International Review of Law and Economics  Culture, Law, and Corporate Governance
Amir N. Licht a,∗, Chanan Goldschmidt a, Shalom H. Schwartz b
留學生畢業dissertationa Interdisciplinary Center Herzliya, Israel
b Department of Psychology, Hebrew University in Jerusalem, Israel
Abstract
This article advances a new framework for investigating a simple yet fundamental question: inwhat ways does the law on the books reflect cultural values? We analyze relations between indices
of investors’ legal rights—as coded by La Porta et al. (LLSV)—and national cultural profiles. Theseindices correlate with cultural priorities that are consistent with societal acceptance of litigation.
Indices of formalism in civil procedure exhibit similar correlations. Such societal stance may berelated to a heritage of British rule. Grouping countries according to legal families—the cornerstoneof LLSV’s legal approach—provides only a partial depiction of the universe of corporate governanceregimes. Our findings cast doubt on the alleged general supremacy of statutes incommonlawcountries.These findings have implications for understanding diversity and convergence in corporate governancesystems and for a systematic analysis of the interface between law and social institutions.
Keywords: Culture; Law; Corporate governance
1. Introduction
This article advances a new framework for investigating a simple yet fundamental question:in what ways does the law on the books reflect cultural values? While the notion that itcould might appear to be a truism, analysis of this question in law and economics is lacking.In 1996, La Porta, Lopez-de-Silanes, Shleifer, and Vishny (LLSV) redefined the analyticalframework for comparative research on corporate governance by introducing an
∗ Corresponding author at: Radzyner School of Law, Interdisciplinary Center Herzliya, Kanfe Nesharim Street,
integrated approach to law and finance.1 LLSV’s operationalization of investors’ legalrights and legal origins enabled statistical tools to be implemented.2 These constructs havebeen shown to correlate with important economic factors, thus suggesting the desirabilityof enhancing investors’ rights through legislative reform.3 LLSV later advanced what theycalled ‘the legal approach’ as the preferred way to understand corporate governance. Thisapproach focused on classification by legal origins more than on particular index scores.4Laterworks by LLSV and others treat the legal origin variable as a proxy for colonial impacton social institutions, with a common law origin generally predicting a beneficial impact.5Several countries formerly under communist regimes implemented legal reforms during
the 1990s with a view to enhancing investor protection. With few exceptions, thesereforms produced outcomes that varied from disappointing to ruinous.6 At the turn of themillennium, commentators came to share the view that simply writing investor rights intothe law is not enough; more fundamental issues must be confronted. Pistor and co-workersthus highlighted the importance of the mode of legal transplantation and the level of lawenforcement (legality).7#p#分頁標題#e#
The insight that changing the law on the books does not guarantee corporate governanceimprovement encouraged researchers to consider additional factors. Rajan and Zingalesrefer to history and politics to call LLSV’s results into question.8 Pagano andVolpin advancea political economy model and evidence that associates minority shareholder protectionwith the prevalence of coalition governments in countries.9 Relatedly, Roe links social-
1 Rafael La Porta et al., Law and Finance, 106 J. Pol. Econ. 1113 (1998), originally appeared as NBERWorkingPaper No. 5661 (1996).
2 The indices have become a standard reference. Early implementations include Ross Levine, Law, Finance, and
Economic Growth, 8 J. Fin. Intermediation 36 (1999); Ross Levine, The Legal Environment, Banks, and Long-Run
Economic Growth, 30 J. Money, Credit & Bank. 596 (1998); Asli Demirguc-Kunt & Vojislav Maksimovic, Law,
Finance, and Firm Growth, 53 J. Fin. 2107 (1998).
3 See La Porta et al., supra note 1; Rafael La Porta et al., Legal Determinants of External Finance, 52 J. Fin. 1131
(1997) (hereinafter Legal Determinants); Rafael La Porta et al., Agency Problems and Dividend Policies AroundtheWorld, 55 J. Fin. 1 (2000). See also Rafael La Porta, Florencio Lopez-de-Silanes,&Andrei Shleifer, CorporateOwnership around theWorld, 54 J. Fin. 471 (1999); Simon Johnson, Rafael La Porta, Florencio Lopez-de-Silanes,
& Andrei Shleifer, Tunneling, 90 Am. Econ. Rev. Papers & Proceedings 22 (2000).
4 Rafael La Porta et al., Investor Protection and Corporate Governance, 58 J. Fin. Econ. 3 (2000).
5 See, e.g. Rafael La Porta, Florencio Lopez-De-Silanes, Andrei Shleifer, & Robert W. Vishny, The Quality of
Government, 15 J. L. Econ. & Org. 222 (1999); Thorsten Beck, Ross Levine, & Norman Loayza, Finance andthe Sources of Growth, 58 J. Fin. Econ. 261 (2000); Ross Levine, Law, Finance, and Economic Growth, 8 J. Fin.Intermediation 36 (1999).
6 Bernard Black, Reinier Kraakman, & Anna Tarassova, Russian Privatization and Corporate Governance:What Went Wrong? 52 Stan. L. Rev. 1739 (2000); Edward Glaeser, Simon Johnson, & Andrei Shleifer, Coaseversus the Coasians, 116 Q. J. Econ. 853 (2001).
7 Daniel Berkowitz, Katharina Pistor,&Jean-Francois Richard, Economic Development, Legality and the TransplantEffect, 47 Eur. Econ. Rev. 165 (2003); Katharina Pistor, Patterns of Legal Change:Shar eholder and CreditorRights in Transition Economies, 1 Eur. Bus. Org. L. Rev. 59 (2000); Katharina Pistor, Martin Raiser, & StanislawGelfer, Law and Finance in Transition Economies, EBRD Working Paper No. 48 (2000).
8 Raghuram Rajan & Luigi Zingales, The Great Reversals:The Politics of Financial Development in the 20thCentury, 69 J. Fin. Econ. 5 (2003).
9 Marco Pagano & Paolo F. Volpin, The Political Economy of Corporate Governance, CSEFWorking Paper No.
29 (2000).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 231democratic governments with the absence of dispersed shareholding in countries.10 Beck,#p#分頁標題#e#
Demirguc¸-Kunt, and Levine present evidence consistent with the legal approach and with atheory of bio-geographic endowment—both of which are connected to the impact of colonialrulers.11 In a remarkable development, Shleifer posited that the practice of justice and the‘structure’ of society rather than the law itself are what matters for investor protection—thatlegal rules are just a reflection of a broader societal stance.12Theorists, policy-makers, and practitioners share the intuition that corporate governancereflects national culture.13 Until recently, however, commentators treated culture either
anecdotally or as a black box. The challenge for economic analysis is to operationalizeculture in ways that permit developing and investigating testable hypotheses. Toward thisend, Stulz and Rohan Williamson and Beck et al. used countries’ predominant religion asa proxy for their national culture.14
Religion is a convenient proxy for culture. Classifying countries by religion, however,fails to capture the richness of cultural differences. This is especially so for the commonlyemployed religious classification of countries as Protestant versus other. This misses boththe complexity of religious variation around the world and the large differences in religiouscommitment within countries. Even more important, it leaves the substantive meaning ofthe cultural differences virtually undefined.We propose an alternative cultural framework for addressing the fundamental questionposed above: in what way do the laws on the books in different societies reflect the culturethat prevails in those societies? Put anotherway: are meaningful, measurable elements of theculture found in different countries manifest in the statutory legal rules of those countries?Like others, we operationalize legal rules by using the LLSV dataset. Notwithstandingits limitations as a measure of investor protection (see below), this dataset provides a uniquesnapshot of certain legal rules in the context of corporate governance. Narrowly interpreted,these indices specify sets of circumstances over which disputing corporate constituenciescan resort to litigation.We derive testable hypotheses that link the LLSV scores of countriesto specific aspects of the cultures of these countries. In doing so, we follow the view thatprotecting investors through statutory legal rights hinges on the practice of justice in thecourt system.
To conceptualize and measure culture in a richer manner, we draw upon the prevailingtheories in cross-cultural psychology—a discipline heretofore untapped by law and economics.
Analyzing cultural value dimensions and generating data that describe national
10 Mark J. Roe, Political Determinants of Corporate Governance (2003).
11 Thorsten Beck, Asli Demirguc-Kunt, & Ross Levine, Law, Endowments, and Finance, 70 J. Fin. Econ. 137
(2003).
12 Andrei Shleifer, Strategies for Protecting Investors, David Horowitz Lectures, Jerusalem and Tel Aviv, Israel,#p#分頁標題#e#
February 2002 (on file with authors).
13 See, e.g. Lucian Arye Bebchuk & Mark J. Roe, A Theory of Path Dependence in Corporate Governance
and Ownership, 52 Stan. L. Rev. 127, 168 (1999); OECD Ad Hoc Task Force on Corporate Governance, OECDPrinciples of Corporate Governance, Document SG/CG(99)5 3 (1999); CalPERS, Global Corporate GovernancePrinciples (1999).
14 Ren´e M. Stulz & Rohan Williamson, Culture, Openness, and Finance, 70 J. Fin. Econ. 313 (2003); Becket al., supra note 11. These works were inspired by David Landes, who argues that religious values may affectthe development of institutions. David Landes, Culture Makes Almost All of the Difference, in Culture Matters 2(Lawrence E. Harrison & Samuel P. Huntington, Eds., 2000).
232 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255cultures is a central focus of this discipline. Models and datasets on cultural value dimensionsallow, for the first time, to derive testable hypotheses without resorting to crude proxieslike religion.
This study reveals correlations, predicted on the basis of theory, between national scoreson cultural value dimensions and indices of shareholder voting rights and of creditor rights.These findings suggest that a national culture that promotes assertiveness in reconcilingconflicting interests and that promotes tolerance for the uncertainty this creates is consistentwith using litigation to deal with economic conflicts. The correlations between nationalculture and legal rules hold regardless of other major characteristics of countries. Moreover,these associations are not due to a reverse causal impact of legal rules on culture. In part, theymay reflect a heritage of British rule. Further evidence suggests that cultural orientationspersist in the face of formal legal reforms. Hence, national culture may impede reform andmay induce path dependence in corporate governance systems.
The final inference from the current study concerns the classification of countries by legalorigin—the cornerstone of LLSV’s legal approach.We find that this classification providesonly a partial depiction of the universe of corporate governance regimes. Shareholder rightsare higher in countries belonging to the English-speaking cultural region. However, thesecountries fare no better than others in protecting creditors. This casts doubt on the allegedgeneral superiority of statutes in common law countries for protecting investors. We conclude
that analyses of corporate governance regimes should combine the legal with thecultural approach.
2. Theory and hypotheses
2.1. Relations between culture and legal rules
The mechanisms that connect culture and law in national societies are mostly terraincognita in economic theory. OliverWilliamson advances a notional model distinguishingfour levels of analysis.15 “Level 1” consists of informal institutions.16 This is where norms,customs, mores, and traditions are located and where religion plays a role. According toWilliamson, “Level 1 is taken as given by most institutional economists.” Level 2, locatedbelow Level 1, consists of formal legal rules, comprising constitutions, law, property rights,etc. North has shown that the definition and enforcement of legal rights are importantfeatures of this level.17 In this model, higher levels impose constraints on the developmentof the levels immediately below.Williamson postulates that Level 1 informal institutions arepervasively linked with complementary institutions, both formal and informal. The resultinginstitutions “have a lasting grip on the way a society conducts itself.” Governance structures(e.g. of firms) and marginal analysis (e.g. of economic outcomes) belong to Levels 3 and15 Oliver E. Williamson, The New Institutional Economics:T aking Stock, Looking Ahead, 38 J. Econ. Lit. 595#p#分頁標題#e#
(2000).
16 Williamson identifies Level 1 with the notion of “embeddedness” proposed by Mark Granovetter, Economic
Action and Social Structure:The Problem of Embeddedness, 91 Am. J. Sociol. 481 (1985). Both concepts must
not be confused with the cultural orientation of Embeddedness.
17 Douglass C. North, Institutions, 5 J. Econ. Perspectives 97 (1991).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 233
4, respectively, and therefore lie beyond the present scope. Although the system is fully
interconnected, feedback among levels is neglected.
Viewing Level 1 informal institutions merely as constraints, that define transaction costs
for alternative formal institutions, does not fully capture their role. The prevailing informal
institutions in a society (e.g. beliefs, norms, and values) also serve as sources of motivation
for and justification of alternative formal institutions.18 Culture—as we conceptualize and
operationalize in the present study—encompasses both facets of Level 1 institutions, namely
as constraints and as motivational factors. Culture operates as a constraint due to its nature as
societies’ “common knowledge” (see Section 2.2). It thus coordinates people’s epistemics
and expectations. This lowers the social costs for developing and sustaining institutions
that are compatible with prevailing cultural values. Culture operates to motivate and justify
action compatible with its values through its impact on organizational policies and on the
values of individual decision-makers.19
The foregoing analysis points to the assumption that underlies our basic hypotheses: in
the long run the content of formal legal rules should be compatible with and partly reflect
the prevailing cultural orientations in a society. The present study considers legal rules that
pertain to reconciling conflicting economic interests through the court system. Specifically,
we focus on litigation-based enforcement of investor rights in corporations. To corroborate
the analysis, we also look at enforcement of simpler contractual rights.
2.2. Value dimensions of culture
Definitions of culture abound. In order to distinguish culture from structural aspects
of society that might influence corporate governance, we define culture in subjective
terms. Culture refers to the complex of meanings, symbols, and assumptions about what
is good or bad, legitimate or illegitimate that underlie the prevailing practices and norms
in a society.20 Value emphases are the essence of culture seen this way. They are the
implicitly or explicitly shared, abstract ideas about what is good, right, and desirable
in a society.21 They justify and guide the ways that social institutions (e.g. the family,
18 Avner Greif, Cultural Beliefs and the Organization of Society:A Historical and Theoretical Reflection on#p#分頁標題#e#
Collectivist and Individualist Societies, 102 J. Pol. Econ. 912 (1994); Victor Nee, A New Institutional Approach
to Economic Sociology, in The Handbook of Economic Sociology 49 (Neil Smelser & Richard Swedberg, Eds.,
2nd ed., 2005).
19 Lilach Sagiv & Shalom H. Schwartz, A New Look at National Culture:Illustr ative Applications to Role Stress
and Managerial Behavior, in Handbook of Organizational Culture and Climate. 417 (Neal M. Ashkanasy, Celeste
P.M.Wilderom, &Mark F. Peterson, Eds., 2000); Peter B. Smith, M. F. Peterson, & Shalom H. Schwartz, with 49
co-authors, Do Cultural Values Predict Managerial Behaviors? A 47 Nation Study, 33 J. Cross-Cultural Psychol.
188 (2002); Shalom H. Schwartz, Mapping and Interpreting Cultural Differences around theWorld, in Comparing
Cultures (Henk Vinken, Joseph Soeters, & Peter Ester, Eds., 2004).
20 Pierre Bourdieu, Outline of a Theory of Practice (1972); Hazel R. Markus & Shinobu Kitayama, A Collective
Fear of the Collective:Implications for Selves and Theories of Selves, 20 Personality & Social Psychol. Bull. 568
(1994).
21 Robin M.Williams, American Society:A Sociological Interpretation (3rd ed., 1970). This definition is similar
to that adopted in studies of the effects of societal development (e.g. Culture Matters: How Values Shape Human
Progress (Lawrence E. Harrison&Samuel P. Huntington, Eds., 2000)) and widespread in cross-cultural psychology
(e.g. Handbook of Cross-Cultural Psychology (John W. Berry, Marshall H. Segall, & Cigdem Kagitcibasi, Eds.,
2nd ed., 1997)).
234 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
education, economic, political, religious systems) function, their goals and modes of
operation. Social actors (e.g. organizational leaders, policy-makers, individual persons)
draw on these cultural value emphases to select actions, evaluate people and events, and
explain or justify their actions and evaluations.22
Many different value emphases could be chosen to characterize cultures. Rather than
using arbitrary intuitions to select values relevant to corporate governance, we utilize the
key dimensions of culture identified in the literature of cross-cultural psychology, the
discipline that specializes in cross-national comparisons of culture. A common postulate
in cross-cultural psychology is that all societies confront similar basic issues or problems
when they come to regulate human activity.23 The key dimensions of culture are derived
from these issues, because the preferred ways of dealing with them are expressed in
different societal value emphases. It is thus possible to characterize the culture of different
societies by measuring the prevailing value emphases on these key dimensions. This yields
unique cultural profiles.
Schwartz has provided the first set of cultural value dimensions we use.24 He derived#p#分頁標題#e#
three bipolar, cultural value dimensions from three basic issues that confront all societies. In
coping with these issues, societies exhibit greater or lesser emphasis on the values at one or
the other pole of each dimension. Analysis of the dimensions yields seven value orientations
on which cultures can be compared. Table 1A briefly describes these basic societal issues,
the three value dimensions dealing with them, and the polar orientations. Hofstede advanced
another pioneering dimensional framework for characterizing culture.25 Table 1B sets forth
Hofstede’s value dimensions and the basic societal problems they address. Here too, each
dimension describes a range of possible stances between two polar extremes.
2.3. Cultural value orientations and investor rights
Corporate governance is the framework that defines the division of wealth and power in
the corporation. Legal rules that shape this division are scattered in various parts of countries’
laws, including specific corporate laws, general commercial codes, bankruptcy codes,
financial institutions regulation, etc.26 Voting rights are a critical aspect of corporate governance,
however, because they define the extent to which shareholders can exert power over
corporate affairs. Voting rights are the essential characteristic of equity, valuable primarily
22 Clyde Kluckhohn, Value and Value Orientations in the Theory of Action, in Toward a General Theory of
Action (Talcott Parsons & Edward Shils, Eds., 1951); Geert H. Hofstede, Culture’s Consequences: International
Differences inWork-Related Values (1980); Shalom H. Schwartz, Cultural Value Differences:Some Implications
for Work, 48 Appl’d Psychol. Int’l Rev. 23 (1999).
23 See Florence R. Kluckhohn & Fred L. Strodtbeck, Variations in Value Orientations (1961).
24 Shalom H. Schwartz, Cultural Dimensions of Values:T owards an Understanding of National Differences,
in Individualism and Collectivism:Theor etical and Methodological Issues 85 (Uichol Kim et al., Eds., 1994);
Schwartz, supra note 19; Schwartz, supra note 22.
25 Geert H. Hofstede, Culture’s Consequences:International Differences in Work-Related Values (1980) (hereinafter
“Culture’s Consequences 1980”); Geert Hofstede, Culture’s Consequences:Comparing Values, Behaviors,
Institutions, and Organizations Across Nations (2nd ed., 2001) (hereinafter “Culture’s Consequences”); Geert H.
Hofstede, Cultures and Organizations:Softwar e of the Mind (1991) (hereinafter “Software of the Mind”).
26 See Mark J. Roe, Strong Managers,Weak Owners:The Political Roots of American Corporate Finance (1994).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 235
Table 1A
The Schwartz cultural value dimensions
Embeddedness/Autonomy#p#分頁標題#e#
Concerns the desirable relationship between the individual and the group. Embeddedness represents a cultural
emphasis on maintenance of the status quo, propriety, and restraint of actions or inclinations that might disrupt
group solidarity or the traditional order. Autonomy describes cultures in which the person is viewed as an
autonomous, bounded entity who finds meaning in his or her own uniqueness. Intellectual Autonomy refers to
a cultural emphasis on the desirability of individuals independently pursuing their own ideas and intellectual
directions; Affective Autonomy to a cultural emphasis on the desirability of individuals independently pursuing
affectively positive experience.
Hierarchy/Egalitarianism
Concerns guaranteeing responsible behavior that will preserve the social fabric. Hierarchy refers to a cultural
emphasis on obeying role obligations within a legitimately unequal distribution of power, roles, and resources.
Egalitarianism refers to an emphasis on transcendence of selfish interests in favor of voluntary commitment to
promoting the welfare of others.
Mastery/Harmony
Concerns the relation of humankind to the natural and social world. Mastery refers to a cultural emphasis on
getting ahead through active self-assertion. Harmony refers to an emphasis on fitting harmoniously into the
social and natural environment.
Table 1B
The Hofstede cultural value dimensions
Individualism/Collectivism
Valuing loosely knit social relations in which individuals are expected to care only for themselves and their
immediate families versus tightly knit relations in which they can expect their wider in-group (e.g. extended
family, clan) to look after them in exchange for unquestioning loyalty.
Power Distance
Accepting an unequal distribution of power in institutions as legitimate or illegitimate.
Uncertainty Avoidance
Feeling uncomfortable or comfortable with uncertainty and ambiguity, and therefore, valuing or devaluing beliefs
and institutions that provide certainty and conformity.
Masculinity/Femininity
Valuing achievement, heroism, assertiveness, and material success versus relationships, modesty, caring for the
weak, and interpersonal harmony.
for large minority shareholders.27 These shareholders can exploit voting rights directly, in
the general meeting, or indirectly, as a credible threat to stand up for their rights in court.
We focus on litigation over voting rights as a means for dealing with conflicting economic
interests in the corporation.28 Countries may differ in the extent to which they direct
investors to resolve their grievances in court. If, as Shleifer proposes, what matters for
investor protection is the practice of justice and the ‘structure’ of society, then these dif-
27 Andrei Shleifer & Robert W. Vishny, A Survey of Corporate Governance, 52 J. Fin. 737, 751–55 (1997).#p#分頁標題#e#
28 We do not view voting rights as an apparatus of corporate democracy, notwithstanding the use of the same
label, since this is not a salient feature of shareholding in most countries.
236 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
ferences should be compatible with the way a society conducts itself in general, namely
with cultural orientations. A similar analysis applies to legal rights granted to creditors,
which become crucial during insolvency. The bankruptcy setting is highly confrontational
as various constituencies try to take as much as they can from a limited pie. The element of
confrontation implies that hypotheses about cultural values and enforcing creditors’ rights
should be similar to the ones advanced with regard to shareholders’ rights.
Our first cultural hypothesis is that greater reliance on concrete legal rules enforceable
in the courts is stronger in nations high on the Schwartz cultural orientation of Mastery and
low on his Harmony orientation. These cultural emphases are compatible with giving power
to investors and encouraging them to stand up and fight for their rights. Cultural Mastery
emphasizes assertiveness, venturing, and active determination of one’s destiny. In contrast,
cultural Harmony opposes head-on confrontation, so it should discourage embodying economic
interests in strict legal form and zealously enforcing them in court.
Our second cultural hypothesis is that investor legal rights are stronger in nations high on
the Hofstede Individualism dimension and low on his Uncertainty Avoidance dimension.
Individualism (versus Collectivism) legitimizes the vigorous pursuit of personal interests
rather than deference to others’ decisions and interests. Uncertainty Avoidance affects the
way power in organizations is exercised. High Uncertainty Avoidance is consistent with
giving power to authorities who control uncertainty and with perceiving conflict in the
corporation as unnatural.29 Low Uncertainty Avoidance is compatible with readiness on the
part of corporate constituencies to challenge one another—in general meetings, in public
media, and in the courts—with indeterminate outcomes.
3. Data
3.1. Cultural variables
3.1.1. The Schwartz data
Respondents from every inhabited continent completed a value survey anonymously in
their native language.30 They rated the importance of 56 single values as “guiding principles
in MY life.” Each value was followed in parentheses by a short explanatory phrase
(e.g. WEALTH [material possessions, money]). Responses ranged from 7 (of supreme
importance) to 3 (important) to 0 (not important) to −1 (opposed to my values). Separate
multidimensional scaling analyses of the 56 values within each of the different nations
established that 45 of the values have near-equivalent meaning across cultures. Only these#p#分頁標題#e#
45 values were, therefore, included in the analyses for testing cultural dimensions. A Similarity
Structure Analysis (SSA) of these values across nations supports the use of the seven
value orientations to represent national cultures.31
29 Hofstede, Culture’s Consequences, supra note 25, at 133, 166.
30 The survey is described in Schwartz, supra note 22.
31 The SSA was performed on data from over 35,000 respondents from 122 samples in 49 nations, gathered
between 1988 and 1993. See Schwartz, supra note 22. On the methods used see Ingwer Borg & James C. Lingoes,
Multidimensional Similarity Structure Analysis (1987); Louis Guttman, A General Nonmetric Technique for
Finding The Smallest Coordinate Space for a Configuration of Points, 33 Psychometrica 469 (1968).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 237
Data for comparing nations might ideally be obtained from representative national samples.
But inferences about national culture from such samples require caution. Differences
in the demographic compositions of national populations (e.g. distributions of age, education,
occupation) affect average value priorities. Consequently, even when comparing the
values of representative national samples, it would be necessary to control for demographic
differences between nations before we could confidently ascribe observed differences in
value priorities to national culture alone. Moreover, many nations contain more than one
sub-cultural group, so a single characterization based on a representative national sample
is still misleading.
The approach taken instead was to obtain samples matched on critical characteristics,
largely from the dominant cultural group in each nation.32 The focal type of sample was
urban school teachers who teach the full range of subjects in grades 3–12 of the most
commontype of school system.Nosingle occupational group represents a culture, but school
teachers may have a number of advantages for characterizing national value priorities. As
a group, they play an explicit role in value socialization, they are presumably key carriers
of culture, and they probably reflect the mid-range of prevailing value priorities in most
societies. By focusing on this single matched group, it was possible to obtain a relatively
pure representation of national differences in value priorities, net of the influences of other
national differences.33
To compute the mean importance of a value orientation in a nation, the importance that
members of the sample from that nation attributed to the set of values that represent that
orientation was averaged. For cross-national comparisons, sample differences in scale use
were eliminated by centering the mean importance of all seven orientations within each
sample around the approximate international mean. The meaningfulness and validity of#p#分頁標題#e#
such subjective, societal-level measures as explanatory variables is supported by associations
between national scores on them and a wide variety of social, economic, political,
demographic, and other characteristics of nations.34
Some of our analyses group nations into regions based on the similarity of their cultural
profiles. These regions emerged when the similarities and differences among nations
on the seven cultural orientations were represented graphically.35 This representation
revealed six distinct groups of nations: English-speaking, West European, East European,
Far Eastern, Latin American, and African. Each cultural region exhibited a distinctive
32 In line with the vast majority of cross-cultural studies, this study compares nations as cultural groups. Where
national boundaries encompassed heterogeneous groups with separate distinctive cultures, their data refer to the
culture of the dominant group.
33 To test the robustness of conclusions from the teacher samples, Schwartz, supra note 19, performed
parallel analyses with data from samples of college students, from a wide variety of majors, in each of
40 nations. The results supported the value dimensions and the locations of the nations on these dimensions.
34 See Marianne Bertrand & Sendhil Mullainathan, Do People Mean What They Say? Implications for Subjective
Survey Data, 91 Am. Econ. Rev. 67 (2001); Hofstede, Culture’s Consequences, supra note 25; Schwartz, supra
note 19.
35 See Yair Goldreich & Adi Raveh, Coplot Display Technique as an Aid to Climatic Classification, 25 Geographical
Analysis 337 (1993) for the statistical technique employed.
238 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
value profile that differed significantly from all other regions on at least one dimension
(see Appendix B).
3.1.2. The Hofstede data
Hofstede’s study originated in an audit of company morale among IBM employees
across the world around 1968 and around 1972. Factor analysis of country mean scores
in 50 countries and three regions produced the four dimensions described above. Hofstede
noted the criticism that “IBMers are very special people, not at all representative
for our country.” He argued, however, that the crucial requirement is that samples be
well matched across countries, not that they be representative. He asserted that comparing
IBM subsidiaries shows national culture differences with unusual clarity because
the samples are so homogeneous in terms of employer, kind of work, and education
level.36
Hofstede used hierarchical cluster analysis to divide countries into culture groups. To
obtain meaningful groupings, however, he had to use his own his judgment to join several
subgroups. This yielded the following culture regions: Anglo, Germanic, Nordic, More#p#分頁標題#e#
Developed Latin, Less Developed Latin, More Developed Asian (consisting only of Japan),
Less Developed Asian, and Near Eastern (see Appendix B).37
3.2. Legal variables
3.2.1. The LLSV data
Our main set of legal variables is LLSV’s dataset covering 49 countries (but no socialist
or transition economies). It includes only countries with at least five domestic, non-financial,
publicly traded firms with no government ownership in 1993.LLSVconstructed two indices,
described below, intending them to capture the degree to which national laws protect outside
investors from insiders.
3.2.2. Anti-director rights
LLSV’s anti-director rights index (ATD) is based on six specific legal rights granted to
shareholders. Scores indicate the number of rights that the company law or commercial
code in a country recognizes. In addition to this overall index, we consider two aspects
that ATD intermixes: control (voting) rights and cash flow (remedial) rights.38 In order to
separate these two aspects, we split ATD into sub-indices. The first sub-index, labeled ATDVote,
includes the four voting-related rights. The other sub-index, labeled ATD-Remedial,
includes the two remedial rights. The two sub-indices correlate positively but moderately
(r = .29).
36 Hofstede, Software of the Mind, supra note 25, at 252; Culture’s Consequences, supra note 25, at 73.
37 Hofstede, Culture’s Consequences 1980, supra note 25, at 333–336. In the 2001 edition, supra note 25, Hofstede
repeats the first step of hierarchical cluster analysis with an extended group of 50 nations and three regions but
fails to repeat the second step of grouping. We, therefore, opted to adhere to Hofstede’s original grouping except
for consolidating the two Asian groups into one.
38 Nonetheless, the ATD index does not include the most important protection against directors and control
persons—namely anti-self-dealing rules. Devising a numerical representation for this aspect is a difficult challenge.
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 239
3.2.3. Creditor rights
LLSV specified four creditor rights in either reorganization or liquidation laws, as these
may substitute for one another. Scores on the creditor rights index (CRD) indicate the
number of rights recognized in the law of a country.39
3.2.4. Legal families
LLSV’s classification of countries into legal families draws on Ren´e David’s taxonomy
of legal systems according to their origin in common law, civil law (with a breakdown into
French, German, and Scandinavian laws), and several other families of law.40
Several researchers, including LLSV have used the indices and the legal family classification
as independent, explanatory variables.41 Only recently have researchers begun#p#分頁標題#e#
to use these indices as dependent variables, as part of a larger trend that seeks to identify
determinants of governance in general. Nonetheless, these indices are not free from criticism.
LLSV acknowledge several problems with their indices: they do not cover merger
and takeover rules, they cover disclosure rules only partially, and they do not cover stock
exchange rules or regulations of financial institutions.42 One might also object to LLSV’s
choice of index components, to the assignment of equal weight to each, and to LLSV’s
reading of particular national laws. Finally, the relevance of legal rules hinges on a general
infrastructure of legality, or rule of law, in each country and especially in transition
economies.43
These limitations imply that the LLSV indices may not reliably capture countries’ levels
of investor protection. These measures nonetheless provide a unique operationalization
of statutory rules in the field of corporate governance for a particularly large number of
nations. These indices can be viewed narrowly as specifying sets of circumstances over
which disputing parties can resort to litigation. Hence, the LLSV indices essentially reflect
the extent to which countries differ in directing investors to resolve their grievances in court.
Accordingly, we hypothesize that investor rights scores for nations correlate positively
with nations’ Mastery scores and negatively with their Harmony scores in the Schwartz
dataset, and positively with Individualism and negatively with Uncertainty Avoidance in
the Hofstede dataset.
39 LLSV, supra note 1, at 1134, state that the CRD index is based on the perspective of senior secured creditors.
Of the components that comprise CRD, however, only two relate to secured creditors while the others apply to
creditors in general.
40 Ren´e David & John Brierly, Major Legal Systems in theWorld Today (1985); 2 Int’l Encyclopedia of Comp. L.
(Ren´e David, Ed., 1972). David’s taxonomy dates from the mid-1960s and has since been criticized, inter alia, as
being “Euro-American-centric”. Ugo Mattei, Three Patterns of Law:T axonomy and Change in the World’s Legal
Systems, 45 Am. J. Comp. L. 5, 10 (1997). Nonetheless, leading authorities on comparative law generally adhere
to this taxonomy. Konrad Zweigert & Hein K¨otz, An Introduction to Comparative Law 63–75 (Tony Weir trans.
3rd ed., 1992); Mary Ann Glendon et al., Comparative Legal Traditions (2nd ed., 1994); Rudolph B. Schlesinger
et al., Comparative Law:Cases-T ext-Materials 1 (5th ed., 1988).
41 See, for example La Porta et al., Legal Determinants, supra note 3, and works cited supra note 2.
42 La Porta et al., supra note 1.
43 Pistor, Raiser, & Gelfer, supra note 7; see also Pistor, supra note 7. It should be noted that LLSV do not neglect#p#分頁標題#e#
this aspect, as they include a measure for the rule of law in their analysis.
240 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
3.2.5. The Djankov et al. data
To corroborate the main analysis, we take advantage of recently available data to examine
relations of national culture to the degree of formalism in the working of court systems
in countries. Djankov, La Porta, Lopez-de-Silanes, and Shleifer capture this feature by
constructing formalism indices for two basic legal procedures: (1) eviction of a residential
tenant in default of rent payments and (2) collection of a check returned for non-payment.44
The legal formalism indices aggregate several rules of procedure, each of which is coded
for the degree of formalism it reflects.
4. Results
4.1. Cultural value orientations and investor rights
Table 2 presents the Pearson correlations between national scores on the cultural value
dimensions and on ATD and ATD-Vote. These variables correlate negatively with Harmony
and Uncertainty Avoidance, as hypothesized.We fail to find the hypothesized relations with
Mastery, however.ATD-Remedial does not correlate with any cultural dimension from either
cultural theory—a result not altogether surprising in light of the very narrow scope of this
index. Thus, we conclude that the voting rights regimes in countries fit with certain cultural
emphases with which they were expected to be compatible, but the (coded) remedial rights
are not related to the surrounding national culture.
Below, we report an analysis of legal families and cultural regions suggesting that Asian
countries with a common law heritage are an outlier sub-group.We, therefore, recomputed
the correlations here after excluding these countries. Table 2 reports the results. The negative
correlations with Harmony and Uncertainty Avoidance remain significant. In the Hofstede
data, Individualism also correlates positively with ATD and ATD-Vote, as hypothesized,
and Power Distance correlates negatively with ATD. The latter finding is consistent with
http://www.400190.com/dissertation_writing/Hofstede’s argument that high Power Distance connotes a negative view of power and
wealth,45 the constitutive elements of corporate governance.
Table 2 further reports Pearson correlations of creditor rights for the full sample of nations
and for a sample from which common-lawAsian countries are excluded. The case for excluding
these countries is at least as strong with regard to creditor rights as for ATD (see below).
Indeed, we advocate considering only the sample excluding these countries, pending further
research on the actual level of creditors’ rights in them. Excluding these countries, the results
for the Schwartz model confirm the hypothesis of a negative correlation with Harmony. The#p#分頁標題#e#
correlation with UncertaintyAvoidance is also negative (at 7%).Aweak negative correlation
between creditors’ rights and Power Distance is consistent with Hofstede’s argument.
4.2. Robustness, consistency, and dynamic aspects
The above results suggest that the content of statutory laws governing the way to reconcile
competing interests in the corporation partly reflects the surrounding national culture.
44 Simeon Djankov et al., Courts:The Lex Mundi Project, 118 Q. J. Econ. 453 (2003).
45 See Hofstede, Culture’s Consequences 1980, supra note 25, at 92.
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 241
Table 2
Pearson correlations between nations’ standings on cultural value dimensions and on investors’ rights
Model Full sample Excluding asian common law
ATD ATD-Vote CRD ATD ATD-Vote CRD
Schwartz
Harmony −.338** −.353** −.360** −.338** −.402** −.327**
Embeddedness .168 .162 .308** .056 .155 .161
Hierarchy .206 .271 414*** .004 .183 .263
Mastery .022 −.059 .102 −.036 −.107 .031
Affective Autonomy −.090 −.007 −.107 .021 .035 .053
Intellectual Autonomy −.306** −.306** −.235 −.172 −.278 −.036
Egalitarianism −.163 −.180 −.369** −.009 −.110 −.198
N 35 35 34 31 31 30
Hofstede
Uncertainty Avoidance −.394*** −.403*** .364*** −.309** −.405*** −.248
Individualism .126 .226 −.111 .288** .321** .040
Power Distance −.092 −.166 −.058 −.277** −.261 −.313**
Masculinity −.015 −.063 −.009 −.039 −.084 −.040
N 43 43 43 38 38 37
Significance levels are one-tailed.
** Significant at 5%.
*** Significant at 1%.
242 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
The content of these laws is compatible with the degree to which the national culture
tolerates or even encourages confrontational processes with indeterminate outcomes. The
present framework allows us rigorously to investigate some general propositions concerning
these relations. This section first examines whether the correlations noted above may be
spurious. In light of the centrality of litigation to our hypotheses, we then look briefly
at the relations between cultural orientations and the degree of formalism in national
court systems. We then address the causal dynamics that may underlie the observed
associations.
4.2.1. Relations with other national characteristics
One might suspect that nations’ cultural orientations, and the scores that represent them,
may serve as proxies for the effects of major socio-economic factors. Such factors must be#p#分頁標題#e#
of sufficient scope and magnitude in order to be considered as “competing” with culture.
We, therefore, consider legality (i.e. “law and order” or “rule of law”), the level of economic
development, and religion. Some degree of mutual influence among these factors and
culture is likely. Nonetheless, it is possible to examine the robustness of the relations of cultural
orientations (here, Harmony) to investor rights by controlling possible effects of such
variables.
Legality may play a role in the actual working of corporate governance systems.46
Higher compliance with legal rules is likely to accompany higher scores on the LLSV
indices because both are associated with greater reliance on the court system for conflict
management. There is no point in writing more rights into the law if people disregard the
law and do not settle their disputes in court.
In theory, the larger size of capital markets in developed economies might induce pressure
on lawmakers to provide investor protection to facilitate theworking of markets. LLSV,
however, find no evidence that investor rights are a consequence of per capita income.47
Moreover, researchers recently excluded GDP per capita from their analyses due to the
endogeneity problems it may entail.48 These limitations notwithstanding, it is still worth
examining whether cultural Harmony predicts the LLSV indices when economic development
is taken into account.
Lastly, we test whether the association of a country’s dominant religion with investor
rights might account for the latter’s association with cultural Harmony. Stulz and
Williamson reported a positive correlation of Protestantism (versus Catholicism) with
creditor rights but not with shareholder rights. Treisman noted a negative correlation
between Protestantism and perceived corruption.49 It is not clear, however, which aspects
of these religions underlie these findings.
Table 3 presents results of regressing ATD, ATD-Vote, and CRD scores on Harmony
and, in turn, the level of legality, economic development, and Protestantism as a dominant
46 See Pistor, Raiser, & Gelfer, supra note 7; Berkowitz, Pistor, & Richard, supra note 7.
47 La Porta et al., supra note 1, at 1139.
48 Beck et al., supra note 11, note that “GDP per capita is endogenous, which causes estimation problems, as
shown by Acemoglu et al. [below note 54].”
49 Daniel Treisman, The Causes of Corruption:A Cross-National Study, 76 J. Pub. Econ. 399 (2000).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 243
Table 3
Regressions of investors’ rights on cultural harmony and other national characteristics
Dependent variable Harmony Rule of law GDP/cap Protestant R2 F N
ATD −39** (2.56) .38** (2.48) .26 5.53*** 35#p#分頁標題#e#
ATD-Vote −.40** (2.57) 33** (2.18) .23 4.86** 35
CRD −.37** (2.21) .09 (.54) .14 2.47* 34
ATD −.35** (2.08) .11 (.63) .13 2.22 34
ATD-Vote −.37** (2.19) .11 (.68) .14 2.46* 34
CRD −.34** (2.00) −.16 (.94) .15 2.74* 33
ATD 39** (2.44) .16 (.97) .17 3.34** 35
ATD-Vote −.48*** (3.25) 32** (2.19) .32 7.35*** 35
CRD −.32* (1.87) .03 (.19) .10 1.76 34
Standardized beta coefficients. t-statistic absolute values are reported in parentheses.
* Significant at 10%.
** Significant at 5%.
*** Significant at 1%.
244 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
Table 4
Pearson correlations between nations’ standings on cultural value dimensions and on legal procedure formalism
Model Evicting non-paying tenants Collecting returned checks
Schwartz
Harmony 346*** .393***
Embeddedness .053 .088
Hierarchy −.211 −.221
Mastery −.111 −.157
Affective Autonomy −.246** −.267**
Intellectual Autonomy .102 .086
Egalitarianism .085 .063
N 49 49
Hofstede
Uncertainty Avoidance .569*** .590***
Individualism −.598*** −.527***
Power Distance .484*** .396***
Masculinity −.052 .074
N 48 48
Significance levels are one-tailed.
** Significant at 5%.
*** Significant at 1%.
religion.50 Consistent with the cultural hypotheses, Harmony robustly predicts all three
measures of corporate governance in the presence of these controls. These potential competitors
with culture failed to account for the relations of cultural Harmony to the LLSV
indices across nations. This strengthens our confidence in the robustness of culture as a
predictor of corporate governance-related laws.
4.2.2. Culture and legal formalism
If cultural value priorities indeed underlie national laws, then national cultural profiles
should relate consistently to laws that govern a variety of subject matters. Particularly
interesting are laws pertaining to litigation, because the LLSV measures relate so closely to
it. Chase argues that the tightly controlled legal procedure in Germany serves the cultural
value of UncertaintyAvoidance. This contrasts with the volatile and relatively unpredictable
quality of the lawyer-dominated American trial.51 We, therefore, expect lower Harmony
and Uncertainty Avoidance to correlate with open-ended, informal litigation processes as
the latter are represented by Djankov et al. (inversely to LLSV’s indices).
Table 4 presents correlations between the formalism indices and the cultural dimensions.
Confirming the hypotheses, both formalism indices—for tenant eviction and check
collection—exhibit the expected positive correlations with Harmony and Uncertainty#p#分頁標題#e#
50 The Rule of Law index, from Daniel Kaufmann, Aart Kraay, & Pablo Zoido-Lobat´on, Governance Matters
(Working Paper 1999), represents the level of perceived legality in countries. Economic development is represented
by the log of GDP per capita in 1993 (the year for which LLSV’s indices were constructed), taken from theWorld
Bank’s World Development Indicators.
51 Oscar G. Chase, Legal Processes and National Culture, 5 Cardozo J. Int’l&Comp. L. 1 (1997). See also Mirjan
Damaska, The Faces of Justice and State Authority:A Comparative Approach to the Legal Process (1974–1975).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 245
Avoidance. They also correlate negatively with Individualism—again, inversely to theLLSV
ATD index and in line with our hypothesis. Finally, formalism correlates positively with
Hofstede’s Power Distance in line with Damaska’s and Chase’s views and in contrast with
the result for ATD. Replicating the regression analyses reported in Table 3 for the formalism
indices yielded consistent results— namely robust positive coefficients for Harmony in the
presence of the control variables (not shown). These findings support the notion that laws
on the books in two different fields—corporate governance and civil procedure—reflect
similar cultural orientations for particular forms of conflict management.
4.2.3. Reverse causality and the role of British heritage
Culture can impact law-making in two different ways: first, cultural values may motivate
lawmakers and interest groups to prefer certain legal arrangements to others; second, culture
may constrain reforms that are not compatible with prevailing value priorities. Culture
has thus been called “the mother of all path dependencies.”52 One might still hesitate to
draw policy implications from the present evidence. If cultural emphases adjust rapidly to
changes in the legal environment, culture may have little relevance to policy formation.
Before addressing this issue formally, a preliminary note is in place.
In order for statutory provisions to affect widespread beliefs about what is right and desirable
(culture), they must either be central and salient or, alternatively, they must influence
people’s day-to-day life even if they are unaware of them. Given the way we operationalize
culture, only a negligible influence from LLSV’s indices to cultural orientations seems
plausible. The data on culture come from respondents unlikely to be familiar with these
legal rules and whose daily practices are unlikely to be affected by them. Hence, there is
little reason to fear reverse causality from the legal to the cultural variables in this study.
Addressing reverse causality between variables commonly relies on using an instrumental#p#分頁標題#e#
variable in two-stage least squares (2SLS) regressions. An ideal instrument would
correlate with the predicting variable but not with the predicted one. Alternatively, laggedperiod
values may be used as instruments as they are not susceptible to feedback from the
predicted variable. The now common practice of using legal family as an instrument in
predicting economic outcomes reflects the latter approach, because many countries’ legal
origins were determined historically by colonial rulers.53 However, the legal family classification
cannot be considered exogenous when studying legal rules because it is a concise
representation of just such rules. This classification also does not distinguish between what
was imported and its evolution into current law.
Many of the variables commonly treated as exogenous are inadequate for controlling
reverse causality in relations between cultural and legal variables. They may have had
reciprocal causal relations with law and/or culture either directly or through economic
factors. The menu of truly exogenous variables is thus limited largely to bio-geographical
factors and to major historical events. Studies that take this approach point to the long-lasting
52 See Amir N. Licht, The Mother of All Path Dependencies:T oward a Cross-Cultural Theory of Corporate
Governance Systems, 26 Delaware J. Corp. L. 147 (2001).
53 See, e.g. Rafael La Porta, Florencio Lopez De Silanes, & Andrei Shleifer, What Works in Securities Laws? J.
Fin. (forthcoming).
246 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
Table 5
Two-stage least squares regressions of investors’ rights on cultural harmony across nations, using a heritage of
British rule as an instrumental variable
Dependent variable Harmony Intercept R2 F N
ATD −3.25*** (−3.14) 16.95*** (3.86) .23 9.87*** 35
ATD-Vote −20.91*** (−2.94) 10.91*** (3.61) .21 8.62*** 35
CRD −2.23** (−2.50) 11.54*** (3.04) .16 6.23*** 34
t-statistic values are reported in parentheses.
** Significant at 5%.
*** Significant at 1%.
implications of episodes of British rule for social and economic outcomes.54 Being ruled
by Britain apparently affected many outcomes, including the law, but as an historical fact, it
could not be affected by them. We, therefore, use episodes of British rule as an instrument
to control for reverse causality in our analyses.
Table 5 reports 2SLS regressions of ATD, ATD-Vote, and CRD scores on Harmony.
The instrumental variable is based on Treisman’s dummy variable for whether a country
had experienced a British rule of any kind.55 These regressions assess the link between
Level 1 and Level 2 institutions in Williamson’s model while isolating potential feedback.
Consistent with the results in Table 2, Harmony has a significant negative coefficient in all#p#分頁標題#e#
the regressions.
Countries in the English-speaking cultural region score lower on Harmony and higher
on Mastery than other countries.56 Hence, the above results might be due to the subgroup
of English-speaking countries among those with a history of British rule. We, therefore,
checked whether non-English-speaking countries with a history of British rule also score
lower on Harmony than other countries. This was indeed the case. The Harmony scores of
non-English-speaking countries with a history of British rule did not differ from those of
English-speaking countries, and they were significantly (at 1%) lower than the Harmony
scores of other countries.
Explanations for the strong emphasis on Mastery and de-emphasis of Harmony in
English-speaking former British colonies call upon the frontier experience, active development,
and the centrality of entrepreneurial business. This applies to the settlement colonies
including Australia, Canada,NewZealand, and the United States. In non-settlement colonies
too (and not without resistance), British rulers engaged in physical development, including
land-use planning and major transportation infrastructures. They also made institutional
reforms including, though not limited to, legal transplantation. Today, these countries share
an enhanced cultural inclination toward venturing and confronting the unknown that, among
other things, supports reliance on court litigation. The results are consistent with the idea
that the experience of British rule left an imprint on the cultures of these countries in tandem
with its influence on their written laws, over and above any reverse causality.
54 See, e.g. Daron Acemoglu, Simon Johnson, & James A. Robinson, The Colonial Origin of Comparative
Development:An Empirical Investigation, 91 Am. Econ. Rev. 1369 (2001).
55 Countries not coded by Treisman were coded according to the CIA’s World Factbook.
56 Shalom H. Schwartz & Maria Ros, Values in the West:A Theoretical and Empirical Challenge to the
Individualism-Collectivism Cultural Dimension, 1 World Psychol. 93 (1995).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 247
4.3. Cultural regions and legal families
The cornerstone of LLSV’s legal approach to corporate governance is the classification
of legal regimes into legal families. Common law regimes rank first in the legal protection
afforded to external investors, whereas the French civil law group ranks last and the
German and Scandinavian groups rank in between. Yet, countries may belong to particular
legal families for many reasons, ranging from voluntary adoption of a legal system
to forced imposition by colonial powers. Countries that belong to the same family (save
Scandinavian countries) are spread across the globe. There is no a priori reason, therefore,#p#分頁標題#e#
to expect an identity between countries’ cultural region and their legal family affiliations.
This section argues that without considering countries’ cultural region affiliation, the legal
family classification alone may yield inaccurate depictions of global patterns in corporate
governance regimes.
We first cross-tabulated the countries in each legal family with each cultural region
for the Schwartz and Hofstede cultural regions.57 The cross-tabulation revealed that the
two taxonomies overlap only partially. Countries in the Schwartz and Hofstede Englishspeaking/
Anglo cultural region stand out as legally homogenous, with a common law
system. Countries in their Latin American/Less Developed Latin region are also homogeneous
with a French civil law origin. The remaining cultural regions each contain countries
from a variety of legal families. Moreover, countries with the same legal pedigree often
belong to different cultural regions. The Far Eastern/Asian region is particularly noteworthy.
It includes relatively equal proportions of countries with common law or civil law
origins.
We use the Goodman–Kruskal lambda statistic LB to assess the relative reduction in error
in predicting legal family when the cultural region is known.58 LB values are .44 (Schwartz
regions) and .62 (Hofstede regions), both statistically significant at 1%. Knowing the cultural
region of a country helps to predict its legal family affiliation, though prediction is far from
perfect (Lb = 1).
The key question is whether the location of a country in a cultural region predicts its
corporate governance regime. Significant differences among cultural regions would suggest
that differences in corporate governance regimes may be associated not only with historical
legal pedigrees but also with concurrent, general societal orientations. Table 6 reports mean
levels of ATD and CRD for each cultural region identified by Schwartz and Hofstede.59
t-tests of the differences between the mean levels of ATD and CRD for each pair of regions
reveal whether regions that differ in their cultural profiles differ systematically in legal rules.
4.3.1. Anti-director rights
The Schwartz cultural region that stands out as different from the others with a high antidirector
rights score is the English-speaking region. For Hofstede’s classification, the Anglo
57 Not shown. See Appendix A. The category “Other” includes legal systems that could not be classified into
LLSV’s categories and several East European countries with a general Civil Law origin.
58 See Sidney N. Siegel & John Castellan, Jr., Nonparametric Statistics for the Behavioral Sciences 298–299
(2nd ed., 1988).
59 We omitted regions for which we have too few observations.
248 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255#p#分頁標題#e#
Table 6
Mean ATD and CRD scores for each cultural region and t-tests for differences between regions
Variable Cultural region Mean English speaking Latin America Western Europe
A. Regions identified by Schwartz
ATD English-speaking 4.29
Latin America 2.80 1.99
Western Europe 2.46 4.31*** .53
Far East 3.75 1.13 1.23 2.89***
Far East-common law 4.50 .49 1.81 3.95***
Far East-civil law 3.00 2.64** .21 1.00
CRD English-speaking 2.14
Latin America 1.00 1.42
Western Europe 1.77 .70 1.49
Far East 2.88 .97 2.36** 2.14**
Far East-common law 3.75 2.09 5.75*** 4.04***
Far East-civil law 2.00 .15 1.10 .37
Variable Cultural region Mean Anglo Germanic Nordic MDL LDL
B. Regions identified by Hofstede
ATD Anglo 4.57
Germanic 2.00 6.39***
Nordic 2.80 4.50*** 1.44
More-developed Latin 2.50 3.17*** .57 .37
Less-developed Latin 2.50 3.90*** .69 .45 .00
Asian 3.55 2.09** 2.33** 1.24 1.50 1.79
Asian-common law 4.60 .09 5.73*** 4.03*** 2.71** 3.36***
Asian-civil law 2.67 5.05*** 1.27 .27 .22 .27
CRD Anglo 2.14
Germanic 2.75 .69
Nordic 2.00 .20 1.14
More-developed Latin 1.33 1.20 2.18 1.43
Less-developed Latin 1.29 1.08 1.64 .98 .07
Asian 3.00 1.32 .34 1.64 2.89** 2.61**
Asian-common law 3.80 2.42** 1.76 4.81*** 6.01*** 3.60***
Asian-civil law 2.33 .24 .49 .49 1.54 1.31
t-statistic values are reported in absolute values.
** Significant at 5%.
*** Significant at 1%.
region also emerges as distinct in its higher level of ATD scores. Cultural regions in which
the dominant legal origins are continental civil law codes (i.e. West European and Latin
American countries) do not differ from one another in their levels of anti-director rights.
Thus, with regard to English-speaking countries, legal uniqueness and cultural uniqueness
overlap and likely reinforce one another. Corporate governance rules in English-speaking
common law countries seem to be grounded in a deep cultural infrastructure.
Differences in ATD scores for the Far Eastern and Asian regions made us suspect that
the results may be misleading, because countries in this region have different legal origins.
We split this region into sub-sets of nations according to legal origin: common law and
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 249
civil law. Comparing the ATD means of Far Eastern nations with common law (n = 5 in the
Schwartz data) versus civil law origins (n = 5) suggests that legal origin dominates cultural
affiliation with regard to anti-director rights. The ATD means of the common law sub-set
of nations are significantly higher (at 1%) both for the nations in the Schwartz and in the
Hofstede studies.
Do the results indicate that Asian countries with a common law heritage protect minority#p#分頁標題#e#
shareholders better? We doubt it. The historical fact that many of these laws were first
enacted by colonial powers means that there was not necessarily a good fit between the
legal structure and the cultural foundation on which it had to stand. Elsewhere, we find that
perceived legality in the East Asian cultural region is significantly lower than in Englishspeaking
and West European countries.60 This finding is prima facie evidence that the law
on the books may play only a minor role in determining shareholder protection in practice
in East Asian countries. A decisive answer will require in-depth empirical research. In
the meantime, we advocate caution in judging East Asian shareholder protection solely
according to legal origin.
4.3.2. Creditor rights
Analysis of creditor rights across cultural regions also yields interesting insights. Apart
from the Far Eastern/Asian cultural region, there are virtually no differences among cultural
regions in mean levels of creditor protection, whether we classify nations according to the
Schwartz or to the Hofstede data. This finding may look puzzling because LLSV do report
that common law regimes score highest and French origin regimes score lowest on CRD.
A cultural analysis may resolve the puzzle and indeed changes the picture substantially. To
see why, consider the Far Eastern cultural region.
The mean CRD level in Far Eastern countries with a common law origin is the highest
among all cultural groups.61 Counter-intuitively, it is even higher than in English-speaking
countries (at 7%). In contrast, the mean CRD level in Far Eastern countries with a civil
law origin does not differ from that in English-speaking countries. These findings question
LLSV’s broad claim that common law regimes protect investors best. Legal rules in the
English-speaking (common law) region offer no more creditors rights than the civil law
systems in Western Europe, Latin America, and even in Asia. Members of the British
Commonwealth and the United States score low or medium on CRD. Most of the common
law countries that score high on CRD are the United Kingdom itself or countries formerly
under British rule in the Far Eastern or African cultural regions.62 It is the latter countries
that elevate the CRD score of the common law family, giving it first rank among legal
families.
60 Amir N. Licht, Chanan Goldschmidt, & Shalom H. Schwartz, Culture Rules:The Foundations of the Rule of
Law and Other Norms of Governance (Working Paper 2003).
61 t-statistic values for the difference in CRD between the two Far Eastern subgroups are 2.05 (significant at 9%)
and 2.53 (significant at 5%) for the Schwartz and Hofstede classifications, respectively.
62 The countries in LLSV’s common law family that received the highest CRD score (4) are: Hong Kong, India,
Israel, Kenya, Malaysia, Nigeria, Pakistan, Singapore, United Kingdom, and Zimbabwe. The following common#p#分頁標題#e#
law countries score only 1: Australia, Canada, Ireland, and the United States. La Porta et al., supra note 1, in
Table 1.
250 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
Different dynamic processes may have led to the patterns we observe. The laws in areas
formerly under British rule may reflect British rulers’ practice of enacting ordinances that
consolidated English statutory and judge-made law. Countries formerly under British rule
may thus amplify a unique feature of English law. On the other hand, the observed difference
could be illusory, reflecting LLSV’s focus only on statutory law. Case law in Englishspeaking
countries other than the United Kingdom might have complemented statutory
law with identical protections, such that these countries offer stronger investor protection.
Alternatively, creditor protection laws may have eroded over time in some English-speaking
countries.
5. Discussion
Legal reform is the primary vehicle in the hands of policy makers for peacefully inducing
social change. The limited scope of legal rules considered here notwithstanding, the
approach employed that yielded the present results suggests some general insights that may
inform policy- and law-making.
In terms of the theoretical framework presented in part II, the value dimension
approach provides a new means for rigorously addressing informal social institutions
located in Level 1 of Williamson’s model. We demonstrate how these deep-seated
institutions need not be “taken as a given” as customarily done. Instead, the value
dimension approach provides a theory-driven, universally validated operationalization of
fundamental societal orientations. This approach thus enables us to derive and empirically
test hypotheses about relations between institutions at Level 1 (culture) and Level 2
(legal rules).
The literature on causal relations among institutions and economic performance is in
flux. Contemporary proponents of the view that “law matters” do not claim that a simple
causal link runs from legal rules to economic outcomes. Likewise, the present evidence
that culture matters does not imply that law does not matter. The observed relations
between culture and legal rules support the proposition that, in the long term, formal
institutions should be consistent with the informal cultural environment. Regarding the
direction of causality, we find evidence that history (colonial rule) may have impacted both
levels.
More specifically, this article considers differences among nations regarding preferred
modes of regulating conflicting economic interests in corporations. The cultural orientation
of Harmony features as a significant factor in this regard.Yet, corporate governance involves
numerous other issues on which nations differ. Such issues include ownership structures,#p#分頁標題#e#
distribution of benefits among corporate constituencies, etc. In contexts like these, other
cultural dimensions may be relevant. The present framework can be applied to assess the
role of culture in such issues.
The foregoing discussion further shows that a comparative analysis of corporate governance
laws cannot rely on any single method of classification. Relying only on the “legal
approach”—as LLSV dubbed it—led them to conclude that “[i]n general, differences among
legal origins are best described by the proposition that some countries protect all investors
better than others, and not by the proposition that some countries protect shareholders
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 251
while other countries protect creditors.”63 The case of Far Eastern countries demonstrates
that combining classifications based on cultural dimensions and on legal families can yield
insights obscured by using one approach alone. The combined approach thus militates
against assessing legal regimes in isolation from their cultural environment.
Legal engineering intended to facilitate socio-economic development in countries has
roots in the 1960s “law and development” movement, which by 1975, had been declared
a failure.64 During the 1990s, the demise of communist regimes in Europe turned the
region into a real-life testing ground for legal and structural reforms aimed to establish
market economies. The idea that “law matters”—reinvigorated in large part by LLSV’s
work—motivated legal reforms that were expected to provide investors with an hospitable
environment. Measured by LLSV’s ATD index, the average level of shareholder rights rose
from substantially below the world average to well above it in just 6 years from 1992 to
1998, surpassing the three civil law families in the LLSV sample.65 Meanwhile, American
academics advised Russia on how to legislate an optimal corporate law that would rely
extensively on litigation over bright-line rules.66 As noted above, these reforms generally
yielded poor results.
As far as we can find, beyond passing comments, the law and economics literature has
paid scant attention to the cultural environment in East-Central Europe as a potential impediment
to reform. There is evidence that culture is such an impediment, however. Shortly
after the collapse of the communist regimes in 1989–1993, Schwartz gathered samples
in nine European countries that had endured communist rule.67 Compared with Western
Europe, former communist countries more strongly endorsed cultural Embeddedness and
Hierarchy. Such value emphases correlate robustly with low perceived legality.68 Stated
otherwise, these countries lacked the cultural foundation that is consistent with widespread,#p#分頁標題#e#
voluntary law-abidingness. Their relatively high average score on Harmony further suggests
that the populace may be inclined to avoid the court system. Engendering beneficial social
change through legal reform thus faces massive hurdles in such societies.
Aclaim is often made that culture adjusts to socio-economic conditions such as democratization
and free markets.69 While such a causal link is plausible,Williamson surmises that
“institutions at this level change very slowly—on the order of centuries and millennia.” Putnam
noted the centuries long impacts of culture in Italy on the emergent social institutions.70
The little evidence regarding historical trends in national culture confirms that cultural
change is very slow. Schwartz, Bardi, and Bianchi revisited several countries in East-Central
Europe 5–9 years after the collapse of communism. During that interval, major transforma-
63 See La Porta et al., supra note 4, at 8.
64 Kevin E. Davis & Michael J. Trebilcock, Legal Reforms and Development, 22 Third World Q. 21 (2001).
65 Pistor, Raiser, & Gelfer, supra note 7.
66 Bernard Black & Reinier Kraakman, A Self-Enforcing Model of Corporate Law, 109 Harv. L. Rev. 911 (1996).
67 Shalom H. Schwartz & Anat Bardi, Influences of Adaptation to Communist Rule on Value Priorities in Eastern
Europe, 18 Pol. Psychol. 385 (1997).
68 Licht, Goldschmidt, & Schwartz, supra note 60.
69 Timur Kuran, Cultural Obstacle to Economic Development:Not Necessarily Real, Often Transitory, in Culture
and Public Action (Vijayendra Rao & Michael Walton, Eds., 2004).
70 Robert Putnam, with Roberto Leonardi & Raffella Nanetti, Making Democracy Work:Civic Traditions in
Modern Italy (1993).
252 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
tions toward democracy and a market economy, including legal reform, took place. Overall,
these changes had no discernable effect on prevailing cultural orientations.71 The World
Values Survey gathered parallel data in a number of countries around the world several
times during the past 15 years. Although values did change somewhat, cultural differences
between nations remained quite stable.72
Thus, absent severe exogenous shocks such as colonial occupation, cultural value adaptation
takes place slowly, mostly in response to change in real life circumstances rather than
to formal reform or indoctrination. Prevailing cultural values generate a path dependence
dynamic that may slow down or block adaptation to change. Both from a narrow corporate
governance context and from a broader perspective, reforms are more likely to succeed
if their designers adjusted them to the cultural environment. Toward this end, the present
framework can be used to assess the suitability of transplanting legal mechanisms from one#p#分頁標題#e#
nation to another.
How to design culture-compatible legal reforms remains a major challenge. The link
between a societal aversion to litigation and high scores on Harmony and Uncertainty
Avoidance implies that in such high scoring countries, implementing a new legal regime
may require alternatives to the court system. For example Gleaser, Johnson, and Shleifer
contrast Poland’s success with the Czech Republic’s failure in establishing functioning
securities markets.73 They argue that active regulation by state authorities may be superior
where investor protection cannot rely on litigation.
Giving regulatory power to state agencies presents problems, however. Cultural
emphases on Embeddedness and Hierarchy—prevalent in many developing and transition
economies—may be conductive to corruption, in parallel to general disregard of the law.74
The challenge for reform designerswould be to find mechanisms for encouraging pro-social
conduct among state regulators and other power holders—a daunting task without doubt.
Research in this direction may benefit from in-depth analysis of Confucian-inspired Asian
societies (e.g. South Korea, China). These societies may have developed norms of social
responsibility that do not rely on court litigation nor on other accountability mechanisms
known in the West.
That said, countries that have not relied on court litigation traditionally need not forego
this mechanism for corporate governance. For instance, there are preliminary signs of somewhat
greater willingness to file shareholder suits in both Korea and Japan, possibly due to
legal reforms that followed economic downturns.75 Nonetheless, the fact that such proceedings
are so celebrated attests to their being exceptional, at least at this point.
71 Shalom Schwartz, Anat Bardi, & Gabriel Bianchi, Value Adaptation to the Imposition and Collapse of Communist
Regimes in East-Central Europe, in Political Psychology: Cultural and Cross-Cultural Foundations 217
(Stanley A. Renshon & John Duckitt, Eds., 2000).
72 Ronald Inglehart &Wayne Baker, Modernization, Cultural Change, and the Persistance of Traditional Values,
65 Am. Sociological Rev. (2000).
73 Glaeser, Johnson, & Shleifer, supra note 6.
74 Licht, Goldschmidt, & Schwartz, supra note 60.
75 See Jill Frances Solomon, Aris Solomna, & Chang-Young Park, A Conceptual Framework for Corporate
Governance Reform in South Korea, 10 Corp. Governance: Int’l Rev. 29 (2002); Zenichi Shishido, Japanese
Corporate Governance:The Hidden Problems of Corporate Law and Their Solutions, 25 Del. J. Corp. L. 189
(2000).
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 253
6. Conclusion
This article presents evidence on relations between statutory law and culture, especially#p#分頁標題#e#
in the context of corporate governance. An analysis of cross-sectional samples of nations
from around the world, drawing on national scores on cultural value dimensions from the
two leading theories in cross-cultural psychology, demonstrates that corporate governance
laws relate systematically to the prevailing culture. Consistent relations between cultural
emphases and the degree of formalism in civil procedure laws also emerge. More generally,
the approach employed to uncover these relations suggests a new way to analyze bases of
social institutions.
Acknowledgements
For helpful comments we thank the Journal’s editors, two anonymous referees, Bernie
Black, Bob Cooter, Mel Eisenberg, Luca Enriques, Hwa-Jin Kim, Paul Mahoney, Paul
Stephan, George Triantis, Detlev Vagts, and participants at the annual meeting of the European
Association for Law and Economics, conference of the Asian Institute of Corporate
Governance, law and economics seminars and workshop at Georgetown, Berkeley, and
Stanford, Law and Finance seminar at NYU, the faculty seminar at Virginia, and the interdisciplinary
workshop at the Judge Institute of Management, University of Cambridge, UK.
Thanks also to Andrei Shleifer. Errors remain our own. The work of Shalom Schwartz on
this paper was supported by Israel Science Foundation Grant No. 921/02-1.
Appendix A
See Table A.1.
Table A.1
Country classifications
Country Schwartz’s regions Hofstede’s regions Legal family LLSVa
Argentina Latin America More-developed Latin French +
Australia English speaking Anglo Common law +
Austria Western Europe Germanic German +
Belgium More-developed Latin French +
Bolivia Latin America French
Brazil Latin America More-developed Latin French +
Canada English speaking Anglo Common Law +
Chile Latin America Less-developed Latin French +
China Far East Other
Colombia Less-developed Latin French +
Cyprus Eastern Europe Common law
Czech Republic Eastern Europe German
Denmark Western Europe Nordic Scandinavian +
Ecuador Less-developed Latin French +
Egypt French +
254 A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255
Table A.1 (Continued )
Country Schwartz’s regions Hofstede’s regions Legal family LLSVa
Estonia Eastern Europe Other (civil law)
Ethiopia African Other
Finland Western Europe Nordic Scandinavian +
France Western Europe More-developed Latin French +
Georgia Eastern Europe Other (civil law)
Germany Western Europe Germanic German +
Ghana African Common law
Greece Western Europe Near Eastern French +
Hong Kong Far East Asian Common law +
Hungary Eastern Europe Other
India Far East Asian Common law +
Indonesia Far East Asian French +
Iran Near Eastern Other#p#分頁標題#e#
Ireland English speaking Anglo Common law +
Israel English speaking Germanic Common law +
Italy Western Europe More-developed Latin French +
Japan Far East Asian German +
Jordan French +
Kenya Common law +
Macedonia Eastern Europe Other (civil law)
Malaysia Far East Asian Common law +
Mexico Latin America Less-developed Latin French +
Namibia African French
Nepal Far East Common law
Netherlands Western Europe Nordic French +
New Zealand English speaking Anglo Common law +
Nigeria Common law +
Norway Western Europe Nordic Scandinavian +
http://www.400190.com/dissertation_writing/Pakistan Asian Common law +
Peru Less-developed Latin French +
Philippines Far East Asian French +
Poland Eastern Europe Other
Portugal Western Europe Less-developed Latin French +
Russia Eastern Europe Other
Singapore Far East Asian Common law +
Slovakia Eastern Europe German
Slovenia Eastern Europe Other (civil law)
South Africa Anglo Common law +
South Korea Asian German +
Spain Western Europe More-developed Latin French +
Sri Lanka Common law +
Sweden Western Europe Nordic Scandinavian +
Switzerland Western Europe Germanic German +
Taiwan Far East Asian German +
Thailand Asian German +
Turkey Eastern Europe Near Eastern French +
Uganda African Common law
United Kingdom English speaking Anglo Common law +
United States English speaking Anglo Common law +
Uruguay Less-developed Latin French +
Venezuela Latin America Less-developed Latin French +
A.N. Licht et al. / International Review of Law and Economics 25 (2005) 229–255 255
Table A.1 (Continued )
Country Schwartz’s regions Hofstede’s regions Legal family LLSVa
Yugoslavia Near Eastern German
Zimbabwe African Common law +
Rafael La Porta et al. Law and Finance, 106 J. Pol. Econ. 1113 (1998).
a “+” denotes that the country is included in LLSV’s study.
Appendix B
See Table B.1.
Table B.1
Mean scores on cultural value dimensions in each cultural region
Region and
number of
countries
Cultural value dimensions
Harmony Embeddedness Hierarchy Mastery Affective
Autonomy
Intellectual
Autonomy
Egalitarianism
A. Regions identified by Schwartz
Africa (5) 3.75 4.17 2.71 4.20 3.04 4.20 4.52
Eastern
Europe
(12)
4.49 4.00 2.31 3.85 3.01 4.29 4.63
Englishspeaking
(7)
3.91 3.66 2.26 4.01 3.64 4.38 4.94
Far East
(10)
4.05 4.02 2.85 4.07 3.09 4.09 4.49
Latin
America
(6)
4.25 3.85 2.24 4.00 3.00 4.40 4.91
Western
Europe
(14)
4.57 3.34 1.90 3.93 3.74 4.86 5.13
Region and number of countries Cultural value dimensions#p#分頁標題#e#
Power
Distance
Uncertainty
Avoidance
Individualism Masculinity
B. Regions identified by Hofstede
Anglo (7) 35.57 44.71 80.57 61.43
Germanic (4) 23.25 68.50 61.00 65.50
Nordic (5) 30.20 42.50 71.40 13.80
More-developed Latin (6) 59.67 83.83 59.50 52.33
Less-developed Latin (8) 69.75 85.25 20.63 51.00
Asian (11) 71.45 53.18 25.45 53.09