned structure, remains as
an open worry that diminishes teachers’ sense of rightfulness. In another school, integrity
is mainly about sense of self-worth and reaffirmation of professionalism in the face of
which accountability judgments are relegated to second place. Implicitly, student needs
are served best by reaffirmed professionals, though this is not subject to explicit
examination. In a third school, integrity is relatively weakly developed. Teachers
maintain a defensive posture against the demands of their principal to align to the system,
not so much in opposition to the system per se, as in opposition to the moral discredit of
their principal’s leadership. But this defensive posture is not public. Being submerged
and informal, it derives its strength from teachers’ sense of being closer to students
than the external agents imposing on them. In the fourth school, opposition to the
system is an articulated philosophical and moral stance and justified on the grounds
that neither professional values nor student needs are served well by the system. But
this is accompanied by a troublesome denial of external obligations and accountability
realities.
The nine schools, selected from a wide performance spectrum within the
accountability system, shed light on the tight constraints within which public school
educators must strive for integrity. None of the schools can afford to ignore the high-
stakes system, and all nine schools but one, which pays for its resistance with
dangerously low test score gains, have responded to the new system controls with
technical alignment of curricular programs and instructional strategies along the lines
of School B’s approach. Across the nine schools, technical alignment and authoritative
consensus come as default reactions. But some schools, more than others, go further by
striving to maintain their educational integrity within this tight structure of control,
most notably one by critically embracing it and one by holding its negative threat at
bay. In each case, integrity was associated with an expansion of agency that combined
moral earnestness with prudent strategizing and actively constructing interpretive
frames that maintained a school’s sense of self-worth.
To be sure, integrity is a fragile quality under these circumstances. Corruption and
fragmentation, the opposites to the moral and integrative dimensions of integrity, are
definite possibilities, as exhibited by the two described schools, respectively, that
function with a compromised moral core (E) or with principled, but fragmented
opposition to the system (D). But fragility is at stake in the schools with higher
integrity as well. Examination of instruction, unease with the narrowness and
tightness of prescriptions, and openness to dissent help School B to remain sensitive to
student needs not easily accommodated in the standardized programmatic structure,
but system rewards that accrue to schools with the highest test score gains could easily
undermine such sensitivities. Reaffirming professional values in the face of negative
accountability sanctions may only temporarily remove the threat of fragmentation
in School C if scores do not continue to go up, or may become a mere justification of
occupational self-interest, particularly if it is coupled with insufficient attention to actual Accountability
student needs. obligations
Integrity challenges leaders to take risks, for example, to deliberately make the
accountability goals a secondary concern or keep questioning the rightfulness of tight
alignment. But the risks may well be worth it. Across the nine schools it appears
that integrity may be a more productive response to external accountability demands
than conformism or strategic alignment. Integrity is associated with an inner strength 717
around values and external obligations and, to a lesser degree to be sure, around
perceived student needs. It is cultivated within a relatively stronger, more open and
more coherent, faculty culture and with stronger leadership, attributes that have
consistently been identified in the school improvement literature as desirable
characteristics of improving schools (Louis, 2007; Bryk et al., 2010; Stoll and Fink,
1996). Across the nine schools, those that bridged accountability obligations, teachers’
goals and values, and their perceptions of student needs with a stronger sense of
integrity tended to fare better in the accountability system. They had a more positive
outlook on the system, by either embedding accountability demands into explicit
concern for student needs or by not privileging accountability obligations at the
expense of internal goals and perceptions of needs.
In this study of nine schools, the relationship between integrity and educational
quality is inconclusive. Two indicators were investigated, API gains over two years
(largely based on standardized test scores) and observed instructional quality. In
absolute terms, neither of these indicators seems to be clearly associated with integrity.
School B appears to be a prime example of relating integrity to strength in both
indicators. But how much of the test score gains can be attributed to the school’s tight
alignment pattern and how much to integrity is unclear. The keen observation of
student learning and instruction which seems to have benefited instructional quality,
however, is more easily attributed to integrity. By contrast, School C exhibited
relatively low instructional quality and its API gains were not higher than those of
schools with a much weaker integrity pattern. On the other hand, the school’s
conscious choice to demote accountability demands to a lower place in its normative
order did not result in lost test score growth relative to the other schools in the sample.
A longitudinal design may have been able to investigate if a school’s added sense of
agency due to better integrity over time contributes to better instructional quality. But
this is beyond the scope of the study. School D is another striking case. This school
exhibited relatively higher instructional quality, but abysmally low API and API
growth[7]. Thus, under these circumstances, any school climate or culture variable can
only have an ambiguous relationship to educational quality. And this ambiguity is
exactly what makes integrity a compelling concept.
After more than a decade of test-based and sanctions-driven school accountability
in the USA, we have accounts of astounding turn-around in schools and painful
distortions (Anagnostopoulos, 2006; Au, 2007; Booher-Jennings, 2005; McNeil, 2000;
Mintrop, 2004; Skrla and Scheurich, 2003; Reyes et al., 1999). In the literature, we read
about schools and districts that are energized and those that are stymied by the
system; those that raise expectations as well as those that retrack their “demotes;”
those in which content becomes fragmented and those in which it is expanded; those
that reinforce care and commitment and those that settle on triage or exclusion. Given
this wide spectrum of responses, it stands to reason that non-systemic factors, beliefs,
and commitments that develop outside of the logic of the system, substantially
influence whether the accountability system produces educationally desirable effects,
JEA and distortions are avoided. One such non-system factor, the nine-school sample
50,5 suggests, may be the degree to which school leaders and school faculties strive toward
collective integrity. Whether integrity develops or survives seems to require a good
dose of educational leaders’ personal strength, but may also depend on the profession’s
insistence to fully exhaust the moral horizon of an institution which obligates
educators to balance equity, system efficiency, child-centeredness, and professionalism
718 with prudence.

Notes
1. The uniqueness of the American constellation becomes apparent when one hypothesizes
high-stakes accountability in the context of educational institutions that clearly submerge
the needs of children under the authority of the state, the leading role of the teacher, or the
reigning fundamentalist ideology, as was the case, for example, in the former east Germany
(Mintrop, 1996).
2. See Appendix 1 for a table displaying student and teacher perception data.
3. See Appendix 1 for a table displaying internal capacity measures.
4. These are perception ratings. This does not mean that these perceptions adequately reflect
reality. Qualitative data show that School D is programmatically less influenced by system
demands than School B, even though perception ratings of system importance are similar.
A similar disconnect between perceptions and reality may occur in School C where high
ratings indicating high regard for student needs in the reality of schools are coupled with
relatively low ratings in instructional quality.
5. School D is an interesting border case. This school lacks this balancing. But here relatively
lower balance and lower cohesion is coupled with higher openness. I will discuss this pattern
in more detail further down with qualitative data.
6. Figures 1-3 are meant as merely illustrative displays. The trend lines are based on only
nine cases or data points. While scatter plots were investigated, they are not shown here. The
trend line display has the purpose of demonstrating the “bundle” of extant relationships, not
to show a calculated correlation.
7. The tenuous relationship between test score gains and other educational quality indicators
has been investigated in a previous article (see Mintrop and Trujillo, 2007a, b).

References
Achinstein, B. and Ogawa, R. (2006), “(In)fidelity: what the resistance of new teachers reveals
about professional principles and prescriptive educational policies”, Harvard Educational
Review, Vol. 76 No. 1, pp. 30-63.
Anagnostopoulos, D. (2006), “ ‘Real students’ and ‘true demotes’: ending social promotion and
the moral ordering of urban high schools”, American Educational Research Journal, Vol. 43
No. 1, pp. 5-42.
Anagnostopoulos, D. and Rutledge, S. (2007), “Making sense of school sanctioning policies in
urban high schools”, Teachers College Record, Vol. 109 No. 5, pp. 1261-302.
Au, W. (2007), “High-stakes testing and curricular control: a qualitative metasynthesis”,
Educational Researcher, Vol. 36 No. 5, pp. 258-67.
Banks, J. and Banks, C. (2009), Multicultural Education: Issues and Perspectives, Wiley,
San Francisco, CA.
Beck, U., Giddens, A. and Lash, S. (1994), Reflexive Modernization: Politics, Tradition And
Aesthetics In The Modern Social Order, Stanford University Press, Stanford, CA.
Benjamin, M. (1990), Splitting the Difference: Compromise and Integrity in Ethics and Politics, Accountability
University Press of Kansas, Lawrence, KS.
obligations
Booher-Jennings, J. (2005), “Below the bubble: “Educational triage” and the Texas accountability
system”, American Educational Research Journal, Vol. 42 No. 2, pp. 231-68.
Bryk, A. and Schneider, B. (2002), Trust in Schools: A Core Resource for Improvement, Russell
Sage Foundation Publications, New York, NY.
Bryk, A., Sebring, P., Allensworth, E. and Luppescu, S. (2010), Organizing Schools for Improvement: 719
Lessons from Chicago, University of Chicago Press, Chicago, IL.
Buchmann, M. (1989), The Script of Life in Modern Society: Entry into Adulthood in a Changing
World, University of Chicago Press, Chicago, IL.
California Department of Education (2006), “API description: overview of the academic
performance index (API)”, available at: www.cde.ca.gov/ta/ac/ap/apidescription.asp (accessed
October 15, 2007).
Callahan, R. (1970), Education and the Cult of Efficiency: A Study of the Social Forces that have
Shaped the Administration of the Public Schools, University of Chicago Press, Chicago, IL.
Campbell, E. (2008), “The ethics of teaching as a moral profession”, Curriculum Inquiry, Vol. 38
No. 4, pp. 357-85.
Carnoy, M., Elmore, R. and Siskin, L. (Eds) (2003), The New Accountability: High Schools and
High Stakes Testing, RoutledgeFalmer, New York, NY.
Carter, S. (1996), Integrity, Basic Books, New York, NY.
Coburn, C. (2006), “Framing the problem of reading instruction: using frame analysis to uncover
the microprocesses of policy implementation”, American Educational Research Journal,
Vol. 43 No. 3, pp. 343-79.
Consortium on Chicago School Research (2003), Survey of Chicago Public School Teachers, Spring
2003, Elementary Teacher Edition, Consortium Report, Chicago, IL.
Cox, D., La Caze, M. and Levine, M. (2003), Integrity and the Fragile Self, Ashgate Publishing Ltd,
Farnham.
Cremin, L. (1961), The Transformation of the American School, Knopf, New York, NY.
Creswell, J. (2003), Research Design: Qualitative, Quantitative, and Mixed Methods Approaches,
2nd ed., Sage Publications, Thousand Oaks, CA.
Diamond, J. and Spillane, J. (2004), “High stakes accountability in urban elementary
schools: challenging or reproducing inequality?”, Teachers College Record, Vol. 106 No. 6,
pp. 1145-76.
Dobel, J.P. (1999), Public Integrity, Johns Hopkins University Press, Baltimore, MD.
Elmore, R. (2004), School Reform from the Inside Out: Policy, Practice, and Performance, Harvard
Educational Publishing Group, Cambridge, MA.
Fuhrman, S. and Elmore, R. (2004), Redesigning accountability systems for education, Teachers
College Press, New York, NY.
Fullan, M. (2003), The Moral Imperative of School Leadership, Corwin Press, Thousand Oaks, CA.
Fuller, B. (2003), “Education policy under cultural pluralism”, Educational Researcher, Vol. 32
No. 9, pp. 15-24.
Gardner, H. (1999), The Disciplined Mind: What All Students Should Understand, Simon & Schuster,
New York, NY.
Gilbert, D. (2008), The American Class Structure in an Age of Growing Inequality, Pine Forge
Press, Newbury Park, CA.
Goldring, E. and Rallis, S. (1993), Principals of Dynamic Schools: Taking Charge of Change,
Corwin Press, Newbury Park, CA.
JEA Graham, P. (1993), “What America has expected of its schools over the past century”, American
Journal of Education, Vol. 101 No. 2, pp. 83-98.
50,5
Green, J., Camilli, G. and Elmore, P. (2006), Handbook of Complementary Methods in Education
Research, Lawrence Erlbaum Associates, Mahwah, NJ.
Hamilton, L., Berend, M. and Stecher, B. (2005), “Teachers’ responses to standards-based
accountability”, working paper, Rand, Santa Monica, CA.
720 Hargreaves, A. (2004), “Distinction and disgust: the emotional politics of school failure”,
International Journal of Leadership in Education, Vol. 7 No. 1, pp. 27-41.
Heclo, H. (2008), On Thinking Institutionally, Paradigm Publishers, Boulder, CO.
Herman, J. (2004), “The effects of testing on instruction”, in Fuhrman, S. and Elmore, R. (Eds),
Redesigning Accountability, Teachers College Press, New York, NY, pp. 141-6.
Hodgkinson, C. (1991), Educational Leadership: The Moral Art, State University of
New York Press, New York, NY.
Honig, M. and Hatch, T. (2004), “Crafting coherence: how schools strategically manage multiple,
external demands”, Educational Researcher, Vol. 33 No. 8, pp. 16-30.
Hursh, D. (2003), “Discourse, power and resistance in New York: the rise of testing and
accountability and the decline of teacher professionalism and local control”, in
Sattersthwaite, J., Atkinson, E. and Gale, K. (Eds), Discourse, Power, Resistance:
Challenging the Rhetoric of Contemporary Education, Trentham, London, pp. 43-56.
Ingersoll, R.M. (2003), Who Controls Teachers’ Work?: Power and Accountability in America’s
Schools, Harvard University Press, Cambridge, MA.
Jacob, B. (2005), “Accountability, incentives, and behavior: the impact of high-stakes testing in
the Chicago schools”, Journal of Public Economics, Vol. 89 Nos 5-6, pp. 761-97.
Johnson, S.M. (1990), Teachers at Work: Achieving Success in our Schools, Basic Books, New York, NY.
Joseph, P. and Burnaford, G. (2001), Images of School Teachers in America, Routledge, Florence, KY.
Kim, J. and Sunderman, G. (2005), “Measuring academic proficiency under the no child left behind
act: implications for educational equity”, Educational Researcher, Vol. 34 No. 8, pp. 3-13.
Kliebard, H. (1987), The Struggle for the American Curriculum, 1893-1958, Routledge, Florence, KY.
Koretz, D. (2008), “Measuring up: what educational testing really tells us”, Harvard Educational
Press, Cambridge, MA.
Labaree, D. (2006), “Power, knowledge, and the rationalization of teaching: a genealogy of the
movement to professionalize teaching”, Teacher Education: Professionalism, Social Justice
and Teacher Education, Vol. 62 No. 2, pp. 123-54.
Lareau, A. (2003), Unequal Childhoods: Class, Race, and Family Life, University of California
Press, Berkeley, CA.
Leithwood, K., Steinbach, R. and Jantzi, D. (2002), “School leadership and teachers’ motivation to
implement accountability policies”, Educational Administration Quarterly, Vol. 38 No. 1,
pp. 94-119.
Louis, K.S. (2007), “Trust and improvement in schools”, Journal of Educational Change, Vol. 8
No. 1, pp. 1-24.
Louis, K.S., Dretzke, B. and Wahlstrom, K. (2010), “How does leadership affect student
achievement? Results from a national US survey”, School Effectiveness and School
Improvement, Vol. 17 No. 1, pp. 1-22.
Louis, K.S., Febey, K. and Schroeder, R. (2005), “State-mandated accountability in high schools:
teachers’ interpretations of a new era”, Educational Evaluation and Policy Analysis, Vol. 27
No. 2, pp. 177-204.
McFall, L. (1987), “Integrity”, Ethics, Vol. 98 No. 1, pp. 5-20.
McLaughlin, M. and Talbert, J. (1993), Contexts that Matter for Teaching and Learning: Strategic Accountability
Opportunities for Meeting the Nation0 s Educational Goals, Center for Research on the
Context of Secondary School Teaching, Stanford, CA. obligations
McNeil, L. (2000), Contradictions of Reform: The educational costs of standardization, Routledge,
New York, NY.
Metz, M.H. (1978), Classrooms and Corridors: The Crisis of Authority in Desegregated Secondary
Schools, University of California Press, Berkeley, CA. 721
Meyer, H. (2006), “The rise and decline of the common school as an institution: taking “myth and
ceremony” seriously”, in Meyer, H.D. and Rowan, B. (Eds), The New Institutionalism in
Education, State University of New York Press, New York, NY, pp. 51-66.
Miles, M. and Huberman, A. (1994), Qualitative Data Analysis: An Expanded Sourcebook, 2nd ed.,
Sage Publications, Thousand Oaks, CA.
Mintrop, H. (1996), “Teachers and changing authority patterns in eastern German schools”,
Comparitive Education Review, Vol. 40 No. 4, pp. 358-76.
Mintrop, H. (2004), Schools on Probation: How Accountability Works (and Doesn’t Work),
Teachers College Press, New York, NY.
Mintrop, H. and Sunderman, G. (2009), “Predictable failure of sanctions-driven accountability for
school improvement – and why we may retain it anyway”, Educational Researcher, Vol. 38
No. 5, pp. 353-64.
Mintrop, H. and Trujillo, T. (2007a), “The practical relevance of accountability systems for school
improvement: a descriptive analysis of California schools”, Educational Evaluation and
Policy Analysis, Vol. 29 No. 4, pp. 319-52.
Mintrop, H. and Trujillo, T. (2007b), School Improvement Under Test-Driven Accountability: A
Comparison of High and Low Performing Middle Schools in California, National Center for
Research on Evaluation, Standards, and Student Testing, Los Angeles, CA.
Mintrop, H. and Trujillo, T.M. (2005), “Corrective action in low performing schools: lessons for
NCLB implementation from first-generation accountability systems”, Education Policy
Analysis Archives, Vol. 13 No. 4, available at: epaa.asu.edu/ojs/
Montefiore, A. and Vines, D. (1999), Integrity in the Public and Private Domains, Routledge, London.
Neckerman, K. (2004), Social Inequality, Russell Sage Foundation, New York, NY.
Newmann, F., Smith, B. and Allensworth, E. (2001), “Instructional program coherence: what it is
and why it should guide school improvement policy”, Educational Evaluation and Policy
Analysis, Vol. 23 No. 4, pp. 297-321.
Pace, J. and Hemmings, A. (2007), “Understanding authority in classrooms: a review of theory,
ideology, and research”, Review of Educational Research, Vol. 77 No. 1, pp. 4-27.
Paine, L. (1994), “Managing for organizational integrity”, Harvard Business Review, Vol. 72 No. 2,
pp. 106-26.
Proefriedt, W. (2008), High Expectations: The Cultural Roots of Standards Reform in American
Education, Teachers College Press, New York, NY.
Popkewitz, T. (1994), “Professionalization in teaching and teacher education: some notes on its
history, ideology, and potential”, Teaching and Teacher Education, Vol. 10 No. 1, pp. 1-14.
Ramsay, H. (1997), Beyond Virtue: Integrity and Morality, St. Martin’s Press, New York, NY.
Reeves, D. (2000), Accountability in Action: A Blueprint for Learning Organizations, Advanced
Learning Press, Denver, CO.
Reyes, P., Scribner, J. and Scribner, A. (1999), Lessons from High-Performing HISPANIC Schools:
Creating Learning Communities, Teachers College Press, New York, NY.
Rorty, S. (1999), “Integrity: political, not psychological”, in Montefiore, A. and Vines, D. (Eds),
Integrity in the Public and Private Domains, Routledge, London, pp. 126-57.
JEA Rutledge, S. (2010), “Contest for jurisdiction: an occupational analysis of principals’ responses to
test-based accountability”, Leadership and Policy in Schools, Vol. 9 No. 1, pp. 78-107.
50,5 Schein, E. (1985), Organizational Culture and Leadership, Jossey-Bass, San Francisco, CA.
Selznick, P. (1994), The Moral Commonwealth: Social Theory and the Promise of Community,
University of California Press, Berkeley, CA.
Sergiovanni, T. (1992), Moral Leadership: Getting to the Heart of School Improvement, Jossey-
722 Bass, San Francisco, CA.
Skrla, L. and Scheurich, J. (2003), “Displacing deficit thinking in school district leadership”,
Education and Urban Society, Vol. 33 No. 3, pp. 235-59.
SRI International, Policy Studies Associates and Consortium for Policy Research in Education (2003),
Evaluation of Title I Accountability Systems and School Improvement Efforts (TASSIE),
2002-03, US Department of Education’s Planning and Evaluation Service, Menlo Park, CA.
Steele, C. (1988), “The psychology of self-affirmation: sustaining the integrity of the self”, in
Berkowitz, L. (Ed.), Advances in Experimental Social Psychology, Vol. 21, pp. 261-302.
Stoll, L. and Fink, D. (1996), Changing our Schools: Linking School Effectiveness and School
Improvement, Open University Press: Buckingham, NY.
Supovitz, J. (2009), “Can high-stakes testing leverage educational improvements? Prospects from
the last decade of testing and accountability reform”, Journal of Educational Change,
Vol. 10 Nos 2-3, pp. 211-27.
Turiel, E. (2005), “Thought, emotions, and social interactional processes in moral development”,
in Killen, M. and Smetana, J. (Eds), Handbook of Moral Development, Erlbaum, Mawah, NJ,
pp. 7-36.
Tyack, D. (1974), The One Best System: A History of American Urban Education, Harvard
University Press, Cambridge, MA.
Valenzuela, A. (Ed.) (2005), Leaving Children Behind: Why “Texas-style” Accountability fails
Latino Youth, State University of New York Press, Albany, NY.
Yin, R. (2003), Case Study Research: Design and Methods, 3rd ed., Sage, Beverly Hills, CA.

Further reading
Cohen, D. and Ball, D. (1999), “Instruction, capacity and improvement”, Research Report No. RR-043,
Consortium for Policy Research in Education, University of Pennsylvania, Philadelphia, PA.
Education Trust (2003), Zap the Gap: Gap Closing Strategies in High-Performing Classrooms,
Schools, Districts and Colleges, Education Trust, Washington, DC.
Lipsky, M. (2010), Street-Level Bureaucracy: Dilemmas of the Individual in Public Service, Russell
Sage Foundation Publications, New York, NY.
Meyer, H. and Rowan, B. (2006), The New Institutionalism in Education, State University of
New York Press, New York, NY.
Weick, K. (1995), Sensemaking in Organizations, Sage, Thousand Oaks, CA.
Appendix 1 Accountability
obligations
F D I C H G E B

Teacher-reported parental support (range: 7-32) 13.9 17.0 17.9 17.7 18.5 20.1 18.6 19.1
Student-reported familial support (range: 6-24) 16.8 18.2 16.9 17.7 16.9 17.3 17.9 17.0 723
Student-reported possession of cultural goods
(range: 1 ¼ none, 4 ¼ all) 2.2 2.2 2.1 2.1 2.1 2.1 2.0 2.0 Table AI.
Student-reported frequency of non-English Teacher and student
home language perceptions of family
(range: 1 ¼ never, 4 ¼ always) 3.0 2.7 2.7 2.9 3.3 3.2 3.0 3.4 background (scale means)

A B C D E F G H I

Total years teaching 12.6 11.5 15.1 9.2 11.1 10.1 9.3 13.4 17.1
Highest degree completed
(percent responding
“above BA”) 32 26 33 41 66 28 46 49 21
Fully certified (percent “yes”) 81 84 90 95 93 96 93 89 89
Sense of preparedness
(percent responding Table AII.
“adequate” or Internal teacher
“very well”) 82 83 98 74 82 100 85 85 89 capacity (means)

A B C D E F G H I
Table AIII.
Positive teacher tone 50 80 59 84 70 50 57 81 59 Classroom observations
Proactive instruction 48 50 15 60 26 27 25 36 47 in percent of observed
Cognitive complexity 40 53 33 51 29 21 12 29 44 snapshots
JEA Appendix 2
50,5
Factor
Integrity pattern loading

724 Good balance

How important should these forces be?
District and state demands
Student needs
Teachers’ values and goals
How important are these forces in reality at your school?
District and state demands
Student needs
Teachers’ values and goals
Scores calculated based on differences between like items
Pulling together
At this school, when it comes to meeting the challenges
of reaching our API or AYP targets, administrators and teachers
are on the same side 0.799
Facing the pressures of school accountability has brought
the faculty together; almost everyone is making a contribution 0.895
The pressures of meeting API or AYP targets have
strengthened the hand of those at the school who are interested
in good teaching 0.836
Reliability (Cronbach’s a) ¼ 0.80
Shared responsibility for performance
In your judgment, how many teachers at this school
Help maintain discipline in the entire school? 0.730
Take responsibility for improving the school? 0.875
Set high standards for themselves? 0.886
Are eager to try new ideas? 0.871
Feel responsible to help each other do their best? 0.861
Feel responsible when students in this school fail? 0.715
Reliability (Cronbach’s a) ¼ 0.90
Collegiality
Most of my colleagues share my beliefs and values about what
the central mission of the school should be 0.763
There is a great deal of cooperative effort among staff here 0.875
I can count on colleagues here when I feel down about my
teaching or my students 0.805
In this school, the faculty discusses major decisions and sees
to it that they are carried out 0.760
Reliability (Cronbach’s a) ¼ 0.81
Learning orientation
My job provides me with continuing professional stimulation and growth 0.657
Teachers in this school continually learning and seeking new ideas 0.812
The staff seldom evaluates its programs and activities (values are reversed) 0.603
Teachers at this school respect those colleagues who are expert at their craft 0.804
The most expert teachers in their field are given leadership roles at this school 0.739
Reliability (Cronbach’s a) ¼ 0.76
Table AIV.
Teacher survey scales (continued)
 Factor
Accountability
Integrity pattern loading obligations
Open communication
Open discussions about the meaningfulness of
the state accountability system and related
district policies are encouraged 0.823 725
Faculty gatherings provide a forum to discuss different perspectives
on school improvement 0.880
It is okay to speak up when you disagree with the powers that be 0.862
Teachers are mainly encouraged rather than told to implement new
programs or policies 0.792
Reliability (Cronbach’s a) ¼ 0.86
Leadership
Managerial leadership
The principal sets priorities, makes plans, and sees
that they are carried out 0.738
The principal puts pressure on teachers to get results 0.715
In this school, the principal tells us what the district and
state expect of us, and we comply 0.856
Reliability (Cronbach’s a) ¼ 0.64
Supportive leadership
The school administration’s behavior toward the staff
is supportive and encouraging 0.929
The principal usually consults with staff members before s/he
makes decisions that affect teachers 0.904
Staff members are recognized for a job well done 0.905
Reliability (Cronbach’s a) ¼ 0.90
Moral leadership
The administration at this school
Places the needs of children ahead of personal and political interests
Models the kind of school they want to create
r ¼ 0.75
Instructional leadership
The administration at this school
Makes clear to the staff their expectations for
meeting instructional goals 0.759
Sets high standards for teaching 0.860
Understands how children learn 0.831
Sets high standards for student learning 0.841
Broadly shares leadership responsibility with the faculty 0.684
Carefully tracks student academic progress 0.751
Monitors and evaluates the quality of teaching in a way
that is meaningful for teachers 0.800
Allocates resources and other supports according
to the school’s goals and standards 0.746
Reliability (Cronbach’s a) ¼ 0.91
Perceptions of accountability system
Guidance/focus
State standards, tests, and performance targets
Provide a focus for my teaching 0.857
Tell us what is important for this school to accomplish 0.883
0.761
(continued) Table AIV.
JEA Factor
50,5 Integrity pattern loading

Have made us concentrate our energy on

instruction and student learning
Reliability (Cronbach’s a) ¼ 0.77
726 Validity
The state assessments assess all of the things I find
important for students to learn 0.788
A good teacher has nothing to fear from the
state accountability system 0.775
The state assessments reflect just plain good teaching 0.843
Reliability (Cronbach’s a) ¼ 0.72
Fairness
For the most part, teachers are unfairly judged by the accountability system
(values are reversed) 0.750
I resent being judged based on school-wide test scores and the performance of other
teachers (values are reversed) 0.679
All schools in California have a fair chance to succeed within the state accountability
system 0.643
The accountability system is stacked against schools located in poor communities
(values are reversed) 0.719
Our students are not behind because of the teachers they have, but because of the
conditions in which they have to grow up (values are reversed) 0.760
Table AIV. Reliability (Cronbach’s a) ¼ 0.75

Corresponding author
Heinrich Mintrop can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints