British Journal of Management, Vol.

00, 1–20 (2021)

DOI: 10.1111/1467-8551.12478

Corporate Culture and Innovation: A Tale

from an Emerging Market
Yaopeng Wang, Hisham Farag and Wasim Ahmad
Birmingham Business School, University House, Edgbaston Park Road, Birmingham, B15 2TY, UK
Corresponding author email: [Link]@[Link]

We examine the influence of corporate culture on innovation using a unique set of data
from Chinese listed companies over the period 2008–2017. Using the competing value
framework, we quantify corporate culture using textual analysis of financial statements.
We find a positive and significant impact of a creation culture on innovation measured by
both patent applications and citations, as well as innovation efficiency. We address endo-
geneity concerns and conduct a battery of robustness tests, including alternative proxies
for both corporate innovation and culture, and conclude that variations in culture have a
significant impact on firm-level innovation. We also show that a strong creation culture
is more likely to spur innovation for firms in more competitive product markets and firms
that are subject to higher managerial career concerns. We provide empirical evidence that
corporate culture is an important driver in enhancing innovation. Our results have clear
implications for directors and shareholders.

Introduction rewards long-term success (Manso, 2011). Culture

is among the primary levers at managers’ disposal
Innovation has been considered a vital factor to maintain organizational viability in the wake
for organizations that aim to achieve and sus- of high uncertainty associated with innovation.
tain a competitive advantage (Baer, 2012; Nelson Several studies intuitively argue that organiza-
and Winter, 1982; Porter, 1992). Undoubtedly, tional culture is a key to innovation success and
companies with better innovative capabilities can corporate performance (see Büschgens, Bausch
enhance their profitability (Atalay, Anafarta and and Balkin, 2013; Sackmann, 2002 for a review).
Sarvan, 2013; Tuan et al., 2016) and can lead to The evidence on the nexus between culture and in-
a country’s economic development (Solow, 1957). novation is based on small sample surveys calling
Therefore, understanding the drivers of a firm’s into question the generalizability and comparabil-
innovative ability is critical from both academic ity of the observed findings. Large sample studies
and industry perspectives. However, innovative beyond developed markets (Fiordelisi et al., 2019)
projects are often risky, unpredictable, lengthy and are scant.
involve significant resources (Holmstrom, 1989). Corporate culture refers to the values and be-
Therefore, stimulating innovation needs a strong liefs that provide norms for expected behaviour by
organizational culture that encourages risk-taking employees (Schein, 1992), and organizations ac-
and tolerance for failure in the short term, and tively promote these values and principles (Guiso,
Sapienza and Zingales, 2015). Culture is consid-
ered a fundamental asset to boost performance1
We are grateful to the Editor and Associate Editor and and a key value driver by business managers
three anonymous reviewers for their constructive feed-
back and suggestions throughout the review process. We
also thank participants of the 5th annual ICGS conference 1
[Link]
in Essex for their constructive comments. All remaining Deloitte/global/Documents/HumanCapital/
errors are the responsibility of the authors. [Link]

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British Academy
of Management. Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main
Street, Malden, MA, 02148, USA.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distri-
bution and reproduction in any medium, provided the original work is properly cited.
2 Y. Wang, H. Farag and W. Ahmad

(Graham et al., 2017). Anecdotal evidence from 2010). Moreover, it promotes agility, adaptability
regulatory reviews and business leaders suggests and creativity. Organizations with a rich creative
that the culture of an organization is set at the culture are likely to be more innovative (Anderson,
top management levels (CEO and board)2 and Potočnik and Zhou, 2014).3
this view is supported by the academic litera- We test the culture–innovation relationship
ture (Kotter and Heskett, 1992). Stakeholders, in on a unique dataset of Chinese listed firms. Our
contrast, play a remarkable role in shaping and firm-level measure of creation culture is based on
influencing corporate culture, particularly in the the competing value framework (CVF) (Cameron
post-global financial crisis era. This results in et al., 2006). We quantify CVF culture dimensions
an ever-increasing emphasis not only on promot- (create, compete, control and collaborate) using
ing ethical and professional standards through textual analysis of financial reports and measure
governance reform, but also on promoting green the intensity of creation culture in an organiza-
innovations for sustainable development. There- tion following Andreou et al. (2019). The basic
fore, understanding the drivers of a firm’s innova- proposition of textual analysis is that the words
tive ability is critical for sustainable developments and language used in official documents reveal the
in an ever-challenging institutional environment. culture of the organization (Levinson, 2003). Our
The competitive conditions in the global busi- choice of the Chinese market is motivated by two
ness environment demand firms pursue growth main reasons. First, China is the largest emerging
through innovation and creativity for long-term market and second largest economy in the world.
sustainability. Such a strategy requires a consis- A large part of China’s rapid economic growth
tent and continuous reliance on creative behaviour could be attributed to the innovative capability
across all levels within the organization (Bednall of Chinese firms.4 China provides an interesting
et al., 2018). Creative culture affects innovation as setting to examine the role of organizational
it shapes the patterns of novelty, individual initia- culture in fostering innovation outside developed
tives and risk-taking behaviours (Kaasa and Vadi, markets. Our findings may have implications for
other emerging economies competing for domi-
2
For instance, a UK Financial Reporting Council (FRC) nance in the global market (Fana, Weib and Xuc,
report: Corporate Culture and the Role of Boards (2016) 2011). Second, there are significant institutional
observes that ‘it is the board’s role to determine the pur- and cultural differences between the Chinese and
pose of the company and ensure that the company’s val-
ues, strategy and business model are aligned to it’. The developed markets (Schneider, 1988). Our study
report further adds that the top leaders, particularly the provides a first-hand test of the generalizability,
Chief Executive, ‘must embody the desired culture, em- comparability and applicability of the CVF in
bedding this at all levels and in every aspect of the busi- an emerging market with significant differences
ness’. Similarly, the UK Corporate Governance Code, in in corporate governance regime compared to the
its preface (supporting principle A.1), states that ‘one of
the key roles for the board includes establishing the cul- developed markets.
ture, values and ethics of the company’. It is important We find overwhelming evidence that creation
that the board sets the correct ‘tone from the top’. The di- culture has a positive effect on firm-level innova-
rectors should lead by example and ensure that good stan- tion for a large sample of Chinese listed firms.
dards of behaviour permeate throughout all levels of the Our results are robust to a number of tests, includ-
organization. ‘This will help prevent misconduct, uneth-
ical practices and support the delivery of long-term suc- ing alternative measures of innovation output, cre-
cess’ (FRC 2016). Greg Medcraft, Chairman of the Aus- ation culture and endogeneity. We also find that
tralian Securities and Investments Commission (ASIC), the effect of culture on innovation is more pro-
states that the role of the board is absolutely critical in nounced for firms that operate in more competitive
setting the culture and tone of the organization and one
of the ways in which the board sets this tone is by selecting
3
a CEO whose values are aligned with the company’s de- Graham et al. (2016) report that more than 50% of CEOs
sired culture. Business leaders have similar views as above and CFOs in their survey identified culture as an influen-
regarding the role of boards in shaping the corporate cul- tial factor for creativity.
4
ture. For example, John Watson, Chairman of Bellway, China filed 1.2 million patents in 2016 – more than
states that ‘The board sets the values and culture (how the combined total of the USA, Japan, The Republic of
people treat each other, how they operate within the sup- Korea and the European Patent Office. It was ranked
ply chain and how they work with employees) and man- among the 20 most innovative economies in the world for
agement is then responsible for implementing this’ (FRC the first time in 2018 ([Link]
2016). articles/2018/article_0005.html).

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
Corporate Culture and Innovation 3

industries and are subject to higher managerial ca- organizational behaviour literature, O’Reilly and
reer concerns. Our findings imply a strong cultural Chatman (1996) define corporate culture as ‘a set
effect on innovation. of norms and values that are widely shared and
We make several contributions to the literature. strongly help throughout the organization’. This
First, we extend recent insights on the role of cul- view suggests corporate culture as a social con-
ture and innovation outside the developed world trol mechanism and is in line with Hofstede (1991),
(Fiordelisi et al., 2019). In doing so, we provide who labels corporate culture as ‘the collective pro-
the first large-sample evidence of the applicability gramming of the mind’ of people in an organiza-
of the CVF to emerging markets and offer a direct tion. Culture acts as a coordinating mechanism to
comparison of our results with those reported enable dealing with unforeseen contingencies and
in the developed countries (Hartnell, Ou and events (Kreps, 1996). Every culture contains un-
Kinicki, 2011). Our results show the generalization spoken beliefs and norms that impact people’s be-
of CVF culture values in countries with different haviour. Schein (2010) ascribes culture as a set of
institutional, cultural and corporate governance basic assumptions among people to evaluate sit-
regimes. Second, we contribute to the innovation uations, human relations and activities, and these
and culture literature by showing that the effect assumptions form the shared norms and beliefs
of culture on innovation might depend on the among a group of people. This concept of culture
industry competitiveness and managerial career is also consistent with Martin’s (1992) view of the
concerns across firms (Chemmanur and Tian, integration perspective, which is based on homo-
2018; Nguyen, 2018). Third, we add to the recent geneity and harmony holding together a diverse
stream of literature studying the determinants of group of people in an organization. The literature
innovation (Amore and Failla, 2018; Kortum and in economics and finance has viewed culture as a
Lerner, 2001; Mazouz and Zhao, 2019).5 Finally, mechanism of explicit and implicit contracts that
we contribute to the recent literature that uses governs behaviour and comprises shared values,
text-based analysis on large datasets to measure preferences and beliefs of individuals in an organi-
corporate culture (Fiordelisi and Ricci, 2014; Jiang zation (Bénabou and Tirole, 2002; Guiso, Sapienza
et al., 2017; Nguyen, Nguyen and Sila, 2019), sen- and Zingales, 2015; Van den Steen, 2010). Culture
timent (Li, 2008; Loughran and McDonald, 2011, facilitates the complex interactions within the or-
2014; Tetlock, 2007), banking competition (Bush- ganization as contracts for each relationship do
man, Hendricks and Williams, 2016) and financial not exist (Audi, Loughran and McDonald, 2016).
constraints (Hoberg and Maksimovic, 2014). Guiso, Sapienza and Zingales (2015) define culture
as ‘principles and values that should inform the be-
havior of all the firm’s employees’. In sum, culture
Theory and hypotheses represents rules based on common values and be-
Conceptualization of corporate culture liefs which determine the interactions within an or-
ganization in order to accomplish certain goals.
There is no universally accepted definition of cor-
porate culture, and culture seems to be denoted
by a variety of meanings and connotations in the The competing value framework
organizational behaviour, economics and finance For our definition of culture, we rely on the CVF
literature (Fiordelisi et al., 2019). Based on the (Cameron et al., 2006), an organizational taxon-
omy widely used in the literature (Hartnell, Ou
5
and Kinicki, 2011). The CVF identifies four cul-
This literature also identifies venture capital (Chemma- ture types (create, compete, control, collaborate)
nur, Loutskina and Tian, 2014; Kortum and Lerner, 2001;
Tian and Wang, 2014), buyout structure (Cumming et al., positioned on four quadrants as in Figure 1. These
2019), CEO characteristics such as age, gender, expe- quadrants also represent two distinct dimensions,
rience, education and overconfidence (Hirshleifer, Low flexibility–discretion vs. stability–control (vertical
and Teoh, 2012), board independence (Balsmeier, Flem- axis) and internal vs. external orientation (horizon-
ing and Manso, 2017), analyst coverage (He and Tian, tal axis). The collaborate and control dimensions
2013), stock liquidity (Fang, Tian and Tice, 2014) and
anti-takeover provisions (Chemmanur and Tian, 2018) as share an internal orientation with a focus on in-
the important determinants of firm-level innovation abil- tegration, collaboration and unity. A collaborate-
ity. oriented culture focuses on teamwork, employee

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
4 Y. Wang, H. Farag and W. Ahmad

Flexibility and discretion

Clan Adhocracy
Thrust Collaborate Thrust Create
Means Cohesion, Means Adaptability, creativity,
participation, communication, agility
empowerment Ends Innovation and cutting-edge,
Ends Morale, people output External focus
Internal focus and development, commitment and
integration Hierarchy Market differentiation
Thrust Control Thrust Compete
Means Capable processes, Means Customer focus,
consistency, process control, productivity, enhancing
measurement competitiveness
Ends Efficiency, timeliness, Ends Market share, profitability,
smooth functioning goal achievement
Stability and control

Figure 1. CVF culture dimensions (Cameron et al., 2006)

development and empowerment, supported by a 2010; Li, 2008; Loughran and McDonald, 2011).
flexible organization structure. A control culture This, however, raises the question of whether text-
type is characterized by values such as efficiency based measures reliably capture corporate culture.
improvement, predictability and conformity, sup- The relevant literature suggests that textual anal-
ported by a stable organizational structure and ysis of financial statements provides valid and ro-
strong internal controls. The compete and create bust measures of industry competition (Li, Lund-
culture types share an external focus based on dif- holm and Minnis, 2013), banking competition
ferentiation, competition and rivalry. A compete (Bushman, Hendricks and Williams, 2016) and
culture is associated with values such as compet- financial constraints (Hoberg and Maksimovic,
itiveness, fast response and goal achievement, and 2014), and these measures reliably predict firm
an organizational structure based on stability and financial outcomes. Recent studies also provide
control. Finally, a create culture focuses on creativ- evidence on the validity of text-based corporate
ity, autonomy and adaptability, and is supported (Fiordelisi and Ricci, 2014; Nguyen, Nguyen and
by a flexible organizational structure. Sila, 2019) and creative (Fiordelisi et al., 2019)
The literature has traditionally relied on survey- culture measures, firms’ market orientation (An-
based instruments using small samples to test CVF dreou, Harris and Philip, 2020) and firms’ thrust
propositions. More recently, Fiordelisi and Ricci to compete (Andreou et al., 2019). Overall, the ev-
(2014) operationalized and quantified the CVF idence suggests that CVF-based text measures re-
culture types by employing a bag-of-words ap- liably capture firms’ corporate culture.
proach and textual analysis of financial reports
(10-K reports). The underlying premise of the tex-
Corporate culture and innovation
tual analysis is that the words used in corporate
filings depict the cultural norms of a company. Culture is manifested in rituals, group norms,
Financial reports are the most important docu- habits of thinking and espoused values (Schein,
ments which could be used by managers to present 1992). These attributes, in turn, help build a strong
major attributes and values of the company to sense of ownership, teamwork and sense of mis-
the outside world (Audi, Loughran and McDon- sion among employees (Fey and Denison, 2003).
ald, 2016). The CVF-based culture measure is Organizational cultures influence employee atti-
broadly supported by the growing literature within tudes, behaviours and commitment, leading to or-
finance and accounting, which provides evidence ganizational effectiveness (Gregory et al., 2009;
that textual analysis can be used to extract im- O’Reilly, Chatman and Caldwell, 1991). Culture,
portant aspects of corporate behaviour, tone and as such, can be a powerful means to elicit de-
sentiment in publicly available official documents sired corporate outcomes (Hogan and Coote,
(Hanley and Hoberg, 2010; Hoberg and Phillips, 2014). Thus, a large body of literature suggests a

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
Corporate Culture and Innovation 5

positive association between culture and perfor- (Büschgens, Bausch and Balkin, 2013; Danneels,
mance (Denison and Mishra, 1995; Sackmann, 2008). By allowing freedom of thought and ac-
2002; Zuckerman, 2002). Innovation is one of the tion among employees, companies rich in creative
key drivers of business performance and value cre- culture aim to develop radical new processes and
ation. Innovation confers a competitive advantage product technologies, innovate in logistics and re-
that helps successful organizations to stay ahead in define entire industries (Fiordelisi and Ricci, 2014;
business. Successful innovation determines a firm’s Fiordelisi et al., 2019). Our primary hypothesis is
future profitability and competitive edge to a large therefore
extent (Ettlie, 1998; Scherer, 1984). Given the im-
portance of innovation for an organization’s suc- H1: Firms with a stronger creation-oriented
cess and performance, previous literature posits culture are more likely to engage in inno-
that culture plays an important role in enhancing vation.
innovation.
The extant literature has identified a variety Research design
of cultural values related to innovation. For in-
stance, innovation-supportive cultures and trans- Sample
formational leadership are likely to enhance in- Our sample includes all Chinese listed companies
novation (Chandler, Keller and Lyon, 2000; Gu- on both the Shanghai and Shenzhen stock ex-
musluoğlu and Ilsev, 2009). Organizations with changes (SHSE, SZSE) between 2008 and 2017.6
supportive cultures, participative decision-making Data on firm-level innovation and accounting vari-
and higher tolerance for failure are found to ables are collected from the China Stock Market &
be more productive at innovation (Abbey and Accounting Research (CSMAR) database. We also
Dickson, 1983; Danneels, 2008; Hurley and Hult, match innovation variables (patents and citations)
1998). While it is widely accepted that culture fos- in CSMAR with the China National Intellectual
ters innovation, some aspects of culture could sti- Property Administration (CNIPA)7 to ensure that
fle innovation (Berson, Oreg and Dvir, 2008; Chat- our innovation variables are consistent and ac-
man and Flynn, 2001; Jaskyte, 2004). Sørensen curate. In case of conflict between CSMAR and
(2002) argues that strong culture firms may find CNIPA, we use the patent data from CNIPA. We
it difficult to engage in exploration learning and use annual reports of Chinese listed firms to con-
adapt to changes in the volatile environment. struct culture-related variables using textual analy-
Dougherty and Heller (1994) suggest that organi- sis. We exclude financial firms and firms with miss-
zational preference for stability may impede prod- ing values from our sample, following Yuan and
uct innovation. Similarly, Staw and Nemeth (1989) Wen (2018). All variables are winsorized at the 1st
argue that strong culture may stifle innovation as and 99th percentiles to mitigate the influence of
cohesion among organizational members leads to outliers. Table A.1 (in the online supporting infor-
less deviation. mation) provides definitions of all variables. Our fi-
Büschgens, Bausch and Balkin (2013), however, nal sample contains 17,259 firm-year observations
argue that the use of a multitude of culture values for 2,583 unique firms.
leads to a fragmented concept of culture and inno-
vation. The authors show that CVF comprehen-
sively describes corporate cultures. We use CVF Innovation measures
and measure the strength of creation culture rel- We construct innovation variables using data from
ative to other culture dimensions. The creation- CSMAR and CNIPA, which contain information
oriented (adhocracy) culture in the CVF (Cameron on patent application date, application identifica-
et al., 2006) suggests that such a culture would tion, grant date and application institution. The
encourage idea-sharing, entrepreneurial thinking Chinese patents are classified into three categories:
and vision-building among employees. This type invention patents, utility model patents and design
of culture promotes agility, adaptability and cre-
ativity, which are key drivers of innovation. The 6
We chose 2008 as our starting year since the Chinese gov-
flexibility orientation of the create dimension en- ernment adopted a new accounting standard in 2007 that
courages acceptance of deviation from norms and required disclosure of R&D information.
7
allows tolerance for failure and greater risk-taking See [Link] for more details.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
6 Y. Wang, H. Farag and W. Ahmad

patents. Invention patents are granted for a new Creation culture measure
solution to a product or technical process. Util-
A firm’s culture is manifested in its symbols, rit-
ity model patents are for a new technical solu-
uals and values, and those ‘symbols are words,
tion or improvement with a lower degree of inven-
gestures, pictures or objects that carry a partic-
tiveness, relating to certain features of a product
ular meaning within a culture’ (Hofstede et al.,
such as shape or structural physical features. De-
1990). Thus, words can be used to convey the val-
sign patents are granted for innovations in external
ues and attributes that are part of a firm’s cul-
features of a product such as shape, pattern and/or
ture. Management’s disclosures in financial docu-
colour, which make the product attractive and fit
ments can provide insights into the value system
for industrial application. Given the limited tech-
of the firm, as the distinctive features of any firm
nological innovation involved in design patents, we
are mirrored in its written documents (Fiordelisi
use only invention and utility model patents to
and Ricci, 2014). Textual analysis is a systematic
construct our innovation measures following rele-
and objective means to examine the specific fea-
vant literature (Fang, Lerner and Wu, 2017; Jiang
tures of a text and is widely supported by re-
and Yuan, 2018; Tan et al., 2015). Consistent with
search in sociology (Atkinson, 1990), communica-
the innovation literature, we use application year
tion theory (Mumby and Stohl, 1991), cultural an-
rather than grant year for our patent counts, be-
thropology (Clifford and Marcus, 1986) and lately
cause application year corresponds more closely to
the finance and management literature (Andreou,
the actual time of the innovation (Fang, Tian and
Harris and Philip, 2020; Antweiler and Frank,
Tice, 2014; He and Tian, 2013; Jiang and Yuan,
2004; Hanley and Hoberg, 2010; Hoberg and
2018).
Phillips, 2010; Li, 2008; Loughran and McDonald,
Our firm-level innovation output is measured in
2011; Nguyen, Nguyen and Sila, 2019). Fiordelisi
four ways. First, Inven is the raw count of invention
and Ricci (2014) apply textual analysis on 10-K fil-
patents filed and eventually granted to a firm in a
ings in the USA to measure CVF culture dimen-
given year. Second, Inven+Utility is the raw count
sions using a two-step procedure. First, they iden-
of invention and utility patents filed and eventu-
tify words related to each culture dimension in the
ally granted to a firm in a given year. Third, Inno-
CVF (Cameron et al., 2006). Second, they use the
vation Efficiency is a measure of how efficiently a
Harvard IV-4 Psychosocial Dictionary to identify
firm transforms its innovation input (R&D expen-
synonyms for words selected in the first step to cre-
diture) into innovation output (patents). We follow
ate four bags of words for each culture dimension
Cao, Cumming and Zhou (2020) and Hirshleifer,
listed in panel A of Table A.2 (in the online sup-
Hsu and Li (2013) to construct innovation effi-
porting information).
ciency, which is measured as the number of patents
We operationalize CVF culture dimensions by
of firm i in year t scaled by firm i’s cumulative R&D
applying textual analysis to annual reports of Chi-
investment in year t−2 through year t.8 Fourth,
nese listed firms following the bag of words de-
Citations is the number of forward citations re-
veloped by Fiordelisi and Ricci (2014). Specifi-
ceived by a firm’s invention patents. While patent
cally, we translate authors’ bags of words into
counts measure the raw output of a firm’s inno-
Chinese using The Oxford English–Chinese Dic-
vative activities, citations capture the technologi-
tionary and The Oxford English Dictionary fol-
cal and economic importance of patents granted
lowing Jiang et al. (2017).9 We further validate
to a firm (Hall, Jaffe and Trajtenberg, 2005). We
our translated keywords for their appropriateness
use the natural logarithm of one plus patent counts
in the Chinese context by using The Contem-
(Inven, Inven+Utility) and Citations in our analysis
porary Chinese Dictionary and only choose the
to address the skewness concerns in patent-related
first explanation of each word in the dictionary
variables (Fang, Lerner and Wu, 2017; Jiang and
to avoid double-counting. To further authenticate
Yuan, 2018).
our approach, we use external verification of our

8
Innovation efficiency is measured following Cao, Cum-
ming and Zhou (2020). The equation is as follows: 9
NumPatent The Chinese ‘bag of words’ can be found in panel B of
Innovation E¡ciencyi,t = R&D + 2 R&D +i,t1 R&D . Table A.2 (in the online supporting information).
i,t 3 i,t−1 3 i,t−2

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
Corporate Culture and Innovation 7

approach by an expert Chinese linguist.10 We across different countries. For instance, van Mui-
quantify each CVF culture dimension, using the jen and Turnipseed (1999) apply the CVF model
Chinese bag of words, by counting the occurrences to construct an international instrument for mea-
of keywords (synonyms) associated with that cul- suring organizational culture across 12 European
ture in each annual report. We obtain four values countries. Other studies prove the validity of the
for each of the culture dimensions for our firm- CVF for Hong Kong (Kwan and Walker, 2004; Lau
years. Using raw counts of each CVF culture di- and Ngo, 2004), South Korea (Dastmalchian, Lee
mension, we construct our Creation Culture vari- and Ng, 2000), Australia (Lamond, 2003), China
able as follows:11 and other Asian countries (Deshpandé and Farley,

Number of keywords describing the creation culture

Creation Culture = (1)
Total number of keywords for all CVF cultures

We scale the frequency of creation culture words

by the total number of words in all culture di- 2004; Lau, David and Zhou, 2002; Ralston et al.,
mensions to capture the relative emphasis or in- 2006). More recently, Zhang, Li and Wei (2008)
tensity of the creation culture compared to other use the CVF to investigate the relationship be-
cultures. Our measure reflects the relative impor- tween corporate culture and performance of Chi-
tance that firms place on their creative values lead- nese firms, and Yu and Wu (2009) argue that the
ing to a creation culture. We also use an alterna- CVF provides better validity and reliability in the
tive approach where the creation culture keyword Chinese context. Overall, the CVF is a widely ap-
count is scaled by the total number of words in the plicable model for corporate culture across differ-
annual report (Creation Culture-Alternative), con- ent countries and institutional contexts, including
sistent with prior literature (Fiordelisi and Ricci, China, and the western origin of the CVF is not
2014; Nguyen, Nguyen and Sila, 2019). The higher a weakness of the model (Büschgens, Bausch and
ratio indicates a higher emphasis on the creation Balkin, 2013).
culture in a firm.12
Although our measurement of creation culture
Firm controls
follows the extant literature (Andreou et al., 2019;
Fiordelisi and Ricci, 2014), there might be con- Following the innovation literature, we control
cerns regarding the applicability of the CVF in the for an array of well-known firm-level character-
Chinese context. The CVF has been developed and istics that may affect firms’ innovation. We in-
applied in western countries (predominantly the clude firm size following Hall and Ziedonis (2001),
USA) and might not be applicable to Chinese firms who argue a positive relation between firm size
due to institutional differences between China and and the number of patents and citations. Firm
the developed countries (You, Zhang and Zhang, age is included to control for differences in the
2017). The relevant literature, however, suggests stages of development across firms (Amore and
the validity and applicability of the CVF model Failla, 2018). We also include leverage and return
on assets (ROA) to control firms’ debt and prof-
itability (He and Tian, 2013; Mazouz and Zhao,
2019). We also include book-to-market ratio to ac-
count for growth opportunities (Lev and Sougian-
10
We thank and appreciate Professor Zhiyang Sun, a lin- nis, 1999) and state ownership (SOE) following
guist from Jiangnan University, for his help in authenti- Cao, Cumming and Zhou (2020). A detailed def-
cating our approach. inition of these control variables is provided in
11
Andreou et al. (2019) use a similar approach for con- Table A.1 (in the online supporting information).
structing their thrust to compete measure.
12
The CVF argues that although all culture dimensions In robustness tests, we also include a number
are typically present in an organization, one or two typi- of board and CEO characteristics as additional
cally are dominant. controls.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
8 Y. Wang, H. Farag and W. Ahmad

Empirical model invention patents each year, which subsequently

receive 5.57 citations.
We use the following model to estimate the re-
In Table 2, we present correlations between
lationship between corporate culture and innova-
variables. All our innovation-related variables
tion:
are significantly positively correlated with
Creation Culture. This indicates that a higher
Innovationi,t = β0 + β1 Creation Culturei,t creative culture increases the innovation output,
 consistent with our expectations.
+ βl Firm Controlsi,t + εi,t
(2)
Multivariate results

where i indexes firms and t indexes time. The Table 3 presents estimation results of our base-
Innovation variables are the natural logarithm line regression in Eq. (2) with all four proxies of
of one plus the number of invention patents innovation in columns (1)–(4), respectively. Cre-
filed (and subsequently granted), the natural log- ation Culture is positively and significantly associ-
arithm of one plus the number of invention ated with all innovation proxies across all the mod-
and utility patents, innovation efficiency and the els. These results suggest that firms with a stronger
natural logarithm of one plus the number of ci- creation culture generate more patents, which sub-
tations per invention patent. Our Innovation vari- sequently receive higher citations and are more ef-
ables account for both the quantity (patent num- ficient in terms of innovation activity. In terms
bers) and quality (citations) of the innovation. of economic significance, a one standard devia-
The key explanatory variable is our Creation Cul- tion increase in creation culture increases invention
ture, measured as in Eq. (1). Firm Controls include patents (invention plus utility) by 6.44% (11.15%)
firm size, firm age, leverage ratio, ROA, book-to- and improves citations by 4.67%. Given that our
market ratio and SOE. Finally, we include industry creation culture measure is based on individual
and province fixed effects to control industry and word counts, the observed effect is economically
province-specific time-invariant heterogeneity and significant.13
year fixed effects to control overall trends in inno- The results for control variables are also consis-
vation. Robust standard errors are clustered at the tent with prior literature. Firm Size is positive and
firm level. statistically significant in models 1, 2 and 4, sug-
gesting that larger firms produce more patents and
are associated with higher patent citations, consis-
Results and discussion tent with Mazouz and Zhao (2019). However, the
coefficient of Firm Size becomes negative when we
Summary statistics
use Innovation Efficiency (Model 3) as the depen-
Table 1 presents various summary statistics. Panel dent variable. This indicates that larger firms’ in-
A provides means of Creation Culture and all In- novation efficiencies decrease, even if they are pro-
novation variables across industry sectors. Panel B ducing more innovation (Janz, Lööf and Peters,
shows the mean value of Creation Culture across 2003) due to economies of scale consistent with
industries and sample years. Panel C reports de- Cao et al. (2016). Firm Age is negatively related
scriptive statistics of the variables used in our re- to innovation in all models, showing that younger
gressions. Panel A shows that manufacturing and firms generate more patents than mature firms
high-tech industries have a more creative culture (Coad, Segarra and Teruel, 2013). Both Leverage
compared to other industries. Consistent with our and ROA indicate a positive and statistically sig-
prediction, these industries also rank higher in nificant relationship with innovation, while Book
terms of innovation outputs. Although there is to Market negatively affects innovation. These re-
a substantial variation in creation culture across sults are largely consistent with previous studies
industries, the variation across sample years is
small (Panel B). This is consistent with the gen- 13
We also re-estimate Eq. (2) using one-period-ahead in-
eral notion that culture is persistent over time and novation measures (Innovationi,t+1 ) and our results are
changes slowly (Fiordelisi et al., 2019). Panel C consistent with those reported in Table 3. Results are not
shows that our sample firms apply, on average, 8.03 reported but are available upon request.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
Corporate Culture and Innovation 9

Table 1. Summary statistics

Panel A. Creation culture and innovation across industry sectors

Creation Culture Inven Inven+Utility Innovative Efficiency Citations

Manufacturing 0.195 8.705 19.756 0.084 5.567

Public Utility 0.147 2.074 5.115 0.014 0.917
Mining 0.149 7.403 17.744 0.023 8.401
Business 0.159 0.886 1.983 0.012 0.218
High-Tech 0.259 7.801 12.901 0.066 5.809

Panel B. Average yearly creation culture across industry sectors

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Manufacturing 0.177 0.179 0.192 0.195 0.197 0.199 0.201 0.205 0.205 0.204
Public Utility 0.138 0.141 0.143 0.141 0.146 0.148 0.153 0.153 0.154 0.153
Mining 0.148 0.148 0.148 0.148 0.149 0.149 0.149 0.150 0.149 0.148
Business 0.157 0.157 0.159 0.158 0.159 0.158 0.161 0.159 0.162 0.162
High-Tech 0.234 0.232 0.254 0.256 0.263 0.266 0.257 0.260 0.288 0.281
Average 0.171 0.171 0.179 0.180 0.183 0.184 0.184 0.186 0.192 0.190

Panel C. Sample summary statistics

Variable Observations Mean Median Min Max Std dev.

Dependent variables
Invention Patent 17,295 8.033 2.000 0.000 110.000 14.227
Invention and Utility Patent 17,295 16.635 5.000 0.000 234.000 30.600
Innovation Efficiency 17,295 0.063 0.010 0.000 0.705 0.110
Citation 17,295 5.574 1.000 0.000 139.000 13.260
Independent variables
Creation Culture 17,295 0.224 0.233 0.000 0.583 0.074
Creation Culture (Alternative) 17,295 0.006 0.005 0.001 0.019 0.002
Firm controls
Firm Size 17,295 8.146 7.947 4.872 14.693 1.291
Firm Age 17,295 2.806 2.833 0.693 4.094 0.316
Leverage 17,295 0.402 0.394 0.007 1.256 0.204
ROA 17,295 0.054 0.047 −1.324 0.598 0.060
Book to Market 17,295 0.808 0.541 0.032 11.005 0.875
SOE 17,295 0.443 0.358 0.276 0.724 0.156

This table presents various summary statistics. Panel A reports the overall Creation Culture and Innovation proxies in five different
industries. Panel B presents average creation culture across industries and years. Panel C presents the summary statistics for the full
sample, which comprises 17,295 firm-year observations with 2,583 unique firms between 2008 and 2017. All variables are defined in
Table A.1.

(David and O’Brien, 2006). Overall, the above re- Robustness tests
sults strongly support our hypothesis that creation
In this section, we check the robustness of our key
culture is positively associated with firms’ innova-
conclusion that a creation culture enhances inno-
tion output.14
vation by conducting additional tests.

14
Additional controls
In order to further strengthen our results, we use simul-
taneous quantile regressions as robustness test to check Although we control for a comprehensive set of
whether the relationship between corporate culture and firm characteristics that may affect firm-level inno-
innovation is persistent. Specifically, we bootstrap the re- vation, the literature suggests that board and CEO
gression 500 times at 50%, 75% and 90% quantile points.
The unreported results are consistent with our baseline re- characteristics are also important determinants of
gression results and support our hypothesis. Unreported innovation (Balsmeier, Fleming and Manso, 2017;
results are available from the authors upon request. Bertrand and Schoar, 2003; Huang and Kisgen,

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
10 Y. Wang, H. Farag and W. Ahmad

Table 2. Correlations matrix

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)

Ln(Inven+1) (1) 1
Ln(Inven+Utility+1) (2) 0.899* 1
Innovation Efficiency (3) 0.577* 0.523* 1
Ln(Citations+1) (4) 0.594* 0.513* 0.348* 1
Creation Culture (5) 0.313* 0.349* 0.226* 0.055* 1
Firm Size (6) 0.091* 0.060* −0.104* 0.097* −0.056* 1
Firm Age (7) 0.016* 0.0042 −0.016* −0.076* 0.075* 0.168* 1
Leverage (8) −0.033* −0.044* −0.115* 0.009 −0.178* 0.543* 0.163* 1
ROA (9) 0.052* 0.044* 0.045* 0.074* 0.021* −0.115* −0.145* −0.373* 1
Book to Market (10) −0.051* −0.061* −0.117* 0.006 −0.159* 0.606* 0.095* 0.582* −0.295* 1
SOE (11) 0.152* 0.164* 0.103* −0.059* 0.346* −0.100* −0.008 −0.075* 0.045* −0.140* 1

This table reports the correlation matrix for the variables used in our base regression model. The full sample comprises 17,295 firm-year
observations with 2,583 unique firms between 2008 and 2017. All variables are defined in Table A.1.
∗ Represents significance at the 5% level.

2013; Islam and Zein, 2020; Simsek, 2007). We Endogeneity

control for CEO characteristics, including gender,
Although the results of our baseline models sug-
age, education, tenure and whether a CEO is po-
gest that creation culture has a positive effect on
litically connected. In addition, we also control for
firm innovation, the endogeneity issues between
board size, board independence and whether the
culture and innovation cannot be ignored, since
CEO is also chairman of the board.15 The results
the association between corporate culture and in-
presented in Table 4 show that the coefficient for
novation may reflect causalities other than those
creation culture is positive and statistically signifi-
hypothesized by us (Hartnell, Ou and Kinicki,
cant across all innovation variables, and therefore
2011). Therefore, our results might be subject to
our baseline results are robust to these additional
two types of endogeneity concerns: omitted vari-
controls.16
ables and reverse causality. The omitted variables
may bias our coefficient estimates if the unobserv-
able firm characteristics are correlated with both
creation culture and innovation output. Further,
our observed relation between creation culture and
15
Gender is a dummy equal to one if the CEO is female, innovation may be spurious if the unobservable
and zero otherwise. Age is the natural logarithm of CEO firm characteristics affect culture and innovation
actual age. Education dummy equals one when the CEO jointly. The reverse causality is another endogene-
has a Master’s degree or above, and zero otherwise. Tenure
is the CEO’s contract period in natural logarithm. Po- ity concern which could result in biased and incon-
litical Connection is a dummy equal to one if the CEO sistent ordinary least squares (OLS) regression es-
is connected to the central government, and zero other- timates. We observe that a more creative culture
wise. Board Size is measured as the total number of board leads to higher innovation output and efficiency,
members. Board Independence is measured as the percent- but some of this observed relation could be at-
age of independent directors on the board. Duality is a
dummy equal to one if the CEO is also the chairman of tributed to the possibility that the level of expected
the board, and zero otherwise. innovation predicts a firm’s creative culture. Al-
16
In addition, to alleviate concerns that other CVF cul- though it is difficult to alleviate endogeneity con-
ture dimensions (compete, control, collaborate) may im- cerns completely, we use a set of robustness tests
pact innovation or weaken the effect of creation culture, to mitigate the potential bias.
we control for these culture dimensions in our baseline re-
gression model. We find a consistent and statistically sig- The omitted variable problem could arise
nificant impact of creation culture on innovation even af- from both time-varying and time-invariant
ter controlling for the other CVF culture dimensions. Re- unobservables. Our earlier test – including a
sults are not reported but available from the authors upon comprehensive set of additional controls for
request. Overall, the presence of competing CVF culture CEO and board characteristics – to a large ex-
dimensions does not attenuate the observed relation be-
tween creation culture and innovation. tent mitigates the effect of time-varying omitted

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
Corporate Culture and Innovation 11

Table 3. Creation culture and innovation

(1) (2) (3) (4)

Dependent variable Ln(Inven+1) Ln(Inven+ Innovation
Utility+1) Efficiency Ln(Citations+1)
Creation Culture 0.843*** 1.428*** 0.100*** 0.617***
(5.188) (7.362) (6.330) (3.730)
Firm Size 0.217*** 0.211*** −0.006*** 0.225***
(19.754) (16.037) (−6.601) (20.868)
Firm Age −0.092*** −0.172*** −0.010*** −0.110***
(−3.030) (−4.790) (−3.608) (−3.595)
Leverage 0.159*** 0.249*** −0.001 0.127**
(2.751) (3.623) (−0.089) (2.221)
ROA 0.794*** 0.775*** 0.031** 1.009***
(5.099) (4.324) (2.156) (6.223)
Book to Market −0.113*** −0.115*** −0.001 −0.091***
(−7.326) (−6.282) (−0.923) (−5.855)
SOE −0.099 −0.166 −0.009 −0.222***
(−1.083) (−1.593) (−1.078) (−3.351)
Constant −1.947*** −1.796*** 0.020 −1.497***
(−14.534) (−11.350) (1.628) (−11.372)
Year Effect YES YES YES YES
Industry Effect YES YES YES YES
Province Effect YES YES YES YES
N 17,295 17,295 17,295 17,295
Adj-R2 0.268 0.315 0.114 0.140

This table presents the baseline OLS estimation results. The dependent variables are: Ln(Inven+1), as the natural logarithm of the
number of invention patent applications plus one in column (1); Ln(Inven+Utility+1), as the natural logarithm of the number of
invention and utility patent applications plus one in column (2); Innovation Efficiency, as the number of invention patents scaled by
firm’s cumulative R&D investment from year t−2 to year t in column (3); and Ln(Citations+1), as the natural logarithm of the number
of invention patent citations plus one in column (4). The full sample comprises 17,295 firm-year observations with 2,583 unique firms
between 2008 and 2017. All variables are defined in Table A.1. All models include year, industry and province fixed effects. Robust
standard errors are clustered at the firm level. t-Values are in parentheses.
* , ** and *** represent significance levels at 10%, 5% and 1%, respectively.

variables, which might be correlated with both and control variables are equally important.17
creation culture and innovation. To mitigate omit- Table A.3 (in the online supporting information)
ted variable concerns caused by time-invariant reports the results, including Industry × Province
firm characteristics, we substitute year, industry and Province × Year fixed effects and the ad-
and province fixed effects in our main regression justment for the effect of unobservables on our
(Table 3) with Industry × Province and Province × creation culture variable. Overall, the results are
Year fixed effects following Cumming et al. (2019) consistent with our baseline findings and robust to
and Fiordelisi et al. (2019). These fixed effects addition of the above fixed effects and adjustment
control for specific industries located in specific for the potential omitted variables. The estimator
provinces and province-level policies, which might δ shows large values which are greater than the 1
simultaneously affect creation culture and inno- cutoff suggested by Oster (2019) across all models.
vation variables over time. We further augment Moreover, the values of β omitted are very close to
this analysis following Fiordelisi et al. (2019) and the coefficients of Creation Culture in all models,
Oster (2019) to examine the potential bias induced suggesting that the potential omitted variables
by omitted variables and add two estimators cause a very small effect on our main regression
(δ and β omitted ) in our regressions. The estimator coefficients.18
δ captures the importance of omitted variables
in relation to control variables and β omitted repre-
sents the effect of creation culture on innovation 17
Assuming δ = 1.
under the assumption that omitted variables 18
We also use Year × Industry and Industry × Province
fixed effects and the unreported results are consistent.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
12 Y. Wang, H. Farag and W. Ahmad

Table 4. Creation culture and innovation with additional controls

(1) (2) (3) (4(

Dependent variable Ln(Inven+1) Ln(Inven+Utility+1) Innovation Efficiency Ln(Citations+1)

Creation Culture 0.834*** 1.409*** 0.097*** 0.616***

(5.144) (7.282) (6.162) (3.733)
Firm Size 0.218*** 0.217*** −0.005*** 0.223***
(19.437) (16.171) (−5.760) (20.461)
Firm Age −0.066** −0.130*** −0.008*** −0.090***
(−2.174) (−3.589) (−2.856) (−2.898)
Leverage 0.185*** 0.286*** 0.001 0.152***
(3.201) (4.170) (0.234) (2.650)
ROA 0.767*** 0.727*** 0.026* 0.992***
(4.925) (4.062) (1.791) (6.124)
Book to Market −0.113*** −0.117*** −0.001 −0.091***
(−7.267) (−6.387) (−1.088) (−5.783)
SOE −0.159* −0.245** −0.016* −0.277***
(−1.725) (−2.333) (−1.751) (−4.129)
CEO Gender −0.165*** −0.183*** 0.000 −0.119***
(−5.044) (−4.632) (0.091) (−3.781)
CEO Age −0.048* −0.041 −0.003 −0.029
(−1.913) (−1.446) (−1.393) (−1.340)
CEO Education 0.148*** 0.232*** 0.008*** 0.127***
(6.542) (8.444) (3.599) (5.996)
CEO Tenure 0.019** 0.009 0.001 0.023***
(2.305) (0.866) (1.025) (2.814)
CEO Political 0.046 −0.092 −0.006 0.135**
(0.868) (−1.475) (−1.345) (2.406)
Board Size 0.022 0.016 −0.001 0.000
(0.838) (0.544) (−0.219) (0.005)
Independent Director −0.048 −0.045 −0.003 0.100
(−0.586) (−0.496) (−0.319) (1.577)
Duality 0.066*** 0.091*** 0.011*** 0.020
(3.235) (3.820) (5.264) (1.009)
Constant −1.989*** −1.946*** 0.019 −1.573***
(−11.772) (−9.847) (1.248) (−10.089)

Year Effect YES YES YES YES

Industry Effect YES YES YES YES
Province Effect YES YES YES YES
N 17,295 17,295 17,295 17,295
Adj-R2 0.272 0.319 0.116 0.142

This table presents the OLS estimation results including additional control variables. The dependent variables are: Ln(Inven+1), as
the natural logarithm of the number of invention patent applications plus one in column (1); Ln(Inven+Utility+1), as the natural
logarithm of the number of invention and utility patent applications plus one in column (2); Innovation Efficiency, as the number of
invention patents scaled by firm’s cumulative R&D investment from year t−2 to year t in column (3); and Ln(Citations+1), as the
natural logarithm of the number of invention patent citations plus one in column (4). The full sample comprises 17,295 firm-level
observations with 2,583 unique firms between 2008 and 2017. All variables are defined in Table A.1. All models include year, industry
and province fixed effects. Robust standard errors are clustered at the firm level. t-Values are in parentheses.
* , ** and *** represent significance levels at 10%, 5% and 1%, respectively.

In order to address reverse causality, we employ however, to note that finding suitable instruments
instrumental variable analysis using a two-stage for culture – and specifically corporate culture –
least-squares (2SLS) approach. It is important, is practically very difficult (Nash and Patel, 2019).
Nonetheless, we rely on the extant literature to
formulate our instrument for the creation culture
Moreover, our unreported results remain qualitatively the (Cumming et al., 2019; Mazouz and Zhao, 2019).
same after including firm fixed effects, as well as including
lagged values (2 and 3) of the creation culture. Following this literature, we use the industry-year

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
Corporate Culture and Innovation 13

average of the creation culture as our main instru- in firms’ annual reports, following Fiordelisi and
mental variable in 2SLS regressions.19 We believe Ricci (2014). Different to our main measure, this
that our instrument is valid because the industry measure shows the relative importance of creation
culture may strongly influence the culture of any words in the whole annual report. The relevant
given firm in the industry. Firms operating in an results presented in Table A.4 (in the online sup-
industry with a creative culture are more likely to porting information) show that our key findings
regard the industry culture as a useful economic persist.
attribute and try to build a creative culture. In Second, it might be argued that not all words
contrast, it is unlikely that these industry-level in our creation bag equally characterize a creation
instrumental variables directly predict firm-level culture.20 In order to mitigate the noise and mea-
innovation that cannot be explained by the surement problems in our Creation Culture vari-
firm-level culture. Furthermore, an industry-level able, we use a reduced bag of words consisting of
variable is less likely to be affected by an individual the most relevant words. We reconstruct our Cre-
firm’s policy, and satisfies both exclusion and rele- ation Culture variable using only words such as
vance conditions. Our approach is consistent with idea*, innovate*, learn*, new*, start* and vision*.
Jiang et al. (2017) and Li et al. (2019), who use the The relevant results reported in Table A.5 (in the
industry average for instrumenting disclosure tone online supporting information) show a consistent
and integrity culture. Table 5 reports results for and significant effect of creation culture across all
the instrumental variable regressions. First-stage measures of innovation.
results are presented in column (1), with Creation Finally, we use a more robust CVF bag of
Culture as dependent variable. Consistent with our words measure for cultural dimensions follow-
expectation, the industry average is positively and ing Andreou et al. (2020). The authors develop
significantly (1% level) associated with creation this new bag of words based on the keywords
culture, suggesting the validity of the instrumental used within the Organizational Culture Assess-
variable. The Kleibergen–Paap rk Wald F statistic ment Instrument (OCAI), a survey developed by
in our 2SLS regression is sufficiently higher than Cameron and Quinn (2011) to measure firms’
the critical value (16.38) for the weak instru- culture through the responses of firms’ employ-
ment test on 2SLS size (Stock and Yogo, 2005), ees.21 The OCAI-based bag of words captures
suggesting that our instrumental variable is not diverse dimensions considered important by the
weak. Columns (2)–(5) (Table 5) present results of firms in measuring their internal culture. Andreou
the second stage using predicted creation culture et al. (2020) demonstrate that the OCAI-based
and all four innovation proxies as the dependent approach provides a relatively better and more
variable. The coefficients of Creation Culture are robust measure of corporate culture using con-
positive and significant for all innovation variables tent, external, dimensionality and predictive va-
except innovation efficiency. Our results from lidity tests. We reconstruct CVF culture dimen-
the instrumental variable analysis are generally sions using the OCAI-based bag of words and re-
consistent with our main results in Table 3. calculate our creation culture measure using Eq.
(1). The results of our baseline regressions using
the OCAI-based creation culture are presented in
Alternative measures of creation culture
Table A.6 (in the online supporting information)
It is possible that our findings are sensitive to how and show our key findings to be unchanged.22
we measure creation culture. We employ three ap-
proaches to minimize the potential measurement 20
Word stems such as dream*, envis*, freedom*, risk*,
issues with our creation culture measure. First, we
thought*, etc. might not be considered relevant to a cre-
scale creation words by the total number of words ative culture.
21
Cameron and Quinn (2011) base OCAI along six dimen-
sions: dominant characteristics, organizational leader-
19 ship, management of employees, organization glue, strate-
We also use the industry median as an instrument for
creation culture in our 2SLS regression and find that the gic emphases and criteria of success.
22
results remain unchanged. Moreover, use of province av- We also test the robustness of our results using an
erage as an instrument yields qualitatively similar results. OCAI-based measure in two ways: (1) using other culture
The unreported results are available from the authors dimensions as controls in the baseline regression model
upon request. and (2) scaling the OCAI creation bag of words by the

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
 14

Table 5. Instrumental variable regressions

First stage Second stage

Model (1) (2) (3) (4) (5)

Dependent variable Creation Culture Ln(Inven+1) Ln(Inven+ Innovation Ln(Citations+1)
Utility+1) Efficiency
Industry Average 0.999***
(34.923)
Creation Culture 0.764** 1.247*** 0.030 0.666***
(2.337) (3.054) (0.900) (3.250)
Firm Size −0.001** 0.217*** 0.211*** −0.006*** 0.225***
(−2.127) (19.774) (16.042) (−6.683) (20.908)
Firm Age −0.018*** −0.093*** −0.176*** −0.011*** −0.109***
(−14.428) (−3.042) (−4.809) (−4.013) (−3.562)
Leverage −0.029*** 0.157*** 0.244*** −0.002 0.129**
(−11.705) (2.690) (3.508) (−0.417) (2.242)
ROA −0.008 0.793*** 0.774*** 0.031** 1.009***
(−1.167) (5.102) (4.323) (2.123) (6.238)
Book to Market 0.000 −0.113*** −0.115*** −0.001 −0.091***
(0.743) (−7.336) (−6.285) (−0.899) (−5.869)
SOE −0.004 −0.100 −0.170 −0.011 −0.221***
(−1.613) (−1.102) (−1.630) (−1.240) (−3.331)
Constant 0.072*** −1.930*** −1.757*** 0.035*** −1.507***
(11.057) (−13.163) (−10.025) (2.598) (−11.070)
YES
Year Effect YES YES YES YES YES
Industry Effect YES YES YES YES YES
Province Effect YES YES YES YES YES
N 17,295 17,295 17,295 17,295 17,295
Adj-R2 0.633 0.268 0.315 0.113 0.140
Kleibergen–Paap rk Wald F 1208.936***
statistic
Critical value − 10% 16.38
maximal IV size
This table presents the 2SLS estimation results where the instrument variable is Industry Average, defined as the industry average of the creation-oriented culture. Column (1) presents
first-stage regression results using Creation Culture as dependent variable. The dependent variables for second-stage regressions are: Ln(Inven+1), as the natural logarithm of the number
of invention patent applications plus one in column (2); Ln(Inven+Utility+1), as the natural logarithm of the number of invention and utility patent applications plus one in column
(3); Innovation Efficiency, as the number of invention patents scaled by firm’s cumulative R&D investment from year t−2 to year t in column (4); and Ln(Citations+1), as the natural
logarithm of the number of invention patent citations plus one in column (5). The full sample comprises 17,295 firm-year observations with 2,583 unique firms between 2008 and 2017.
The Kleibergen–Paap rk Wald F statistic is used in our regression. All variables are defined in Table A.1. All models include year, industry and province fixed effects. Robust standard
errors are clustered at the firm level. t-Values are in parentheses.
* , ** and *** represent significance levels at 10%, 5% and 1%, respectively.

Academy of Management.
Y. Wang, H. Farag and W. Ahmad

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Corporate Culture and Innovation 15

Overall, the results from tests using alternative more competitive environment compared to firms
measures of culture show that our key findings in low market competition. The chi-square test val-
are not sensitive to the measurement of creation ues from columns (1)–(4) indicate that the differ-
culture. ence in culture coefficients in low and high prod-
uct market competition subsamples is statistically
significant. These results suggest that creation cul-
Alternative estimation method
ture exerts a more pronounced effect on innovation
Since our patent and citation variables are count- in firms operating in a highly competitive product
based, with many observations having zero patents market, and are consistent with our prediction.
and citations, we use Poisson and Tobit regres- A creation-oriented culture focuses on creating
sion models following Jiang and Yuan (2018) and or generating new resources through innovation by
Mazouz and Zhao (2019). The results reported in fostering independent thinking, autonomy, flexi-
Table A.7 (in the online supporting information) bility and risk-taking (Hartnell, Ou and Kinicki,
are qualitatively similar to the main results. The 2011). However, experimentation and creativity in
creation culture is significantly related to all in- new products and services is associated with fail-
novation measures across both Poisson and Tobit ures and costs. Organizations that instil a creative
models. culture provide a flexible structure with adapt-
ability, where failure is accepted, tolerated and
considered part of the cost of doing novel work
Further analysis (Hutchison-Krupat and Chao, 2014). Moreover,
given the uncertainty and long-term nature of in-
In this section, we use cross-sectional variation in novative projects, creative cultures shield managers
firms’ product market competition and managers’ from short-term performance evaluations by pro-
career concerns to further explore the effect of cre- viding higher tolerance for failure. A strong cre-
ation culture on innovation. Firms in a competitive ative culture is likely to reduce managers’ career
product market strive to achieve and maintain a concerns, thereby facilitating innovation, and we
competitive advantage which is necessary for sur- expect that the effect of creation culture on inno-
vival in the market. Investment in R&D and in- vation varies with the extent of managers’ career
novation is vital for firms to attain a sustainable concerns within a cross-section of firms. We fol-
competitive advantage. If a creation-oriented cul- low Kim, Park and Song (2019) and use firm-level
ture leads to higher innovation, we expect to have profitability growth as a proxy for managers’ ca-
a stronger effect of creation culture in a highly reer concerns. Firms are split into high and low ca-
competitive product market. In order to test this reer concerns using the change in ROA growth rate
conjecture, we use the median value of the during a given year compared to the previous year.
Herfindahl Hirschman Index (HHI) to split our Firms with a ROA growth that changes from top to
sample into a high competitive product market bottom tertile are defined as firms with high career
(HHI below sample median) and a low competi- concern. Sample firms that do not belong to the
tive product market (HHI above sample median) high career concern subsample are defined as firms
following Chemmanur and Tian (2018) and Ma- with low career concern. The results of our anal-
zouz and Zhao (2019). Columns (1)–(4) of Table 6 ysis using subsamples based on career concerns
report the results using high and low product mar- are reported in columns (5)–(8) of Table 6. While
ket competition.23 The results show that the mag- significant in both subsamples, the coefficient esti-
nitude of coefficient estimates on creation culture mates of creation culture are much larger in high
is much larger and highly significant for firms in a career concern firms compared to low career con-
cern ones for both invention patents and citations.
total number of words in the annual report. The unre- Moreover, the difference in culture coefficients be-
ported results remain unchanged. tween low and high career concern subsamples
23
We use our baseline regression specification (Eq. (2)) is statistically significant. Consistent with our ex-
augmented with additional controls for board and CEO pectations, creation culture has a larger effect on
characteristics. For brevity, we only report results using
invention patents (Inven) and Citations as dependent vari- innovation in firms with high managerial career
ables. Unreported results are mainly consistent if we use concerns.
other innovation proxies.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
 16

Table 6. Product market competition and managerial career concerns

Ln(Inven+1) Citations Ln(Inven+1) Ln(Citations+1)

Dependent variable Low competition High competition Low competition High competition Low career concern High career concern Low career concern High career concern

(1) (2) (3) (4) (5) (6) (7) (8)

Creation Culture 0.586*** 0.981*** 0.024 1.099*** 0.634*** 1.132*** 0.386*** 1.184***
(2.564) (4.313) (0.103) (4.706) (3.342) (3.628) (1.986) (3.863)
Firm Controls YES YES YES YES YES YES YES YES
CEO Characteristics YES YES YES YES YES YES YES YES
Board Characteristics YES YES YES YES YES YES YES YES
Year Effect YES YES YES YES YES YES YES YES
Industry Effect YES YES YES YES YES YES YES YES
Province Effect YES YES YES YES YES YES YES YES
N 8,646 8,649 8,660 8,635 12,437 4,858 12,439 4,856
Adj-R2 0.320 0.274 0.167 0.161 0.276 0.316 0.152 0.165
χ 2 test statistics 6.157*** 5.644*** 8.157*** 7.215***

This table presents estimation results for various subsamples. Columns (1) and (3) presents the results for the low competitive product market subsample, while columns (2) and (4) show
the results for the high competitive product market subsample. Columns (5) and (7) present the results for the subsample of firms with low career concern, while columns (6) and (8)
show the results for the subsample of firms with high career concern. The dependent variables are: Ln(Inven+1), as the natural logarithm of the number of invention patent applications
plus one and Ln(Citations+1), as the natural logarithm of the number of invention patent citations plus one. All variables are defined in Table A.1. Firm controls and characteristics for
CEO and board are included. All models include year, industry and province fixed effects. Robust standard errors are clustered at the firm level. t-Values are in parentheses.
* , ** and *** represent significance levels at 10%, 5% and 1%, respectively.

Academy of Management.
Y. Wang, H. Farag and W. Ahmad

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Corporate Culture and Innovation 17

Conclusion Amore, M. D. and V. Failla (2018). ‘Pay dispersion and execu-

tive behaviour: evidence from innovation’, British Journal of
Management, 31, pp. 487–504.
In this study, we investigate whether, and to what Anderson, N., K. Potočnik and J. Zhou (2014). ‘Innovation
extent, corporate creation culture influences firm and creativity in organizations: a state-of-the-science review,
innovation in the Chinese market. We use textual prospective commentary, and guiding framework’, Journal of
analysis of annual reports following Fiordelisi and Management, 40, pp. 1297–1333.
Andreou, P. C., T. Harris and D. Philip (2020). ‘Measuring
Ricci (2014) and Andreou et al. (2019) to quantify
firms’ market orientation using textual analysis of 10-K fil-
a creation culture. Using all listed firms on both the ings’, British Journal of Management, 31, pp. 872–895.
Shanghai and Shenzhen stock exchanges between Andreou, P. C., F. Fiordelisi, T. Harris and D. Philip (2019).
2008 and 2017, we find strong and consistent evi- ‘Institutional ownership and firms’ thrust to compete’, SSRN
dence that a creation culture positively affects in- Electronic Journal. Available at [Link]
3120677.
novation. Our results are robust to using alterna-
Andreou, P. C., T. Harris, D. Philip and W. Zhang (2020). ‘The
tive proxies of innovation and creation culture, as OCAI textual measure of organizational culture: construct
well as to the alternative model specifications in- validation documentation’, SSRN Electronic Journal. Avail-
cluding tests related to endogeneity issues. We also able at [Link]
find that firms in more competitive markets and Antweiler, W. and M. Z. Frank (2004). ‘Is all that talk just noise?
The information content of internet stock message boards’,
subject to higher managerial career concerns ben-
The Journal of Finance, 59, pp. 1259–1294.
efit most from the cultural effect. Atalay, M., N. Anafarta and F. Sarvan (2013). ‘The relationship
We contribute to the growing stream of liter- between innovation and firm performance: an empirical evi-
ature on corporate culture as well as the deter- dence from Turkish automotive supplier industry’, Procedia -
minants of innovation. Recent studies have doc- Social and Behavioral Sciences, 75, pp. 226–235.
Atkinson, D. (1990). ‘Discourse analysis and written discourse
umented corporate culture as an important factor
conventions’, Annual Review of Applied Linguistics, 11, pp. 57–
in determining firm-level economic outcomes such 76.
as risk-taking and CEO turnover (Fiordelisi and Audi, R., T. Loughran and B. McDonald (2016). ‘Trust, but ver-
Ricci, 2014; Nguyen, Nguyen and Sila, 2019). We ify: MD&A language and the role of trust in corporate cul-
add to this stream of literature by showing that ture’, Journal of Business Ethics, 139, pp. 551–561.
Baer, M. (2012). ‘Putting creativity to work: the implementation
corporate culture is also an important factor for
of creative ideas in organizations’, Academy of Management
the innovative performance of firms. Corporate in- Journal, 55, pp. 1102–1119.
novation is the key to firms’ long-term competi- Balsmeier, B., L. Fleming and G. Manso (2017). ‘Independent
tiveness and sustainability, and finding factors that boards and innovation’, Journal of Financial Economics, 123,
determine innovation is equally important for aca- pp. 536–557.
Bednall, T. C., A. E. Rafferty, H. Shipton, K. Sanders and C.
demic experts as well as practitioners. Our findings
J. Jackson (2018). ‘Innovative behaviour: how much transfor-
suggest that a creation culture is essential to spur mational leadership do you need?’, British Journal of Manage-
innovation in firms, and these results have impor- ment, 29, pp. 796–816.
tant practical implications for managers, boards, Bénabou, R. and J. Tirole (2002). ‘Self-confidence and personal
shareholders and policymakers alike. motivation’, The Quarterly Journal of Economics, 117, pp. 871–
915.
Our study is not without limitations. First, al-
Berson, Y., S. Oreg and T. Dvir (2008). ‘CEO values, organiza-
though we have carried out a battery of robustness tional culture and firm outcomes’, Journal of Organizational
tests and provide evidence suggesting a causal in- Behavior, 29, pp. 615–633.
terpretation, we cannot rule out that unobserved Bertrand, M. and A. Schoar (2003). ‘Managing with style: the
characteristics could bias our findings. Second, we effect of managers on firm policies’, The Quarterly Journal of
Economics, 118, pp. 1169–1208.
rely on patent application data, and while patent-
Büschgens, T., A. Bausch and D. B. Balkin (2013). ‘Organiza-
based proxies have several advantages over other tional culture and innovation: a meta-analytic review’, Journal
innovation-related metrics such as R&D, they are of Product Innovation Management, 30, pp. 763–781.
imperfect measures of innovation and do not cap- Bushman, R. M., B. E. Hendricks and C. D. Williams (2016).
ture innovations that are not patented by the firms. ‘Bank competition: measurement, decision-making, and risk-
taking’, Journal of Accounting Research, 54, pp. 777–826.
Cameron, K. S. and R. E. Quinn (2011). Diagnosing and Chang-
References ing Organizational Culture: Based on the Competing Values
Framework. Chichester: Wiley.
Abbey, A. and J. W. Dickson (1983). ‘R&D work climate and in- Cameron, K. S., R. E. Quinn, J. DeGraff and A. V. Thakor
novation in semiconductors’, Academy of Management Jour- (2006). Competing Values Leadership. Cheltenham: Edward
nal, 26, pp. 362–368. Elgar.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
18 Y. Wang, H. Farag and W. Ahmad

Cao, X., D. Cumming and S. Zhou (2020). ‘State ownership and Fang, V. W., X. Tian and S. Tice (2014). ‘Does stock liquidity
corporate innovative efficiency’, Emerging Markets Review, 44, enhance or impede firm innovation?’, The Journal of Finance,
art. 100699. 69, pp. 2085–2125.
Cao, X., D. J. Cumming, S. Zhou and L. Zhou (2016). ‘State own- Fey, C. F. and D. R. Denison (2003). ‘Organizational culture and
ership and corporate innovation efficiency’, SSRN Electronic effectiveness: can American theory be applied in Russia?’, Or-
Journal. Available at [Link] ganization Science, 14, p. 686.
Chandler, G. N., C. Keller and D. W. Lyon (2000). ‘Unrav- Fiordelisi, F. and O. Ricci (2014). ‘Corporate culture and CEO
eling the determinants and consequences of an innovation- turnover’, Journal of Corporate Finance, 28, pp. 66–82.
supportive organizational culture’, Entrepreneurship Theory Fiordelisi, F., L. Renneboog, O. Ricci and S. S. Lopes (2019).
and Practice, 25, pp. 59–76. ‘Creative corporate culture and innovation’, SSRN Electronic
Chatman, J. A. and F. J. Flynn (2001). ‘The influence of demo- Journal. Available at [Link]
graphic heterogeneity on the emergence and consequences of Graham, J. R., C. R. Harvey, J. Popadak and S. Rajgopal (2017).
cooperative norms in work teams’, Academy of Management ‘Corporate culture: evidence from the field’, NBER Working
Journal, 44, pp. 956–974. Paper 23255.
Chemmanur, T. J. and X. Tian (2018). ‘Do antitakeover provi- Gregory, B. T., S. G. Harris, A. A. Armenakis and C. L. Shook
sions spur corporate innovation? A regression discontinuity (2009). ‘Organizational culture and effectiveness: a study of
analysis’, Journal of Financial and Quantitative Analysis, 53, values, attitudes, and organizational outcomes’, Journal of
pp. 1163–1194. Business Research, 62, pp. 673–679.
Chemmanur Thomas J., E. Loutskina and X. Tian (2014). Cor- Guiso, L., P. Sapienza and L. Zingales (2015). ‘The value of cor-
porate venture capital, value creation, and innovation. Re- porate culture’, Journal of Financial Economics, 117, pp. 60–
view of Financial Studies, 27 (8), 2434–2473. [Link] 76.
10.1093/rfs/hhu033. Gumusluoğlu, L. and A. Ilsev (2009). ‘Transformational leader-
Clifford, J. and G. E. Marcus (1986). Writing Culture: The Po- ship and organizational innovation: the roles of internal and
etics and Politics of Ethnography: A School of American Re- external support for innovation’, Journal of Product Innovation
search Advanced Seminar. Berkeley, CA: University of Cali- Management, 26, pp. 264–277.
fornia Press. Hall, B. H. and R. H. Ziedonis (2001). ‘The patent paradox re-
Coad, A., A. Segarra and M. Teruel (2013). ‘Like milk or wine: visited: an empirical study of patenting in the US semicon-
does firm performance improve with age?’, Structural Change ductor industry, 1979–1995’, RAND Journal of Economics, 32,
and Economic Dynamics, 24, pp. 173–189. pp. 101–128
Cumming, D., R. Peter, A. Sannajust and M. Tarsalewska (2019). Hall, B. H., A. Jaffe and M. Trajtenberg (2005). ‘Market value
‘Pre-going private ownership around the world’, British Jour- and patent citations’, RAND Journal of Economics, 36, pp. 16–
nal of Management, 30, pp. 692–711. 38.
Danneels, E. (2008). ‘Organizational antecedents of second- Hanley, K. W. and G. Hoberg (2010). ‘The information content
order competences’, Strategic Management Journal, 29, of IPO prospectuses’, The Review of Financial Studies, 23, pp.
pp. 519–543. 2821–2864.
Dastmalchian, A., S. Lee and I. Ng (2000). ‘The interplay be- Hartnell, C. A., A. Y. Ou and A. Kinicki (2011). Organizational
tween organizational and national cultures: a comparison of Culture and Organizational Effectiveness: A Meta-analytic In-
organizational practices in Canada and South Korea using the vestigation of the Competing Values Framework’s Theoretical
competing values framework’, International Journal of Human Suppositions. Washington, D.C: American Psychological As-
Resource Management, 11, pp. 388–412. sociation.
David, P. and J. O’Brien (2006). ‘Capital structure, R&D invest- He, J. J. and X. Tian (2013). ‘The dark side of analyst coverage:
ment, and firm performance: effects of heterogeneity among the case of innovation’, Journal of Financial Economics, 109,
lenders’. Academy of Management Proceedings, pp. JJ1–JJ6. pp. 856–878.
Denison, D. R. and A. K. Mishra (1995). ‘Toward a theory of or- Hirshleifer, D., P.-H. Hsu and D. Li (2013). ‘Innovative efficiency
ganizational culture and effectiveness’, Organization Science, and stock returns’, Journal of Financial Economics, 107, pp.
6, pp. 204–223. 632–654.
Deshpandé, R. and J. U. Farley (2004). ‘Organizational culture, Hirshleifer D., A. Low and S. H. Teoh (2012). Are overconfident
market orientation, innovativeness, and firm performance: an ceos better innovators?. The Journal of Finance, 67 (4), 1457–
international research odyssey’, International Journal of Re- 1498. [Link]
search in Marketing, 21, pp. 3–22. Hoberg, G. and V. Maksimovic (2014). ‘Redefining financial con-
Dougherty, D. and T. Heller (1994). ‘The illegitimacy of success- straints: a text-based analysis’, The Review of Financial Stud-
ful product innovation in established firms’, Organization Sci- ies, 28, pp. 1312–1352.
ence, 5, pp. 200–218. Hoberg, G. and G. Phillips (2010). ‘Product market synergies and
Ettlie, J. E. (1998). ‘R&D and global manufacturing perfor- competition in mergers and acquisitions: a text-based analy-
mance’, Management Science, 44, pp. 1–11. sis’, The Review of Financial Studies, 23, pp. 3773–3811.
Fana, J. P. H., K. C. J. Weib and X. Xuc (2011). ‘Corporate fi- Hofstede, G. (1991). Organizations and Cultures: Software of the
nance and governance in emerging markets: a selective review Mind. New York: McGraw-Hill.
and an agenda for future research’, Journal of Corporate Fi- Hofstede, G., B. Neuijen, D. D. Ohayv and G. Sanders (1990).
nance, 17, pp. 207–214. ‘Measuring organizational cultures: a qualitative and quanti-
Fang, L. H., J. Lerner and C. Wu (2017). ‘Intellectual property tative study across twenty cases’, Administrative Science Quar-
rights protection, ownership, and innovation: evidence from terly, 35, pp. 286–316.
China’, The Review of Financial Studies, 30, pp. 2446–2477.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
Corporate Culture and Innovation 19

Hogan, S. J. and L. V. Coote (2014). ‘Organizational culture, in- Language in Mind: Advances in the Study of Language and
novation, and performance: a test of Schein’s model’, Journal Thought, pp. 25–[Link], MA: MIT Press.
of Business Research, 67, pp. 1609–1621. Li, F. (2008). ‘Annual report readability, current earnings, and
Holmstrom, B. (1989). ‘Agency costs and innovation’, Journal of earnings persistence’, Journal of Accounting and Economics,
Economic Behavior & Organization, 12, pp. 305–327. 45, pp. 221–247.
Huang, J. and D. J. Kisgen (2013). ‘Gender and corporate fi- Li, F., R. Lundholm and M. Minnis (2013). ‘A measure of com-
nance: are male executives overconfident relative to female petition based on 10-K filings’, Journal of Accounting Re-
executives?’, Journal of Financial Economics, 108, pp. 822– search, 51, pp. 399–436.
839. Li, M., H. Wu, M. Xiao and J. You (2019). ‘Beyond cheap talk:
Hurley, R. F. and G. T. M. Hult (1998). ‘Innovation, market ori- management’s informative tone in corporate disclosures’, Ac-
entation, and organizational learning: an integration and em- counting & Finance, 59, pp. 2905–2959.
pirical examination’, Journal of Marketing, 62, pp. 42–54. Loughran, T. and B. McDonald (2011). ‘When is a liability not a
Hutchison-Krupat, J. and R. O. Chao (2014). ‘Tolerance for fail- liability? Textual analysis, dictionaries, and 10-Ks’, The Jour-
ure and incentives for collaborative innovation’, Production nal of Finance, 66, pp. 35–65.
Loughran, T. and B. McDonald (2014). ‘Measuring readability
and Operations Management, 23, pp. 1265–1285.
in financial disclosures’, The Journal of Finance, 69, pp. 1643–
Islam, E. and J. Zein (2020). ‘Inventor CEOs’, Journal of Finan-
cial Economics, 135, pp. 505–527. 1671.
Manso, G. (2011). ‘Motivating innovation’, The Journal of Fi-
Janz, N., H. Lööf and B. Peters (2003). ‘Firm level innova-
nance, 66, pp. 1823–1860.
tion and productivity – is there a common story across coun-
Martin, J. (1992). Cultures in Organizations: Three Perspectives.
tries?’, SSRN Electronic Journal. Available at [Link]
Oxford: Oxford University Press.
abstract=416444. Mazouz, K. and Y. Zhao (2019). ‘CEO incentives, takeover pro-
Jaskyte, K. (2004). ‘Transformational leadership, organizational tection and corporate innovation’, British Journal of Manage-
culture, and innovativeness in nonprofit organizations’, Non- ment, 30, pp. 494–515.
profit Management and Leadership, 15, pp. 153–168. Mumby, D. K. and C. Stohl (1991). ‘Power and discourse in or-
Jiang, X. and Q. Yuan (2018). ‘Institutional investors’ corporate ganization studies: absence and the dialectic of control’, Dis-
site visits and corporate innovation’, Journal of Corporate Fi- course & Society, 2, pp. 313–332.
nance, 48, pp. 148–168. Nash, R. and A. Patel (2019). ‘Instrumental variables analysis
Jiang, F., K. A. Kim, Y. Ma, J. R. Nofsinger and B. Shi (2017). and the role of national culture in corporate finance’, Financial
‘Corporate culture and investment–cash flow sensitivity’, Jour- Management, 48, pp. 385–416.
nal of Business Ethics, 154, pp. 425–439. Nelson, R. R. and S. G. Winter (1982). An Evolutionary Theory
Kaasa, A. and M. Vadi (2010). ‘How does culture contribute to of Economic Change. London: Belknap Press.
innovation? Evidence from European countries’, Economics of Nguyen, D. D., L. Nguyen and V. Sila (2019). ‘Does corpo-
Innovation and New Technology, 19, pp. 583–604. rate culture affect bank risk-taking? Evidence from loan-level
Kim, H.-D., K. Park and R. K. Song (2019). ‘Do long-term in- data’, British Journal of Management, 30, pp. 106–133.
stitutional investors foster corporate innovation?’, Accounting Nguyen, T. (2018). ‘CEO incentives and corporate innovation’,
& Finance, 59, pp. 1163–1195. Financial Review, 53, pp. 255–300.
Kortum, S. and J. Lerner (2001). ‘Does venture capital spur O’Reilly, C. A. and J. A. Chatman (1996). ‘Culture as social con-
innovation?’. In G. D. Libecap (ed.), Entrepreneurial Inputs trol: corporations, cults, and commitment’, Research in Orga-
and Outcomes: New Studies of Entrepreneurship in the United nizational Behavior, 18, pp. 157–200.
States, pp. 1–[Link]: Emerald Group Publishing. O’Reilly, C. A., J. Chatman and D. F. Caldwell (1991). ‘People
Kotter, J. P. and J. L. Heskett (1992). Corporate Culture and Per- and organizational culture: a profile comparison approach to
formance. New York: The Free Press. assessing person–organization fit’, The Academy of Manage-
Kreps, D. M. (1996). ‘Corporate culture and economic theory’. ment Journal, 34, pp. 487–516.
In P. J. Buckley and J. Michie (eds), Firms, Organizations and Oster, E. (2019). ‘Unobservable selection and coefficient sta-
Contracts, pp. 221–[Link]: Oxford University Press. bility: theory and evidence’, Journal of Business & Economic
Kwan, P. and A. Walker (2004). ‘Validating the competing values Statistics, 37, pp. 187–204.
model as a representation of organizational culture through Porter, M. E. (1992). ‘Capital disadvantage: America’s fail-
inter-institutional comparisons’, Organizational Analysis, 12, ing capital investment system’, Harvard Business Review, 70,
pp. 21–37. pp. 65–82.
Lamond, D. (2003). ‘The value of Quinn’s competing values Ralston, D. A., J. Terpstra-Tong, R. H. Terpstra, X. Wang and
model in an Australian context’, Journal of Managerial Psy- C. Egri (2006). ‘Today’s state-owned enterprises of China: are
chology, 18, pp. 46–59. they dying dinosaurs or dynamic dynamos?’, Strategic Man-
Lau, C. M. and H. Y. Ngo (2004). ‘The HR system, organiza- agement Journal, 27, pp. 825–843.
tional culture, and product innovation’, International Business Sackmann, S. A. (2002). ‘Culture and performance’. In N. M.
Review, 13, pp. 685–703. Ashkanasy, C. P. M. Wilderom and M. F. Peterson (eds),
Lau, C.-M., K. T. David and N. Zhou (2002). ‘Institutional Handbook of Organizational Culture and Climate, pp. 188–
forces and organizational culture in China: effects on change [Link] Oaks, CA: Sage.
schemas, firm commitment and job satisfaction’, Journal of In- Schein, E. H. (1992). Organizational Culture and Leadership, 2nd
ternational Business Studies, 33, pp. 533–550. edn. Chichester: Wiley.
Lev, B. and T. Sougiannis (1999). ‘Penetrating the book-to- Schein, E. H. (2010). Organizational Culture and Leadership, 4th
market black box: the R&D effect’, Journal of Business Finance edn. Chichester: Wiley.
& Accounting, 26, pp. 419–449. Scherer, F. M. (1984). New Perspectives on Economic Growth and
Levinson, S. C. (2003). ‘Language and mind: let’s get the is- Technological Innovation. Washington, D.C: Brookings Insti-
sues straight’. In D. Gentner and S. Goldin-Meadow (eds), tution Press.

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.
20 Y. Wang, H. Farag and W. Ahmad

Schneider, S. C. (1988). ‘National vs. corporate culture: impli- Tetlock, P. C. (2007). ‘Giving content to investor sentiment: the
cations for human resource management’, Human Resource role of media in the stock market’, The Journal of Finance, 62,
Management, 27, pp. 231–246. pp. 1139–1168.
Simsek, Z. (2007). ‘CEO tenure and organizational performance: Tuan, N., N. Nhan, P. Giang and N. Ngoc (2016). ‘The effects
an intervening model’, Strategic Management Journal, 28, pp. of innovation on firm performance of supporting industries in
653–662. Hanoi, Vietnam’, Journal of Industrial Engineering and Man-
Solow, R. M. (1957). ‘Technical change and the aggregate pro- agement, 9, p. 413.
duction function’, The Review of Economics and Statistics, 39, Van den Steen, E. (2010). ‘On the origin of shared beliefs (and
pp. 312–320. corporate culture)’, The RAND Journal of Economics, 41, pp.
Sørensen, J. B. (2002). ‘The strength of corporate culture and the 617–648.
reliability of firm performance’, Administrative Science Quar- van Muijen, J. J. and D. L. Turnipseed (1999). ‘Organizational
terly, 47, pp. 70–91. culture: the focus questionnaire’, European Journal of Work
Staw, C. and C. Nemeth (1989). ‘The tradeoffs of social and Organizational Psychology, 8, pp. 551–568.
control and innovation in groups and organizations’, Ad- You, J., B. Zhang and L. Zhang (2017). ‘Who captures the power
vances in Experimental Social Psychology, 22, pp. 175– of the pen?’, The Review of Financial Studies, 31, pp. 43–96.
210. Yu, T. and N. Wu (2009). ‘A review of study on the competing
Stock, J. H. and M. Yogo (2005). Testing for Weak Instruments in values framework’, International Journal of Business and Man-
Linear IV Regression. Andrews, DWK, Identification and Infer- agement, 4, pp. 37–42.
ence for Econometric Models. New York: Cambridge Univer- Yuan, R. and W. Wen (2018). ‘Managerial foreign experience and
sity Press, 80–108. corporate innovation’, Journal of Corporate Finance, 48, pp.
Tan, Y., X. Tian, X. Zhang and H. Zhao (2015). ‘The real effects 752–770.
Zhang, M., H. Li and J. Wei (2008). ‘Examining the relation-
of privatization: evidence from China’s split share structure re-
ship between organizational culture and performance: the per-
form’, SSRN Electronic Journal. Available at [Link]
spectives of consistency and balance’, Frontiers of Business Re-
abstract=2433824.
search in China, 2, p. 256.
Tian X. and T. Y. Wang (2014). Tolerance for failure and corpo-
Zuckerman, A. (2002). ‘Strong corporate cultures and firm per-
rate innovation. Review of Financial Studies, 27 (1), 211–255.
formance: are there tradeoffs?’, Academy of Management Per-
[Link]
spectives, 16, pp. 158–160.

Yaopeng Wang is a PhD student in the Finance Department at Birmingham Business School. His
current research areas are corporate governance, corporate innovation, corporate culture and board
characteristics. He holds an MSc in Financial Management from the University of Birmingham.

Hisham Farag is a Professor of Finance at Birmingham Business School, where he chairs the Depart-
ment of Finance. His current research areas are corporate finance, corporate governance, sustainable
finance and FinTech, in which he has published widely. Hisham is an Associate Editor of the European
Journal of Finance and Research in International Business and Finance.

Wasim Ahmad is a Lecturer in Finance at Birmingham Business School. His research focuses mainly
on initial public offerings, corporate innovation, venture capital and private equity. He has published in
journals such as Journal of Corporate Finance, International Review of Financial Analysis and Journal
of Business Finance and Accounting.

Supporting Information
Additional supporting information may be found online in the Supporting Information section at the end
of the article.
Table A.1. Definition of variables
Table A.2. CVF bag of words
Table A.3. Industry × Province and Year × Province fixed effects
Table A.4. Alternative measure of creation culture
Table A.5. Reduced creation culture ‘bag of words’ and innovation
Table A.6. OCAI-based creation culture measure and innovation
Table A.7. Poisson and Tobit regressions

© 2021 The Authors. British Journal of Management published by John Wiley & Sons Ltd on behalf of British
Academy of Management.