The Impact of Cultural Diversity in Corporate Boards on Firm Performance

Bart Frijns, Olga Dodd, Helena Cimerova

PII: S0929-1199(16)30094-3
DOI: doi: 10.1016/j.jcorpfin.2016.07.014
Reference: CORFIN 1067

To appear in: Journal of Corporate Finance

Received date: 4 January 2015

Revised date: 28 July 2016
Accepted date: 29 July 2016

Please cite this article as: Frijns, Bart, Dodd, Olga, Cimerova, Helena, The Impact
of Cultural Diversity in Corporate Boards on Firm Performance, Journal of Corporate
Finance (2016), doi: 10.1016/j.jcorpfin.2016.07.014

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT

The Impact of Cultural Diversity in Corporate Boards on Firm Performance§

T
R IP
SC
NU
Bart Frijnsa,*
MA
Olga Dodda

Helena Cimerovaa
D
P TE

a
Department of Finance, Auckland University of Technology, Auckland, New Zealand
CE
AC

§This paper has benefitted from feedback and discussions with Mario Daniele Amore, Sjoerd
Beugelsdijk, Sadok El Ghoul, Miguel Ferreira, Ray Friedman, Jarrad Harford, Thanh Huynh,
William Mayew, and Alireza Tourani-Rad. We also thank participants of the 2015
Conference on Culture and Finance, Winston-Salem, US; the 2015 FMA Conference,
Orlando, US; the 2015 SFA Annual Meetings, Florida, US; the 6th Financial Markets and
Corporate Governance Conference, Perth, Australia; the 2015 New Zealand Finance
Colloquium, Hamilton, New Zealand; the 2014 Auckland Finance Meeting, Auckland, New
Zealand, and seminar participants at the Auckland University of Technology, ISCTE
Business School, Maastricht University, Vanderbilt University, Victoria University of
Wellington, University of Auckland, University of Otago, University of Queensland, and the
University of Technology Sydney for useful comments and suggestions. Olga Dodd and Bart
Frijns acknowledge the financial support from Auckland University of Technology through
Contestable Research Grants (Ref: RP2014-13).
*Corresponding author. Department of Finance, Auckland University of Technology, Private
Bag 92006, 1142 Auckland, New Zealand. Tel.: +64 9 921 9999 (ext. 5706); fax: +64 9 921
9940; email: bart.frijns@aut.ac.nz.

1
 ACCEPTED MANUSCRIPT

The impact of cultural diversity in corporate boards on firm performance

T
IP
Abstract

R
We examine the impact of cultural diversity in boards of directors on firm performance. We

SC
construct a measure of national cultural diversity by calculating the average of cultural
distances between board members using Hofstede’s culture framework. Our findings indicate
that national cultural diversity in boards negatively affects firm performance measured by

NU
Tobin’s Q and ROA. These results hold after controlling for potential endogeneity using firm
fixed effects and instrumental variables regressions. Further, the results are robust to
MA
controlling for a wide range of board and firm characteristics, including various measures of
“foreignness” of the firm, alternative culture frameworks, and other measures of culture. The
negative impact of cultural diversity on performance is mitigated by the complexity of the
D

firm and the size of foreign sales and operations. In addition, we find that the negative effects
TE

of cultural diversity are concentrated among the independent directors. Finally, we find that
not all aspects of cultural differences are equally important and that it is mainly the diversity
P

in individualism and masculinity that affects the effectiveness of boards of directors.

CE
AC

Keywords: cultural diversity, cultural distance, board diversity, firm performance

JEL classification: G3

2
 ACCEPTED MANUSCRIPT

1. Introduction

Numerous studies have documented that culture affects financial decision-making and

financial outcomes. In corporate finance, this evidence relates mostly to strategic decision-

T
making and firm-level outcomes.1 Studies in this area either use culture as a national trait that

IP
can explain cross-country differences in corporate practices (e.g., Bryan et al., 2015; El

R
Ghoul and Zheng, 2016; Zheng et al., 2012), or focus on cultural differences and how they

SC
affect financial outcomes (e.g., Ahern et al., 2015; Beugelsdijk and Frijns, 2010; Karolyi,

NU
2016). Whereas previous research on cultural differences examines the impact of cultural

differences between groups, in this paper, we focus on cultural diversity, which measures
MA
cultural differences within groups – a novel concept in the culture and finance literature. We

introduce a measure of within-group cultural diversity to examine the impact of national

D

cultural diversity in corporate boards on firm performance and document a robust negative
TE

relation between cultural diversity in corporate boards and firm performance. We make two

important contributions in this paper. First, we contribute to the culture and finance literature
P
CE

by providing a novel metric that can be used to assess the impact of culture within an

organization. 2 Second, we contribute to the literature on board diversity (which has

AC

predominantly focused on board independence and gender diversity) by introducing cultural

diversity, a thus-far overlooked aspect of board diversity.

Given the importance of boards as the key governing body in a corporation and one of the

principal groups to make corporate decisions, many studies have focused on the connection

1
See Aggarwal and Goodell (2009), Aggarwal et al. (2012), and Zheng et al. (2012) on financing decisions;
Ahern et al. (2015), Ferris et al. (2013), and Frijns et al. (2013) on mergers and acquisitions; Dodd et al. (2015)
on cross-listing decisions, Li et al. (2013) and Lievenbrück and Schmid (2014) on corporate risk-taking and
hedging decisions. Bryan et al. (2015), Burns et al. (in press), and Tosi and Greckhamer (2004) show the
importance of culture for CEO compensation, Chui et al. (2002) for capital structure, and Shao et al. (2010) for
dividend policy.
2
We apply this metric to measuring the impact of cultural diversity in corporate boards on firm performance, but
this metric can easily be adapted to other settings.

3
 ACCEPTED MANUSCRIPT

between board effectiveness and different board characteristics, including various aspects of

board diversity such as independence, gender, ethnicity, education, experience, and tenure

(e.g., Adams and Ferreira, 2009; Anderson et al., 2011; Ferreira, 2010).3 Despite a significant

T
number of studies conducted on the issue, the debate on the effects of diversity on boards is

IP
still ongoing.4 The importance of understanding the role of diversity is also recognized by

R
regulators. In particular, the 2014 UK Corporate Governance Code articulates this as follows:

SC
NU
“One of the ways in which constructive debate can be encouraged is through having

sufficient diversity on the board. This includes, but is not limited to, gender and race.
MA
Diverse board composition in these respects is not on its own a guarantee. Diversity

is as much about differences of approach and experience, and it is very important in

D

ensuring effective engagement with key stakeholders and in order to deliver the
TE

business strategy.” (Financial Reporting Council, 2014, p. 2)

P
CE

Although diversity on boards has been extensively studied, to our knowledge, national
AC

cultural diversity within the board has not been considered in the corporate finance literature.

Outside the corporate finance literature, however, there is a long-standing theoretical debate

on cultural diversity and how it affects group outcomes. The management literature generally

identifies cultural diversity as a “double-edged sword” (Milliken and Martins, 1996),

3
Boards work as corporate monitors and advisors, they are involved in the company’s most important strategic,
investment, and financing decisions as well as in hiring, assessing, and firing top managers. Both the
management and economic literature recognize the two main roles for boards – the advisory and the monitoring
role – and both strands of literature agree that firms choose directors according to their needs (Ferreira, 2010).
4
Empirical evidence on the effects of diversity on boards is inconclusive. For instance, in terms of the
relationship between gender diversity on boards and firm performance, Farrell and Hersch (2005) report no
effects, Carter et al. (2003) report positive effects, and Adams and Ferreira (2009) negative effects.

4
 ACCEPTED MANUSCRIPT

recognizing both positive and negative aspects of cultural diversity.5 On the positive side,

cultural diversity engenders information elaboration, offering a diverse range of knowledge

and perspectives (Nederveen Pieterse et al., 2013). In addition, foreign nationals can bring in

T
specific knowledge of their home countries, which may benefit the firm if it has operations in

IP
that market (Maznevski, 1994). This aspect of cultural diversity could explain, for instance,

R
the findings of Masulis et al. (2012) who demonstrate that firms with foreign independent

SC
directors make better cross-border acquisitions when the targets are from their home country.

On the negative side, cultural diversity imposes frictions. In culturally diverse groups,

NU
coordination is more difficult, communication is slower, more confused, and more frequently
MA
a source of misunderstanding (Anderson et al., 2011; Doney et al., 1998). In addition, cultural

diversity can lead to lower levels of intragroup trust (Bjørnskov, 2008). These negative

aspects of cultural diversity can be linked to the findings of Ahern et al. (2015), who present
D
TE

strong evidence that cultural distance between an acquiring firm and its target reduces the

likelihood of a successful acquisition.

P
CE

In this paper, we evaluate the impact of cultural diversity in corporate boards on firm
AC

performance as measured by Tobin’s Q and ROA. We focus on a sample of UK firms that

constitute about 95% of the market capitalization of the London Stock Exchange between

2002 and 2014.6 We introduce a new measure of cultural diversity of a firm’s board, defined

as the average of cultural distances between board members. For our main analysis, this

measure of cultural diversity is based on the culture framework of Hofstede (2001). We

document a significant negative relation between cultural diversity and firm performance,

5
Positive and negative aspects of diversity on boards are discussed also in the business press. For example,
Manzoni et al. (2010) give a practitioners’ account of the workings of diverse boards.
6
We choose a UK sample as there are considerably more boards with foreigners in UK firms than in US firms.
In our sample, more than 62% of the firms have at least one foreigner on the board, whereas in the US sample of
Masulis et al. (2012), only 13% of firms have at least one foreign independent director on the board.

5
 ACCEPTED MANUSCRIPT

which suggests that the frictions imposed by cultural diversity, on average, outweigh the

positive aspects. This result is robust to the inclusion of various firm and board

characteristics, the use of alternative culture frameworks in the computation of cultural

T
diversity (House et al., 2004 – GLOBE; Schwartz, 2006; Tang and Koveos, 2008), and

IP
alternative measures for culture (bilateral trust as in Guiso et al., 2009, language and religion

R
similarity, and genetic and somatic diversity). Our findings are economically significant.

SC
Firms at the 75th cultural diversity percentile achieve a Tobin’s Q that is 0.23 (about 11.5% of

the average Tobin’s Q) less than firms at the 25th percentile. Similarly, ROA is 1.43% (about

NU
13% of the average ROA) lower for firms at the 75th percentile compared with firms at the

25th percentile. We address the issue of endogeneity by estimating fixed effects regressions
MA
and by implementing an instrumental variables approach. In both specifications, the negative

effect of cultural diversity on firm performance remains, lending support to a causal

D
TE

interpretation of our findings.

P
CE

We also consider other factors that may be related to board composition and firm

performance. We examine whether cultural diversity is merely a proxy for the degree of a
AC

firm’s foreign orientation, that is, a firm’s “foreignness.” At the board level, our results are

not just a consequence of having foreign directors on the board. Further, the inclusion of a

dummy for the presence of a foreign independent director or the percentage of foreign

independent directors, which are the two main variables of interest in Masulis et al. (2012),

does not affect our results. At the firm level, neither the inclusion of the firm’s foreign sales

or foreign assets, nor controlling for whether the firm has a listing on the NYSE or in other

foreign markets explains away our findings.

6
 ACCEPTED MANUSCRIPT

The notion that cultural diversity is a “double-edged sword” with both positive and negative

consequences suggests that not all firms may be affected equally by cultural diversity.

Therefore, we distinguish first between firms based on the level of complexity of the firm’s

T
operations, measured by the number of business segments in which the firm operates. We

IP
find that for complex firms, i.e., firms that operate in more than three business segments,

R
performance is not affected by cultural diversity, whereas for non-complex firms, i.e., firms

SC
that operate in three or less business segments, performance is negatively affected. This may

be due to the positive aspects of diversity (such as a more diverse range of knowledge)

NU
outweighing the negative aspects in complex firms. Second, we consider the extent of the

firm’s presence in foreign markets. We find that for firms with a strong presence in foreign
MA
markets, as measured by the proportion of foreign sales or assets, the negative relation

between performance and cultural diversity disappears. This could be attributed to market-
D
TE

specific knowledge that foreign directors bring to the board and corroborates the finding of

Masulis et al. (2012) on the role of foreign directors in cross-border acquisitions and firm
P

performance.
CE
AC

We further examine whether the negative relation between cultural diversity and firm

performance is more prevalent among dependent or independent directors, and find that the

negative effect is concentrated among the independent directors. In the context of the

organizational learning and culture literature (see, e.g., Ouchi, 1979), this result suggests that

dependent directors may be indoctrinated by organizational culture, which may mitigate the

negative effect of national cultural diversity. Likewise, this finding is consistent with

dependent directors being in more frequent contact with one another, thus being able to learn

to deal with cultural differences between one another. Lastly, given that dependent and

independent directors have different roles in the board (advising versus monitoring), where

7
 ACCEPTED MANUSCRIPT

one of the key roles of independent directors is to monitor, our finding suggests that cultural

diversity may be more disruptive to the monitoring task of the independent directors.

T
IP
Finally, we assess the importance of each of the individual cultural dimensions of Hofstede

(2001) separately. Our findings suggest that differences in the individualism-collectivism and

R
SC
masculinity-femininity dimensions of culture most strongly affect firm performance. The

greatest degree of friction arguably arise along these two dimensions as these two dimensions

NU
relate to people’s perception of how the group process should work. Our findings are in line

with the observations of Gudykunst and Bond (1997) and Kirkman et al. (2006), who report
MA
that the individualism dimension of culture is the most salient dimension of cultural

heterogeneity in intragroup processes. These findings are also in line with Lim et al. (2016)
D

who find that differences in individualism and masculinity are the most important cultural
TE

dimensions that affect the bid premium in cross-border takeovers.

P
CE

Overall, our paper makes two important contributions. First, we contribute to the growing
AC

literature on culture and finance by introducing a new measure of cultural diversity of

corporate boards. While previous studies in finance use culture mainly to explain cross-

country differences in financial outcomes and decision-making, or use cultural distance to

test the impact of differences between groups on firm outcomes, we explore cultural diversity

as an avenue of assessing the impact of differences within groups on firm outcomes. We

show that cultural diversity represents an important aspect of board diversity. To our

knowledge, this is the first study that examines the role of cultural diversity of the board on

firm performance. Second, we expand on insights from the board diversity literature and, in

line with the well-established finance studies in the area, caution against “romanticizing”

8
 ACCEPTED MANUSCRIPT

cultural diversity when it comes to company outcomes. Our baseline results suggest that the

negative aspects of cultural diversity prevail over the positive aspects. However, factors such

as firm complexity and presence in foreign markets mitigate the negative effects of cultural

T
diversity. Our study on the relation between board diversity and firm performance offers a

IP
more nuanced story of board effectiveness.7

R
SC
This paper is related to several studies in corporate finance. First, our study relates to Masulis

NU
et al. (2012), who examine the role of foreign independent directors in corporate governance

and firm performance. Masulis et al. (2012) find that having a foreign independent director
MA
on the board is detrimental to firm performance and attribute this to a decreased ability of

foreign independent directors to effectively monitor the firm due to higher coordination costs.
D

We confirm the finding that foreign directors have a detrimental effect on firm performance,
TE

but show that this is primarily due to cultural differences. Another study to which our work is
P

related is that of Anderson et al. (2011). To evaluate the effect of board heterogeneity on firm
CE

performance, Anderson et al. (2011) construct a board heterogeneity index that considers

occupational and social heterogeneity. Overall, they find a positive relation between board
AC

heterogeneity and firm performance, which can be attributed to the benefits of overall

diversity outweighing the costs. However, when they examine the role of board heterogeneity

in complex versus non-complex firms, they find that board heterogeneity is negatively related

to the performance of non-complex firms. We extend their work by considering the cultural

diversity on boards. Although we do not find a significant positive relation between cultural

7
An important question is why firms adopt culturally diverse boards if this results in lower performance, and
why they do not quickly realize that diversity lowers performance. One argument is that diversity, in public
perception, is generally deemed to be a good thing. The focus on diversity may be promoted by media,
shareholders, and regulatory bodies, or companies adopt more diverse boards in an attempt to improve their
public image. Further, the negative consequences of cultural diversity may not be easily observed. As
Kirchmeyer and Cohen (1992) point out, conflict in multicultural teams may be difficult to identify.

9
 ACCEPTED MANUSCRIPT

diversity and firm performance,8 we do find a negative relation between cultural diversity of

the board and firm performance for non-complex firms.

T
IP
The paper proceeds as follows. Section 2 offers a discussion on diversity in boards and on our

measurement of cultural diversity. Section 3 describes the data and discusses the results from

R
SC
our univariate analysis. In Section 4, we present our baseline results as well as the results

from robustness tests, and discuss the findings. Section 5 concludes.

NU
MA
2. Background

2.1. Diversity on boards

D

If rather than a single entity, we view boards as “groups of diverse individuals who have
TE

different biases and prejudices and whose behavior is affected by social constraints and
P

power relations” (Ferreira, 2010, p. 225), diversity on boards emerges as a significant

CE

governance issue. Indeed, the current board diversity debate raises numerous contentious

issues and extends to a wide array of stakeholders. The importance of diversity is often
AC

highlighted in terms of gender, age, experience, etc. The question remains, however, whether

different types of diversity on corporate boards have an actual economic impact and enhance

shareholder value, or whether striving for diversity simply reflects overall social trends

towards assuring equity (Carter et al., 2003).

8
This could be due to the fact that we focus on one particular aspect of diversity whose benefits may be smaller.
Indeed, Anderson et al. (2011) find that the positive impact of social heterogeneity is much weaker than that of
occupational heterogeneity.

10
 ACCEPTED MANUSCRIPT

In examining the role of board diversity and its impact on firm value, many studies consider

director independence, that is, they distinguish between dependent and independent directors

on boards (e.g., Crespí-Cladera and Pascual-Fuster, 2014; Linck et al., 2008), or gender

T
diversity on corporate boards (e.g., Adams and Ferreira, 2009; Farrell and Hersch, 2005).

IP
Other studies tackle board diversity by simultaneously taking into account a number of

R
director characteristics such as gender, age, ethnicity, race, education, experience, function,

SC
rank, and profession (Anderson et al., 2011; White et al., 2014).

NU
One factor that thus far has been ignored in the board diversity literature is national cultural
MA
diversity. In this paper, we view national culture as a set of values and beliefs that guides how

people select or evaluate actions, policies, events, or other people (Schwartz, 2012), and
D

argue that culture represents one of the bases for decision-making. Although the output of a
TE

board’s work is entirely cognitive, the effective functioning of a board largely depends on
P

social and psychological processes (Forbes and Milliken, 1999). One of the factors to impact
CE

on such group processes is culture. Culture influences how all individuals, even members of

elite decision-making groups, such as corporate boards, perceive and interpret information
AC

and relate to others (Maznevski, 1994; Schneider and De Meyer, 1991). Hence, directors are

not immune to cultural biases, and cultural diversity may introduce “negative externalities”

that might undermine board effectiveness (e.g., Manzoni et al., 2010; Nederveen Pieterse et

al., 2013). In addition, culture is intrinsic to the person, practically unchangeable, and not

necessarily public or obvious (e.g., Hofstede, 2001).

Cultural diversity can affect the decision-making of the board in two ways: through the

degree of task-related and relationship conflict it creates and/or through the impact it has on

11
 ACCEPTED MANUSCRIPT

intragroup trust. The degree of task-related and relationship conflict represents the advantages

and disadvantages of diversity that make up the two sides of the double-edged sword

(Milliken and Martins, 1996). The advantages of cultural diversity to group decision-making

T
can be general or specific (Maznevski, 1994). General advantages are referred to as task-

IP
related or cognitive conflict, and relate to different (world) views, ideas, opinions, and

R
different ways of perception and interpretation of information by the group (Nederveen

SC
Pieterse et al., 2013; Simons and Peterson, 2000). In this context, diversity reduces, for

example, groupthink. Task-related conflict is found to be particularly beneficial to group

NU
performance when tasks are complex. Specific advantages of cultural diversity exist when a
MA
person becomes a valuable information source about a foreign culture due to their

communication networks, linguistic resources, and familiarity with a country’s customs

(Maznevski, 1994).
D
P TE

The disadvantages of diversity are associated with relationship or affective conflict (an
CE

awareness of interpersonal incompatibilities). Many studies have documented the detrimental

effects of relationship conflict on group commitment and group decision quality, as it can
AC

lead to a reduced effort to resolve the group’s cognitive tasks due to increased levels of stress

and conflict within the group (e.g., De Wit et al., 2012; Jehn and Mannix, 2001). In culturally

diverse groups, communication is slower, more difficult, more confused, and more frequently

a source of misunderstanding. Differences in style and attribution of meanings curtail conflict

resolution (Anderson et al., 2011; Doney et al., 1998). As Kirchmeyer and Cohen (1992)

point out, conflict in multicultural teams may be difficult to identify and even more difficult

to resolve. Furthermore, Simons and Peterson (2000) show that an increase in task-related

conflict goes hand in hand with an increase in relationship conflict, making it difficult to

12
 ACCEPTED MANUSCRIPT

increase the benefits of diversity without increasing the disadvantages (for instance, people

may interpret a discordant view as a personal attack or promotion of a hidden agenda).

T
IP
Ferreira (2010) recognizes the potential benefits and costs of diversity. In addition to the

benefits of increased creativity, different perspectives, and greater access to resources and

R
SC
connections, diversity benefits the company also through demonstrating equal opportunities

to lower-level workers and conforming to societal expectation. The potential costs of

NU
diversity according to Ferreira (2010) include conflict, lack of cooperation, insufficient

communication, conflicts of interest, and people pushing their own agendas.

MA
D

The second way in which cultural diversity can affect the functioning of the board is through
TE

its negative effect on the level of intragroup trust.9 Bjørnskov (2008) introduces a theoretical

framework in which people have a given trust radius; they trust people who are inside the
P

trust radius and distrust people who are outside this radius. In more diverse groups, many
CE

people fall outside the trust radius and, therefore, there is less trust in a more heterogeneous
AC

group. Differences in how people perceive the world has a detrimental effect on trust building

within that group. Thus, in a culturally diverse board, intragroup trust may be lower, which

can negatively affect the performance of the board. In addition, Simons and Peterson (2000)

argue that intragroup trust has an intensifying effect on the relation between task-related and

relationship conflict, which can further emphasize the negative aspects of cultural diversity.

Based on the arguments presented above, cultural diversity can have both positive and

negative effects on firm performance. The net effect of cultural diversity on firm

9
We thank one of the referees for pointing out this argument.

13
 ACCEPTED MANUSCRIPT

performance, therefore, depends on whether the positive aspects outweigh the negative

aspects. This is the empirical question we address in this paper.

T
IP
The potential negative effects of cultural diversity may be mitigated when there are particular

benefits of having culturally diverse boards. For example, in more complex firms (firms that

R
SC
operate in many business segments) or firms with higher percentages of foreign sales or

foreign operations, directors from different national cultural backgrounds are more likely to

NU
represent an important information source (see also Anderson et al., 2011, and Masulis et al.,

2012). We thus expect that firms with complex operations or a strong presence in foreign
MA
markets would benefit more from cultural diversity.
D
TE

2.2. Measuring cultural diversity

According to Hofstede (2001), “culture is the collective programming of the mind

P
CE

distinguishing the members of one group or category of people from others.” There are

systematic differences in people’s beliefs and values, which generate far-reaching economic
AC

and social implications (e.g., Guiso et al., 2006; Tabellini, 2008a, b). Although culture itself

is not observed, several initiatives have been undertaken to quantify different aspects of

culture. The most well-known initiative is by Hofstede (2001), who constructs various

cultural dimensions that characterize different cultural traits of a nation.10 Initially, Hofstede

introduced four dimensions: individualism-collectivism, masculinity-femininity, power

distance, and uncertainty avoidance, but later added a fifth dimension (long-term orientation)

10
Hofstede analyzed value scores obtained from a large respondent group of IBM managers from 70 countries
between 1967 and 1973. The first version of the model covered value scores from 40 countries, later from 50
countries, and currently from 76 countries. The importance of Hofstede’s cultural dimensions is highlighted by
Kirkman et al. (2006), who document 180 empirical studies that rely on Hofstede’s cultural dimensions and
were published in leading journals between 1980 and 2002.

14
 ACCEPTED MANUSCRIPT

and a sixth dimension (indulgence). We focus on the initial four dimensions of Hofstede as

these have been used most frequently in prior studies and are based on the original surveys

conducted by Hofstede (Kirkman et al., 2006).11, 12 Each country is given a score on each of

T
the dimensions. The individualism score indicates how much value members of a society

IP
place on taking care of themselves and their close families. The masculinity score measures

R
the importance societies’ members place on achievement, assertiveness, and material reward

SC
for success. The power distance score captures the acceptance of societies’ members of an

unequal distribution of power amongst people. Finally, the uncertainty avoidance score

NU
represents the degree to which people from that country feel uncomfortable with uncertainty
MA
and ambiguity.
D

To construct our measure of cultural diversity of the board, we first compute the cultural
TE

distance between directors. We do this by taking the cultural scores on the above mentioned
P

individual cultural dimensions of the country of nationality for each director,13 and compute
CE

cultural distance following Kogut and Singh (1988):

AC

11
In finance, several studies employ the Hofstede culture framework, including Aggarwal and Goodell (2009),
Aggarwal et al. (2012), Beugelsdijk and Frijns (2010), Bryan et al. (2015), Burns et al. (in press), Dodd et al.
(2015), Frijns et al. (2013), Karolyi (2016), Lievenbrück and Schmid (2014), and Zheng et al. (2012).
12
The fifth and sixth dimension were computed ex post based on different surveys and among a different set of
participants. We confirm that our results are similar to those reported in the paper when we use either five
(including long-term orientation) or six (including long-term orientation and indulgence) dimensions.
13
An alternative would be to use country of residence instead of nationality. However, nationality may reflect a
person’s cultural values and norms more accurately than country of residence. According to Hofstede (2001),
culture represents values and norms that are transmitted from parents to children, who would carry those values
with them through life. When a person relocates to another country, this person does not immediately adopt the
cultural values of the new country of residence. Acculturation is a gradual, complex, and lengthy process (Berry,
2005). Even if a person has lived in a foreign country for several years, it is questionable whether that person
has identical norms and values to a native from that country, or whether some cultural norms and values from
the home country still remain. Use of nationality makes a clear distinction in this case. If a person has lived in a
country for a substantial period of time and is a naturalized citizen of that new country, it is most likely that that
person will report her/his nationality as a citizen of that country. No matter how we capture culture, however, it
will be measured with error. Noisy proxies lead to an error-in-variables problem, which generally biases
regression coefficients towards insignificance, and suggests that the true relation between board diversity and
firm performance may actually be stronger than what we document.

15
 ACCEPTED MANUSCRIPT

4
CDij  {( I
k 1
ki  I kj ) 2 / Vk }  i  j, (1)

T
where CDij is the cultural distance between each two directors (i, j), Iki is the culture score on

IP
dimension k for a director i, Ikj is the cultural score on dimension k for a director j, and Vk is

R
the in-sample variance of the score for the specific cultural dimension. This measure of

SC
cultural distance has been applied in various finance studies, for example, Beugelsdijk and

NU
Frijns (2010), Dodd et al. (2015), and Karolyi (2016).MA
Based on the cultural distance scores, we compute the firm-level cultural diversity of the

board, similarly to Hutzschenreuter and Voll (2008).14 We define cultural diversity as the
D
TE

average of cultural distances in all pairs of board members:

P
CE

 CD ij ,nt

CD BOARDnt  i, j
 i j, (2)
m(m  1) / 2
AC

where CD BOARDnt is the measure of cultural diversity of the board of firm n in year t, and m

is the number of board members. The measure of cultural diversity is scaled by the number of

pairs of board members, so that the measure is normalized for the size of the board.

14
Hutzschenreuter and Voll (2008) examine the impact of added cultural distance in the expansion path of
multinational enterprises (MNEs) on firm performance. They employ the cultural diversity variable to control
for the effect of the average degree of cultural differences within the MNE. Our study differs from theirs in that
we measure the cultural diversity of the board of directors and assess its impact on firm performance.

16
 ACCEPTED MANUSCRIPT

3. Data

For our empirical analysis, we focus on a sample of large British firms, as the UK market

provides reasonable variation in terms of cultural diversity in corporate boards and foreign

T
directors mostly sit on the boards of large companies. In our sample, which constitutes 95%

IP
of the UK stock market capitalization, foreign directors are present in 62.55% of the firm-

R
year observations. This compares with Masulis et al. (2012), who in their US sample of S&P

SC
1,500 firms (representing approximately 90% of US stock market capitalization) have at least

one foreign independent director in 13% of their firm-year observations.15 We thus expect to

NU
see more variation in terms of cultural diversity in a UK sample.16
MA

We obtain our sample firms from Datastream, excluding firms from the finance industry.
D

After checking for data availability on directors and firm characteristics, we have a sample of
TE

243 firms. We collect director- and firm-level data annually for the period from 2002 to 2014.
P

Our final sample has 2,852 firm-year observations.17

CE
AC

3.1. Director-level data

We obtain information on board membership from the Orbis and Osiris databases maintained

by Bureau van Dijk. These databases contain name, gender, age, and nationality for many

directors. Where data on directors are missing, we search through annual reports and Internet

15
In our sample, 51.65% of our firm-year observations have at least one foreign independent director.
16
An international sample could potentially provide more variation in terms of cultural diversity on boards.
However, it would also confound the analysis due to large cross-country variations (e.g., due to institutional
differences).
17
Out of the 243 firms in our sample, 26 firms (10.7%) belong to the Basic Materials industry, 27 firms (11.1%)
Consumer Goods, 66 firms (27.2%) Consumer Services, 10 firms (4.1%) Health Care, 71 firms (29.2%)
Industrials, 17 firms (7.0%) Oil & Gas, 12 firms (4.9%) Technology, 7 firms (2.9%) Telecommunications, and 7
firms (2.9%) Utilities. Given that utilities companies operate in a regulated environment and potentially have
different board structures (see, e.g., Hoi and Wollan, 2010), as a robustness test, we estimate our models for a
subsample that excludes Utilities. We find that excluding utilities companies does not affect the results.

17
 ACCEPTED MANUSCRIPT

sources such as Bloomberg Businessweek or LinkedIn. Our final sample comprises 25,511

director-year observations, with directors from 52 different countries.

T
IP
Table 1 provides a breakdown of the sample of directors from each country by year, and in

the last column, reports the total percentage of directors coming from a specific country.

R
SC
Considering the total percentages, about 78.5% of directors come from the UK. The second

largest group of directors is from the US, representing about 7% of our sample. Further, we

NU
note that there are relatively high percentages for other Anglo-Saxon countries (Australia,

Canada, and Ireland) and the developed countries with close economic ties to the UK
MA
(France, Germany, and the Netherlands). Over time, there appears to be an increase in the

percentage of foreign directors, from a low of 18.75% in 2002 to a high of 26.05% in 2014.
D
TE

INSERT TABLE 1 HERE

P
CE

In Table 2, we compare the characteristics of UK directors with foreign directors, and report
AC

the percentage of males, the average age of directors, and the percentage of independent

directors. On average, 91% of UK directors are male versus 86% of foreign directors. This

difference of about 5% is significant at the 1% level, indicating that there are significantly

more female foreign directors than female UK directors. Foreign directors are significantly

older than UK directors; on average, UK directors are 55 years old, whereas foreign directors

are 56.8 years old. Finally, we note that 59.19% of foreign directors versus 53.92% of UK

directors are independent. The difference is, again, highly statistically significant.

18
 ACCEPTED MANUSCRIPT

INSERT TABLE 2 HERE

3.2. Firm-level data

T
IP
3.2.1. Board characteristics

R
Based on the director-level data, we can determine various firm-level board characteristics. In

SC
Table 3, we present the distribution of the number of foreign directors on the board over time.

In total, there is at least one foreigner on the board in 62.55% of the firm-year observations.

NU
This percentage increases over time, from 56.7% in 2002 to 71.8% in 2014. Panel A presents
MA
the distribution of the number of foreigners on boards, and shows that for 1,068 firm-year

observations there are no foreigners on the board. The greatest number of foreigners on a
D

board is 12. In Panel B, we present the distribution of the number of different nationalities
TE

present on the board. The maximum number of nationalities represented on a board is 7. In

66% of the firm-year observations, there is either none or one foreign national on the board.
P
CE

INSERT TABLE 3 HERE

AC

In Table 4, we report summary statistics on firm-level board diversity measures. For our

cultural board diversity measure CD BOARD, the mean is 0.5261, with a median of 0.2307

and a maximum of 3.3603. If all directors have the same nationality, then the cultural

diversity of such boards equals zero. Overall, we observe a steady increase in cultural

diversity from 0.4386 in 2002 to 0.6455 in 2014 (reported in Panel B of Table 4).18

18
Our measure of cultural diversity is based on cultural differences within pairs of directors on a board. To give
some indication of what drives the cultural diversity scores, we provide examples of cultural distances relative
to the UK. The cultural distance relative to a UK board member is small for foreigners from, for example, the

19
 ACCEPTED MANUSCRIPT

INSERT TABLE 4 HERE

T
IP
Although our focus is on cultural diversity, we also consider other important board

R
characteristics such as board size, gender diversity, board independence, director age, and

SC
CEO duality. We provide a full description of each variable in Appendix A. In Panel A of

Table 4, we report summary statistics for the board characteristics, and in Panel B,

NU
distributions of their values over time. We observe that the average number of directors on a

board is 8.89, with a variation from 2 to 20 directors, and this number is relatively stable.
MA
Firms with foreign directors tend to have larger boards than firms without. The average

number of foreign directors on a board is 1.90, with a maximum of 12. This number increases
D

steadily from 1.60 in 2002 to 2.24 and 2.22 in 2013 and 2014, respectively. Males, on
TE

average, constitute 90.2% of directors. We observe a downward trend in male representation

P

on boards with the percentage decreasing from 95.28% in 2002 to 81.35% in 2014. Firms
CE

with foreign directors tend to have fewer males on the board than firms without. The average
AC

percentage of independent directors is 54.39%, with an upward trend from 44.78% in 2002 to

62.94% in 2014. There is a significant positive difference in the percentage of independent

directors for firms with and without foreign directors. The average age of directors is 55.41

years; it increases from 53.38 in 2002 to 57.12 in 2014. In line with Table 2, firms with

foreign directors have, on average, older directors than firms without. In contrast to the

increase in average age over time, the age range (the age difference between the oldest and

youngest director) decreases from 25.5 years in 2002 to 22.21 years in 2014. The age range

US, Australia, and Ireland, moderate for foreigners from Germany, Italy, and Switzerland, and large for
foreigners from Malaysia, Chile, and Russia. We stress, however, that our measure is computed as an average
for all pairs of board members and it is not UK-centered. Cultural diversity is a function of both the distance
between any two directors and the number of pairs with different nationalities.

20
 ACCEPTED MANUSCRIPT

tends to be wider in firms with foreign directors. Finally, 10.4% of firms have a CEO who is

also the chairman of the board; this number steadily decreases from 10.61% in 2002 to 6.41%

in 2014. Firms with foreign directors tend to have a higher occurrence of CEOs serving also

T
as the chairman of the board.

R IP
SC
3.2.2. Firm performance measures

In our empirical analysis, we examine the impact of cultural diversity of boards on firm

NU
performance. We employ two commonly used measures of firm performance: Tobin’s Q and

ROA. Tobin’s Q is calculated as the book value of total assets minus the book value of equity
MA
plus the market value of equity, all divided by the book value of total assets. ROA is

calculated as operating income divided by the year-end book value of total assets. We
D

winsorize Tobin’s Q and ROA at 1% on each side of the distribution. Panel A of Table 5
TE

reports summary statistics of our firm performance measures. The mean Tobin’s Q is 2.01,
P

with a median of 1.61 and a range from 0.65 to 8.29. The average ROA is 10.94%, with a
CE

median of 9.49 and range from -14.02 to 39.55. The difference in means of Tobin’s Q
AC

between firms with and without foreign directors is not statistically significant, whereas firms

with foreign directors have significantly lower ROA than firms without foreign directors.

INSERT TABLE 5 HERE

3.2.3. Firm characteristics

We account for a number of firm characteristics that may affect firm performance and board

composition such as firm size, leverage, firm age, firm complexity, return volatility, and sales

21
 ACCEPTED MANUSCRIPT

growth. Panel B of Table 5 reports summary statistics for firm-level variables for the full

sample and mean values for subsamples of firms with and without foreign directors. The

average size of firms is 5,117.6 million GBP, with a significant difference between firms with

T
and without foreign directors: 7,473.2 versus 1,221.2 million GBP, respectively. Leverage, on

IP
average, is 0.23, and firms with foreign directors have lower leverage than firms without

R
(0.226 vs. 0.235, respectively). The average firm age is 70.3 years and the average number of

SC
business segments (our measure of firm complexity) is 2.46. Firms with foreign directors are

significantly older than firms without (74.76 vs. 62.82 years) and operate in a significantly

NU
greater number of business segments (2.60 vs. 2.26 segments). The average daily return
MA
volatility is 2.15%, and the average sales growth is around 10%, with no significant

differences between firms with or without foreign directors.

D
TE

4. Results
P

4.1. Cultural diversity of boards and firm performance: main results

CE

We start by estimating several regressions of firm performance (Tobin’s Q and ROA) on

AC

cultural diversity of the board (CD BOARD), controlling for a range of board and firm

characteristics. We control for board size as it reflects the firm’s advising and monitoring

needs. Other board characteristics we control for are gender diversity, board independence,

average age, age differences between board members, and CEO duality (e.g., Adams and

Ferreira, 2009; Anderson et al., 2011). Regarding firm characteristics, we control for firm

size and age, leverage, firm complexity, return volatility, and sales growth. Finally, we

include industry and year fixed effects, and compute robust standard errors clustered at the

firm level (see Petersen, 2009).

22
 ACCEPTED MANUSCRIPT

INSERT TABLE 6 HERE

In the first column of Table 6, we report the results for Tobin’s Q as the dependent variable,

T
IP
and CD BOARD and firm characteristics as explanatory variables. We observe that the

coefficient on cultural diversity is negative and significant at the 1% level. The negative

R
impact of cultural diversity on Tobin’s Q suggests that the costs and frictions inherent in

SC
cultural diversity outweigh the potential benefits of having culturally diverse boards. The

NU
impact of CD BOARD is economically significant: a firm at the 75th percentile of the CD

BOARD distribution achieves a Tobin’s Q that is 0.23 lower than a firm at the 25th percentile.
MA
D

Firm size is significant and positive, suggesting that larger firms achieve higher valuations
TE

(this is a common observation in studies on performance of UK firms; see, e.g., Guest, 2009;

Short and Keasey, 1999). Leverage, firm age, firm complexity, and return volatility are all
P

negative and significant, suggesting that more leveraged, older, more complex, and riskier
CE

firms achieve lower valuations. The relation between Tobin’s Q and sales growth is positive
AC

and significant.

Given that cultural diversity on the board may reflect other board characteristics, it is

important to control for these. The inclusion of board characteristics (Column 2 of Table 6)

does not alter the significant negative relation between cultural diversity of the board and

Tobin’s Q. Board size is significant and negative, suggesting that firms with larger boards

achieve lower valuations. The average age of directors is a negative determinant of Tobin’s

Q, whereas the directors’ age range is a positive determinant, although their significance is

marginal.

23
 ACCEPTED MANUSCRIPT

Columns 3 and 4 of Table 6 report the results for ROA as the dependent variable. The results

for ROA are similar to those for Tobin’s Q. We observe a negative and significant relation

T
between cultural diversity of the board and ROA, irrespective of controlling for other board

IP
characteristics. This result is economically significant: firms at the 75th percentile of the CD

R
BOARD distribution achieve a 1.43% lower ROA than firms at the 25th percentile. Firm size

SC
remains significantly positive, as does sales growth. Firm complexity and return volatility

NU
have a negative effect on ROA. Board size is a negative and significant determinant of ROA.
MA
4.2. Endogeneity of cultural diversity in boards
D

We address the issue of a potentially endogenous relation between firm performance and
TE

cultural diversity in two ways. First, we estimate specifications using firm fixed effects,

which control for any time-invariant firm-specific factors related to both firm performance
P

and cultural diversity, and address the concerns of a potentially omitted variables bias. 19
CE

Second, we employ an instrumental variables approach.

AC

In Table 7, we report the results of the regressions that include firm fixed effects. Columns 1

and 2 report results for Tobin’s Q. In the first regression, which excludes board

characteristics, the coefficient on the cultural diversity variable is negative and significant at

19
We also perform a test for omitted variables, following Oster (2015). Based on assumptions about the
importance of potentially omitted variables relative to observed variables and potential improvement in R 2, one
can assess the impact of potentially omitted variables on the coefficient of interest. As suggested by Oster
(2015), we assume that: (a) omitted variables and the included variables are of equal importance; (b) the
inclusion of omitted variables can lead to a maximum R2 of 1.3 times the estimated R2 in the full specification in
Table 6. Under these settings, we observe that potentially omitted variables could change the coefficient on
cultural diversity by 0.056 and 0.155 for Tobin’s Q and ROA, respectively. These changes are well within the
confidence intervals of the estimated coefficients, suggesting that potentially omitted variables do not
significantly affect the results.

24
 ACCEPTED MANUSCRIPT

the 5% level. After controlling additionally for board characteristics, cultural diversity

continues to be significant, but only at the 10% level.

T
IP
INSERT TABLE 7 HERE

R
SC
Columns 3 and 4 of Table 7 report the results for ROA. For the regression that excludes

board characteristics, we find a negative relation between ROA and CD BOARD, significant

NU
at the 5% level. However, after the inclusion of board characteristics, the significance of the
MA
coefficient on CD BOARD drops just below the 10% level (p = 0.101). Despite some loss in

statistical significance, the fixed effects estimates suggest that our results are not driven by an
D

omitted variables bias.20

P TE

Another potential source of endogeneity is simultaneity, that is, the presence of foreign
CE

directors on boards may be determined by the firm’s need for foreign directors, which, in

turn, is potentially related to firm performance. A solution to this problem is to use an

AC

instrumental variables approach, which employs instrumental variables that correlate highly

with the variable of interest, but have no effect on the dependent variable after controlling for

all other effects. The use of instrumental variables also addresses potential errors-in-variables

issues.

20
Such a reduction in significance is often observed in studies that focus on board characteristics (e.g., Masulis
et al., 2012) and may be a consequence of the relatively low power of firm fixed effects to detect the effects of
variables that differ mostly in the cross-section rather than over time (Zhou, 2001).

25
 ACCEPTED MANUSCRIPT

The instrument we employ is a dummy variable for whether a firm is headquartered outside

of a large metropolitan area (i.e., in a town with a population less than 250,000 people). One

of the motivations for selecting this instrument is that we expect firms headquartered in these

T
areas to be less culturally diverse than firms headquartered in large metropolitan areas. This

IP
argument is in line with Anderson et al. (2011), who use a measure of county heterogeneity

R
as their instrument of board heterogeneity. In addition, one can argue that for foreign

SC
directors, firms headquartered outside of large metropolitan areas are more difficult to reach;

these firms may therefore employ fewer foreign directors on the board. This argument is in

NU
line with the reasoning of Masulis et al. (2012) concerning their instrument – a dummy
MA
variable for whether a firm is headquartered within 100 km from a large US airport.
D

In Table 8, we present the results of the instrumental variables regressions estimated by two-
TE

stage least squares. In the first stage regression of CD BOARD (Column 1), the instrumental
P

variable has the expected negative sign and is highly significant. Firms located outside of
CE

large metropolitan areas have significantly less culturally diverse boards. Board size, gender,

board independence, and average directors’ age are all significant and positive. In addition,
AC

CEO duality is negatively related to CD BOARD. Firm size, leverage, and return volatility are

significantly positive; that is, larger, more leveraged, and riskier firms tend to have culturally

more diverse boards. Further, we find that firm age and firm complexity are negative and

significant, suggesting that older and more complex firms have less culturally diverse boards.

To assess the strength of our instrumental variable, we perform a standard test for weak

instruments by computing the F-statistic of the instrument. This F-statistic is 82.15, and well

exceeds the threshold value of 10 suggested by Staigler and Stock (1997). This statistic also

26
 ACCEPTED MANUSCRIPT

well exceeds the Stock and Yogo (2005) critical value of a 5% Wald test if the desired

maximum size of the test is 10%. In addition, the partial R2 of the instrumental variable is

2.64%, that is, it adds about 10% to the R2 in the first stage regression. Taken together, these

T
tests suggest that our instrument is not weak.

R IP
SC
INSERT TABLE 8 HERE

NU
In Columns 2 and 3, we report the results of the second stage regressions for Tobin’s Q and
MA
ROA, respectively. CD BOARD is negative and significant in both regressions. This suggests

that the relation found with the OLS regression can be interpreted in a causal way, that is, the
D

more culturally diverse a board is, the lower the performance of the firm will be. We note that
TE

the estimated coefficients increase in magnitude relative to the OLS regressions. This can be

attributed to the instrumental variables approach reducing the errors-in-variables bias, and
P

reconfirms the strong negative relation between cultural diversity and firm performance.
CE
AC

4.3. Further analysis

4.3.1. Other measures of “foreignness” of the board

Our results thus far suggest a strong negative relation between cultural diversity and firm

performance. However, cultural diversity may just be a proxy for the degree of “foreignness”

of the board. Indeed, Masulis et al. (2012) demonstrate that firm performance is negatively

affected when firms have foreign independent directors on the board. They explain this

negative relation by the lesser effectiveness of foreign directors, whose physical distance

from the firm limits their ability to attend board meetings and effectively monitor

27
 ACCEPTED MANUSCRIPT

management. It may be the case that our measure of cultural diversity merely captures the

presence of foreigners on boards. To address this issue, we include a number of alternative

variables that measure the degree of the “foreignness” of the board.

T
IP
INSERT TABLE 9 HERE

R
SC
In Panel A of Table 9, we report the results for different measures of “foreignness” of the

NU
board. A straightforward test would be to include a dummy variable that equals one if a firm
MA
has one or more foreign directors on the board, and observe how this impacts our results.

However, this dummy variable is perfectly correlated with the degree of cultural diversity of
D

the board. Hence, we first run a regression where we exclude firms without foreign
TE

directors.21 In Column 1.1, we show that after the exclusion of firms without foreign directors

the coefficient on cultural diversity remains negative and highly significant.

P
CE

We include several measures for the “foreignness” of the board: the proportion of foreign
AC

directors on the board (Foreign directors ratio), the ratio of different nationalities present on

the board (Nationalities ratio), a dummy variable that equals one if the firm has at least one

foreign independent director on the board (FID dummy), 22 and the ratio of foreign

independent directors to the total number of directors on the board (FID ratio). We note that

Foreign directors ratio and FID dummy are significantly related to firm performance on their

21
If our results are driven by whether a company has foreigners on the board or not, then a regression that
excludes firms without foreigners should no longer yield a significant coefficient on CD BOARD.
22
This is the main variable of interest in Masulis et al. (2012). Note that our definition of foreign directors is
based on directors’ nationality, whereas Masulis et al. (2012) define foreign directors based on their physical
location, irrespective of their nationality.

28
 ACCEPTED MANUSCRIPT

own. However, once we add cultural diversity of the board to these regressions, the variables

become insignificant, while the significance of CD BOARD remains.

T
IP
Columns 2.1–2.9 in Panel A of Table 9 report the results for ROA. We observe that the

exclusion of firms without foreign directors does not affect the negative relation between

R
SC
cultural diversity and firm performance. None of the alternative measures of foreignness of

the board are significant determinants of ROA. CD BOARD remains negative and statistically

NU
significant in all regressions.
MA
These results indicate that our cultural diversity measure is not merely a proxy for the
D

“foreignness” of boards, but it explains, in fact, why “foreignness” of the board matters for
TE

firm performance. These results extend the findings of Masulis et al. (2012) and suggest that

cultural diversity plays an important role that extends beyond the limited effectiveness of
P

foreign independent directors due to physical distance from the firm.

CE
AC

4.3.2. Other measures of “foreignness” of the firm

Cultural diversity could also be related to the degree of the firm’s presence in foreign product

and financial markets. Firms with higher degrees of foreign orientation are likely to have a

greater share of foreign directors and, accordingly, greater cultural diversity within the board

(Ferreira, 2010; Masulis et al., 2012; Oxelheim and Randøy, 2003). The extent of foreign

operations, in turn, may be related to firm performance. Hence, the observed impact of

cultural diversity on firm performance could be a consequence of a firm’s presence in foreign

markets, and our CD BOARD measure could be a proxy for this.

29
 ACCEPTED MANUSCRIPT

We consider a firm’s foreign sales and a firm’s foreign assets to capture the degree of a

firm’s presence in foreign product markets and the degree of foreign operations. In our

T
sample, firms have, on average, 42.5% of their total sales coming from foreign markets. This

IP
number is 55.4% (23.5%) for firms with (without) foreign directors. Similarly, foreign assets

R
constitute, on average, 25.8% of total assets for the full sample, but 33.9% (13.8%) for the

SC
subsample of firms with (without) foreign directors. We also control for a firm’s presence in

NU
foreign financial markets. We consider whether a firm is listed on the New York Stock

Exchange (NYSE). Doidge et al. (2004) report that non-US firms listed on the NYSE have
MA
significantly higher valuations than firms from the same country that do not list on the NYSE.

Also, a listing on the NYSE calls for foreign expertise on the board and could therefore result
D

in a more culturally diverse board. Around 8.5% of our sample firms have an NYSE listing;
TE

this number is 13.5% (1.2%) for firms with (without) foreign directors. We also account for a

firm having its shares listed on any foreign stock exchange outside of the home market (UK).
P
CE

Firms with foreign directors are more likely to have a foreign listing than firms without

foreign directors (25.1% vs. 4.3%, with a sample average of 16.7%). We construct dummy
AC

variables to control for NYSE listing and listing on a stock exchange outside the UK.

In Panel B of Table 9, we report the results where we control for different measures of

“foreignness” of the firm. The ratio of foreign sales is not related to Tobin’s Q (Columns 1.2

and 1.3) or ROA (Columns 2.2 and 2.3) and its inclusion does not affect the negative effect of

CD BOARD. Foreign assets are a negative and significant determinant of Tobin’s Q

(Columns 1.4 and 1.5) and ROA (Columns 2.4 and 2.5), suggesting that firms with higher

proportions of assets located overseas have lower valuations. Nevertheless, CD BOARD

30
 ACCEPTED MANUSCRIPT

remains negative and significant even after controlling for the extent of a firm’s foreign

operations. Further, neither the NYSE listing variable (Columns 1.6 and 1.7), nor the foreign

listing variable (Columns 1.8 and 1.9) affects Tobin’s Q. Firms listed on the NYSE have a

T
significantly higher ROA. CD BOARD remains unaffected by the inclusion of these two

IP
variables.

R
SC
These results demonstrate that CD BOARD is not a proxy for other measures of a firm’s

NU
“foreignness,” and that the negative impact of cultural diversity on firm performance goes

beyond what can be attributed to the firm’s presence in foreign markets.

MA
D

4.3.3. Cultural diversity and firm operations

TE

The observed negative effect of cultural diversity on firm performance may be mitigated
P

when firms have more complex operations or significant foreign operations. This is because
CE

such firms have a greater need for broader expertise and foreign market specific knowledge

on their boards. Anderson et al. (2011) show that for firms with more complex operations,
AC

board heterogeneity has a positive impact on firm performance, whereas for firms with less

complex operations, board heterogeneity has a negative impact. In addition, Masulis et al.

(2012) show that the negative impact of foreign independent directors (FIDs) on performance

is mitigated if the firm has a stronger presence in the FID’s country (measured by foreign

sales). In this section, we extend the analysis of Anderson et al. (2011) and Masulis et al.

(2012), and examine how the degree of complexity of a firm’s operations and the degree of a

firm’s foreign operations affect the relation between cultural diversity and firm performance.

31
 ACCEPTED MANUSCRIPT

INSERT TABLE 10 HERE

To assess the moderating role of firm complexity, we split the sample into two subsamples.

T
IP
We classify a firm as having complex operations if it operates in more than three business

segments (the 75th percentile value of this variable), and as having less complex operations if

R
SC
it operates in three or less business segments. Panel A of Table 10 reports the results. For

Tobin’s Q and ROA, cultural diversity is negative and significant in the subsample of less

NU
complex firms and insignificant in the subsample of more complex firms. This finding is in

line with Anderson et al. (2011) and suggests that the negative effect of cultural diversity is
MA
mitigated in complex firms that require a wider range of expertise.
D
TE

Next, we consider the extent of a firm’s export orientation. We split the sample based on a

firm’s foreign sales. We classify a firm as having a strong export orientation if more than
P

75.59% of its sales come from foreign markets (the 75th percentile value of this variable).
CE

Panel B of Table 10 reports the results. For both Tobin’s Q and ROA, cultural diversity is
AC

insignificant for firms with a strong export orientation. This result supports the findings of

Masulis et al. (2012) and suggests that firms that require higher levels of foreign expertise are

not negatively affected by cultural diversity.

Lastly, we consider the extent of a firm’s foreign operations measured by the firm’s foreign

assets, and classify a firm as having substantial foreign operations if more than 50.28% of its

assets are located abroad (the 75th percentile value of this variable). We report the results in

Panel C of Table 10. Cultural diversity is negative and significant for both Tobin’s Q and

ROA in the subsample of firms that have less foreign assets, and marginally significant for

32
 ACCEPTED MANUSCRIPT

Tobin’s Q and insignificant for ROA in the subsample of firms with substantial foreign

operations. Similar to the results for export orientation, the negative effect of cultural

diversity is mitigated in firms that require foreign market expertise.

T
IP
4.3.4. Dependent versus independent directors

R
SC
An important question to address is whether the effect of cultural diversity on firm

performance is equal when this diversity is measured among the dependent or independent

NU
directors. There are several reasons why we might expect to see differences in the impact of

cultural diversity on firm performance when this is measured among the two types of
MA
directors. First, there is a long-standing literature on organizational learning and culture (e.g.,

Ouchi, 1979), which promotes the development of a set of organizational values and beliefs
D

as a mechanism to manage subordinates to achieve a company’s goals. Dependent directors

TE

may be indoctrinated with this organizational culture, which may dilute the effect of national
P

culture and their cultural differences. 23 Independent directors are not exposed to this
CE

organizational culture, thus, we may expect that the negative effect of cultural diversity on
AC

performance is more pronounced in the subset of independent directors.

Second, we may expect that foreign dependent directors overcome frictions introduced by

cultural differences as they are in regular contact and meet on a more frequent basis than

independent directors. This regular contact can diminish the negative effects of cultural

differences as well as increase the level of intragroup trust, and therefore mitigate the

observed negative impact of cultural diversity on firm performance in the group of dependent

directors.

23
We thank the referee for pointing out this argument.

33
 ACCEPTED MANUSCRIPT

A third reason has to do with the different roles of directors. A key task of independent

directors is to monitor and ensure that decisions are made in the best interests of shareholders,

T
while dependent directors have mainly an advisory role. Since the advisory role relies more

IP
on the specific expertise of a director, the positive aspects of diversity may play a more

R
important role than the negative aspects.

SC
NU
INSERT TABLE 11 HERE
MA

In Table 11, we report the results for the models where we calculate the cultural diversity for
D

the dependent (CD BOARD DEP) and independent (CD BOARD INDEP) directors for each
TE

firm. For both Tobin’s Q and ROA, only the coefficients on CD BOARD INDEP are
P

significantly negative. These results suggest that organizational culture dilutes the negative
CE

effects of national cultural diversity, or that dependent directors learn to overcome the

frictions imposed by cultural diversity through more frequent meetings. Alternatively,

AC

cultural diversity may impact the different functions of the board (advising versus

monitoring) differently. Given that independent directors have a stronger focus on

monitoring, the findings suggest that cultural diversity is more detrimental to that role.

4.4. Alternative culture frameworks and other measures of culture

In this section, we conduct additional robustness tests by, first, employing several alternative

culture frameworks and, second, by examining other measures of culture.

34
 ACCEPTED MANUSCRIPT

We employ three alternative culture frameworks. First, we use the culture scores of Tang and

Koveos (2008), who update Hofstede’s culture scores by controlling for changes in economic

conditions such as GDP per capita. Second, we employ culture scores from the GLOBE

T
project (House et al., 2004).24 Although some of the dimensions of Hofstede and the GLOBE

IP
are comparable, there are conceptual and methodological differences between the two culture

R
frameworks (see, e.g., Smith, 2006). Third, we use the culture framework of Schwartz

SC
(2006), which contains seven value orientations that are based on three culture dimensions:

embeddedness vs. autonomy,25 hierarchy vs. egalitarianism, and mastery vs. harmony. For

NU
each alternative culture framework, we recalculate the cultural diversity measure as in Eq.
MA
(2). We then estimate our main model for the full sample and for the subsample of firms that

have at least one foreign director on the board.

D
TE

INSERT TABLE 12 HERE

P
CE

In Table 12, we report the results for Tobin’s Q in Panel A and for ROA in Panel B. In Panel
AC

A, we observe that the coefficients on CD BOARD are negative and significant, with the

exception of the relation between the GLOBE- and Schwarz-based measures of cultural

diversity and Tobin’s Q when using the full sample. For ROA, cultural diversity is negative

and significant for all alternative culture frameworks (Panel B). Overall, the results suggest

that irrespective of the choice of culture framework, cultural diversity has a significant

negative impact on firm performance.

24
The GLOBE project relies on an alternative framework and produces nine different culture dimensions:
performance orientation, assertiveness orientation, future orientation, humane orientation, institutional
collectivism, family collectivism, gender egalitarianism, power distance, and uncertainty avoidance.
25
Note that Schwartz (2006) divides autonomy into two classifications: affective autonomy and intellectual
autonomy.

35
 ACCEPTED MANUSCRIPT

Next, following Guiso et al. (2009), we employ five alternative measures of culture: bilateral

trust, language similarity, religion similarity, genetic diversity, and somatic diversity (i.e.,

T
differences in physical features). We obtain data on bilateral trust from Guiso et al. (2009).

IP
Similar to our cultural diversity measure, we construct a measure of bilateral trust between

R
each two directors on the board:

SC
 BT ij ,nt

NU
TRUSTnt  i, j
 i j, (3)
m(m  1) / 2
MA

where BTij,nt is the bilateral trust between each two directors (i, j).
D
TE

For language and religion similarity, we obtain the main language and main religion of the
P

country of nationality of a director, and construct similarity measures based on Herfindahl

CE

indexes (as in Anderson et al., 2011). For genetic and somatic diversity, we compute board-
AC

level diversity scores similar to the trust score in Eq. (3). We obtain genetic distance data

from Spolaore and Wacziarg (2009) and somatic distance data from Paola Sapienza’s

website.26 Bilateral trust, genetic distance, and somatic distance scores are only available for

a subset of countries in our sample.

INSERT TABLE 13 HERE

26
Enrico Spolaore’s website with the genetic distance data from Spolaore and Wacziarg:
http://sites.tufts.edu/enricospolaore/category/personal-webpage/. Paola Sapienza’s website with somatic distance
data from Guiso et al. (2009): http://www.kellogg.northwestern.edu/faculty/sapienza/htm/research.htm.

36
 ACCEPTED MANUSCRIPT

Table 13 reports the results of the regressions with the alternative measures of culture. In line

with our findings on cultural diversity, higher levels of bilateral trust, as well as greater

language and religion similarity, are positively related to firm performance, whereas greater

T
genetic and somatic diversity are negatively related to firm performance.

R IP
SC
4.5. Individual culture dimensions

In this section, we consider each Hofstede cultural dimension separately in the calculation of

NU
the cultural diversity measure to investigate whether specific cultural traits are more

important than others. As Shenkar (2001) and Karolyi (2016) point out, not all aspects of
MA
cultural differences may matter equally, and some culture dimensions may be more important

in explaining the impact of board diversity on firm performance. Specifically, two of the
D

dimensions, individualism and masculinity, relate to cultural values that affect group
TE

processes. Thus, differences in these dimensions could be harmful for relationship conflict
P

(Elron, 1998).
CE
AC

The individualism dimension considers whether societies focus on the promotion of the self,

or whether the goals and objectives of the group are deemed more important than the

promotion of the individual. This dimension is linked to different treatment of people that

belong to the in-group versus people that belong to the out-group (Triandis et al., 1988).

Gudykunst and Bond (1997) and Kirkman et al. (2006) show that the individualism

dimension is the most salient dimension of cultural heterogeneity in intragroup processes.

Different views on these intragroup processes could lead to relationship conflict.

37
 ACCEPTED MANUSCRIPT

The masculinity dimension considers whether a society focuses more on the “masculine”

aspects of being assertive and performance-driven versus the “feminine” aspects of

cooperation and minding the quality of life. This dimension is also referred to as the

T
cooperation-performance orientation index (Hofstede et al., 2006). According to Hofstede

IP
(2003) and Hofstede et al. (2006), trust is relatively low in performance-oriented societies,

R
whereas trust is relatively high in cooperation-oriented societies. Differences in the level of

SC
trust can affect the intragroup processes, which may lead to relationship conflict. In addition,

Hofstede (2003) points out that in masculine societies, people might lie or bribe someone in

NU
order to win. If directors are aware of such a possibility, intragroup trust may deteriorate.
MA

The power distance dimension relates to the way people in a society view the power relations
D

between superiors and subordinates. In a high power distance society, people tend to behave
TE

submissively in the presence of their superior and would not disagree or provide alternative
P

views. In a low power distance society, people are often encouraged to share their alternative
CE

views. These differences suggest that directors may perceive the group dynamics of the board

differently, which may result in relationship conflict. Finally, the uncertainty avoidance
AC

dimension focuses on how comfortable people are in dealing with uncertainty or uncertain

outcomes. People from high uncertainty avoidance societies prefer structure and rules

(Hofstede, 2001). Although differences in uncertainty avoidance may lead to different views

on corporate strategy (and this disagreement may lead to conflict), we do not expect

differences in uncertainty avoidance to lead to different views on the group dynamics of the

board.

38
 ACCEPTED MANUSCRIPT

Overall, we expect differences in individualism and masculinity to be more disruptive and

more likely to cause relationship conflict than differences in the other two dimensions.

However, power distance may play a role as well, as it may affect directors’ willingness to

T
share their views. We expect differences in uncertainty avoidance to play a lesser role.

R IP
SC
INSERT TABLE 14 HERE

NU
We report the results for the individual measures in Table 14. We find that all measures
MA
produce negative coefficients for both Tobin’s Q (Panel A) and ROA (Panel B). However, we

only observe significant coefficients for cultural diversity based on individualism and
D

masculinity. This suggests that differences in the perceived importance of self-assertion,

TE

group cohesiveness, and group integration (examples of differences that characterize the

individualism dimension), and also differences in the perceived importance of assertive

P

behavior, competition, material success, and interpersonal relationships (examples of

CE

differences pertaining to the masculinity dimension) seem to be the principal source of

AC

relationship conflict on boards. These results are in line with Lim et al. (2016), who find that

these dimensions play a significant role in explaining the asymmetric impact of cultural

distance on premiums paid in cross-border mergers and acquisitions. We find marginal

evidence that diversity in terms of power distance affects firm performance. Diversity in

terms of uncertainty avoidance among directors in corporate boards is insignificant.

5. Conclusion

39
 ACCEPTED MANUSCRIPT

We examine the impact of cultural diversity in corporate boards of directors on firm

performance. We find that cultural diversity in boards negatively affects firm performance,

indicating that the frictions imposed by cultural diversity outweigh the potential benefits. Our

T
results hold after controlling for firm fixed effects, and after implementing an instrumental

IP
variables approach. Our results are further robust to various measures of “foreignness” of the

R
board and “foreignness” of the firm, as well as to the use of alternative culture frameworks

SC
and alternative measures of culture. We find that not all firms are equally affected by cultural

diversity. The negative impact of cultural diversity is mitigated by the complexity of the firm

NU
and the extent of the firm’s presence in foreign markets. The negative relation between
MA
cultural diversity and firm performance is concentrated among the independent directors.

Finally, not all aspects of cultural differences are equally important and it is mainly the

diversity in individualism and masculinity that affects the effectiveness of boards.

D
P TE

Cultural diversity in modern boardrooms is a hot topic. In discussions about the economic
CE

value of cultural diversity, practitioners rarely acknowledge “where diversity goes awry” (see

Manzoni et al., 2010). Our study uncovers the potential for employing culture measures in
AC

corporate finance and highlights the considerable explanatory power of sophisticated cultural

diversity measures. We contribute to the discussion by documenting that cultural diversity in

boards is economically important. Our results suggest that realizing the positive effects of

cultural diversity on corporate boards is not straightforward. The relevance of foreign

directors’ knowledge and experience to the firm’s needs seems to be the key to making

cultural diversity an asset for firms. However, unlocking the potential of cultural diversity

might also require dealing with its disruptive, negative consequences and may require

initiatives that improve communication and promote group integration on boards (see, e.g.,

Nederveen Pieterse et al., 2013, or for a practitioner’s perspective, see Manzoni et al., 2010).

40
 ACCEPTED MANUSCRIPT

For practical purposes, our results provide a warning about romanticizing cultural diversity

and highlight the need for selecting board members on the basis of their expertise and

experience with foreign operations.

T
R IP
SC
NU
MA
D
P TE
CE
AC

41
 ACCEPTED MANUSCRIPT

References

Adams, R., Ferreira D., 2009. Women in the boardroom and their impact on governance and
performance. J. Financ. Econ. 94, 291–309.

T
Aggarwal, R., Goodell J.W., 2009. Markets and institutions in financial intermediation:

IP
national characteristics as determinants. J. Bank. Financ. 33, 1770–1780.

Aggarwal, R., Kearney, C., Lucey, B., 2012. Gravity and culture in foreign portfolio

R
investment. J. Bank. Financ. 36, 525–538.

SC
Ahern, K., Daminelli, D., Fracassi, C., 2015. Lost in translation? The effect of cultural values
on mergers around the world. J. Financ. Econ. 117, 165–189.

NU
Anderson, R.C., Reeb, D.M., Upadhyay, A., Zhao, W., 2011. The economics of director
heterogeneity. Financ. Manag. 40, 5–38.
MA
Berry, J.W., 2005. Acculturation: living successfully in two cultures. Int. J. Intercult. Rel. 29,
697–712.
D

Beugelsdijk, S., Frijns, B., 2010. A cultural explanation of the foreign bias in international
TE

asset allocation. J. Bank. Financ. 34, 2121–2131.

Bjørnskov, C., 2008. Social trust and fractionalization: a possible reinterpretation. Eur.
P

Sociol. Rev. 24, 271–283.

CE

Bryan, S.H., Nash, R.C., Patel, A., 2015. The effect of cultural distance on contracting
decisions: the case of executive compensation. J. Corp. Financ. 33, 180–195.
AC

Burns, N., Minnick, K., Starks, L., in press. CEO tournaments: a cross-country analysis of
causes, cultural influences and consequences. J. Financ. Quant. Anal.

Carter, D.A., Simkins, B.J., Simpson, W.G., 2003. Corporate governance, board diversity,
and firm value. Financ. Rev. 38, 33–53.

Chui, A.C.W., Lloyd, A.E., Kwok, C.C.Y., 2002. The determination of capital structure: is
national culture a missing piece to the puzzle? J. Int. Bus. Stud. 33, 99–127.

Crespí-Cladera, R., Pascual-Fuster, B., 2014. Does the independence of independent directors
matter? J. Corp. Financ. 28, 116–134.

De Wit, F.R.C., Greeg, L.L., Jehn, K.A., 2012. The paradox of intragroup conflict: a meta-
analysis. J. Appl. Psychol. 97, 360–390.

42
 ACCEPTED MANUSCRIPT

Dodd, O., Frijns, B., Gilbert A., 2015. On the role of cultural distance in the decision to
cross-list. Eur. Financ. Manag. 21, 706–741.

Doidge, C., Karolyi, G.A., Stulz, R.M., 2004. Why are foreign firms listed in the U.S. worth
more? J. Financ. Econ. 71, 205–238.

T
Doney, P.M., Cannon, J.P., Mullen, M.R., 1998. Understanding the influence of national

IP
culture on the development of trust. Acad. Manag. Rev. 23, 601–620.

R
El Ghoul, S., Zheng, X., 2016. Trade credit provision and national culture. J. Corp. Financ.,

SC
forthcoming.

Elron, E., 1998. Top management teams within multinational corporations: effects of cultural

NU
heterogeneity. Leadership Quart. 8, 393–412.

Farrell, K.A., Hersch, P.L., 2005. Additions to corporate boards: the effect of gender. J. Corp.
MA
Financ. 11, 85–106.

Ferreira, D., 2010. Board diversity. In: Anderson, R., Baker, H.K. (Eds.), Corporate
Governance: A Synthesis of Theory, Research, and Practice. John Wiley & Sons,
D

Hoboken, NJ, 225–242.

TE

Ferris, S.P., Jayaraman, N., Sabherwal, S., 2013. CEO overconfidence and international
merger and acquisition activity. J. Financ. Quant. Anal. 48, 137–164.
P

Financial Reporting Council, 2014. The UK Corporate Governance Code (September 2014).
CE

Financial Reporting Council, London, UK.

Forbes, D.P., Milliken, F.J., 1999. Cognition and corporate governance: understanding boards
AC

of directors as strategic decision-making groups. Acad. Manag. Rev. 24, 489–505.

Frijns, B., Gilbert, A., Lehnert, T., Tourani-Rad, A., 2013. Uncertainty avoidance, risk
tolerance and corporate takeover decisions. J. Bank. Financ. 37, 2457–2471.

Gudykunst, W.B., Bond, M.H., 1997. Intergroup relations across cultures. In: Berry, J.W.,
Segall, M.H., Cagitçibasi, C. (Eds.), Handbook of Cross-Cultural Psychology, Volume 3:
Social Behavior and Applications. Allyn and Bacon, Needham Heights, MA, 119–161.

Guest, P.M., 2009. The impact of board size on firm performance: evidence from the UK.
Eur. J. Financ. 15, 385–404.

Guiso, L., Sapienza, P., Zingales, L., 2006. Does culture affect economic outcomes? J. Econ.
Perspect. 20, 23–48.

43
 ACCEPTED MANUSCRIPT

Guiso, L., Sapienza, P., Zingales, L., 2009. Cultural biases in economic exchange? Q. J.
Econ. 124, 1095–1131.

Hofstede, G., 2001. Culture’s Consequences: International Differences in Work-Related

Values, second ed. Sage Publications, Beverly Hills, CA.

T
Hofstede, G.J., 2003. Transparency in netchains. In: Harnos, S., Herdon, M., Wiwczaroski,

IP
T.B. (Eds.), Information Technology for a Better Agri-Food Sector, Environment and

R
Rural Living, Proceedings of the EFITA 2003 Conference. Debrecen University,
Debrecen, Hungary, 17–29.

SC
Hofstede, G.J., Jonker, C.M., Meijer, S., Verwaart, T., 2006. Modelling trade and trust across
cultures. In: Stølen, K., Winsborough. W.H., Martinelli, F., Massacci, F. (Eds.), Trust

NU
Management, Proceedings of the iTrust 2006 Conference. Springer, Berlin and
Heidelberg, 120–134.
MA
Hoi, C., Wollan, P., 2010. Effects of firm complexity on the adaption of board structure:
evidence from U.S. electric utilities following deregulation. Working Paper. Rochester
D

Institute of Technology.
TE

House, R., Hanges, P.J., Javidan, M., Dorfman, P., Gupta, V., 2004. GLOBE, Cultures,
Leadership, and Organizations: GLOBE Study of 62 Societies. Sage Publications,
P

Newbury Park, CA.

CE

Hutzschenreuter, T., Voll, J.C., 2008. Performance effects of “added cultural distance” in the
path of international expansion: the case of German multinational enterprises. J. Int. Bus.
AC

Stud. 39, 53–70.

Jehn, K.A., Mannix, E.A., 2001. The dynamic nature of conflict: a longitudinal study of
intragroup conflict and group performance. Acad. Manag. J. 44, 238–251.

Karolyi, A., 2016. The gravity of culture for finance. J. of Corp. Financ., forthcoming.
Kirchmeyer, C., Cohen, A., 1992. Multicultural groups – their performance and reactions
with constructive conflict. Group Organ. Manag. 17, 153–170.

Kirkman, B., Lowe, K., Gibson, C., 2006. A quarter century of culture’s consequences: a
review of empirical research incorporating Hofstede’s cultural values framework. J. Int.
Bus. Stud. 37, 285–320.

Kogut, B., Singh, H., 1988. The effect of national culture on the choice of entry mode. J. Int.
Bus. Stud. 19, 411–432.

44
 ACCEPTED MANUSCRIPT

Li, K., Griffin, D., Yue, H., Zhao, L., 2013. How does culture influence corporate risk-
taking? J. Corp. Financ. 23, 1–22.

Lievenbrück, M., Schmid, T., 2014. Why do firms (not) hedge? — Novel evidence on
cultural influence. J. Corp. Financ. 25, 92–106.

T
Lim, J., Makhija, A.K., Shenkar, O., 2016. The asymmetric relationship between national

IP
culture distance and target premiums in cross-border M&As. J. Corp. Financ.,

R
forthcoming.

SC
Linck, J., Netter, J., Yang, T., 2008. The determinants of board structure. J. Financ. Econ. 87,
308–328.

NU
Manzoni, J.-F., Strebel, P., Barsoux, J.-L., 2010. Why diversity can backfire on company
boards. The Wall Street Journal, 25 September 2010.

Masulis, R.W., Wang, C., Xie, F., 2012. Globalizing the boardroom – the effects of foreign
MA
directors on corporate governance and firm performance. J. Account. Econ. 53, 527–554.

Maznevski, M.L., 1994. Understanding our differences: performance in decision-making

D

groups with diverse members. Hum. Relat. 47, 531–552.

TE

Milliken, F.J., Martins, L.L., 1996. Searching for common threads: understanding the
multiple effects of diversity in organizational groups. Acad. Manag. Rev. 21, 402–433.
P

Nederveen Pieterse, A., van Knippenberg, D., van Diererdonck, D., 2013. Cultural diversity
CE

and team performance: the role of team member goal orientation. Acad. Manag. J. 56,
782–804.
AC

Oster, E., 2015. Unobservable selection and coefficient stability: theory and evidence.
Working Paper. Brown University and NBER.

Ouchi, W.G., 1979. A conceptual framework for the design of organizational control
mechanisms. Manag. Sci. 25, 833–848.

Oxelheim, L., Randøy, T., 2003. The impact of foreign board membership on firm value. J.
Bank. Financ.27, 2369–2392.

Petersen, M., 2009. Estimating standard errors in finance panel data sets: comparing
approaches. Rev. Financ. Stud. 22, 435–480.

Schneider, S.C., De Meyer, A., 1991. Interpreting and responding to strategic issues: the
impact of national culture. Strategic Manag. J. 12, 307–320.

45
 ACCEPTED MANUSCRIPT

Schwartz, S.H., 2006. A theory of cultural value orientations: explication and applications.
Int. Stud. Sociol. Soc. Anthropol. 104, 137–182.

Schwartz, S.H., 2012. An overview of the Schwartz theory of basic values. Online Read.
Psychol. Cult. 2(1). Accessed from http://dx.doi.org/10.9707/2307-0919.1116.

T
Shao, L., Kwok, C.C., Guedhami, O., 2010. National culture and dividend policy. J. Int. Bus.

IP
Stud. 41, 1391–1414.

R
Shenkar, O. (2001). Cultural distance revisited: towards a more rigorous conceptualization

SC
and measurement of cultural differences. J. Int. Bus. Stud. 32, 519–535.

Short, H., Keasey, K., 1999. Managerial ownership and the performance of firms: evidence

NU
from the UK. J. Corp. Financ. 5, 79–101.

Simons, T.L., Peterson, R.S., 2000. Task conflict and relationship conflict in top management
MA
teams: the pivotal role of intragroup trust. J. Appl. Psychol. 85, 102–111.

Smith, P.B., 2006. When elephants fight, the grass gets trampled: the GLOBE and Hofstede
projects. J. Int. Bus. Stud. 37, 915–921.
D

Spolaore, E., Wacziarg, R., 2009. The diffusion of development. Q. J. Econ. 124, 469–592.
TE

Staigler, D., Stock, J., 1997. Instrumental variables regression with weak instruments.
Econometrica 65, 557–586.
P

Stock, J.H., Yogo, M., 2005. Testing for weak instruments in linear IV regressions. In:
CE

Andrews, D.W.K., Stock, J.H. (Eds.) (2005). Identification and Inference for
Econometric Models: Essays in Honor of Thomas Rothenberg. Cambridge University
AC

Press, New York, 80–108.

Tabellini, G., 2008a. Presidential address: Institutions and culture. J. Eur. Econ. Assoc. 6,
255–294.

Tabellini, G., 2008b. The scope of cooperation: norms and incentives. Q. J. Econ. 123, 905–
950.

Tang, L., Koveos, P.E., 2008. A framework to update Hofstede's cultural value indices:
economic dynamics and institutional stability. J. Int. Bus. Stud. 39, 1045–1063.

Tosi, H.L., Greckhamer, T., 2004. Culture and CEO compensation. Organ. Sci. 15, 657–670.

Triandis, H.C., Bontempo, R., Villareal, M.J., Asai, M., Lucca, N., 1988. Individualism and
collectivism: cross-cultural perspectives on self-ingroup relationships. J. Pers. Soc.
Psychol. 54, 323–338.

46
 ACCEPTED MANUSCRIPT

White, J.T., Woidtke, T., Black, H.A., Schweitzer, R.L., 2014. Appointments of academic
directors. J. Corp. Financ. 28, 135–151.

Zheng, X., El Ghoul, S., Guedhami, O., Kwok, C.C.Y., 2012. National culture and corporate
debt maturity. J. Bank. Financ. 36, 468–488.

T
Zhou, X., 2001. Understanding the determinants of managerial ownership and the link

IP
between ownership and performance: comment. J. Financ. Econ. 62, 559–571.

R
SC
NU
MA
D
P TE
CE
AC

47
 ACCEPTED MANUSCRIPT

Table 1
Number of directors from each country by year.
Country 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Total % Total

T
Argentina 1 1 1 1 1 1 1 1 0 0 0 0 0 8 0.03%

IP
Australia 20 23 23 19 22 23 25 23 24 28 32 32 35 329 1.29%

CR
Austria 2 1 1 1 2 3 3 3 3 7 6 7 6 45 0.18%
Belgium 5 6 5 6 6 5 7 6 6 5 7 7 6 77 0.30%

US
Brazil 1 2 2 2 2 2 1 1 1 1 2 3 3 23 0.09%
Canada 8 9 10 9 9 13 13 13 20 19 23 25 24 195 0.76%

N
Chile 2 3 5 7 7 7 7 7 6 5 5 6 6 73 0.29%

MA
China 1 1 1 1 1 1 1 1 1 1 1 4 4 19 0.07%
Colombia 0 0 0 1 1 1 1 1 1 1 1 1 1 10 0.04%
Czech Republic 1 1 0 0 0 0 0 1 0 4 4 4 2 17 0.07%

D
Denmark 1 0 3 4 5 2 2 3 4 3 3 2 4 36 0.14%

TE
Finland 0 0 0 0 1 2 2 1 1 1 1 1 1 11 0.04%
France 23 29 30 26 27 24 31 30 33 38 38 40 39 408 1.60%

P
Germany 18 21 22 19 21 32 34 28 23 27 28 27 28 328 1.29%
Ghana
Greece
1
3
0
4
0
4
0
4
CE 0
5
0
5
0
5
1
4
1
4
1
4
1
4
1
3
1
3 52
7 0.03%
0.20%
AC
Hong Kong 1 1 1 1 1 1 1 0 0 0 0 1 1 9 0.04%
India 1 1 3 8 8 8 9 9 13 10 15 16 14 115 0.45%
Indonesia 0 0 0 0 0 0 0 0 0 0 4 3 3 10 0.04%
Ireland 24 26 27 26 30 26 25 23 26 26 31 31 31 352 1.38%
Israel 1 1 1 1 1 0 0 1 0 0 0 0 0 6 0.02%
Italy 5 5 4 7 5 4 4 5 6 7 9 11 10 82 0.32%
Jordan 0 0 0 0 3 4 4 4 4 4 4 3 3 33 0.13%
Kazakhstan 0 0 0 0 3 3 3 4 4 3 4 4 3 31 0.12%
Kenya 0 0 0 0 0 0 0 0 0 0 1 1 2 4 0.02%
Lebanon 0 0 1 2 2 1 0 0 0 0 0 0 0 6 0.02%
Luxembourg 0 0 0 0 0 0 0 0 1 1 1 1 1 5 0.02%

48
 ACCEPTED MANUSCRIPT

Malaysia 3 3 2 2 2 2 3 2 2 2 2 3 3 31 0.12%
Mauritius 0 0 0 0 0 0 0 0 2 2 2 2 0 8 0.03%

T
Mexico 0 0 0 0 0 0 5 4 4 4 5 5 6 33 0.13%

IP
Netherlands 18 15 25 23 30 38 37 33 34 36 34 34 37 394 1.54%
New Zealand 2 2 2 1 0 2 2 5 6 6 7 7 5 47 0.18%

CR
Nigeria 0 0 0 0 2 3 3 2 2 2 2 2 2 20 0.08%
Norway 0 0 1 1 3 2 2 2 3 3 2 1 3 23 0.09%

US
Pakistan 0 1 1 1 1 0 0 0 1 1 1 1 1 9 0.04%
Paraguay 0 0 0 0 0 0 0 0 0 0 0 1 1 2 0.01%

N
Philippines 0 0 0 0 1 1 1 1 1 1 1 0 0 7 0.03%

MA
Poland 0 0 0 0 0 0 1 1 0 0 0 0 0 2 0.01%
Portugal 1 1 1 1 1 1 1 1 2 3 3 3 3 22 0.09%
South Korea 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0.00%

D
Russia 3 3 3 3 4 4 5 5 3 3 8 9 9 62 0.24%

TE
Singapore 4 5 5 4 4 4 5 5 5 4 6 5 6 62 0.24%
South Africa 27 20 22 22 26 27 24 26 28 25 28 29 22 326 1.28%

P
Spain 2 5 5 5 7 6 6 6 6 6 4 4 5 67 0.26%
Sweden
Switzerland
14
2
13
2
15
2
16
3
CE 14
3
10
4
11
7
13
8
15
9
16
8
17
8
17
8
16
6
187
70
0.73%
0.27%
AC
Tanzania 0 0 0 0 0 0 0 0 0 0 1 1 1 3 0.01%
Turkey 1 1 1 1 1 0 0 0 0 0 0 0 0 5 0.02%
Ukraine 0 0 0 0 0 0 2 2 2 3 3 3 2 17 0.07%
United Arab Emirates 0 0 0 1 1 1 2 2 2 2 2 2 2 17 0.07%
United Kingdom 1,287 1,337 1,458 1,520 1,534 1,607 1,617 1,608 1,663 1,672 1,647 1,588 1,482 20,020 78.48%
United States 101 113 112 129 135 144 146 139 136 139 161 169 161 1,785 7.00%

Total 1,584 1,656 1,799 1,878 1,933 2,024 2,059 2,035 2,108 2,134 2,169 2,128 2,004 25,511
% of foreign directors 18.75% 19.26% 18.95% 19.06% 20.64% 20.60% 21.47% 20.98% 21.11% 21.65% 24.07% 25.38% 26.05% 21.52%
This table reports the number of directors from different countries by year. Data are reported annually for the period 2002–2012. Column “% Total” reports the percentage of
all directors coming from a particular country.

49
 ACCEPTED MANUSCRIPT

Table 2
Characteristics of domestic versus foreign directors.
Nationality Gender (male) Directors’ age Directors’ independence
All directors 89.81% 55.46 55.05%
UK 90.89% 55.10 53.92%
Foreign 85.87% 56.77 59.19%

T
Difference -5.03%*** 1.67*** 5.27%***

IP
This table reports differences in Gender, Age and Independence for the UK directors versus foreign (non-UK)
directors. Gender (male) is the percentage of males in the total number of directors. Directors’ Age is the
average age of directors in years. Directors’ Independence is the percentage of independent directors in the total

R
number of directors. The last row of the table reports the differences in characteristics between foreign versus
UK directors and their statistical significance based on a t-test. *** indicates significance at the 1% level.

SC
NU
MA
D
P TE
CE
AC

50
 ACCEPTED MANUSCRIPT

Table 3
Foreign directors and nationalities in boards.
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Total

T
Total # of firms 178 186 197 202 216 221 229 229 239 237 243 241 234 2,852

IP
% of firms with 56.74% 57.53% 56.85% 55.94% 57.41% 60.18% 62.01% 63.76% 62.76% 63.29% 68.31% 71.37% 71.79% 62.55%
foreign
Panel A:directors

CR
Number of foreign directors on the board
0 77 79 85 89 92 88 87 83 89 87 77 69 66 1,068
1 37 34 31 36 36 39 41 48 54 53 52 59 53 573

US
2 21 23 28 22 29 31 35 35 27 31 38 37 36 393
3 13 19 19 12 14 19 17 16 22 14 19 19 25 228

N
4 14 14 14 16 14 14 16 18 12 19 16 14 14 195

MA
5 1 3 6 11 15 12 11 11 16 12 15 13 11 137
6 4 1 4 7 5 6 7 8 8 9 8 6 10 83
7 7 8 4 4 4 5 6 2 5 3 7 10 12 77

D
8 2 3 3 2 3 3 6 4 1 2 5 7 2 43

TE
9 2 2 3 2 1 2 2 2 2 1 1 4 3 27
10 0 0 0 0 2 1 0 0 1 4 3 2 1 14

P
11 0 0 0 0 1 0 1 2 0 1 2 1 1 9
12 0 0 0
Panel B: Number of nationalities represented on the board
CE
1 0 1 0 0 2 1 0 0 0 5

0 77 79 85 89 92 92 87 83 89 87 77 69 66 1,068
AC

1 55 55 51 55 56 56 62 63 68 68 74 79 65 813
2 25 29 33 27 36 36 40 45 39 39 43 43 54 489
3 15 14 20 19 17 17 23 20 23 22 24 26 26 268
4 5 7 5 11 11 11 11 14 14 12 17 14 11 141
5 1 1 2 0 3 3 6 4 5 6 4 5 7 49
6 0 0 0 1 1 1 0 0 1 3 2 4 5 19
7 0 1 1 0 0 0 0 0 0 0 2 1 0 5
This table reports the distribution of firm-year observations annually for the sample period 2002–2012. Panel A reports the number of firms with a particular number of
foreign directors on the board. Panel B reports the number of firms with a particular number of different nationalities on the board.

51
 ACCEPTED MANUSCRIPT

Table 4
Firm-level characteristics of boards.
# of
CD Board foreign Gender Board Directors’ Directors’ CEO
BOARD size directors (male) independence avg. age age range duality
Panel A. Summary statistics

T
Mean 0.5261 8.89 1.90 90.20% 54.39% 55.41 23.97 10.40%
Median 0.2307 9.00 1.00 90.00% 55.56% 55.44 23.00 0.0

IP
Min 0.0 2.00 0.0 40.00% 0.0% 38 1.00 0.0
Max 3.3603 20.0 12.0 100.0% 100.0% 70.33 50.00 100.0%

R
Firms

SC
with FDs 0.8399 9.57 3.04 89.71% 56.61% 55.75 24.28 11.12%
without FDs 0.0 7.74 0.0 91.02% 50.69% 54.53 23.45 9.18%
Difference 0.840*** 1.83*** 3.04*** -1.31%*** 5.92%*** 1.22*** 0.82** 2.70%**
Panel B. Average value by year

NU
2002 0.4386 8.65 1.60 95.28% 44.78% 53.38 25.50 10.61%
2003 0.4555 8.78 1.68 94.59% 46.99% 53.70 25.31 10.75%
2004 0.4749 8.96 1.68 94.23% 47.17% 54.14 25.33 11.56%
MA
2005 0.4413 9.22 1.74 93.33% 49.78% 54.36 25.64 10.71%
2006 0.4989 8.87 1.81 91.68% 51.95% 54.28 24.40 11.39%
2007 0.4797 9.09 1.85 91.54% 53.74% 54.64 24.27 12.67%
2008 0.5191 8.99 1.93 91.45% 53.64% 54.97 23.93 12.66%
D

2009 0.5266 8.89 1.86 91.42% 55.53% 55.54 23.71 11.35%

TE

2010 0.5321 8.85 1.87 90.93% 56.44% 55.91 23.40 9.24%

2011 0.5443 8.93 1.93 89.93% 57.66% 56.35 23.37 9.21%
2012 0.5956 8.93 2.15 86.46% 58.92% 56.40 23.06 9.88%
P

2013 0.6221 8.93 2.24 84.04% 61.39% 56.66 22.66 8.71%

0.6455 8.55 2.22 81.35% 62.94% 57.12 22.21 6.41%
CE

2014
Panel A of the table reports summary statistics of firm-level characteristics of boards. CD BOARD is the measure of cultural board
diversity computed as in Eq. (2) (see Section 2.2). All other variables are defined in Appendix A. Panel A also reports means separately
for firms that have foreign directors on the board (with FDs) and firms that do not have foreign directors on the board (without FDs), and
AC

the differences in the means and their statistical significance based on a t-test. ***, **, and * indicate significance at the 1%, 5%, and
10% level, respectively. Panel B reports the average values of firm-level characteristics of boards by year.

52
 ACCEPTED MANUSCRIPT

Table 5
Summary statistics: firm performance and firm characteristics.
Mean for subsamples
Firms Firms
Variable Mean Median Min Max with FDs without FDs Difference
Panel A. Firm performance

T
Tobin’s Q 2.01 1.61 0.65 8.29 2.03 1.98 0.05
ROA 10.94 9.49 -14.01 39.55 10.60 11.50 -0.90***

IP
Panel B. Firm characteristics

R
Firm size 5,117.6 1,095.3 9 127,867 7,473.2 1,221.2 6,252.0***
Leverage 0.23 0.21 0 1.33 0.226 0.235 -0.009**

SC
Firm age 70.3 49.0 0 299.0 74.76 62.82 11.94***
Firm complexity 2.46 2.00 1 10 2.60 2.26 0.34***
Return volatility 2.15 1.85 0.41 12.66 2.14 2.17 -0.03
Sales growth 0.10 0.06 -0.42 1.10 0.09 0.10 -0.001

NU
This table reports summary statistics for firm performance (Panel A) and firm characteristics (Panel B). The table also reports
means separately for firms that have foreign directors on the board (firms with FDs) and firms that do not have foreign
directors on the board (firms without FDs), and the differences in the means and their statistical significance based on a t-test.
All variables are defined in Appendix A. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.
MA
D
P TE
CE
AC

53
 ACCEPTED MANUSCRIPT

Table 6
Cultural diversity and firm performance.
Tobin’s Q ROA
(1) (2) (3) (4)
CD BOARD -0.26*** -0.22*** -1.64*** -1.40**
(-3.57) (-3.04) (-2.95) (-2.52)

T
Board characteristics
Board size -0.54*** -2.71**

IP
(-2.66) (-2.15)
Gender (male) 0.15 3.42

R
(0.29) (0.93)
Board independence -0.24 -3.23

SC
(-0.67) (-1.43)
Director age -1.36* -6.42
(-1.93) (-1.41)
Directors’ age range 0.23* 0.95

NU
(1.74) (0.88)
CEO duality 0.14 0.87
(0.87) (0.79)
MA
Firm characteristics
Firm size 0.10** 0.18*** 0.62** 1.07***
(2.29) (3.57) (2.04) (3.20)
Leverage -0.76** -0.72** -2.52 -2.24
(-2.12) (-2.10) (-0.97) (-0.92)
D

Firm age -0.13** -0.12** -0.04 0.04

(-2.30) (-2.11) (-0.12) (0.11)
TE

Firm complexity -0.26*** -0.22** -1.87*** -1.68***

(-2.96) (-2.56) (-2.99) (-2.74)
Return volatility -0.08** -0.07* -1.34*** -1.25***
P

(-2.08) (-1.71) (-4.38) (-4.17)

Sales growth 0.64*** 0.53*** 4.72*** 4.04***
CE

(4.10) (3.44) (4.35) (3.76)

Constant 2.14*** 7.30*** 14.76*** 37.24**
(4.66) (2.74) (4.47) (2.15)
AC

Industry fixed effects YES YES YES YES

Year fixed effects YES YES YES YES

Observations 2,646 2,646 2,646 2,646

R-squared 0.17 0.19 0.11 0.12
This table reports OLS regression estimation results of firm performance measures on board and firm
characteristics. CD BOARD is the measure of cultural board diversity computed as in Eq. (2) (see Section 2.2).
All other variables are defined in Appendix A. Standard errors are clustered at the firm level and t-statistics are
reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

54
 ACCEPTED MANUSCRIPT

Table 7
Tobin’s Q & ROA: firm fixed effects.
Tobin’s Q ROA
(1) (2) (3) (4)
CD BOARD -0.11** -0.09* -0.71** -0.59
(-2.30) (-1.86) (-1.97) (-1.64)

T
Board characteristics
Board size -0.39*** -2.00***

IP
(-4.79) (-3.31)
Gender (male) 0.41** 5.34***

R
(1.99) (3.51)
Board independence 0.07 -1.20

SC
(0.49) (-1.17)
Director age -0.31 -4.91**
(-1.11) (-2.33)
Directors’ age range 0.11* 1.65***

NU
(1.95) (3.96)
CEO duality 0.03 0.68
(0.41) (1.36)
Firm characteristics
MA
Firm size 0.65*** 0.66*** 2.10*** 2.17***
(23.53) (24.01) (10.21) (10.52)
Leverage 0.28* 0.32** -1.99* -1.67
(1.89) (2.13) (-1.77) (-1.49)
D

Firm age -0.37*** -0.37*** -0.17 -0.16

(-4.10) (-4.06) (-0.25) (-0.24)
TE

Firm complexity -0.10*** -0.09** -0.48* -0.38

(-2.60) (-2.27) (-1.66) (-1.32)
Return volatility 0.11*** 0.11*** -0.19 -0.17
P

(5.71) (5.95) (-1.32) (-1.17)

Sales growth 0.09 0.07 4.44*** 4.33***
CE

(1.36) (1.09) (8.66) (8.47)

Constant -1.35*** -0.20 -1.80 11.33
(-3.47) (-0.18) (-0.62) (1.35)
AC

Firm fixed effects YES YES YES YES

Year fixed effects YES YES YES YES

Observations 2,646 2,646 2,646 2,646

R-squared 0.32 0.33 0.14 0.15
This table reports firm fixed effects regression estimation results of firm performance measures on board and
firm characteristics. CD BOARD is the measure of cultural board diversity computed as in Eq. (2) (see Section
2.2). All other variables are defined in Appendix A. ***, **, and * indicate significance at the 1%, 5%, and 10%
level, respectively.

55
 ACCEPTED MANUSCRIPT

Table 8
Instrumental variables regressions.
First stage Second stage
CD BOARD Tobin’s Q ROA
(1) (2) (3)
Instrumental variables

T
UK small town -0.187***
(-9.06)

IP
CD BOARD -0.74*** -3.65**
(-2.86) (-2.22)

R
Board characteristics
Board size 0.26*** -0.40*** -2.13**

SC
(4.60) (-2.93) (-2.57)
Gender (male) 0.28** 0.27 3.92**
(2.20) (1.00) (2.12)
Board independence 0.33*** -0.08 -2.54**

NU
(4.28) (-0.47) (-2.29)
Director age 0.46*** -1.09*** -5.26**
(2.70) (-3.08) (-2.21)
Directors’ age range 0.005 0.23*** 0.98*
MA
(0.11) (3.37) (1.89)
CEO duality -0.10*** 0.10 0.69
(-3.28) (1.36) (1.33)
Firm characteristics
D

Firm size 0.09*** 0.23*** 1.30***

(7.79) (6.23) (5.58)
TE

Leverage 0.15*** -0.64*** -1.91

(2.56) (-3.64) (-1.61)
Firm age -0.05*** -0.15*** -0.09
P

(-4.61) (-5.53) (-0.54)

Firm complexity -0.073*** -0.26*** -1.85***
CE

(-3.94) (-6.17) (-6.52)

Return volatility 0.045*** -0.05 -1.13***
(3.17) (-1.42) (-5.32)
Sales growth -0.008 0.53*** 4.03***
AC

(-0.14) (4.13) (4.11)

Constant -2.23*** 5.96*** 31.47***
(-3.38) (4.19) (3.31)

Industry fixed effects YES YES YES

Year fixed effects YES YES YES

Observations 2,646 2,646 2,646

R-squared 0.29 0.13 0.10
This table reports instrumental variables regression estimation results. Column 1 reports the first-stage results of
the 2SLS regressions with CD BOARD as the dependent variable. CD BOARD is the measure of cultural board
diversity computed as in Eq. (2) (see Section 2.2). UK small town is a dummy variable equal to one if a firm is
headquartered outside of a large metropolitan area (i.e., in a town with population less than 250,000 people). All
other variables are defined in Appendix A. Columns 2 and 3 report the second-stage results from 2SLS
regressions for Tobin’s Q and ROA, respectively. Standard errors are clustered at the firm level and t-statistics
are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

56
 ACCEPTED MANUSCRIPT

Table 9
Alternative measures of “foreignness” of board and firm.
Tobin’s Q ROA

T
(1.1) (1.2) (1.3) (1.4) (1.5) (1.6) (1.7) (1.8) (1.9) (2.1) (2.2) (2.3) (2.4) (2.5) (2.6) (2.7) (2.8) (2.9)

IP
Panel A. “Foreignness” of board
Foreign directors -0.49* 0.14 -2.17 2.95

CR
ratio (-1.76) (0.27) (-1.08) (0.94)
Nationalities ratio -0.24 1.37** -3.60 3.66
(-0.60) (2.02) (-1.28) (0.79)
FID dummy 0.40* 0.36 0.88 0.60

US
(1.78) (1.63) (0.54) (0.37)
FID ratio 0.25 0.25 2.10 2.10
(0.78) (0.78) (0.96) (0.96)

N
CD BOARD -0.36*** -0.25* -0.45*** -0.22*** -0.22*** -1.40** -2.05** -2.00** -1.40** -1.40**

MA
(-3.52) (-1.84) (-3.53) (-3.01) (-3.04) (-2.52) (-2.40) (-2.14) (-2.51) (-2.52)

Control variables YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
Industry FEs YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES

D
Year FEs YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES

TE
Observations 1,636 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 1,636 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646
R-squared 0.18 0.18 0.19 0.18 0.19 0.18 0.19 0.18 0.19 0.12 0.12 0.12 0.12 0.12 0.11 0.12 0.11 0.12
Panel B. “Foreignness” of firm

P
Foreign sales -0.16 -0.05 -0.38 0.38

Foreign assets
(-0.93) (-0.29)
-0.52*** -0.45***
CE (-0.35) (0.36)
-2.11* -1.58
(-3.25) (-2.92) (-1.94) (-1.50)
AC
NYSE listing 0.13 0.16 2.92** 3.14**
(0.59) (0.72) (2.06) (2.20)
Foreign listing -0.15 -0.12 -0.20 0.01
(-1.18) (-0.93) (-0.22) (0.01)
CD BOARD -0.21*** -0.18*** -0.22*** -0.21*** -1.46*** -1.24** -1.48*** -1.40**
(-3.18) (-2.61) (-3.05) (-2.93) (-2.64) (-2.27) (-2.76) (-2.52)

Control variables YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
Industry FEs YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
Year FEs YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES

Observations 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646 2,646
R-squared 0.18 0.19 0.19 0.20 0.18 0.19 0.18 0.19 0.11 0.12 0.12 0.12 0.12 0.13 0.11 0.12

57
 ACCEPTED MANUSCRIPT

This table reports OLS regression estimation results of firm performance measures on board and firm characteristics. CD BOARD is the measure of cultural board diversity
computed as in Eq. (2) (see Section 2.2). All other variables are defined in Appendix A. Standard errors are clustered at the firm level and t-statistics are reported in
parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

T
IP
CR
N US
MA
D
P TE
CE
AC

58
 ACCEPTED MANUSCRIPT

Table 10
Cultural diversity, firm performance, and firm’s operations.
Tobin’s Q ROA
Panel A. By firm complexity
# of business # of business # of business # of business
segments <=3 segments >3 segments <=3 segments >3

T
CD BOARD -0.22** -0.15 -1.37** -0.81
(-2.68) (-1.28) (-2.30) (-0.89)

IP
Control variables YES YES YES YES

R
Industry fixed effects YES YES YES YES
Year fixed effects YES YES YES YES

SC
Observations 2,084 568 2,084 568
R-squared 0.20 0.18 0.11 0.20
Panel B. By foreign sales

NU
foreign sales foreign sales foreign sales foreign sales
<=75.585% >75.585% <=75.585% >75.585%
CD BOARD -0.23** -0.14 -1.43** 0.10
MA
(-2.29) (-1.57) (-2.22) (0.12)

Control variables YES YES YES YES

Industry fixed effects YES YES YES YES
Year fixed effects YES YES YES YES
D

Observations 1,857 642 1,857 642

TE

R-squared 0.21 0.29 0.13 0.33

Panel C. By foreign assets
foreign assets foreign assets foreign assets foreign assets
P

<=50.275% >50.275% <=50.275% >50.275%

CE

CD BOARD -0.25** -0.14* -1.26* -0.90

(-2.44) (-1.75) (-1.84) (-1.29)

Control variables YES YES YES YES

AC

Industry fixed effects YES YES YES YES

Year fixed effects YES YES YES YES

Observations 1,862 637 1,862 637

R-squared 0.20 0.31 0.14 0.30
This table reports OLS regression estimation results of firm performance measures on board and firm
characteristics. CD BOARD is the measure of cultural board diversity computed as in Eq. (2) (see Section 2.2).
All other variables are defined in Appendix A. Standard errors are clustered at the firm level and t-statistics are
reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

59
 ACCEPTED MANUSCRIPT

Table 11
Dependent versus independent directors.
Tobin’s Q ROA
(1) (2) (3) (4)
CD BOARD DEP -0.08 -0.08 0.37 0.35
(-1.31) (-1.33) (0.67) (0.66)

T
CD BOARD INDEP -0.21*** -0.18*** -1.71*** -1.45***
(-3.26) (-2.65) (-3.54) (-3.00)

IP
Control variables YES YES YES YES

R
Industry fixed effects YES YES YES YES
Year fixed effects YES YES YES YES

SC
Observations 2,447 2,447 2,447 2,447
R-squared 0.17 0.19 0.11 0.13
This table reports OLS regression estimation results of firm performance measures on board and firm

NU
characteristics. CD BOARD DEP and CD BOARD INDEP are the measures of cultural board diversity
computed for dependent and independent directors, respectively. All other variables are defined in Appendix A.
Standard errors are clustered at the firm level and t-statistics are reported in parentheses. ***, **, and * indicate
significance at the 1%, 5%, and 10% level, respectively.
MA
D
P TE
CE
AC

60
 ACCEPTED MANUSCRIPT

Table 12
Alternative culture frameworks.
Panel A: Tobin’s Q
Tang & Koveos GLOBE Schwartz
Firms with Firms with Firms with
All firms FDs All firms FDs All firms FDs

T
IP
CD BOARD -0.19** -0.37*** -0.10 -0.26*** -0.39 -0.95**
(-2.10) (-2.88) (-1.43) (-2.96) (-1.60) (-2.58)

R
Control variables YES YES YES YES YES YES
Industry fixed effects YES YES YES YES YES YES

SC
Year fixed effects YES YES YES YES YES YES

Observations 2,536 1,526 2,471 1,461 2,372 1,362

NU
R-squared 0.19 0.18 0.19 0.17 0.19 0.17
Panel B: ROA
Tang & Koveos GLOBE Schwartz
MA
Firms with Firms with Firms with
All firms FDs All firms FDs All firms FDs

CD BOARD -1.36** -1.58* -0.91* -1.03* -2.57* -3.72*

(-2.04) (-1.72) (-1.87) (-1.72) (-1.66) (-1.79)
D

Control variables YES YES YES YES YES YES

TE

Industry fixed effects YES YES YES YES YES YES

Year fixed effects YES YES YES YES YES YES
P

Observations 2,536 1,526 2,471 1,461 2,372 1.362

R-squared 0.12 0.13 0.12 0.11 0.12 0.13
CE

This table reports OLS regression estimation results of firm performance measures on board and firm characteristics.
CD BOARD is the measure of cultural board diversity computed as in Eq. (2) (see Section 2.2) using different culture
frameworks. All other variables are defined in Appendix A. Standard errors are clustered at the firm level and t-
statistics are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.
AC

61
 ACCEPTED MANUSCRIPT

Table 13
Other measures of culture.
Panel A: Tobin’s Q

Bilateral trust 1.54**

(2.52)

T
Language similarity index 0.65**
(2.77)

IP
Religion similarity index 0.84***
(3.37)

R
Genetic diversity -0.56**
(-2.36)

SC
Somatic diversity -0.27***
(-2.80)
Control variables YES YES YES YES YES
Industry fixed effects YES YES YES YES YES

NU
Year fixed effects YES YES YES YES YES

Observations 1,401 2,646 2,646 1,715 1,401

R-squared 0.26 0.19 0.19 0.23 0.26
MA
Panel B: ROA

Bilateral trust 9.96***

(2.69)
D

Language similarity index 4.12**

(2.36)
TE

Religion similarity index 3.81**

(2.20)
Genetic diversity -6.43**
P

(-2.29)
Somatic diversity -2.06***
CE

(-3.34)
Control variables YES YES YES YES YES
Industry fixed effects YES YES YES YES YES
Year fixed effects YES YES YES YES YES
AC

Observations 1,401 2,646 2,646 1,715 1,401

R-squared 0.17 0.12 0.12 0.15 0.18
This table reports OLS regression estimation results of firm performance measures on board and firm
characteristics. Bilateral trust, genetic diversity, and somatic diversity are calculated as in Eq. (3) (see Section
4.4), that is, to obtain the diversity measure, all differences between each two directors are averaged. The
language and religion similarity indexes are based on the main language and religion of a country, and are
computed based on a Herfindahl index. All other variables are defined in Appendix A. Standard errors are
clustered at the firm level and t-statistics are reported in parentheses. ***, **, and * indicate significance at the
1%, 5%, and 10% level, respectively.

62
 ACCEPTED MANUSCRIPT

Table 14
Individual culture dimensions.
Panel A: Tobin’s Q
All firms Firms with FDs
(1) (2) (3) (4) (5) (6) (7) (8)

T
CD BOARD (IDV) -0.19** -0.38***
(-2.73) (-3.48)

IP
CD BOARD (MAS) -0.12* -0.24***
(-1.70) (-2.98)
-0.17*

R
CD BOARD (PDI) -0.09
(-1.08) (-1.93)

SC
CD BOARD (UAI) -0.04 -0.07
(-0.49) (-0.93)

Control variables YES YES YES YES YES YES YES YES

NU
Industry fixed effects YES YES YES YES YES YES YES YES
Year fixed effects YES YES YES YES YES YES YES YES

Observations 2,646 2,646 2,646 2,646 1,636 1,636 1,636 1,636

MA
R-squared 0.18 0.18 0.18 0.18 0.18 0.16 0.16 0.15
Panel B: ROA
All firms Firms with FDs
(1) (2) (3) (4) (5) (6) (7) (8)
D

CD BOARD (IDV) -1.34*** -1.62**

TE

(-2.66) (-2.36)
CD BOARD (MAS) -0.89* -0.89
(-1.78) (-1.61)
P

CD BOARD (PDI) -0.78 -0.54

(-1.43) (-0.91)
CE

CD BOARD (UAI) -0.58 -0.30

(-1.14) (-0.51)

Control variables YES YES YES YES YES YES YES YES
AC

Industry fixed effects YES YES YES YES YES YES YES YES
Year fixed effects YES YES YES YES YES YES YES YES

Observations 2,646 2,646 2,646 2,646 1,636 1,636 1,636 1,636

R-squared 0.13 0.12 0.12 0.12 0.14 0.13 0.13 0.13
This table reports OLS regression estimation results of firm performance measures on board and firm
characteristics. CD BOARD is the measure of cultural board diversity computed as in Eq. (2) (see Section 2.2),
where the cultural distance between each two directors is the difference in one of the following individual
cultural dimensions: individualism (IDV), masculinity (MAS), power distance (PDI) and uncertainty avoidance
(UAI). All other variables are defined in Appendix A. Standard errors are clustered at the firm level and t-
statistics are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% level,
respectively.

63
 ACCEPTED MANUSCRIPT

Appendix A
Table A1
Variable definitions and data sources.
Variable Source Definition
Panel A: Cultural diversity

T
CD BOARD Calculated The average of cultural distances between each two directors; cultural
distances are based on value scores from the Hofstede culture framework,

IP
or alternative culture frameworks.
Panel B: Firm performance

R
Tobin’s Q Calculated Market value of assets over book value of assets: [(Book value of total
based on data assets - Book value of equity + Market value of equity)/Book value of total

SC
from assets] – all components are measured at the end of the calendar year;
Datastream winsorized at 1% at each end of the distribution.
ROA Calculated Return on assets (in %): (Operating income / Book value of total assets);
based on data all components are measured at the end of the calendar year; winsorized at

NU
from 1% at each end of the distribution.
Datastream
Panel C: Firm characteristics
Firm size Datastream Market value in millions of GBP measured at the end of the calendar year;
MA
log-transformed.
Leverage Calculated Total debt scaled by total assets, measured at the end of the calendar year.
based on data
from
Datastream
D

Firm age Datastream The number of years since the company was established; log-transformed.
Firm complexity Datastream The number of business segments (i.e., product lines); log-transformed.
TE

Return volatility Calculated The standard deviation of daily stock returns during a calendar year (in %).
based on data
from
P

Datastream
Sales growth Calculated The annual growth rate of the firm’s total sales (in %); winsorized at 1% at
CE

based on data each end of the distribution.

from
Datastream
Panel D: Board characteristics
AC

Board size Orbis, annual The number of directors on the board; log-transformed.
reports
Gender (male) Orbis, annual The proportion of male directors (in %).
reports
Board Orbis, annual The proportion of independent directors (in %); independence
independence reports encompasses formally independent directors as declared in company
documents and non-executive directors with no ties to the company’s
management.
Director age Orbis, annual The average age (in years) of all directors; log-transformed.
reports
Directors’ age Orbis, annual The age difference (in years) between the oldest and youngest directors on
range reports the board; log-transformed.
CEO duality Orbis, annual An indicator variable that equals one if the CEO serves also as the
reports Chairman of the board of directors, and zero otherwise.
Panel E: Measures of “foreignness”
Foreign directors Orbis, annual The proportion of directors that have a foreign nationality, i.e. the number
ratio reports of non-UK directors on the board divided by the total number of directors.
Nationalities ratio Orbis, annual The number of different nationalities represented on the board divided by
reports the total number of directors.
FID dummy Orbis, annual An indicator variable that equals one if the firm has at least one foreign
reports independent director on the board, and zero otherwise.

64
 ACCEPTED MANUSCRIPT

FID ratio Orbis, annual The number of foreign independent directors on the board divided by the
reports total number of directors.
Foreign sales Datastream A firm’s foreign sales as a percentage of total sales.
Foreign assets Datastream A firm’s foreign assets as a percentage of total assets.
NYSE listing Datastream An indicator variable that equals one if a firm is listed on the New York
Stock Exchange (NYSE), and zero otherwise.
Foreign listing Datastream An indicator variable that equals one if a firm is listed on a stock exchange

T
outside of the UK, and zero otherwise.

IP
The table contains definitions for all variables employed in our empirical analysis and principal sources of data used
to compute their values.

R
SC
NU
MA
D
P TE
CE
AC

65
 ACCEPTED MANUSCRIPT

Highlights

 We study the impact of cultural diversity in boards of directors on firm performance.

 Cultural diversity is a scaled average of cultural distances between board directors.

T
 Cultural diversity in corporate boards negatively affects firm performance.

IP
 The negative impact is mitigated by a firm’s complexity and foreign operations.

R
 Diversity in individualism and masculinity affect the effectiveness of boards.

SC
NU
MA
D
P TE
CE
AC

66