Chapter 20

Behavioral Approaches to
Business and Industrial
Problems: Organizational
Behavior Management
William B. Abernathy

Applied behavior analysis is based on B. F. Skin- performances improved and others’ did not. The
ner’s operant learning principles. Skinner’s seminal remaining 10% of the articles reported no effect.
work in this regard was Science and Human Behav- Of these articles, feedback alone was the most
ior (Skinner, 1953). Applied behavior analysis common OBM application. Examples of feedback
encompasses many settings and types of applica- include supervisor vocal comments, written narra-
tions, especially clinical and educational applica- tives, and graphic feedback. Performance improved
tions. Organizational behavior management (OBM) more when the feedback was graphic and delivered by
is a subfield of applied behavior analysis that supervisors or managers. Feedback was most effective
focuses on applications in the workplace. I begin when combined with rewards, goal setting, or both.
this chapter with a broad-strokes summary of the Alvero, Bucklin, and Austin (2001) replicated
outcomes of OBM projects as reported in two liter- and expanded Balcazar et al.’s (1986) review by ana-
ature reviews and a new review conducted for this lyzing OBM projects in 68 organizational settings on
chapter. Next, I provide a brief history of OBM, the the basis of articles in the Journal of Organizational
learning principles underlying OBM applications, Behavior Management for the period 1985 through
the OBM implementation process, performance 1998. Consistent effects were found in 58% of the
analysis strategies, the organizational behavior sys- projects, mixed effects in 41%, and no effects in 1%.
tems perspective, and future directions and collab- Combining both reviews, the most frequently
orations for OBM. used method was feedback alone. Feedback was
most commonly provided by the supervisor or man-
ager. Individual feedback was used most, but group
Organizational Behavior
feedback yielded the most consistent results in both
Management Results
reviews. In the Balcazar et al. (1986) review, graphic
Balcazar, Hopkins, and Suarez (1986) reviewed arti- feedback was the most common and yielded the
cles for the period 1974 through 1984 and found 69 most consistent results. In the Alvero et al. (2001)
OBM applications in organizational settings. The review, written feedback was most common,
articles were found in four journals: Academy of whereas a combination of written and graphic feed-
Management Journal, Journal of Applied Behavior back yielded the most consistent results. In the first
Analysis, Journal of Applied Psychology, and the review, feedback was most often provided daily, and
Journal of Organizational Behavior Management. All the most consistent results were obtained with daily
projects measured individual performance. Of the or weekly feedback. In the second review, feedback
articles, 41% reported consistent performance was most commonly delivered weekly and produced
improvements in all individuals. Another 49% more consistent positive results (80% of studies)
reported mixed effects in which some individuals’ than daily feedback (71%). In Alvero et al., monthly

DOI: 10.1037/13938-020
APA Handbook of Behavior Analysis: Vol. 2. Translating Principles Into Practice, G. J. Madden (Editor-in-Chief)
501
Copyright © 2013 by the American Psychological Association. All rights reserved.
William B. Abernathy

feedback also yielded consistent results in 80% of team member was awarded $100 if performance on
the studies reviewed. a target measure improved at least 30% over base-
For the purpose of this chapter, I reviewed a line. The results are depicted in Figure 20.1. On
sample of 68 OBM studies published in the Journal average, performances improved by 58%, with no
of Organizational Behavior Management from 1984 to performances deteriorating at the end of the 90-day
2009. Studies selected were conducted in organiza- intervention. Thus, across three literature reviews,
tional settings and reported improvement statistics. and in the analyses depicted in Figures 20.1 and
Table 20.1 shows that across all 68 studies, the OBM 20.2, varied types of evidence have supported the
interventions improved performance by a median of position that OBM interventions are effective in
54.5%. The settings for the interventions included improving employee performances.
33 in manufacturing, 24 in retail, and 11 in not-for-
profits. The types of interventions included 16 based
Beginnings of Organizational
on antecedent stimuli (e.g., training), 28 with feed-
Behavior Management
back only, one that arranged nonmonetary conse-
quences, and 23 that arranged monetary The beginnings of OBM are often traced back to
consequences. As in the prior reviews, feedback was Owen Aldis’s 1961 article “Of Pigeons and Men,” in
generally effective in improving performance. which he made the case for applying operant psy-
In a chapter in the book Organizational Change chology principles to the workplace. In 1969, Wal-
(Abernathy, 2001), I examined 2,195 performance ter Nord expanded the case for OBM in his article
measures in client organizations that began using “Beyond the Teaching Machine: The Neglected Area
feedback and performance pay. He found a composite of Operant Conditioning in the Theory and Practice
trend of a 33% increase over a 12-month period. The of Management.” One of the first articles to describe
composite was created by converting the raw data for an actual application of OBM in business was
each measure to standard scores, then averaging the authored by Brethower and Rummler (1966).
standard scores. In the same era, the application of operant princi-
More recently (Abernathy, 2010), I compared 68 ples to training and instruction led to the founding
project baseline performances in a medium-sized bank of the National Society for Programmed Instruction,
with performance levels at the end of 90 days. Target now the International Society for Performance
measures were selected by manager–subordinate Improvement, with more than 1,000 members.
teams from previously developed performance Much of this early performance-based employee
matrices. Examples of these measures included per- training was developed at the University of Michigan
centage of days balanced, percentage of accounts between 1961 and 1969 (Dickinson, 2001). Rum-
past due, accounts opened per hour, and percentage mler left the University of Michigan in 1969 to
of audits completed on time. The intervention con- found the consulting firm Praxis Corporation with
sisted of an incentive pay plan linked to the perfor- Tom Gilbert, who later authored Human Competence:
mances specified in the matrices. In addition, each Engineering Worthy Performance (Gilbert, 1978).

Table 20.1

Summary of Percentage Gains by Project Intervention

Intervention No. of studies Average (%) SD (%) Minimum (%) Maximum (%)
Overall 68 90.4 110 1.8 625
Antecedents 16 78.1 110 1.8 418
Feedback 28 120 133 3.0 625
Nonmonetary 1 89
Incentive pay 23 63.3 69.6 3.0 300

502
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

700
600

% Improvement
Over Baseline
500
400
300
200
100
0
1 10 20 30 40 50 60
OBM Projects ordered by Percent Improvement

Figure 20.1.  Baseline-to-intervention percentage improvements

in 68 organizational behavior management (OBM) projects. From “A
Comprehensive Performance Analysis and Improvement Method,” by W. B.
Abernathy, 2010, Performance Improvement, 49, p. 15. Copyright 2010 by
the International Society for Performance Improvement. Adapted with per-
mission of John Wiley & Sons, Inc.

Dale Brethower introduced the behavior systems 1972). Frederiksen and Johnson (1981) have argued
perspective with his 1972 book Behavioral Analysis that Feeney likely “opened the door for the wide-
in Business and Industry: A Total Performance Sys- spread application of OBM” (p. 193). Examples of the
tem. Paul Brown founded the consulting firm work of Feeney and his consultants can be found in
Instructional Design Associates, Inc., in 1971. He the handbook Industrial Behavior Modification
provided OBM-oriented management workshops at (O’Brien, Dickinson, & Rosow, 1982).
IBM and coauthored the book Behavior Modification In 1977, the Journal of Organizational Behavior
in Business, Industry and the Government (Brown & Management was first published by the previously
Presbie, 1976). mentioned consulting firm Behavioral Systems, Inc.,
In 1972, Aubrey Daniels, a clinical psychologist, with Aubrey Daniels serving as editor. In 1982, the
and Fran Tarkenton, quarterback of the Minnesota Organizational Behavior Network was founded, and
Vikings and owner of Learning Foundation, founded in 1988 the first conference devoted exclusively to
Behavioral Systems, Inc. Their early work was with OBM was held in cooperation with the Florida Asso-
Milliken Textiles and other southern textile firms ciation of Behavior Analysis.
and focused on employee attendance and turnover.
In 1978, Daniels founded Aubrey Daniels and Asso-
Basic Learning Principles and
ciates, which is now Aubrey Daniels International.
Organizational Behavior
In 1973, Ed Feeney founded Edward J. Feeney
Management: Measurement,
and Associates, a consulting firm for businesses.
Feedback, and Consequences
Feeney was a sales manager for Emery Air Freight
and attended a University of Michigan workshop on As an applied psychology, OBM is unique in that
the management of behavior change. Feeney then all of its principles and practices can be directly
implemented OBM programs throughout Emery Air traced to the operant psychology of B. F. Skinner
Freight. Feeney’s (1976) seminar workbook was (1953). OBM is conceptually and methodologically
titled Behavioral Engineering and Systems Training. grounded in behaviorism and the science of
These applications were reported widely, including behavior.
in Business Week (“Where Skinner’s Theories
Work,” 1972); Organizational Dynamics (“At Performance Measurement
Emery Air Freight,” 1973); and in the film Business, In OBM, the dependent variable is observable
Behaviorism and the Bottom Line (CRM Productions, employee behavior. Skinner (1938) recommended

503
William B. Abernathy

that the basic measure of behavior be its frequency or produce the desired result. For example, a banking
rate, although measuring duration, computing client wanted to increase the number of customer
latency, and reporting other dimensions of observable accounts opened by new accounts clerks (a result).
behavior in the workplace have also become common. The bank’s management thought the clerks could
best improve sales by increasing their use of cross-
Behavior versus results.  A critical performance selling techniques. For example, when a customer
measurement issue in OBM relates to what Skinner opened a checking account, the clerk would suggest
(1953) termed the free operant. In Skinner’s original also opening a savings account. As new account
work with animals, response rate was not recorded. clerks focused on these specific behaviors (rather
Rather, the rate of bar-press switch closures was than on results), the use of cross-selling techniques
recorded. The topography of the rat’s response (e.g., increased, but the number of new accounts opened
pressing with the right paw, left paw, or chin or sit- declined at several branch banks. With the emphasis
ting on the bar) was typically not recorded—only on the behavior of cross-selling, many clerks
the actual closure of the switch. Skinner referred to ignored established customers and others who
this class of functionally related, but topographically appeared unable to open multiple accounts.
distinct, responses as an operant because they oper- A second example was in bank collections. The
ate on the environment to produce a common result result of interest was to reduce the delay in collect-
(in this case, the switch closure). ing payments on outstanding loans and credit cards.
This distinction between operants, or response Management thought the problem was that the col-
classes, and specific responses is a key measurement lectors were not making enough phone calls. This
issue in the workplace. In most cases, the results of focus on the specific behavior thought to underlie
an employee’s performance are more important to the result led management to measure the number
the organization than are the details of the behavior of calls made by collectors each month. As a conse-
that produced the result. Edward J. Feeney’s (1976) quence, the collectors began to call friends and fam-
Behavioral Engineering and Systems Training recom- ily rather than customers with past due accounts.
mended that results should be defined as a “noun Needless to say, this increase was not accompanied
followed by a past tense verb” (p. 20). For example, by a decrease in the number of days that loan collec-
letter typed, machine repaired, or call completed are tions were outstanding. Table 20.2 is excerpted from
results that should be measured rather than the Daniels and Rosen (1983) and summarizes the dis-
behaviors of typing, repairing, or calling. tinctions between behavior and results measures.
Measuring results avoids targeting behaviors that There are exceptions to the principle of measur-
may have unintended effects and may not ultimately ing results. For example, behaviors must be directly

Table 20.2

Distinguishing Results From Behaviors

Behavior Results
What people are doing What people have produced
What you see people do when they are working What you see after people stop working
Must see people working Not necessary to see people working
Tends to be expressed in present tense, verbs ending in ing Tends to be expressed in the past tense by noun–adjective
pairings: “document filed”
Cue words: by, through, to Cue words and phases: in order to, so that, to achieve, to be able to
Commonly used terms: input, process, activity, means Commonly used terms: output, result, outcome, achievement, ends

Note. From Performance Management: Improving Quality and Productivity Through Positive Reinforcement (p. 80), by A. C.
Daniels and T. A. Rosen, 1983, Tucker, GA: Performance Management Publications. Copyright 1983 by Aubrey Daniels
International. Adapted with permission.

504
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

measured when prevention is the goal. Much of the The Balanced Scorecard: Translating Strategy Into
excellent OBM work in the area of safety involves Action. Balancing addresses problems associated
counting safe behaviors (in which a checklist is used with one-dimensional performance measurement
to identify and quantify specific response topogra- plans. If feedback and recognition, or especially
phies, e.g., lifting a heavy object with a straight monetary incentives, are made contingent on one
back) rather than counting the resulting accidents performance dimension, performances on other
or injuries (see review by Grindle, Dickinson, & dimensions may deteriorate. For example, measur-
Boettcher, 2000). One could measure pilot safety by ing productivity may cause quality or safety to suf-
the number of airplane crashes, but a more reason- fer. Conversely, exclusively measuring error rates
able approach would be to monitor pilot behavior may result in a decline in production.
via a preflight checklist. A solution is balanced measurement, in which
A second exception is behavioral style. Adminis- management, supervisors, employees, or all of these
trators of an organization may want customer ser- identify potential conflicting behaviors and include
vice agents to greet people with specific behavior them in a performance scorecard (Kaplan & Norton,
topographies such as greeting them by name, smil- 1996). An example of one such approach is the per-
ing, or making eye contact. Salespeople may be formance matrix (Felix & Riggs, 1986) shown in
expected to use techniques that have proven effec- Figure 20.2. Unlike other scorecard formats, the
tive in the past or that comply with company poli- matrix mathematically balances multiple perfor-
cies. Under these conditions, measures of ongoing mances by assigning a percentage weight to each.
behavior would be preferable to measures of the The weights are expressed as percentages (summing
results of those behaviors. to 100%) and reflect the relative importance of each
measure in the overall performance of the employee
Measure actionable results.  Another critical
or unit. Balance is also ensured by capping the per-
measurement issue in workplace settings is to mea-
centage scores that may be earned for any measure.
sure results that are actionable by the employee.
For example, if the deadlines-met goal for the
Measures become less controllable by the individual
month was 95% and the employee completed 100%
employee as more variables influence the result or
of the deadlines assigned, he or she would still
as more employees are involved in producing the
receive only the percentage score indicated in the
result. Profit is generally a poor performance mea-
weight column of the performance matrix (40% in
sure because it is made up of several income and
Figure 20.2). Capping prevents an employee from
expense items, many of which a specific employee’s
focusing on one performance and driving it above
behavior may have little or no influence on. In addi-
the goal to allow him or her to ignore other
tion to the individual employee, all other employees
performances.
in the organization have some influence on this
measure, which significantly decreases the contin-
gent relationship between the employee’s behavior
Performance Feedback
and the result.
Although the concept of feedback is not used much
Balanced measurement.  A final measurement in operant psychology, the term is common in
issue is balanced measurement, as popularized by OBM. Performance feedback refers to information
Kaplan and Norton (1996) in their influential book that follows a performance and describes the level or

Measure Name Base Goal Weight

Items / Hour 20 40 15%
Accuracy % 90% 100% 25%
% Deadlines Met 80% 95% 40%
Safety Checklist % 70% 100% 20%

Figure 20.2.  Sample performance matrix.

505
William B. Abernathy

quality of the performance. Positive feedback may 0.06

serve as a discriminative stimulus that may occasion

Standard Score
the performance under similar conditions in the 0.04

future because it describes the performance that is

0.02
likely to produce reinforcement. It may also
strengthen the behavior on which it is contingent.
0.00
When feedback strengthens the behavior on which A B C D E F G H
it is contingent through the process of reinforce-
-0.02
ment, the positive feedback is probably functioning Organization
as a generalized or conditioned reinforcer, because
positive feedback is often followed by back-up rein- Figure 20.3.  Effects of monthly versus quar-
terly performance feedback on 12-month stan-
forcers such as public recognition, preferred job
dard score trends. From Pay for Profit: Creating
assignments, preferred work shifts, annual perfor- an Organization-Wide Performance System
mance reviews, promotions, and pay. The reinforc- (p. 173), by W. B. Abernathy, 2011, Atlanta,
ing value of feedback probably depends on the GA: Performance Management Publications
(ISBN 0-9655276-1-1). Copyright 2011 by
feedback’s being followed by other reinforcers. If Aubrey Daniels International, Inc. Adapted
this fails to occur, then positive feedback is likely to with permission.
lose its reinforcing effect. If this interpretation is
correct, then providing employees feedback that is 0.025
not linked to primary reinforcers will ultimately fail
0.020
Standard Score

to sustain performance.
0.015
Immediacy of Feedback
0.010
The immediacy of feedback is a key concern for
organizational behavior analysts. Many organiza- 0.005
tions provide formal feedback to employees only
0.000
once a year through annual performance reviews. In Team Individual
basic and applied research, reinforcers are generally
more effective when they occur immediately after Figure 20.4.  Effects of providing feedback
to teams or individual employees. Twelve-month
the behavior. In contrast, if feedback is viewed pri- standard score trends are shown. Adapted from
marily as a prompt for future reinforcers, then feed- Pay for Profit: Creating an Organization-Wide
back can be effective even when somewhat delayed. Performance System (p. 183), by W. B. Abernathy,
I studied feedback across 2,195 performance mea- 2011, Atlanta, GA: Performance Management
Publications (ISBN 0-9655276-1-1). Copyright
sures in eight organizations (Abernathy, 2011). 2011 by Aubrey Daniels International, Inc.
Figure 20.3 shows these organizations’ composite Adapted with permission.
monthly performance trends for the first 12 months
after performance matrix feedback was provided to Individual Versus Team Feedback
individual employees. These composite measures Figure 20.4 (Abernathy, 2011) shows a comparison
were developed by converting all measures to stan- of performance improvements when feedback was
dard scores and averaging them. Organizations A, D, provided about the team’s collective results versus
E, G, and H in Figure 20.3 provided feedback to results produced by individual employees. Again,
employees on a monthly basis, whereas Organiza- the constraints of conducting research in organiza-
tions B, C, and F provided feedback quarterly. tional settings do not allow for randomized controls,
Although the study was not randomly controlled, but these findings suggest that providing feedback
these findings suggest that monthly feedback was about individual employees’ performance has a big-
superior to quarterly feedback in its impact on ger impact than does providing feedback based on
employee performance trends. team performance.

506
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

Performance Consequences Negative reinforcement.  A key mission of many

Skinner’s (1938) operant psychology is grounded in OBM practitioners is to transition organizations
the general theory of selectionism (see Chapter 10, this from a reliance on negative reinforcement (coercion;
volume). Unlike cause-and-effect models that focus on “do it or else”) to the use of positive reinforcement.
motives and attitudes that goad the individual into Although most people think of wages and salary as
action, selectionism finds that behavior is controlled positive reinforcers, Skinner (1953) suggested that
by events that occur after the behavior. According to they might actually function as negative reinforcers.
the operant account, voluntary behavior is not driven Hourly wages and monthly salaries do not meet the
or elicited but rather emitted and then selected by the definition of positive reinforcement because they are
environment through its consequences. Behavior that not paid contingent on employee performance. They
produces positive consequences is selected by the are paid on the basis of a predetermined time period
environment, behavior with no consequences is not (e.g., a paycheck every 2 weeks). The “dead man rule”
selected by the environment, and behavior with aver- applies: If an employee died, and no one informed
sive consequences is selected against. This selectionist payroll, he or she would still receive a paycheck.
philosophy is equally at home in the free market econ- According to Skinner’s (1953) analysis, employ-
omy (e.g., A. Smith, 1776) and in evolutionary biology ees do not work to receive a wage or salary; they
(e.g., Darwin, 1859). work to avoid losing it. The employee need not
Taking a selectionist approach to behavior means experience this directly. The contingent relation
the OBM technician will be especially interested in may be stated in a contract or observed from the
determining the consequences of employee behav- consequences of other employees’ poor perfor-
iors. If target performances (e.g., completing tasks mance. Many managers do not see a problem with
more quickly) have no consequences (e.g., manage- negative reinforcement. “If they don’t perform, we
ment never acknowledges them) or aversive fire them!” is an all-too-common response to the
consequences (e.g., fellow employees chastise the recommendation that the organization arrange posi-
employee for making them look bad), then improve- tive reinforcement contingencies to manage
ment techniques that do not alter these consequences employees.
(e.g., training, work methods improvements) may The case for businesses to transition from nega-
fail to sustain performance or prove only partially tive to positive reinforcement is good. First, in busi-
successful. ness settings dominated by negative reinforcement
As outlined in Table 20.3, Skinner (1938) orga- (i.e., avoiding getting fired), management fails to
nized the consequences of behavior into four catego- measure performance regularly or precisely. Instead,
ries: positive reinforcement, negative reinforcement, management emphasizes time spent working rather
positive punishment, and negative punishment. than results. When a company begins discussing tra-
Each consequence is defined operationally in terms ditional compensation arrangements, it is amazing
of whether the event is added or removed after the how much the conversation concerns time: over-
behavior and whether the performance increases or time, undertime, time clocks, time logs, time in
decreases as a consequence of this operation. grade, vacation time, sick time, compensatory time,

Table 20.3

Consequences of Behavior (Skinner, 1938)

Term Action Effect

Positive reinforcement Event added Responding increases
Negative reinforcement Event removed Responding increases
Positive punishment Event added Responding decreases
Negative punishment Event removed Responding decreases

507
William B. Abernathy

and so on. Little attention is given in these meetings early. Emotional responses such as fear or anger
to discussing actual employee performance. A slo- may interfere with customer interactions or preci-
gan that describes this issue is “If you pay for time, sion and creative work. Murray Sidman (1989)
you get time. If you pay for results, you get results.” discussed problems associated with negative rein-
In a time-based wage and salary system, the forcement in his book Coercion and Its Fallout.
employee learns to fill the day and look busy in an A third issue, created by an exclusive reliance on
attempt to avoid a host of aversive consequences negative reinforcement, is poor performance. When
(e.g., being yelled at by the boss, being fired). As I a performance standard is adopted, it is in the
often tell clients, however, “Busyness isn’t business.” employee’s best interests to meet the standard but
For example, a banking client of mine was experi- not to exceed it. Meeting the standard allows the
encing long delays in processing credit card applica- employee to avoid aversive events, but exceeding
tions. The management of the area was unwilling to the standard leads to no further gain (i.e., no posi-
consider performance pay, so we suggested that tive reinforcers) and may often result in aversive
when all the credit cards scheduled for processing in events (e.g., management may increase the standard,
a given day were completed, the employees could go and fellow employees may chide the employee for
home and still receive a full day’s pay. Management making them look bad). A similar process occurs if
initially objected that this practice would be too after employees achieve a goal, it is increased. Dan-
expensive until it was pointed out that the compen- iels and Daniels (2004) have referred to these phe-
sation cost per card would be the same. Manage- nomena as losses of discretionary effort.
ment reluctantly agreed to try the plan, and it was A final issue with negative reinforcement is
subsequently explained to a group of somewhat expense. Managing through negative reinforcement
skeptical employees. requires a high level of expensive direct supervision,
On the following Monday, all the cards were pro- whereas pay for results requires less. In a negative-
cessed by noon, and everyone went home (i.e., effi- reinforcement workplace, the supervisors must
cient credit card processing was reinforced). This spend their time directly observing employees to
continued for a month, and the delay in card pro- ensure they are performing while on the clock. By
cessing was cut by two thirds with a reduction in contrast, paying employees for results may be used
overtime pay. At the end of the month, despite its without ever observing the ongoing performance,
success, management terminated the program, argu- making it perfect for employees who work at home
ing that the employees had been intentionally filling or in distant locations. The result of managing
their days to keep their jobs and receive overtime. through positive reinforcement is often that manag-
This observation was correct, but management ers may supervise more employees, thereby reduc-
blamed the employees rather than the faulty rein- ing management expenses (Van Fleet & Bedian,
forcement system for it. 1977). For example, Lincoln Electric, a manufac-
A second issue, created by the use of negative turer of arc welders in Cleveland, Ohio, was
reinforcement–based wage schemes, relates to founded with a performance pay system rather than
behaviors that are natural responses to aversive con- a wage and salary system. Lincoln Electric’s manag-
trol (recall that negative reinforcement involves ers’ supervision span of control averaged 100
escape from or avoidance of aversive events). These employees per manager (Jackson & Schuler, 2003).
reactions to aversive control include aggression,
unsanctioned instances of escape and avoidance, Positive reinforcement.  One example of positive
and emotional responses (Azrin, Hutchinson, & reinforcement in organizational settings is perfor-
McLaughlin, 1967). Possible aggressive responses mance pay. There is much confusion in the busi-
include complaining, lawsuits, conflicts, sabotage, ness world about performance pay. Many business
and theft. Unsanctioned escape and avoidance may managers believe that the annual pay raise (merit
be evidenced by failures to carry out orders, absen- increase) is performance pay. However, most raises
teeism, tardiness, extended breaks, and leaving are annual and typically determined by subjective,

508
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

after-the-fact appraisals based on the manager’s rec- subjective evaluations of management had previ-
ollection of the employee’s performance. As such, ously indicated.
these appraisals are often based on more recent Partly because some organizations are resistant
performance or on employee attitude rather than to performance pay, many OBM projects use non-
results. Likewise, annual bonuses are also thought monetary positive reinforcers. Another reason is that
to be examples of performance pay. However, some early OBM practitioners objected to using
these payments are delayed and fail to prescribe in money as a positive reinforcer. As mentioned earlier,
advance what the employee must do to earn them. in the 1970s Edward J. Feeney was one of the pio-
Profit sharing is sometimes viewed as performance neers in applying behavior analysis to the work-
pay, but it is an annual group plan whose measure is place. As a consultant with Edward J. Feeney and
largely out of individual employees’ control except Associates at that time, I was instructed that money
in very small organizations. Studies of profit sharing was not to be used as a positive reinforcer, only
in larger organizations have found no convincing supervisor recognition. The latter, it was argued,
relation between profit sharing and employee pro- was always accessible to management, which
ductivity (Binder, 1990). ensured that positive reinforcers would be delivered
Examples of effective performance pay plans immediately.
include piece rates, commissions, and goal sharing. Today, many OBM consultants provide manager
Piece rates and commissions pay for each unit of training in behavioral principles (e.g., operant con-
production or sales. Goal sharing establishes goals ditioning). The training focuses, in part, on how
for one or more measures. If the goal, or in some managers can use nonmonetary reinforcers to man-
cases a percentage of the goal, is achieved, the age their subordinates. Nonmonetary reinforcers
employee receives a payout. used in OBM have included supervisor and peer rec-
Unfortunately, many organizations are unwilling ognition, public postings of performance, recogni-
to attempt the transition from salaries to perfor- tion awards, and preferred work assignments or
mance pay. This unwillingness may be because (a) shifts, to name a few (see Nelson & Spitzer, 2003).
managers do not fully understand the benefits of Merchandise is frequently used as a more tangi-
positive reinforcement, (b) the wage and salary sys- ble reinforcer. Many plans use vendors that provide
tem is more familiar, or (c) the performance mea- gift catalogs with point requirements listed for
sures and tracking system required to implement a each gift. The employee earns points for improve-
performance pay system are viewed as too difficult ments or goal accomplishment that are then
to develop and implement. The reasons for this exchanged for gifts. In other cases, employees are
resistance may also be personal. In my experience, given time off with pay for good performance. An
managers are more resistant than workers to adopt- important issue occurs in the use of nonmonetary
ing performance pay systems; perhaps because they positive reinforcement. If the organization continues
are already receiving a good salary, they have more to use negative reinforcement as the primary means
financial risk than their subordinates. Moreover, of compensating employee performance, it may
managers may experience a loss of personal discre- prove difficult to sustain performance with nonmon-
tionary control in the transition from subjective etary programs because the behaviors reinforced by
evaluations and negative reinforcement toward the negative reinforcement may compete with those
objective measurements and positive reinforcement. that are positively reinforced. Moreover, the negative
When there is resistance from nonmanagement reinforcement allows the manager to revert back to
employees, it often comes from those subjectively conventional practices that do not require continu-
judged by management to be good performers. ous performance measurement or feedback.
These employees (e.g., those who get by on a good
attitude toward management) may worry that their Leveraged pay.  A key issue in performance pay is
status will fall if objective measurement of their per- whether it should be in addition to base pay (wage
formance reveals that they are not as good as the or salary) or should replace some or all of the base

509
William B. Abernathy

pay. Many plans add incentive pay on top of the 1. Conduct Performance Management
employee’s wage or salary. This approach has sev- Workshop
eral drawbacks. First, payroll expense may increase Performance management workshops are usually
without a concomitant increase in profits. This is presented onsite or at the consulting firm’s head-
especially the case with safety, customer service, and quarters. Workshops typically last between 1 and 5
other measures that may have no immediate impact days and focus on a behavioral approach to manag-
on profitability. As a result, the incentives offered ing employee performance. Specifically, the
may be minimal and, therefore, may have a limited workshops cover techniques for identifying and
impact on performance. Second, when incentives are measuring important outcomes, how to set up a
added to base pay, the employees continue to receive feedback system, goal setting, managing the conse-
market-comparable pay whether they participate in quences of performance, and dealing with unwanted
the incentive plan or not. The effort–reward ratio performances.
may be insufficient to encourage participation if the Workshop materials are often customized to the
incentive amount is negligible. client’s requirements and supplemented by standard
A different approach to performance compensa- instructional materials. For example, at the Aubrey
tion is leveraged pay. In a leveraged pay system, a Daniels consulting firm, the basic materials are pre-
portion of each employee’s pay is indexed to per- sented in Performance Management: Improving Quality
sonal performance and organizational profitability. and Productivity Through Positive Reinforcement (Dan-
As a result, employee pay becomes more of a vari- iels & Rosen, 1983) and in the later revised version
able expense that increases only with organizational (Daniels & Daniels, 2004). In some cases, the OBM
profitability. The economist Martin Weitzman consultant follows up the workshop by observing
(1984) pointed out that indexing pay to profit managers’ interactions with subordinates and coach-
increases the organization’s survivability, increases ing them on better applications of the techniques.
employee pay over competitors in good business Performance management focuses management’s
cycles, and reduces the need for layoffs in poor busi- attention on three events in the sequence of
ness cycles. In Abernathy (2011), I described a per- employee performance: events happening before the
formance pay system that indexes performance pay behavior (antecedents), the behavior, and events
opportunity to organizational profitability while dis- after the behavior (consequences). Daniels and
tributing the opportunities on the basis of individual Rosen (1983) referred to this as an A–B–C analysis,
and small-team performances. and it is known more broadly in behavioral psychol-
ogy as the three-term contingency.
Organizational Behavior
Antecedents.  Absent, infrequent, or poorly
Management Implementation
constructed prompts from the supervisor are
Process
often a performance constraint for subordinates.
I have used the OBM process enumerated next in Antecedents prompt behavior, guide behavior, and
more than 170 organizations in seven countries rep- allow for more precise behavior. They inform the
resenting a diverse array of sizes and industry types. employee as to the who, what, when, where, how,
The process closely mirrors the typical process used and sometimes why of a performance. A three-term
in OBM: contingency analysis suggests that antecedents
derive their ability to influence performances from
1. Conduct performance management workshop. their close association with consequences. Thus, the
2. Develop performance measures. employee may have a wonderful set of instructions
3. Identify critical improvement results and but never follow them because doing so does not
behaviors. result in positive reinforcement.
4. Link positive consequences to performances. A second type of antecedent discussed in perfor-
5. Conduct a performance analysis. mance management workshops is a performance

510
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

goal. Goals for the employee may be set as part of a 2. Develop Performance Measures
manager’s performance coaching. These goals may In Abernathy (2011), I performed a cluster analysis
change as the coaching process continues. As with on 2,195 objective client organizations’ performance
any antecedent, a goal’s ability to influence perfor- measures and found that they could be usefully par-
mance is determined by the association between titioned into seven measurement categories. The
meeting the goal and the consequences of doing so. categories were sales, expense control, productivity,
Thus, the most effective goal-setting techniques will cash flow, regulatory compliance, customer service,
tie goal accomplishment to performance evaluations and special projects. Measures for these categories
or performance pay. In these instances, the goal is can be defined for the organization and cascaded
typically set by management and is not negotiated. down through management to workers. Alterna-
Unlike a performance coaching goal, goal setting tively, measures for the categories can be defined for
done in connection with performance evaluation or a specific job position targeted by survey findings,
pay is usually consistent across performances and PIPs, or the manager’s personal objectives.
job positions. Goals may be statistically derived on All measurement categories may not apply to a
the basis of performance variability. Examples are given organization or job position. Moreover, some
the performance improvement potential (PIP) analy- categories may require two or more measures to
sis, in which the top performance becomes the goal. accurately reflect the category. Measurements in a
The goal may also be defined as a standard deviation category may be a simple count of the result or
or some percentage above historical performance. behavior per period of time (rate). In a few cases,
Goals may also be defined using a top-down variations in the topography of the behavior are
approach in which the sum of the goals contributes counted. The percentage of accuracy, duration,
to the achievement of a higher level goal of the latency, and turnaround time of the behavior or
organization. result are also potential measurement dimensions.
In many cases, a measure may be expressed as a
Feedback.  Feedback may be provided by the
ratio when differences in the opportunity to perform
supervisor through performance postings, coach-
exist. Percentage of accuracy ([outputs − errors] /
ing of individual employees, team meetings, or all
outputs) is typically a better measure than the num-
of these. Feedback may also be self-reported by the
ber of errors. For example, employee A has made 10
employee, who manually collects and organizes the
errors this month, and employee B has committed
data. Ineffective performance feedback from the
five errors. Which employee is the better performer?
supervisor may be a key constraint on employee
Given only this information, one would say that B
performance. Issues may be that feedback is absent,
was the better performer. However, one then finds
delayed, not actionable, imprecise, or negative (tells
that employee A processed 100 outputs, and
the employee what he or she is doing wrong rather
employee B processed 20. Employee A’s accuracy is
than what he or she needs to do to improve). As
(100 − 10)/100 = 90%, whereas employee B’s error
mentioned previously, feedback may serve as an
rate is (20 − 5)/20 = 75%.
antecedent that signals possible reinforcement or
Results may be expressed in dollars, volumes,
may function directly as a social reinforcer.
ratios, percentages, standard time, delays, or other
Consequences.  Ineffective use of performance formats. In some cases, ratios may be a poor mea-
consequences may also be a critical performance surement choice. For example, a measurement sys-
constraint. Consequences delivered by the supervi- tem was requested for bank loan officers. The bank
sor may be absent, aversive, delayed, or unpredict- was having trouble securing profitable loans. There-
able and can be social, nonmonetary, or monetary. In fore, management preferred using the average
some cases, simply removing aversive or conflicting income percentage per loan as a performance mea-
consequences improves performance. In other cases, sure for the officers. The problem with this ratio, or
reducing the delay and improving the certainty of any ratio, is that it is not sensitive to volume. For
the consequence is the solution. example, one loan officer might make two loans that

511
William B. Abernathy

total $1,000 and produce an income of $100, and Table 20.4

another officer might make 10 loans for $10,000
that produce an income of $500. The income per- Summary of Characteristics of Good Performance
centage of the first loan is $100 / $1,000, or 10%. Measures
The income percentage on the second loan is
$500/$10,000, or 5%. The income from the second Good measures Poor measures
loan portfolio is $500 − $100, or $400 more than Objective Subjective
the portfolio with the higher percentage income. Results Activities
Which is more important to the bank—the percent- Individual or small team Group
Actionable No action likely
age income or the total income? The answer might Related to strategy or profit Related to job description or
depend on the bank’s funds availability. local problem
Other alternatives would be to track both the Balanced Unbalanced
percentage income per loan and the total loan Aligned Conflicts with other measures
Immediate feedback Delayed feedback
income. These two measures would be included in One element Composite
the loan officer’s scorecard or performance matrix Reported by third party Self-reported
with appropriate weights attached to each result.
Alternatively, total net income could be tracked,
matrices on the basis of declining trends or subgoal
which is influenced by both the dollar volume of the
performances. As mentioned previously, one strategy
loan and the percentage income of the loan. Loan
for measuring the many components of these perfor-
net income is simpler to track and leaves the strat-
mances is to design and implement organizationwide
egy for increasing it up to the individual loan officer.
performance matrices using the method of cascading
The officer can focus on high-volume, low-income
objectives proposed by Kaplan and Norton (1996;
loans or lower volume, high-income loans.
see Figure 20.5). First, an organizational scorecard
The point of these examples is to illustrate that
(matrix) is developed by the executive team that lists
designing or selecting a measure for a results pin-
the organization’s two to seven key objectives for the
point includes many considerations and that mea-
coming year. These objectives are then stated for the
suring different aspects of a result will likely drive
entire organization as performance measures with
different employee behaviors. Many organizational
goals and priority weights. These measures then
incentive plans have failed to produce the desired
serve as a guide for designing the executive team’s
results simply because the behavior or result mea-
measures. This phase is followed by developing the
sure was defined incorrectly. Table 20.4 summarizes
executive team’s subordinate’s measures and so on
the characteristics of good performance measures.
down through the organization to the worker. Once
baseline data have been collected, the measures with
3. Identify Critical Improvement Results
the most deficient levels or trends are targeted for
and Behavior
improvement. The advantages of this approach are
In the sections that follow, I describe four techniques
that it is objective, it will reveal linked and depen-
for identifying critical improvement opportunities:
dent performances, and everyone will participate.
(a) an organization-wide performance measurement
The disadvantage is the time and effort involved in
system, (b) an organization-wide management prac-
an organization-wide initiative.
tices survey, (c) calculation of performance improve-
ment potentials, and (d) determination of critical Organization-wide management practices survey. 
improvement opportunities by the manager. A second strategy for identifying performances in
need of improvement is to have all employees in the
Organization-wide measurement system.  In this organization complete a management practices sur-
strategy, an organizationwide performance measure- vey (Abernathy, 2010). The survey asks employees
ment system is implemented before any improve- to rate the effectiveness of a variety of management
ment projects. Projects are then selected from the practices. Practices that are rated low are targeted for

512
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

Organizational
Scorecard

Customer
Sales Operations Finance HR
Service
Department Department Department Department
Department
Scorecard Scorecard Scorecard Scorecard
Scorecard

Receiving Collections Payroll

Sales Team Inquiry Team
Team Team Team
Scorecard Scorecard
Scorecard Scorecard Scorecard

Order Entry Shipping Payables Benefits

Sales Support
Team Team Team Team
Scorecard
Scorecard Scorecard Scorecard Scorecard

Purchasing Bookkeeping HR Rep

Team Team Team
Scorecard Scorecard Scorecard

Figure 20.5.  Method of cascading objectives. HR = human resources.

performance improvement projects. Although this meet with subordinates to identify a target result
strategy minimizes time and effort compared with and measure. Although this strategy minimizes the
the cascading strategy just discussed, it is based on costs associated with the other methods, the disad-
perceptions rather than actual data. It will also likely vantages are the manager training time and expense.
miss critical linkages and dependencies across job Another potential issue is that managers and their
positions. For example, the manager may be con- subordinates may select results or behaviors that can
cerned with better scheduling in the department, be improved easily rather than those that are of high
but changing the schedule may affect other perfor- value to the organization. As with the management
mances in the department or in other departments. practices survey discussed earlier, this technique
may also miss the linkages and dependencies across
Performance improvement potential.  A third
job positions.
improvement identification strategy was developed
by Tom Gilbert (1978). Target performances are
4. Link Consequences to Performance
identified through the computation of the PIP, which
Positive consequences should be made contingent
is the exemplar (best) performance divided by the
on performance to both improve and sustain it. As
average performance. For example, if the exemplar
previously discussed, these reinforcers may be
performance is 20, and the average performance is
social, tangible, or monetary. Examples of social
10, the PIP is 20/10 or 2.0. The underlying assump-
reinforcers are praise and recognition. Social rein-
tion is that the exemplar’s performance is a reason-
forcers likely function best when they are linked to
able potential because it has actually been achieved
tangible or monetary reinforcers. Examples of tangi-
by someone in the job. PIPs may be computed for
ble reinforcers are time off, lunches, and merchan-
all performances in a job position, department, or
dise. As discussed previously, the effectiveness of
organization, and the highest PIP ratios are targeted
monetary reinforcers can be enhanced by leveraging
for improvement.
base pay.
Manager determination.  A final strategy is to On the basis of her review of the literature,
allow managers trained in performance analysis to Dickinson (2005) suggested that “the critical

513
William B. Abernathy

determinant of performance is the contingent rela- procedures here because each brings important
tionship between pay and performance, rather than insights to the task of determining the cause of
the amount or percentage of incentive pay” (p. 22). underperformance.
Because pay cannot practically be immediately
delivered after most performances, the solution has Behavior engineering model.  In the behavioral
been to introduce immediate feedback that informs engineering model, the OBM consultant uses the
the employee of the forthcoming amount of perfor- six cells shown in Table 20.5 to identify probable
mance pay. causes of employee underperformance. Three of the
Dierks and McNally (1987) reported an applica- cells relate to the employee’s environment, and three
tion with bank proof operators. They introduced relate to the individual. The environment and indi-
daily supervisor feedback that increased the item- vidual cells are organized with respect to the ante-
processing rate from 980 items per hour to 1,650 cedent stimulus, behavior, and consequence issues
items per hour (net of errors). However, with a discussed earlier.
change in supervisors, the new supervisor failed to Gilbert (1978) recommended that the environ-
calculate and distribute the feedback. Performance ment cells be considered before the individual cells
dropped almost to its original level. To prevent this when attempting to determine the probable cause of
problem in the future, performance pay was underperformance. Therefore, an examination of
installed, and the rate increased to 2,300 items per missing or incomplete antecedents (instructions and
hour for a 3-year period. The final program was a guides), behaviors (resources, tools, processes), and
pay-per-item (piece rate) plan that replaced the consequences (incentives) would precede an analy-
hourly pay. Performance rose to 3,100 items per sis of individual knowledge, selection procedures, or
hour, where it remained for 12 years until the bank employee motives.
was acquired. Both piece rate and its analog, sales When Dean (1994) asked approximately 1,000
commissions, have the advantage of providing participants from business and industry to identify
immediate earnings to the employee, thus combin- which one of Gilbert’s (1978) six cells was the big-
ing feedback and monetary reinforcement. gest obstacle to their optimal performance, 75.6% of
the participants selected one of the three environ-
5. Conduct a Performance Analysis mental cells. This finding is interesting because it
Performance analysis is a procedure for determining indicates that training, employee selection, and gen-
the causes of performance deficiencies. Several pro- eral employee motivation are perceived as less com-
cedures have been used throughout OBM’s evolu- mon performance constraints than are information
tion. I present many of these performance analysis (prompts), resources, and incentives. This finding

Table 20.5

Gilbert’s (1978) Behavior Engineering Model

Source Stimulus Behavior Consequences

Environment Information: Job descriptions, Resources: Tools, time, Incentives: Monetary and
guides, feedback materials, access, staffing, nonmonetary incentives,
processes career development,
consequences for poor
performance
Individual Knowledge training Capacity: Selection procedures, Motives: Workers’ willingness to
flexible scheduling, job aids work for available incentives,
recruitment process

Note. From Human Competence: Engineering Worthy Performance (p. 88), by T. F. Gilbert, 1978, New
York, NY: McGraw-Hill. Copyright 1978 by John Wiley & Sons, Inc. Adapted with permission.

514
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

tentatively supports OBM’s primary focus on the investigations to gather information at all vantage
environment rather than the individual employee’s levels, (b) use the information to analyze problems,
capabilities and personal motivation. and (c) develop interventions consistent with the
values at all levels of vantage.
Performance analysis flowchart.  Mager and Pipe
(1984) developed a decision tree to assist the perfor- PIC/NIC analysis.  Daniels and Daniels’s (2004)
mance analyst in determining the causes of a defi- PIC/NIC analysis prompts the consultant to con-
cient performance. The key decision points in the sider three characteristics of effective consequences.
decision tree are much abbreviated in Figure 20.6. First, and for reasons discussed in detail earlier,
The order of the questions supports the previous positive (P) consequences are preferred over nega-
view that environmental performance constraints tive (N) consequences. Second, any consequence,
should be addressed before factors related to the be it positive or negative, will be more effective if
individual. it is delivered immediately (I) rather than after a
delay. Finally, consequences are more effective when
Vantage analysis chart.  The vantage analysis
they are certain (C) rather than probabilistic. Thus,
chart (J. M. Smith & Chase, 1990) was designed to
performances are least likely to reach optimal levels
help performance engineers analyze organizational
when consequences are negative, future, and uncer-
systems in an orderly fashion. The vantage analysis
tain. Behaviors with many antecedents (prompts)
chart has six different vantage points from which
are more likely to occur than behaviors with few
to analyze an organization: philosophical, social,
antecedents.
organizational, departmental, individual outcomes,
and individual activities. It is used to demonstrate Opportunity–capability–context analysis.  The
how performance engineers can (a) target their first step in this analysis (Abernathy, 2009) is to

Expectations clear? (Yes/No) Clarify expectations.

Resources sufficient? (Yes/No) Improve resource availability.

Feedback present? (Yes/No) Implement feedback.

Desired performance punished? (No/Yes) Remove punishment.

Poor performance rewarded? (No/Yes) Remove reward.

Effective consequences? (Yes/No) Implement positive reinforcement.

Skill deficiency? (Yes/No) Improve skills.

Task simplification? (Yes/No) Simplify task.

Person’s potential? (Yes/No) Improve employee selection procedure.

Figure 20.6.  Performance analysis decision tree.

515
William B. Abernathy

define organizational measures for each of the seven target performances. Analysis methods have varied
measurement categories previously described. The from simply asking the supervisor what the prob-
historical trends of these organizational measures lems are to more formal analyses, such as described
are computed. Declining trends, or trends con- earlier. Once the analysis is completed, a solution
sistently below goal, determine which categories that removes the constraint is designed and
should receive further analysis. The trends of lower implemented.
level (division, department, job position) measures The design, implementation, and administration
in the category are then reviewed for declines, or of projects are termed performance management.
being consistently below goal, to pinpoint target OBM solutions have focused on Daniels and Daniels’s
measures for improvement projects. (2004) antecedents and consequence of the behav-
The second step is determining the constraints ior, Gilbert’s (1978) environmental stimulus and
that are causing the target performance’s declining consequences, Mager and Pipe’s (1984) feedback
trend or suboptimal level. These constraints are cat- and consequences, Harless’s (1994) environmental
egorized as a lack of opportunity to perform, capa- and motivation and incentives, and Abernathy’s
bility issues, or context issues (see Figure 20.7). The (2009) context constraints.
constraint categories are considered in order from
left to right, because if there is no opportunity to
Organizational Behavior Systems
perform, capability is irrelevant, and if there is no
Management
capability, then context is irrelevant. Once the con-
straint category is determined, the most appropriate Organizational behavior systems management is an
method within the category is selected. For exam- alternative perspective and methodology to conven-
ple, if capability is the targeted constraint category, tional OBM. OBM tends to focus on the individual
then the subcategories competence, resources, and employee and the specific job context, whereas orga-
processes are examined. If it is determined that nizational behavior systems management examines
worker competence is the constraint, then either the total organizational contingencies. Two of the
worker selection or worker training methods are earliest presentations of this perspective were those
improved. of Dale Brethower (1972) in Behavioral Analysis in
The preceding performance analyses are con- Business and Industry and Harshbarger and Maley
ducted to pinpoint what factors are constraining the (1974) in Behavior Analysis and Systems Analysis.

Target
Measure

Opportunity Capability Context

Time Work Input Competence Resources Processes Prompts Feedback Consequences

Work Scheduling Work Flow Worker Supplies Job Manager Manager Manager
Selection Design Prompts Feedback Consequences

Work Utilization Worker Tools Job Organizational Organizational Organizational

Distribution Training Methods Prompts Feedback Consequences

Figure 20.7.  Performance constraint analysis. From “A Comprehensive Performance Analysis and
Improvement Method,” by W. B. Abernathy, 2010, Performance Improvement, 49, p. 9. Copyright 2010 by the
International Society for Performance Improvement. Adapted with permission of John Wiley & Sons, Inc.

516
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

Other behavior system advocates have been Krapfl If an organization is to successfully adopt
and Gasparato (1982), Senge (1990), Rummler and and implement OBM principles, it must provide
Brache (1995), Kaplan and Norton (1996), Mawhin- a management structure that supports a new
ney (2001), Redmon and Mason (2001), Malott reliance on positive reinforcement. This is best
(2003), and Abernathy (2007). accomplished by a transition from a traditional
Behavior systems have been described laterally closed, hierarchical management system (Figure
as inputs that are necessary for processes (e.g., 20.9) that relies on direct supervision and nega-
employee performance) that produce the company’s tive reinforcement to an open, network system
outputs (Brethower, 1972; Brethower & Rummler, (Figure 20.10) organized more as self-managed
1966; Krapfl & Gasparato, 1982; Malott, 2003; teams that are in direct contact with the conse-
Rummler & Brache, 1995). In other cases, behavior quences of their performances.
systems have been described vertically as job perfor-
mances that drive department performances, which Conclusions
then drive organizational performances (Kaplan & In this chapter I have outlined a general OBM
Norton, 1996). Figure 20.8 illustrates how both per- implementation process that includes setting up
spectives can be integrated (Abernathy, 2011). performance management workshops, developing
The integrated behavior system model considers performance measures, identifying improvement
the lateral work processes of an organization and its opportunities, linking consequences to performance,
vertical management hierarchy. The ultimate con- and conducting a performance analysis. Five issues
cern of management is with the organizational out- have been explored and are recapped in this con-
puts. If one of these outputs is deficient, the source cluding section of the chapter: (a) OBM effective-
is traced to a department or departments and ulti- ness, (b) issues with the three-term contingency,
mately to one or more job positions. The lateral (c) performance analysis and improvement,
inputs of the department or job may constrain the (d) behavior systems, and (e) OBM training.
employees’ opportunity to perform.
Human resources functions affect the organiza- Organizational Behavior Modification
tional behavior system. These functions include Effectiveness
organizational management structure, job defini- OBM is rooted in the operant learning principles
tions, orientation and training, performance evalua- of B. F. Skinner (1938) and grew out of applied
tion, promotion and termination policies, and behavior analysis to become an effective behavioral
compensation. The performance analyst must technology in the 1970s. OBM has demonstrated
ensure that these functions are aligned with positive results in 41% to 58% of the applications
employee, departmental, and organizational perfor- reviewed from 1974 through 1999 (Alvero et al.,
mances objectives. 2001; Balcazar et al., 1986). I found a median

ORGANIZATIONAL
PERFORMANCES OUTPUTS

INPUTS DEPARTMENT
PERFORMANCES OUTPUTS

JOB PERFORMANCES OUTPUTS

Figure 20.8.  Integrated behavior system model.

517
William B. Abernathy

Figure 20.9.  Closed, hierarchical behavior system.

Figure 20.10.  Open, network behavior system.

improvement of 54.8% across 68 studies from 1984 provided as vocal comments by the manager but
through 2009. These findings indicate that OBM may be graphic or written.
projects are generally quite effective in improving Behavior may be measured directly, or the result
employee performances in organizational settings. of the behavior may be measured. With some excep-
tions (e.g., safety, customer service), I recommend
Issues With the Three-Term Contingency measuring results first. If this proves ineffective,
The three-term contingency (antecedents–behaviors– then the behaviors that produce the result should be
consequences) has been used in many, if not most, measured. When results are measured, they should
applications. Antecedents have included behavior be actionable for the employee and balanced to
checklists, instructions, and guides, and conse- ensure improvement in one result is not detrimental
quences have included feedback, social reinforcers, to other dimensions of the result (e.g., quality vs.
tangible reinforcers, and performance pay. The most quantity) or other related results (e.g., sales vs.
common method used in OBM has been the intro- collections).
duction of performance feedback (Alvero et al., Feedback has been a common technique in OBM
2001; Balcazar et al., 1986). Feedback is typically partly because it can be applied immediately and

518
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

targeted to individuals more so than other types of organizational psychology. They concluded that
consequences. However, its effective delivery OBM and industrial–organizational psychology both
depends on the cooperation of the manager over focus on improving organizational effectiveness, but
long periods of time, which may not always occur their strengths and weaknesses differ. OBM tends
(Dierks & McNally, 1987). Moreover, if feedback to offer solutions to specific problems, whereas
does not reliably predict backup reinforcers (money, industrial–organizational psychology is more expan-
time off, promotions, etc.), it may be able to sive and conducts research on varied and complex
improve performances in the short term but prove organizational topics (e.g., job analysis, employee
unable to sustain them over the long term. selection techniques, survey research). OBM could
An important consideration for OBM technolo- also benefit greatly from sharing the techniques of
gists is that many traditional wage and salary com- industrial engineering, for example, work standards,
pensation systems tend to rely more on negative workflow, work processes, work scheduling, staff-
reinforcement to manage employee performance ing, and quality assurance.
than on positive reinforcement. As a result, man-
agers attend to exceptions or errors rather than to Organizational Behavior Systems
improvement or performance at goal. A reliance Management
on managing exceptions and negative reinforce- Ludwig Von Bertalanffy (1950) was an early propo-
ment allows the organization to manage employ- nent of general systems theory, which has since
ees without objectively measuring or consistently been applied in several academic fields such as
tracking employee performances. This culture archeology, anthropology, political science, bio-
must be overcome before OBM applications can be chemistry, ecology, and economics. Applications
truly effective over the long term. As a result, have included telecommunications, computer sci-
most large-scale OBM interventions include, or ence, control systems, electronics, environmental
should include, manager training in operant prin- engineering, and many others. With the emergence
ciples. Moreover, the compensation system should in OBM of the behavior systems perspective (Breth-
transition from conventional wages and salaries to ower, 1972; Harshbarger & Maley, 1974; Krapfl &
pay for performance to change the culture of Gasparato, 1982; Malott, 2003; Mawhinney, 2001;
exceptions and sustain the performance improve- Rummler & Brache, 1995), a collaboration between
ment initiative. systems engineering and OBM could prove quite
valuable.1 A first step would be to apply correla-
Performance Analysis and Improvement tional techniques such as factor analysis, spectral
Several performance analysis models were presented. analysis, and neural network analysis to organiza-
Most of the models have in common an expansion of tional systemwide data. Such analyses would assist
improvement interventions beyond the three-term in identifying critical behavioral interactions across
contingency model. For example, Gilbert (1978) departments as well as hierarchical levels of the
referred in his behavior engineering model to job organization.
description, tools, materials, staffing, employee
scheduling, processes, employee selection, and job
training; Abernathy (2010) added work flow, work
References
distribution and cross-training to this list. Abernathy, W. B. (2001). An analysis of the results and
structure of twelve organizations’ performance score-
Two disciplines that provide these additional card and incentive pay systems. In L. Hayes (Ed.),
tools are industrial–organizational psychology and Organizational change (pp. 240–272). Reno, NV:
industrial engineering. Bucklin, Alvero, Dickinson, Context Press.
Austin, and Jackson (2000) conducted an extensive Abernathy, W. B. (2008). Implications and applica-
comparison of the disciplines of OBM and industrial– tions of a behavior systems perspective. Journal of

A partial list of universities that offer advanced degrees in applied behavior analysis and OBM can be found on the OBM Network’s website (http://
1

www.obmnetwork.com).

519
William B. Abernathy

Organizational Behavior Management, 28, 123–138. Daniels, A. C., & Rosen, T. A. (1983). Performance man-
doi:10.1080/01608060802100980 agement: Improving quality and productivity through
Abernathy, W. B. (2009). Walden Two revisited: positive reinforcement. Tucker, GA: Performance
Optimizing behavioral systems. Journal of Management Publications.
Organizational Behavior Management, 29, 175–192. Darwin, C. (1859). On the origin of species by means
doi:10.1080/01608060902874567 of natural selection, or the preservation of favoured
Abernathy, W. B. (2010). A comprehensive performance races in the struggle for life. London, England: John
analysis and improvement method. Performance Murray.
Improvement, 49, 5–17. doi:10.1002/pfi.20146 Dean, P. J. (Ed.). (1994). Performance engineering at work.
Abernathy, W. B. (2011). Pay for profit: Creating an Batavia, IL: International Board of Standards for
organization-wide performance system. Atlanta, GA: Training, Performance, and Instruction.
Performance Management Publications. Dickinson, A. M. (2001). The historical roots of organi-
Aldis, O. (1961). Of pigeons and men. Harvard Business zational behavior management in the private sector:
Review, 39, 59–63. The 1950s–1980s. Journal of Organizational Behavior
Management, 20, 9–58. doi:10.1300/J075v20n03_02
Alvero, A. M., Bucklin, B. R., & Austin, J. (2001). An
objective review of the effectiveness and essen- Dickinson, A. M. (2005). Are we motivated by money? Some
tial characteristics of performance feedback in results from the laboratory. Performance Improvement,
organizational settings (1985–1998). Journal of 44(3), 18–24. doi:10.1002/pfi.4140440306
Organizational Behavior Management, 21, 3–29. Dierks, W., & McNally, K. (1987). Incentives you can
doi:10.1300/J075v21n01_02 bank on. Personnel Administrator, 32, 61–65.
At Emery Air Freight: Positive reinforcement boosts Feeney, E. J. (1976). Behavioral engineering and sys-
performance. (1973). Organizational Dynamics, 1, tems training. Redding, CT: Edward J. Feeney &
41–50. doi:10.1016/S0090-2616(73)80017-8 Associates.
Azrin, N. H., Hutchinson, R. R., & McLaughlin, R. Felix, G. H., & Riggs, J. L. (1986). Results by the objectives
(1967). Attack, avoidance, and escape reactions to matrix. Corvallis: Oregon Productivity Center.
aversive shock. Journal of the Experimental Analysis of
Behavior, 10, 131–148. doi:10.1901/jeab.1967.10-131 Frederiksen, L. W., & Johnson, R. P. (1981).
Organizational behavior management: Perspectives
Balcazar, F., Hopkins, B. L., & Suarez, Y. (1986). A on an emerging field. Journal of Organizational
critical, objective review of performance feedback. Behavior Management, 2, 193–203.
Journal of Organizational Behavior Management, 7,
65–89. doi:10.1300/J075v07n03_05 Gilbert, T. F. (1978). Human competence: Engineering wor-
thy performance. New York, NY: McGraw-Hill.
Binder, A. (Ed.). (1990). Paying for productivity: A look at
the evidence. Washington, DC: Brookings Institute. Grindle, A. C., Dickinson, A. M., & Boettcher, W. (2000).
Behavioral safety research in manufacturing settings:
Brethower, D. M. (1972). Behavioral analysis in business A review of the literature. Journal of Organizational
and industry: A total performance system. Kalamazoo, Behavior Management, 20, 29–68. doi:10.1300/
MI: Behaviordelia. J075v20n01_03
Brethower, D. M., & Rummler, G. A. (1966). For Harless, J. H. (1994). Performance quality improvement
improved work performance accentuate the positive. system. Newnan, GA: Harless Performance Guild.
Personnel, 43, 40–49.
Harshbarger, D., & Maley, R. F. (1974). Behavior analysis
Brown, P. L., & Presbie, R. J. (1976). Behavior modifica-
and systems analysis: An integrative approach to men-
tion in business, industry and government. New Paltz,
tal health programs. Kalamazoo, MI: Behaviordelia.
NY: Behavior Improvement Associates.
Jackson, S. E., & Schuler, R. S. (2003). Managing human
Bucklin, B. R., Alvero, A. M., Dickinson, A. M., Austin, J.,
resources through strategic partnerships (8th ed.).
& Jackson, A. K. (2000). Industrial-organizational
Cincinnati, OH: Southwestern.
psychology and organizational behavior man-
agement: An objective comparison. Journal of Kaplan, R. S., & Norton, D. P. (1996). The balanced score-
Organizational Behavior Management, 20, 27–75. card: Translating strategy into action. Cambridge, MA:
doi:10.1300/J075v20n02_03 Harvard Business Press.
CRM Productions. (Producer). (1972). Business, behaviorism Krapfl, J., & Gasparato, G. (1982). Behavioral systems
and the bottom line [VHS]. Carlsbad, CA: CRM Films. analysis. In L. W. Fredericksen (Ed.), Handbook of
Daniels, A. C., & Daniels, J. E. (2004). Performance man- organizational behavior management (pp. 21–38).
agement: Changing behaviors that drive organizational New York, NY: Wiley.
effectiveness. Atlanta, GA: Performance Management Mager, R. F., & Pipe, P. (1984). Analyzing performance
Publications. problems. Belmont, CA: Lake.

520
 Behavioral Approaches to Business and Industrial Problems: Organizational Behavior Management

Malott, M. E. (2003). Paradox of organizational change: Senge, P. (1990). The fifth discipline: The art and practice
Engineering organizations with behavior systems anal- of the learning organization. New York, NY: Currency
ysis. Reno, NV: Context Press. Doubleday.
Mawhinney, T. C. (2001). Organization-environment Sidman, M. (1989). Coercion and its fallout. Boston, MA:
systems as OBM intervention context: Minding Authors Cooperative.
you metacontingencies. In L. J. Hayes, J. Austin, R. Skinner, B. F. (1938). The behavior of organisms: An exper-
Houmanfar, & M. C. Clayton (Eds.), Organizational imental analysis. New York, NY: Appleton-Century-
change (pp. 137–165). Reno, NV: Context Press. Crofts.
Nelson, B., & Spitzer, D. (2003). The 1001 rewards and Skinner, B. F. (1953). Science and human behavior. New
recognition fieldbook: The complete guide. New York, York, NY: Free Press.
NY: Workman.
Smith, A. (1776). An inquiry into the nature and causes of
Nord, W. R. (1969). Beyond the teaching machine: The the wealth of nations. London, England: W. Stahan &
neglected area of operant conditioning in the theory T. Cadell.
and practice of management. Organizational Behavior
and Human Performance, 4, 375–401. doi:10.1016/ Smith, J. M., & Chase, P. N. (1990). Using the vantage
0030-5073(69)90017-8 analysis chart to solve organization-wide problems.
Journal of Organizational Behavior Management, 11,
O’Brien, R. M., Dickinson, A. M., & Rosow, M. P. (Eds.). 127–148. doi:10.1300/J075v11n01_09
(1982). Industrial behavior modification. New York,
NY: Pergamon Press. Van Fleet, D., & Bedian, A. (1977). A history of the span
of management. Academy of Management Review, 2,
Redmon, W. K., & Mason, M. A. (2001). Organizational 356–372.
culture and behavioral systems analysis. In C. M.
Johnson, W. K. Redmon, & T. C. Mawhinney (Eds.), Von Bertalanffy, L. (1950). An outline of general systems
Handbook of organizational performance: Behavior theory. British Journal for the Philosophy of Science, 1,
analysis and management (pp. 437–456). New York, 134–165.
NY: Haworth Press. Weitzman, M. L. (1984). The share economy: Conquering
Rummler, G. A., & Brache, A. P. (1995). Improving perfor- stagflation. Boston, MA: University Press.
mance: How to manage the white space on the organiza- Where Skinner’s theories work. (1972, December 2).
tional chart (2nd ed.). San Francisco, CA: Jossey-Bass. Business Week, pp. 64–65.

521