2.1.1 – Culture-definitions: definitional tactics

18Schematically four main categories of meaning can be retained for culture, only one of them being of interest here: the sociological meaning [3]. It broadly refers to values, norms or beliefs, shared by some human group, reflecting its specificity. Precise definitions abound, and it is customary to point to the seminal book by Kroeber and Kluckhohn (1952), listing 251 definitions and statements about culture. In appendix 1, an illustration is provided about the variety in definitions, as derived from definitions obtained from the authors cited in this Survey.

19Reuter (2010) emphasizes how the understanding of “culture” is dependent upon a defined geographical or historical area. It is suggested that “definitional tactics” are an underlying issue when defining culture: the definitions of culture retained by specific actors, in specific communities, at specific times, reflect who actors are, how communities interface, and the epistemological, political or social posture of these actors, rather than providing definitional content, per se. As a preliminary illustration, the next sections focus on how the cultural definitions recently proposed in economics [4], are resting upon fundamental intellectual and epistemological postures.

2.1.2 – Recent definitions of culture in economics: illustrating epistemological stakes

20Three recent contributions about “culture and economics”, including two books, attest of the timeliness (Throsby, 2001; Guiso, et al., 2007; De Jong, 2009^ΔΔΔ) of the interest for this concept in economics. The culture-conceptualizations retained by these scholars reflect the difficulty in agreeing on a definition and underlying intellectual orientations.

21Guiso, et al., focus on the idea of temporal stability to identify cultural elements (see their definition in appendix 1). Doing so, they echo a conception expressed earlier by Williamson (2000), who defines culture as one of the elements of the more general “institutions of embededdness”, elements that are resulting in lasting constraints on the economy: they operate “on the order of centuries or millennia” (Williamson, 2000: p. 596 [5]).

22Relating to a second definition of “culture in economics”, Throsby characterizes culture as a process, delineated by the specificity of some human group vis-à-vis other groups (appendix 1). So, he retains a conceptualization akin to those in the North-American Sociology (Cuche, 2004). With his conceptualization, he allows for fluidity and change, elements that are, by definition, excluded with Guiso, et al., or Williamson. Process-wise, instead of focusing on durability to identify cultural elements, Throsby focuses on cross-sectional homogeneity. Behavior is important here, as well as meaning. Both can be observed, and their relative homogeneity (within a group) and heterogeneity (across groups) lead to the identification of a “culture”.

23Now, concerning the third definition, De Jong (2009)^ΔΔΔ retains a motivational aspect, through the concept of “values”. He follows the path chosen in cross-cultural psychology (as in the dimensionalist approach, see section 3.1) [6]. In this sense, cultures are enduring, as in the first definition.

24Yet another important distinction emerges, between “constraining” and “enabling” views. DiMaggio (1994, 1997) reminds us how the cognitive aspects of culture provide ground for “enabling” views on culture: culture not only constrains, but structures and fosters behaviors, communication, decision-making, etc. DiMaggio emphasizes how this distinction is central in the epistemological structuring of scientific disciplines and in the choice of methodologies [7]. In fact, he concludes that economics itself can be seen as a “cultural system” (1994: p. 46 and 47), a point that has been recently transposed to finance with considerable force (MacKenzie and Millo, 2003, published a seminal article, espousing what is becoming known as the Social Studies of Finance).

25Larry Summers (1985: p. 1)° can be quoted to illustrate the point: “the fields of economics and finance are allied. Indeed finance is sometimes defined as a subfield of economics [… ] yet we are increasingly witnessing the development of two cultures [emphasis added]“. Larry Summers’ statement may seem anecdotic, because it is not clear whether it assimilates finance into something that could be described as a “cultural system”. Yet, it provides ground for more than just familiarity, or knowledge of the financial field and theory. The distinction with “economic culture” suggests some form of relative coherence that distinguishes the two sub-fields in theory and in practice. This coherence excludes anecdotic views. In fact, this statement may seem anecdotic precisely because of epistemological postures. The idea of a coherent whole, guiding and structuring behaviors, is an “enabling” view of culture, so that the idea of a financial culture may seem straightforward to a sociologist, while it is less clear and apparently anecdotic for a “constraining” view of culture, as in economics. These issues are further explored in sections 4.2, 4.3 and 5.4. [8]

26So, provided the complexity and stakes involved in a definition of “culture”, the primary purpose of this article is to identify which views of culture are currently emerging in finance. So, rather than providing an upfront definition, it is hoped that it emerges from the field, and it is expected that this process will, in itself, be informative in many ways. So, a systematic and scientific process has been established to do so; and it is this process that is described next.

2.2.1 – Choosing journals

27As the purpose has been to favor a grounded emergence of cultural definitions from the discipline, the social boundaries have been identified and used, in turn, to identify the financial discipline: peer-reviewed journals of a certain prominence are rested upon, i.e. those followed by the Journal of Citation Report (JCR). The JCR is a Thomson-Reuter product, part of the ISI-Web of Science database, that produces and publicizes the “impact factors”, now largely used in evaluating scholars and institutions. The database provides a category labeled “Finance and Accounting” containing 52 journals, which need to be untangled.

28Discriminating between the accounting and financial disciplines is essential. In accounting, the interest in cultural approaches is strong, with an established history. It is characterized by a plurality of accepted cultural frameworks, and debates and controversies among varying streams of research. Besides, the sheer volume of references that the screening would uncover in accounting-reviews forbids the mere possibility of reviewing the literature in this article. Last, provided the strong differentiation between the two disciplines, the suggested process aiming at letting definitions emerge from the field would become ambivalent.

29To resolve this issue, the JCR’s “Finance and Accounting” category is crosschecked against available national rankings, that then provide the required distinction. Those categories available for Germany (by VHB), France (by CNRS) and the U.K. (by ABS) are used pragmatically, because they all provide this distinction between accounting and finance [9]. Overall, 23 journals were identified for the financial research community stricto sensu (table 1). As a robustness check, and to account for potential “regional biases” (Charreaux and Schatt, 2006), at a later stage, the screening was extended to financial journals that are mentioned in all these three national rankings simultaneously, while they were not included in the J.C.R. list. This extension provided only one additional article, and attests to the robustness of the screening-process.

2.2.2 – Time-line and identification (main screenings)

30No specific timeline is set, because the survey mainly utilizes the EBSCO-database to manually screen selected journals; thus, the timeline depends on the number of years included in EBSCO, for each journal. The Journal of Finance is included since 1946, so 1946 is the de facto starting point for the screening of the Journal of Finance. This is incidentally the longest time-period for all financial journals in the list surveyed. Table 1 provides the number of articles screened for each journal (see, third column).

Table 1

List of journals and results of the screenings

This includes the extended list of journals followed by the Journal of Citation Report plus those followed by the three national rankings simultaneously (in Germany, France and the U.K.- details in the fourth column). Rankings are provided accordingly, and if an impact factor is available, national rankings are omitted (JCR-list journals are generally followed in the three national rankings). Classifications. Relating to the re-categorization of the JCR journals into accounting and finance separately, few journals were ambiguously classified (25%). The general criterion to classify a journal as “financial” is that at least 2 of the 3 national rankings use this label while the third one does NOT use the “accounting label”. So potential ambiguity is to be derived from confusion with other disciplines than between finance and accounting: some rankings differentiate other sub-fields such as insurance, real estate or tax from either finance or accounting. All rankings differentiate finance from management, economics or psychology. The observed ambiguity in our re-classification involves either “macro-economy”, or “insurance”, or “general management”. All journals mentioned here, in the national rankings, are followed simultaneously in the three rankings, and are classified as "financial" three times. Notes. 1: discussion, book review, conference announcement. 2: criteria are: a very low occurrence combined with the fact that cultural variables are not used in the empirical analysis. 3: in these articles, the reference to societal culture is significant, but it is not related to the study of a financial subject. 4: this is pertaining to a prolonged debate (3 à’s, 40’s…) on accountants’ training and whether it should focus on cultural or technical contents. 5: the impact factor or the national rankings if the journal is not followed by the J.C.R.. 6: formerly Geneva Papers on Risk and Insurance Theor y. 7: formerly European Finance

List of journals and results of the screenings

31The process follows a procedure used elsewhere; for instance, in strategic management (Wiltbank, et al., 2006 for a recent example), or in finance (Charreaux and Schatt, 2006). The process rests on the screening of abstracts, or titles, of articles, based on some key words. Words such as ‘culture’, ‘cultures’, ‘cultural’(shortened “cultur^*”) have been searched for, specifically. This screening results in the identification of a number of other related words that could be used for further screening, such as ‘religion’, ‘trust’ or ‘language’.

Table 2

General Database as resulting from the main screenings

Authors Year Journal Classification Section Sub-topic Definition Occurrences of "cultur*" Cross-national focus? Dimensionalist reference? Aggarwal & Goodell*** 2009 JBF Financial macro-structures 3.2.2 Bank vs. market systems no 17 yes yes Alessandrini, Calgagnini & Zazzaro*** 2008 JBF International financial flows 4.1.2 M&As no 15 rel. no Chui, Titman & Wei*** 2010 JF Stock market functioning 3.3.3 Momentum profits yes 76 yes yes Conn & alii*** 2005 JBFA International financial flows 3.1.1 M&As yes 24 yes yes Focarelli & Pozzolo*** 2008 JBF International financial flows 4.1.1 M&As no 6 yes no Gande & Parsley*** 2003 JFE International financial flows 4.1.1 Spillovers in the sovereign debt market no 6 yes no Gleason, McNulty & Pennathur*** 2005 JBF International financial flows 4.1.1 4.1.2 M&As no 10 rel. no Grinblatt & Keloharju*** 2001 JF Investment 4.2.2 Investing: stockholding no 29 rel. no Grinold, Rudd & Stefek*** + 1989 JPM Investment 4.2.2 Institutional investment in international context no 1 yes no Hilary & Hui*** 2009 JFE Corporate finance policies 4.3.3 Corporate investment, growth, returns no 17 no no Mian*** 2006 JF International financial flows 4.1.1 Corporate lending in foreign countries no 27 rel. no O’Bar & Conley*** + 1992 FAJ Investment 4.3.3 Institutional investment management yes 14 no no Outreville*** 2008 JRI International financial flows 3.1.1 M&As yes 27 yes yes Rosati & Secola*** 2006 JBF International financial flows 4.1.1 Interbank payment flows no 6 yes no Sarkissian & Schill*** 2004 RFS International financial flows 4.1.1 Overseas listing no 11 yes no Schmeling** 2009 JEF Stock market functioning 3.3.3 Stock returns and sentiment yes 17 yes yes Stonehill & alii*** + 1975 FM Corporate finance policies 4.2.2 Financial policy no 3 yes no Stulz & Williamson*** 2003 JFE Financial macro-structures 4.1.1 Legal infrastructure yes 75 yes no Marking. + these three articles have either a very low occurrence of "cultur*" (in two cases), or a very specific approach (in all three cases), and their inclusion in the database is subject to a special discussion (section 4.4). Stars indicate results from the main screenings, with three stars (***) indicating results from screenings of the JCR-journal-list, and two stars (**) indicating the additional result obtained from the screenings of national journals. The protocol followed for the main screenings is outlined in section 2.2.2 on the next page.

General Database as resulting from the main screenings

32However, we restrict ourselves to “cultur^*” here for two reasons; first, because additional words should be subjected to preliminary discussions, and, second, because we precisely wish to let these connections emerge from the process [10].

33This process results in the identification of sixty hits, of which 11 are not research-articles, 13 articles can be considered as having an anecdotic use of “cultur^*”, 3 articles refer to the “cultural affinity hypothesis” (section 4.3.1), and 15 articles use specific expressions that are not related to the core meaning of culture understood in this research, such as pop-culture, corporate culture, technical cultures, and in particular “financial cultures” (sections 4.3.– and 5.4). Table 1 provides details on this process. The result is eighteen articles (including the specific result from the regional screening), which are presented in table 2, with some details.

2.2.3 – The double review process

34As the dimensionalist framework emerges as the only consistent framework spanning several articles in the database, a “double review process” is undertaken, completing the systematic screening with an ad hoc survey of dimensionalism in finance, which is based on the systematic exploitation of cross-citations (received, made) within the 5 dimensionalist articles initially obtained, and in any article subsequently identified. Eventually, the base of dimensionalist articles is extended from 5 to 29 references [11]. They are detailed in table 3, and presented with the review of dimensionalism in finance just before section 3.2.

35With this “double review process”, the intention is to propose a complete picture of the dimensionalist framework in finance – its coverage, its results and its relative successes. This should be helpful in two ways. For researchers specifically interested in dimensionalism, this provides a more comprehensive picture of the empirical coverage and strengths. For researchers who are more critical about the approach, this provides a contrast between articles that made their way into peer-reviewed journals in finance, and other articles, either published in related disciplines, or not (yet) published.

3.1.1 – The label

43The label dimensionalism is proposed in a book edited by Vinken, et al. (2004^ΔΔΔ: chapter 1). This book compiles contributions, at the occasion of a gathering of “well-known dimensionalist researchers”, in Holland, in 2004 (including in particular: Hofstede, Inglehart, Schwartz, or still Triandis).

44Dimensionalism is defined as aiming at “finding the ultimate, most frugal, and yet most meaningful basic set of axes, with which to explain the broad range of attitudes, beliefs, life styles and the diversity of practices among large populations and/or organizations across societies” (Vinken, et al., 2004^ΔΔΔ: p.8).

45The main focus is on cross-national comparisons and the main tool is the development of numerical scales, to distinguish nations’ cultures, based on synthetic and abstract “dimensions”, the national cultural indices.

3.1.2 – Main sources and indices

46The dimensionalist large-scale surveys that were mainly identified in the screening, are those of Hofstede (2001^ΔΔΔ: the revised edition includes references to meta-studies and critiques), and Schwartz (1994)^ΔΔΔ [14]: within our extended database of 29 dimensionalist articles, Hofstede’s indices are used in 24 cases (twice in combination with Schwartz’s); Schwartz’s indices are used in 6 cases (three times in combination, twice with Hofstede and once with GLOBE); two articles use another reference (GLOBE, in combination, Smith, et al.).

47The tentative measurement of national culture, with indices, can be traced back to Hofstede (1980)^ΔΔΔ. He used the results of a large-scale survey of IBM employees, in the 70’s, to identify differences between countries. He collected a large number of statements on the value-orientation of IBM employees. He averaged them by country, conducting a factorial analysis on the aggregate scores, still by country. He identified 4 dimensions on which cross-national differences were clustered. He labeled them “individualism / collectivism” (strong vs. loose ties between individuals in a society), “power distance” (acceptance of inequality generated by differences in power), “uncertainty avoidance” (a normative assessment of uncertain situations), and “masculinity / feminity” (gender roles and relative importance of self-assertiveness and success vs. harmony in society). More precise definitions are provided in appendix 2.

48Building on Hofstede, Schwartz proposed a complementary approach, theoretically-grounded [15] and oriented by three polarizations: “mastery-harmony”, “egalitarianism-hierarchy” and “conservatism-autonomy” (appendix 2 for a definition).

49A comparison of Schwartz’s indices with those of Hofstede and Inglehart is available in Schwartz (2004)^ΔΔΔ, and an empirical assessment of pair wise correlations between Hofstede’s and Schwartz’s indices is available in Schwartz (1994^ΔΔΔ: p. 109). De Jong (2009^ΔΔΔ: appendix 1: p. 197-206) provides a more general overview of most national culture and/or dimensionalist indices, including those by Hofstede, Schwartz, the GLOBE consortium, Trompenaars and the World Value Survey.

3.1.3 – Cultural distance

50It is important to mention a specific operationalization, called “cultural distance”. This approach aggregates the four original national culture indices, from Hofstede, into one value. Specifically it sums up squared differences across all four cultural indices, which results in one score for each pair of countries [16]. The approach was initially proposed by Kogut and Singh (1988)^ΔΔΔ. This cultural distance measure has been extensively used in management, and Siegel, et al., (2008) provide, for instance, a detailed review on the use of this measure in relation to the determinants of FDIs. More recent operationalizations sum up squared differences, using Schwartz’s, or, still, other indices. The measure is implemented in five articles in this database; based four times on Hofstede, and once on Schwartz.

3.1.4 – Antecedents and disciplinary path

51Triandis (2004)^ΔΔΔ provides an account on how Hofstede’s intellectual heritage influenced psychologists at the turn of the 90’s, and gave ground to the rise of a new discipline, known today as: “cross- cultural psychology”.

52Scholarly antecedents are ancient in human sciences. Hofstede (2001)^ΔΔΔ provides an extensive outline of these, in his first chapter, tracing them back to post-war influences by Kluckhohn, et al., or Rokeach (references therein). It is important to note that the key underlying mechanisms are values, through its motivational grip on individual and collective behaviors (Hofstede, 2001^ΔΔΔ: p. 11; Oyserman, 2002). De Jong, 2009^ΔΔΔ re-emphasizes how other cultural dimensions, such as rituals, heroes and symbols, all rest essentially on the same core dimension: that of values. Hence the emphasis on measuring values as expressed by individuals, to aggregate them “within the boundaries of defined social groups”, which has generally meant countries. Breuer and Quinten (2009: table 1, p.5)^* provide an outline on the varying cultural models that preceded, and followed, that of Hofstede.

53No further elaboration is made here, but multiple alternatives to values are available, in order to investigate cultures [17].

54The relative success of Hofstede’s research has been uneven across disciplines. Baskerville (2003)^ΔΔ shows how it was much more successful in management related disciplines (particularly in general management, International Business, organization studies, marketing or communication) than in economics, sociology or anthropology. So the relative success must be tackled, discipline-by-discipline. Recently Kirkman, et al. (2006)^ΔΔΔ provide a meta-study of all research incorporating Hofstede’s dimensions in management. Ramirez and Tadesse (2007: appendix)^*++ provide a review and a table including cross-references to Hofstede in management, operational research and economics/finance; they confirm Baskerville^ΔΔ’s suggestions, above, and highlight that cross-referencing into economics/finance is very recent, if peaking at all.

3.2.1 – Financial flows across nations and international investment

55In this section, we focus on financial flows across borders including debt (bonds, sovereign bonds) and equity (portfolio management, M&As, FDIs), and we include the study of the “home bias”. Five articles, in seven, use Hofstede’s indices, and five articles use the cultural distance approach, corresponding to all our articles using the cultural distance approach, but one.

56The general reasoning is straightforward, cultural distance between two nations hinders financial flows between them. Two articles apply this reasoning to flows of debt and then to flows of equity: the higher the distance, the lower the flow (Aggarwal, et al., 2009^*++ using all Hofstede’s indices separately and summed up; Siegel, et al., 2006^*++ using egalitarian distance, a measure derived from Schwartz). Relatedly, Beugelsdijk and Frijns (2009)^*++ note that the home bias is higher in foreign countries, with a higher cultural distance relative to the home country. The reasoning is applied to flows of M&As in two articles. Distance, as measured by egalitarian distance, has a negative effect on volumes of cross-border M&As (Siegel, et al., 2006^*++, but the effect is not robust for some specifications). Distance has not only a negative effect on volumes for M&As, but on profitability as well: cumulative abnormal returns are lower for U.K. companies, which are acquiring other companies abroad, when the acquisition is made in countries with higher cultural distance (Conn, et al., 2005^*** using all Hofstede’s indices aggregated). Lastly, it applies to flows of investment in Outreville (2008)^***: insurance groups choose to locate in countries where the cultural distance is lower.

Table 3

Dimensionalist database resulting from the main screenings and then the double review

Markings: *** for the 4 articles resulting from the main screenings and overlapping with the main database. ** for the regional article. * for the articles obtained through the double review process (dimensionalist articles). Details on both processes in section II.2.2. ++ is a marking for articles, where at least one of the co-authors has an established track record, in any of the journals screened as part of the main screenings (the J.C.R. list). It results that *** and ** stands for articles from the financial community stricto sensu, while *++ stands for articles from the financial community lato sensu. Eventually * stands either for aticles from financial researchers, with no track record in the J.C.R.-financial-journal-list (yet), or for established researchers in proximate fields (management mainly). Any unpublished article is notes "WP" for Working Paper in the column for the source. Other journals’ names are abridged by their initials. Further notes: 1) clustering of countries based on Hofstede. 2) cultural distance based on Smith, Peterson & Schwartz. 3) Globe measures

Dimensionalist database resulting from the main screenings and then the double review

57A marked difference between studies that overlap with economics is noted, and those that overlap with management, are noted as differentiated by specific subjects. On the one hand, following a tradition in trade economics, the introduction of cultural distance is very recent (Aggarwal, et al., 2009^*++). It is investigated in the section 4.1 of this article, because apart from Aggarwal, et al., there is no relation to dimensionalism. Additionally, it is noted that this study is mainly dealing with working papers, the conclusion is made that, on the economic side, and the use of cultural distance has not yet made its way into the peer-reviewed literature in finance.

58On the other hand, an overlap with management is identified when the focus is on either M&As or FDIs – two topics with a voluminous research track record in management journals (as identified through the J.C.R. list). There, the use of cultural distance is fairly ancient (Siegel, et al., 2008, for a review). Furthermore, it has been controversial; so, for instance, Morosini, et al. (1998), more than a decade ago, question what was then a decade of consensual results, and show that cultural distance enhances, rather than reduces, cross-border acquisition performance. This thread, however, was not followed any further, provided by the use of dimensionalism, in connection to M&As or FDIs, outside finance journals, stricto sensu, because the sheer volume of literature forbade it.

59However, an inclusion in the database has been made of Weber, et al. (1996)^* [18]. They show that corporate culture interferes with national culture, in its effects on cross-border acquisitions. They highlight that measures of cultural distance need further theorization, before anything can be concluded about the relation between cultural distance and cross-border M&As; so, for instance, cultural distance can foster rather than hinder cross-border acquisitions, because partners complement one-another through their differences (i.e. through their cultural distance).

60Another aspect, worth emphasizing here, is the large number of institutional control-variables, in many articles, in the database. Table 3 provides a rough overview, and appendix 3 lists categories for these control variables, as identified in the extended dimensionalist database. A further analysis of these references, in relation to national cultures, and the varying aspects of the institutional context, was initiated by the author, and will be the object of a specific article. Appendix 3 conveys a sense of the extremely wide range of context dimensions, which were considered simultaneously with culture, in most of the dimensionalist research-papers.

61Lastly, Anderson, et al. (2007)^* provide an innovative approach to culture, in relation to the home bias. They also provide the only article in this section not related to cultural distance. Interpreting the definitions provided by Hofstede, Anderson, et al., suggests that the relative magnitude of the home biases should be correlated with national cultural indices, namely the absolute value of the uncertainty avoidance and the masculinity/feminity indices. The reasoning is the following: investors from countries where uncertainty avoidance is higher (less acceptance of uncertainty) will diversify less, because diversification involves investing in more uncertain stocks. Further, investors from masculine countries will have higher potential overconfidence and self-attribution biases, which will drive higher exposures abroad, a lower home bias, ceteris paribus.

3.2.2 – Financial macro-structures

62This broad subject provides us with a variety of sub-fields, including the determinants of financial macro-structures (the relative importance of banks and markets: 2 articles), the development of stock markets (1), the determinants of corporate governance structures (governance indices; ownership concentration: 2), an article is concerned with growth in opaque industries (1), another with the consumption of life-insurance (1) and the last article investigates the connection between inflation, central bank independence and national cultures (1). All studies refer to Hofstede’s indices, but one, which uses Schwartz’s indices. Three articles, in eight, use only one of the Hofstede’s indices, while five of them implement indices simultaneously (either Hofstede’s or Schwartz’s indices).

63Hofstede’s uncertainty avoidance index features prominently in this section, because it is associated, on a stand-alone basis, with three articles, and it is associated with other indices in three other articles. Moreover, it is used in the article that results from the main screenings: Aggarwal and Goodell (2009)^***, extending a previous article by Kwok and Tadesse (2006)^*++, establish that countries characterized by greater uncertainty avoidance, favor bank-based systems [19]. Importantly, their focus, unlike that of Kwok and Tadesse, is not specifically dimensionalist, and they show more generally that “national preferences for market financing increase with political stability, societal openness, economic inequality, equity market concentration, and they decrease with regulatory quality and ambiguity aversion [i.e. uncertainty avoidance]” (ibid: p. 1770). They provide specific definitions, neither of culture, nor of national cultures. This is important because it provides contextual evidence about the way the financial research community receives, and uses, dimensionalist frameworks: Kwok and Tadesse published their research in the Journal of International Business Studies (excluded from the J.C.R. list of financial journals), while Aggarwal and Goodell published in the Journal of Banking and Finance (further commentaries in section 4.3).

64So, uncertainty avoidance is negatively associated with stock market financing. Furthermore, it is negatively associated with growth in “opaque” industries (i.e. industries characterized by informational opacity: Huang, 2009)^*. It is correlated with stock market development (negatively), with investor’s protection (negatively), with ownership concentration (positively), with the tandem that associates central bank independence and limited inflation (positively), with the development of pension schemes (positively). However, Chui and Kwok (2008)^*++ fail to identify a significant relation with life insurance consumption (potentially because it is mediated by the development of pension schemes).

65In the three articles, with simultaneous implementation of three or four Hofstede’ indices, results are mixed regarding the significance of the indices. Chui and Kwok (2008)^*++ identify a negative relation between life insurance consumption and the three other indices, collectivism, masculinity and power distance (the higher the three indices, the lower the life insurance consumption). De Jong and Semenov (2002)^* combine uncertainty acceptance and masculinity to abstractly define an “ability to accept competition”; they show that acceptance of competition is related to the development of stock markets (so stronger market development, with higher uncertainty acceptance and masculinity). De Jong and Semenov (2006)^* use the same two dimensions to relate them to ownership dispersion. They show that ownership concentration increases with masculinity and uncertainty avoidance (so the reverse relationship, as compared with stock market development). Relating to central bank independence and inflation, it is power distance, which combines successfully with uncertainty avoidance: power distance is negatively related to both limited inflation and central bank independence (De Jong, 2002^*).

66In a fourth article, with simultaneous implementation of several indices, Breuer and Salzmann (2008)^*++ show that Schwartz’s indices systematically correlate with governance indicators, and with the underlying macro-financial structure. So, countries with a stronger emphasis on autonomy, hierarchy, or mastery, tend to have market based systems.

67The large use of complementary legal, political and cultural variables is, again, very noticeable (table 3 and appendix 3). All articles, but one, use complementary political and legal variables. Three articles use complementary cultural variables (religion in three cases, ethnic fractionalization once). Overall, the need to complement the study of cultural determinants to financial macro-structures, by a wide range of institutional variables, is, once again, very strong.

3.2.3 – Stock-market functioning

68This domain of research belongs here, due to the recent Journal of Finance publication, relating momentum profits (on stock markets) to Hofstede’s individualism index (Chui, et al., 2010^***). This section also includes an article obtained through the “regional” screening (Schmeling, 2009^**), and related to the one by Chui, et al.

69All articles use Hofstede’s indices. Extending the findings from the preceding section, the use of indices is more parsimonious here [20]: three articles use only one of the four indices, while three of them use two indices in combination with one another. The study of financial macro-structures may be more dependent on the institutional context of nations than that of stock market functioning. It results that: the “country puzzle” is less prevalent, and that Williamson’s “institutions of embededdness” are less relevant (further commentaries in sections 5.1 to 5.4 and Reuter, 2010). This results in approaches that have fewer comprehensive indices and more specificity. In fact, it is noted that the inclusion of institutional variables is much more limited here, as compared to the previous two sections (table 3, in relation to appendix 3).

70Furthermore, it is noted that, in this section, most articles are published articles; in two cases in financial journals, and in four cases in journals at the intersection with law, economics and financial accounting. This relative success may be supported by the parsimony in approach, and second, by a more specific characterization of the relationship between national cultures and the financial subjects being investigated. This echoes the commentary in footnote 20, and these insights are built upon for the recommendations that are made in sections 4.2 and 4.3.

71Chui, et al. (2010^***) note that momentum profit strategies are strong in US stock markets, and weak in many Asian countries. They hypothesize that the ability to derive momentum profits may be, partly, driven by the individual mechanisms of over-confidence and self-attribution biases, which in turn are both fostered by (and positively associated with) the degree of individualism in a given country. Schmeling (2009)^**, extending this result, investigates the tendency of national stock markets to be more or less prone to “over-reaction and herd-like behaviors”. He shows how these behaviors may be related not only to individualism, but to collectivism (its opposite) as well, and to uncertainty avoidance. While individualism triggers individual behaviors such as over-confidence and self-attribution, collectivism triggers a collective behavior: “herd-like behavior”.

72This distinction is particularly interesting, culture-wise, for two reasons. Firstly it provides specifics on the mechanics of enabling views of culture: here the polarity involves not only constraints, but generates specific behaviors (it is enabling, see section 2.1 and 5.4). Secondly, culture would be precisely what characterizes collective behaviors, as opposed to individual behaviors (and related to psychology). Overall, Schmeling shows that the relation between sentiment and stock market returns are magnified by collectivism, and uncertainty avoidance. At the same time, they are moderated by market integrity, a further measure derived from Chui, et al., which is correlated with the two dimensionalist indices. So the relationships are complex, involving first and second-order effects, combined with contradictory influences.

73Providing consistent results with the ones just mentioned, Clement, et al. (2003)^* show that the individual experience of market analysts has more leverage on the forecast accuracy in individualistic countries than in collectivistic countries. Similarly, Pirouz (2004)^* combines individualism with power distance to define what he labels a “relationship-orientation”. He shows that there is a correlation between relationship orientation (high relationship orientation meaning high collectivism and power distance), stock market volatility and limited openness.

74Kimbro (2002)^* investigates the relation between corruption, on the one hand, accounting, and financial statement quality, on the other. He shows how the previously identified positive relationship, between corruption and two cultural variables, collectivism and power distance, turns negative when the level of economic development and accounting standards are controlled. So, more individualism drives better monitoring and information, which leads, in turn, to less corruption opportunities (primary and indirect effect). However, more individualism leads to a stronger willingness to take advantage of corruption opportunities, so that it drives more corruption (secondary but direct effect of a generally smaller magnitude). Kimbro concludes that the smaller, direct, effect of individualism was previously overseen, because it was covered by the collective externality provided by a higher quality in accounting and financial statements. This line of analysis is reminiscent of the one by Schmeling (2009)^**, because it shows that the general embededdness of institutional and cultural phenomena drives first and second order-phenomena, which easily lead to false assumptions on the basis of over-simplified analyses.

75Overall, this section features a large diversity in subjects, provides a large number of articles featured in peer-reviewed financial journals, and distinguishes itself by a flexible dimensionalist approach. Further a number of syncretic concepts emerge, such as “relationship orientation”, or “herd-like behavior”. Last, fine-tuned propositions with second order causality mechanisms appear, such as Kimbro’s analysis.

3.2.4 – Dimensionalist antecedents to financial laws

76One article is positioned at the intersection of the two previous sections (stock-market functioning, financial macro-structures). The article provides a comprehensive approach, culture-wise, with far-reaching consequences for the practical relevance of legal determinism in financial research. Licht, et al. (2005)^*++ study the antecedents of financial laws (investors’ rights, procedural formalism), which they relate to a large range of cultural indices.

77They show how legal variables are robustly correlated to the cultural constructs, from both Hofstede (uncertainty avoidance and power distance) and Schwartz (all of them), and are controlling for other cultural and legal variables (namely, the rule of law, religion), and economic development. In particular, they identify a robust correlation between harmony, both shareholders’ and creditors’ rights (negative), and procedural formalism (positive). Furthermore, formalism is also negatively correlated to affective autonomy. In addition, they identify a general significance of uncertainty avoidance with shareholders’ rights (negative), and with formalism (positive), echoing other articles in section 3.2.2 (financial macrostructures). A wider range of Hofstede’s indices are powerful in the formalism regressions: uncertainty avoidance, collectivism and power distance, all three indices correlate positively to formalism. In further regressions on shareholders’ rights, where they exclude Asian and Australasian Common Law countries (Australia, India, Malaysia, Nepal, Singapore, Thailand), they find that the same three indices (uncertainty avoidance, collectivism and power distance) are negatively correlated to the importance of shareholders’ rights. Their argument is that, in these countries, the letter of the law does not reflect the effective legal practice, so that countries have to be treated practically as “outliers”.

78Licht, et al., conclude that their study casts doubt on the “alleged supremacy of statutes” and on the ability to change the letter of the law, to foster economic and/or financial development, with effective consequences on behaviors and institutions. Further reflections on these conclusions can be seen in sections 4.2.1 and 5.

3.2.5 – Corporate finance policies

79Provided that there was no result from the main screenings, for this topic, this section will be brief. Besides, five of the seven articles identified, here, through the double review process, are working papers. Three articles use Schwartz’s indices, while five of them use Hofstede’s measures (one article overlaps). Cultural indices are sometimes used on a stand-alone basis, and sometimes used in combination. Generally, few other institutional variables are used here, in stark difference to the previous sections; and apart from legal origin and investor’s protection (they both feature prominently: four times each. This needs to be considered in the light of the studies relating culture to investors’ rights, in sections 3.2.4 and 4.2.1). No other cultural variable is used.

80The topics covered in this section include capital structure, dividend policies and cash holdings (twice each, generally one with Hofstede’s and the other with Schwartz’s indices). A seventh paper is related to earnings management (table 3).

4.1.1 – “Familiarity breeds investment” [21]

82Six articles are reviewed in this section and one is left for the next sub-section. These articles contribute to a research-stream in the adjacent field of trade-economics, which investigates the determinants of bilateral trade-flows, sometimes using “gravity models”. These models traditionally investigate trade-flows patterns, and focus on the relative size of trading partners and on their geographical proximity (hence the ‘gravity’ label). They were recently expanded to include the idea of “psychological geography”, among other determinants. Portes, et al. (2001) consider the importance of the volume of information, available on trading-partners, as a determinant for economic exchange. While distance is often prejudicial to both information-exchange and physical-trade, Portes, et al., extend the idea of geographical proximity, conceptually, to the idea of “cultural affinity”, because affinities reduce informational frictions and facilitate information exchange. While the exact nature of the relation between distance, information and “cultural affinity” remains unexplored in their article, their research triggered a wave of empirical testing of culture-related proxies in trade-economics.

83Overall, in conclusion of this brief excursus on gravity models, it is noted that if the empirical investigation of culture-related proxies is recent, references to “cultural similarity” or “psychic distance” are fairly ancient, as they were proposed as early as the 1950s (Beckerman, 1956).

84The six articles included in this section investigate a range of financial flows including foreign corporate lending by banks (1 article), spillovers on the sovereign debt market (1), international cross-listings (1), payments on the interbank market (1), and, cross-country flows of M&As (2).

85While Rosati and Secola (2006)^*** identify a strong proximity effect in a gravity model of inter-bank payment flows in the E.U., Gande and Parsley (2005)^*** identify neither cultural, nor any institutional effect, on the “spillover” of a country’s credit rating onto the credit rating of other countries. They note that credit-rating spillovers are highly influenced by a common correlation to financial flows to-and-from the U.S. Sarkissian and Schill (2004)^***, studying the determinants of overseas-listing, conclude that “the same proximity constraints that are believed to lead to home bias … also exert a profound influence on financing decision” [22]. Mian (2006)^*** shows that cultural distance limits the efficiency of lending to corporations by foreign banks, in Pakistan. In his study, he dichotomizes cultural distance by a distinction between Asian- and non-Asian foreign banks, and he finds that Asian banks do better than non-Asian foreign banks. Focarelli and Pozzolo (2008)^*** show that the amount of cross-border M&As depends on a range of proxies, which characterize proximity between two countries: they use a range of culture-related proxies, including, in particular, distance, language, and thirdly, the similarity in financial and economic macro-structures. Gleason, et al. (2005)^*** have a related approach, and they show that cumulative abnormal returns for the acquisition of foreign privatized banks, by British banks, is influenced by cultural proximity.

86Those articles, here, focusing on the occurrence and treatment of culture-related concepts, have a low occurrence of “cultur^*” even through the idea of national cultures, and the underlying idea of “psychic distance”, is central to the argumentation. In general, “cultur^*” is associated with specific expressions, such as “cultural distance”, “geographic, economic, cultural, and industrial proximity”, “cultural similarity”, “cultural and, or institutional similarities”, “geographical and cultural proximity”, “cultural and institutional linkage”, “cultural barriers”, or “cultural integration”. These expressions convey the idea that distance between countries has many complementary sources, culture being only one of them. Moreover, as no definition of culture is ever provided, the precise mechanism by which culture influences, and the interrelation between varying dimensions: cultural, institutional, etc., are left undefined.

87As a result, the overall picture about the respective contribution, to international financial flows, of cultures, and institutions, is fuzzy, even confusing. Gande and Parsley (2005)^***, or Rosati and Secola (2006)^***, combine national cultures and institutions together, and analyze them in concert, through common proxies. Gleason, et al. (2005)^*** define “cultural similarity” more narrowly, by geographical distance and a commonality in language. So they clearly distinguish culture from other “institutional” dimensions, and they incidentally define institutions with legal variables. Sarkissian and Schill (2004)^*** explicitly emphasize that a distinction between cultural and institutional linkages should be made because “greater psychological tolerance” comes as a complement to information exchange, but they emphasize this important distinction, to be made, in conclusion, as a promising research direction. Eventually some authors treat proximity as purely generic across a wider range of sources: Focarelli and Pozzolo (2008)^*** consider “geographical distance and economic and cultural integration” in combination, through combined proxies.

88The research, reviewed in this section, aims at uncovering the quantitative impact of proximity on exchange. It largely foregoes to specify the mechanisms linking proximity and exchange; their inspiration is largely macro-economic (hence, the relation to gravity models).

89Language is most consensual as a proxy for culture and/or institutional differentiation. A general framework, spanning most of this research, is that culture operates through information exchanges, so that it is mediated by the use of a common language. Yet, this link is usually implicit, and most authors believe, following Portes, et al., or Sarkissian and Schill, that the two mechanisms, i.e. cultural affinity and information availability, are separate and complementary. Yet they depend, inter alia, upon a common language.

90Other sources for cultural proximity are often considered in this literature. They include commonality in religion, in law or in colonial history, common borders, geographic distance, similarity in financial or socio-economic systems, etc. However there is no consistent use of these proxies, and a poor specification dominates this research-field, as to whether these variables are included as variables, per se, or as proxies; or whether these variables are included in combination, with, or in contrast, to one-an-other, and to cultural variables.

4.1.2 – Corporate culture, “social capital” and merger flows

91A seventh article, on inter-regional exchange flows, is by Alessandrini, et al. (2008)^***. They develop a typology of M&As in the Italian banking industry, distinguishing what they call cleaning strategies from portfolio strategies. A cleaning strategy corresponds to the situation, where, in the situation of one bank acquiring another, the acquiring bank cleans the portfolio of the acquired bank of inefficient borrowers. A portfolio strategy corresponds to the situation, where the acquiring bank completely re-orients lending in the acquired bank, depending on its specific know-how, its organizational processes and its preferred clientele. Alessandrini, et al., show that cleaning strategies dominate in what they call in-market acquisitions, while portfolio strategies dominate in out-market acquisitions.

92It is through this distinction between in- and out-market, that the cultural variables enter into play. Alessandrini, et al., characterize in- from out-market by “geographical and cultural distance”. They implement cultural distance as the difference in relative “social capital” between varying Italian regions. They combine two definitions for defining “social capital”: first, the number of blood-bags donated by citizens in any given Italian region, and, second, the average regional voter-turnover at a specific and important referendum. While providing no definition of the “cultural”, they use it very specifically, with diverging meaning-categories. First, they use it in connection to a region-specific “culture of mutual trust”, which they label “social capital”, following a stream of literature in economics (see section 4.1.3). Second, they use it in another way, related to “corporate culture”.

93So, their article departs from those in the preceding section (4.1.1) in three significant ways. First it clearly distinguishes geographical from cultural distance: so a bank can be “culturally close” to the working environment found in a “distant province”. Second, short of a specific definition, they provide an extent of references and definitions for their two categories of meaning.

94Third, they implement two contrasting views of culture, simultaneously. The expression “corporate culture” (five occurrences) denotes a construct that is separate from regional “social capital”, related in particular to the construct of “societal trust”. So the two meaning categories are conceptually quite separate [23]. The “corporate culture” is in particular related to a specific know-how, which serves as basis for the portfolio strategy (ibid, p. 700 and 701). In this sense, the corporate culture of the bank is intimately related to its lending strategy, and it clearly evokes the expression “credit culture”, that can be found in Gleason, et al. (2005)^***. This conceptual category is, in fact, very common in the material initially collected through the screening (the initial sixty hits). Credit cultures describe this technical knowhow, from which banks derive superior performance, in itself, or from acquisitions, when they transfer it to acquired banks. In the initial screening, this use of culture was classified as technical and brushed it aside, as anecdotic. However, further analysis is required, and provided in sections 4.3.3, and 5.4.

95Eventually, it should be noted that in this research article, a concrete example of two types, two kinds of cultures that inter-relate and overlap, exist. The first is akin to national cultures, here, in a regional sense, while the second is to be related to a mix of professional (technical) and organizational types, blended together into the specific bank’s culture. This is a practical illustration of what will be labeled as cultural layers, later, in section 5.4.

4.1.3 – Culture, trust and financial flows

96The previous paragraph accounts for the emergence of a specific understanding of culture as civic culture, or societal trust. This specific view, emerging from the screening, has depth and precedence in economics, and beyond. Given the extended survey of dimensionalism in finance (resulting in table 3), we wish to provide, here, a brief, and complementary, outline on culture as trust. Luigi Zingales provides, in particular, 10 research articles on his webpage [24], mostly co-authored by Guiso and Sapienza, all of them on the subjects of “culture and economics”. These articles explore, further, this aspect of cultures, called, alternatively, generalized trust, civic capital, social capital, and societal trust. The research-stream identified here, through this connection with research by Guiso, Sapienza and Zingales, has antecedents in the political sciences (Banfield, 1958; Putnam, 1993) and in economics (Arrow, 1972: see Guiso, et al., 2009 for details).

97A recent article by these co-authors (Guiso, et al., 2009) is of interest here, because it explores a topic extensively covered in this research: the antecedents of financial flows, trade-, portfolio-, investment-flows as well as M&As and FDIs. Besides, this topic is explored with respect to dimensionalism (section 3.2.1) and includes most of the other identified articles in the screening (the two preceding sections 4.1.1 and 4.1.2) [25]. This article is of particular interest for four reasons. Firstly, it intends to unpack the source of relatedness between nations, separating out rational, geographical, institutional and cultural sources. Secondly, it extends the range of topics tackled at once, investigating three endogenous variables simultaneously. Thirdly, it provides many details, as well as sophisticated definitions of the culture-related variable, trust. Fourthly, it brings forward some new methodological guidance for the treatment of culture-related concepts in economics and/or finance, in particular with respect to causality and the use of instrumental variables.

98Guiso, et al. (2009), characterize culture, either as the stable aspect of individuals’ priors, and preferences, or as the stable social constraints in which individuals make judgments and choices. They operationalize it through “interpersonal trust”. They do not define this concept, rather, they split it into three categories: “personalized”, “mutual”, or “generalized” trust. They provide an extended discussion on how trust relates to the “culture” of individuals, as opposed to “rational expectations”. They suggest a relation to Arrow’s seminal works, because they view trust as either derived from repeated and successful interpersonal exchanges, so it is a “rational expectation”, or as related to stable priors and preferences, hence it is “cultural”. In this sense, an extended conceptual and empirical discussion about varying aspects, components, of trust, rational and cultural ones, overturns the need for a definition.

99It is very interesting to compare this article by Guiso, et al., to a dimensionalist article discussed earlier (section 3.2.1), by Siegel, et al. (2006)^*++. Siegel, et al., use egalitarianism to investigate a very similar range of international investment flows. The comparison of the two articles teaches us that, in their methodological proceedings, both articles instrumentalize their culture variable, trust for Guiso, et al., egalitarianism for Siegel, et al., with identical instruments: the number of years countries were at war across centuries, ethnic fractionalization within countries and commonality in dominant religion.

100This illustrates a point brought forward by this survey: the dimensionalist approach and other culture-approaches, for instance, trust, need not be opposed. Here, the two articles pursue identical objectives (explaining three types of quasi-identical, or related, financial flows), they define culture in seemingly opposing ways (Guiso, et al., refer to the constraining conceptualization, while Siegel, et al., refer to the enabling one, as emphasized in the Introduction), but both articles instrumentalize it with the same set of instruments, and with comparable success. So taken together, it seems that egalitarianism and mutual trust between European countries (pair by pair) explain the same set of phenomena, while they are caused by the same underlying factors.

101This opens up a number of questions about what Williamson labels the “institutions of embededdness”. Are trust and egalitarianism spurious in this investigation? Is the one causing the other? Which one causes the other? Are they both driven by a third underlying variable [26]? This illustrates the kind of questions that will be raised in section 4.3.

102Additionally, the fact that neither article has found its way into a peer-reviewed journal in finance is interesting. So Guiso, et al., have published in an economic journal while Siegel, et al., have cast their article in a new light, and are successfully positioning it in the management field. We suggest that this reflects disciplinary inbreeding, as discussed in the Introduction, the Synthesis and Conclusion. Culture is to be understood and accepted differently, by different academic fields. Relating to a financial angle, as considered through peer-reviewed journals, the discipline still seems reluctant to accept either type of approach; whether the dimensionalist one, or the one based on trust, as exemplified by Zingales’ web-site, be positioned as an economic rather than a financial research-contribution.

4.2.1 – National cultures as antecedents of investors’ rights

104Investors’ rights, along with legal origin, rule of law and enforcement, have been extensively used, in the law and finance approach, to investigate a number of financial issues, spanning from the quality of corporate governance, to the development of financial markets, and beyond to economic and political freedom (La Porta, et al. 2004). This stream of research considers investor’s rights as exogenous. It provides policy recommendations in consequence [27]. In this context, the article by Stulz and Williamson (2003)^***, along with the one by Licht, et al. (2005)^*++ considered earlier, show how financial laws can correlate with enduring and stable underlying socio-cultural characteristics, casting doubt on whether changes to the letter of the law can have effective consequences in practice.

105Stulz and Williamson provide a detailed definition for culture (appendix 1) and they use two proxies: religion and language. They note that religion has to be considered as a cultural variable, per se (not as a proxy), and they provide extensive scholarly references drawn from the fields of economics, economic sociology, law, economic history and religious history.

106In contrast, they note that the second variable, language, proxies for “world-views” or “world-systems” and they contrast an Anglo-American worldview (“Anglo-Saxon”, operationalized as the English-speaking countries), with Hispanic and other worldviews.

107Importantly, Stulz and Williamson (2003)^*** provide a discussion on how Protestantism relates to decentralization and individualism (conceptually, not in relation to Hofstede’s individualism index), with a suggested relation to societal trust as well [28]. Further they exemplify how the religious spirit and related social norms have interacted with national institutions, to foster cross-national differentiation across Europe. This is in fact one of the reasons why they retain the “dominant national religion”, as the variable of choice, precisely because the dominant religion best reflects this process of path-dependant maturation processes within nations. Stulz and Williamson further provide references for the relation between religious faith, and lending, an/or investing activities.

4.2.2 – Concluding on international flows and the antecedents of investors’ rights

108Overall, we wish to underline three key issues. First, the two articles considered here, like the two articles about international capital flows (section 4.1.3), the dimensionalist ones, and the ones on trust, religion, or worldviews, all provide evidence on the mechanisms that relate culture with financial outcomes. So these comparisons exemplify that culture must be examined in connection with a specific mechanism relating it to the outcomes considered.

109Second, taken together, these articles, emphasize the strong underlying, and implicit, relations between varying cultural, and institutional dimensions. In the last two sections (4.1.3 and 4.2.1), a dimensionalist approach, and an alternative approach, achieve similar results, on similar issues, with competing cultural approaches. This illustrates both the disciplinary stakes that are at play, and the potential for a middle ground, not yet achieved.

110Last, in many cases, and in particular in Stulz and Williamson (2003)^***’s article, the authors emphasize potential transversal linkages between views of culture.

4.2.3 – Culture vs. language and distance in portfolio holdings

111Grinblatt and Keloharju (2001)^*** investigate the importance of investor’s native cultures, understood here as ethnicity, as a determinant of investors’ stockholding patterns. They do not provide a definition of culture, but rather “classify the culture of the firm based on the name and native language of the CEO (Finnish/Swedish)”. Then they compare the ethnic origin of investors with that of the shares they are holding.

112Here culture, unlike section 4.1 on financial flows, is conceptualized and operationalized separately from language and distance: the language of the company is that of the annual reports, and the culture of the company is the native ethnic culture of the CEO, either Swede or Finnish. So the authors provide a test of the relative importance of culture as opposed to language or distance, on the nature of portfolio-holdings. They show that culture matters.

113Most importantly, they show that the culture-influence is moderated by distance and by education (or training): the culture-effect disappears for larger distances [29], and it disappears for savvy and institutional investors.

114This underlines an important and specific aspect of “national cultures”: not every citizen, in a given country, is subject to the same cultural influence, and a national culture’s influence can be individually circumscribed through education, experience or other circumstances. This relation between a synthetic index of national culture (as in dimensionalism), and the individuals considered in any empirical investigation, is problematic. It is, indeed, one of the key issues in the debates surrounding dimensionalism in management; it is one of the pillars of the critiques (see sections 5.1 and 5.2 for further developments and references).

4.2.4 – Corporate culture, religiosity and investment

115The article by Hilary and Hui (2009)^*** shows that, in effect, higher religiosity has an implication for the variance in equity returns (lower), ROA (lower), investment (lower), growth (lower) and break-even point for investment (higher).

116This article stands out, in this survey, in many respects. First, its focus is on comparing behaviors across local communities within the U.S., not across nations: the comparison is conducted at the county level (read county, like borough).

117Second, it investigates the influence of religiosity, rather than the effect of religious affiliation, i.e. its focus is procedural rather than substantial (we use this terminology, because it is standard in the sociology of culture, e.g. DiMaggio, 1994). Third, and consequently, it shows that the observed firms’ outcomes are related to the relative intensity of religious practice (i.e. the religiosity of executives, a procedural aspect), rather than on any specific religious affiliation, i.e. any specific cultural norm (a substantial aspect). This focus on the intensity of a given cultural (religious) practice is unique in the dataset obtained from the screenings. However, it reflects a long-standing scientific proceeding in anthropology or sociology – as well as in insurance research [30].

118Fourth, the reason this article has been included in the database is not related to a use of words based on “religion”, even though it is often associated with culture-research as exemplified above. This article was distinguished in the screening, because of its use of “corporate culture”. In fact the study aims at “examining the influence of community religion on corporate decisions”. The implicit link is that the narrow geographical focus on counties accounts for a direct link between the “corporate culture”, that of the CEO and that of the surrounding environment.

119Thus, this article underlines the prominence of a number issues acknowledged earlier. If nations, and related national cultures, provide the strongest common link within the database, there are other cultural sources to be considered, in particular corporate cultures. Second, the focus on a specific norm, or a specific definition, does not need to be the most appropriate way to proceed. As can be seen, in the previous section (4.2.2), an emphasis on mechanisms can help to specify cultural definitions. Moving one step further, the mechanism itself can be the key dimension, independent of the content: the procedural supersedes the substantial. By extension, the dimensionalist literature focuses on values as the building block to operationalize “national cultures” (section 3.1). Hence, it rests on the motivational aspect of cultures. But what is the importance of the relative motivational intensities to these values, and their interactions?

4.3.2 – The “country puzzle”, as an underlying theme, in many articles

123A large number of articles, initially obtained, refer to culture as a key source for country differentiation. Yet, most articles do not pursue the analysis further: the frequency of use of “cultur^*” is very low, often limited to a single occurrence in the abstract.

124This is echoing the idea that cultural concerns societal, or “societal” (O. Williamson, 2000) analysis, but not economic analysis. These articles, considered in particular together with those included in section 4.1, provide us with an inconsistent, and confusing view on national cultures, national institutions, and on the way they interrelate. They reflect what Stulz (2005) [31] labeled the “country puzzle” in finance: countries have a significant empirical effect on many financial issues, yet no financial theory proposes any framework to date.

125A few articles with a low occurrence of “cultur^*” in the body of the text were included in the database, because the occurrence seemed connected to a central argument made in the article. This included in particular Stonehill, et al. (1975)^***, Grinold, et al. (1989), and O’Bar and Conley (1992)^***.

126The two former articles exemplify the “country puzzle”, each in its own way. So Grinold, et al. (1989: p.1)^*** investigate the stakes involved with managing global portfolios. They emphasize the fact that “cultural differences among countries raise some important questions regarding investor preferences, market efficiency, and security analysis”. They mention that the relative importance of countries, vs. industries, must guide investment-management, because countries’ influence is more prevalent than that of industries. The focus of their effort, in this article, relates to an empirical split of both sources, national and industrial, in the global risk-return profile of the portfolio.

127Stonehill, et al. (1975)^*** present the results of a global survey on the practice of financial policies in five developed countries. Back in 1975, they conclude, that: “looking ahead, it seems unlikely that financial executives outside the U.S would adopt the stockholder wealth maximization model”. Again, this article is somewhat secondary to the main purpose of this research thsis, and it has a strong historical connotation. Yet, it is important in helping to replace the financial discipline and financial practice upon specific trajectories. It emphasizes, further, the relative prevalence of national backgrounds, about four decades ago; while it exemplifies how something akin to a “financial culture” can be situated historically, and geographically.

4.3.3 – The “culture of investing”

128The article by O’Bar and Conley (1992)^*** must receive, like the one by Hilary and Hui (2009)^***, a special place in this survey, because it is difficult to classify, specifically. It provides a financial ethnography, on the motives guiding investment decisions by managers, in institutional investment-funds, in the U.S. It concludes the ethnographic approach with the idea that corporate culture and corporate politics dominate more technical and rational “financial criteria” in managing investment-funds [32].

129This article, unlike the two other just reviewed, offers a high frequency of occurrence for “cultur^*”, and it provides a specific definition (appendix 1), which accounts, in particular, for both “corporate-” and “national-cultures”. Further, “cultur^*” is used in combination with a number of other expressions, or layers, including “business culture”, “culture of capital”, “civil service culture” and “public fund culture”. This extension in the way culture is conceived, and utilized, may seem puzzling. It reflects a reality that emerges from the screening undertaken, here. In fact, the question of how anecdotic these, and other uses, are should be considered seriously. It is arguable that they do not just reflect a way of speech, but a deeper reality, engrained in the notion of “culture”. Besides they convey shades of meaning in line with ancient academic knowledge throughout a range of more qualitative disciplines (Cuche, 2004).

130It must be noted that there is, in particular, a connection between the developments in section 4.1.2 about a “credit culture”, and a number of expressions emerging from the screening, and substantiating the idea that financial techniques constitute a specific culture [33]. Further, they echo Larry Summers’s (1985)° statement about the emerging specificity of a “financial culture”, provided in section 2.1. Last, they mirror the expression “equity culture”, used by Guiso, et al. (2001) to describe a increasing willingness, among the middle and upper classes in continental Europe, to invest their savings in stock markets.

131Overall, this crossing path, does not come out as a simplification of the survey, but it reflects research in other fields, and, in particular, in sociology, anthropology, and more; and more frequently, in International Business. These remarks are central to providing justification for the developments in section 5.4, below.

5.1.2 – Recent trajectory in International Business

137To provide the reader with a sense of the current trajectory of dimensionalism, a specific screening has been conducted, over one decade, centered on the main peer-reviewed journal for International Business: the Journal of International Business Studies.

138Overall, it seems that there is a consensus emerging on the need to “re-balance” cultural research, by shifting the approach away from that of mean-values by country (focusing for instance on variance as well), to using sources other than Hofstede’s, or still, to favoring approaches other than the dimensionalist one. In particular the emphasis is on contrasting varying cultural sources (community, gender, ethnicity, professions, organization, and so on), and on accounting for the need to better specify the effects of each cultural source on the phenomena considered.

139More specifically, Tung (2008)^ΔΔ(Δ) re-emphasizes the need to consider intra-country cultural variation (see earlier propositions by Au, 1999^ΔΔ). She insists, as well, on the need to consider cultural change. Smith (2006)^ΔΔ, emphasizes, among other things, the need to verify that relations at the individual levels (as operationalized in surveys), do hold in aggregate (at the level of the national culture index). He stresses that a better analysis of the effect of economic development is needed, and that wealth-related effects must be disentangled from cultural effects, stricto sensu. Lenartowicz and Roth (1999)^ΔΔ recall that the validity of a country grouping must be established beforehand, and that the use of a national culture indices must be made, depending what is being studied, and how it relates to culture (a focus on mechanisms). They propose several alternative methods. In a related enterprise, Sivakumar and Nakata (2001)^ΔΔ(Δ) examine how countries can be grouped depending on Hofstede’s indices. They show how the dimensions retained, and the weighting of these dimensions, can lead to a large volatility in the way countries are grouped together, compared and opposed.

140Leung, et al. (2005)^ΔΔ as well as Earley (2006)^ΔΔ emphasize how dimensionalist approaches should be supplemented by other approaches in the future. In particular Leung, et al. (2005)^ΔΔ re-emphasize the importance of varying cultural “layers” (Global, national, organizational, group-culture and then individual behaviors); they re-emphasize the processes of cultural change, the interaction between cultural layers and the varying ways in which values can influence behaviors (either causal links, mediating or moderating effects). They conclude: “much of previous research has adopted what we view as a simplistic view of culture, which tends to examine the static influence of a few cultural elements in isolation from other cultural elements and contextual variables…Hofstede dimensions fall into this category…” (p. 374). Earley underlines what approaches, other than those based on values, are seen to be promising. He recalls that “meaning” moderates the relation between values and behaviors. He concludes his review with the idea that “it may well be time that this form of large-scale, multi- country survey [i.e. Hofstede’s or Globe’s] be set aside” (p. 922).

5.3.1 – Opening the black box (opening up venues for culture in finance)…

145The survey, presented here, outlines the importance of nations, and national cultures, for many research articles. Relating to the dimensionalist papers, it is noted that the articles do not come without extensive additional references to institutions (appendix 3). Conversely, in other approaches, much confusion between culture and institutions has been observed (sections 4.1, 4.2.1 and 4.3). Overall, it seems that the emergence of the culture theme in financial research corresponds to a renewed interest in trying to untangle the sources of national embededdness: cultural, political, institutional, structural (networks), etc. It is noted, in passing, that this effort has started some time ago, in other disciplines. For instance Guillen and Suarez (2005) highlight the coexistence of five schools of thought, in relation to the exploration of the broader institutional context, in Organization Theory; dimensionalism being only one of them, and law and finance another. Further, the conclusions of the essays by Maseland (2009), or by De Jong (2009^ΔΔΔ: chapter 2), on the recent re-emergence of culture in economics, applies very well to financial research, but with a potential time lag.

146Stulz (2005)’s proposal of “the twin agency” model is an inspiring theoretical step to explore the “country puzzle”. It opens up venues for exploring the black box of national embededdness (The expression is original to the author). It echoes Williamson’s call to pay attention to the much-needed alignment between the “institutions of embededdness” and the varying levels of economic analysis (levels of societal analysis 1, 2, 3 and 4 in his terminology). It corresponds to the primary importance given by North (see appendix 1) to “informal institutions” and “culture”, in economic research.

147The contributions by Stulz and Williamson (2003)^***, and by Licht, et al. (2005)^*++ cast a new light on the successes of the law and finance research stream. The empirical regularities observed in law and finance research articles may be robust, but if the laws are related to enduring and stable cultural characteristics, country by country (understood as world-views, national religions, egalitarianism and harmony, and still other dimensionalist indices), then how can the results of law and finance be normatively interpreted? And how can recommendations be derived? There are many robust empirical regularities, which have been evidenced in law and finance, but short of clear conceptual mechanism, are there reliable predictions?

148For these reasons, and provided the formal, and less formal attention, now given to cultures in financial research, it is arguable that culture and finance is set to become a full-blown research-area soon. This answers to the second research-question proposed in introduction.

5.3.2 – Using all the tools?

149Future research may pragmatically draw from existing frameworks, from adjacent disciplines, in management (in particular with emphasis on dimensionalism), in economics and development-economics (in particular with special emphasis on trust or religion), or in economic sociology (see below). It is suggested that this should not be done without detailed cross-references. In fact, a detailed analysis of the database shows that successful cultural research in finance generally comes with very specific references. For example Stulz and Williamson (2003)^*** provide extensive scholarly references drawn from the fields of economics, economic sociology, law, economic and religious history. Guiso, et al. (web-site references provided earlier) provide a stream of articles on trust, including experimental investigation in trust-games, exploitation of survey questionnaires, comparison of varying methods from varying disciplines, and they derive conceptual mechanisms from ancient scholarship in political economy and economics (references provided above). Furthermore, a number of authors provide an extensive analysis of risk-taking behaviors by individuals, of their acceptance of uncertainty and of related concepts. They draw from a range of sources, from psychology to economics, to investigate the behaviors of market-participants and investors (see for instance Statman, 2007, 2008; Schmeling, 2009**; Chui, et al., 2010^***).

150All of these approaches (trust, religion, risk-appetite) intend to unpack specific culture influences, through specific cultural lenses. They need not be opposed, and in fact, they present large overlaps, often in relation to national and institutional influences.

151Stulz and Williamson provide a discussion on how Protestantism relates to decentralization and individualism, with a suggested relation to societal trust. Furthermore, they show how the religious spirit and the related norms interact with national institutions to foster the development of cross-national differentiation. In relation to the opposite connection, Guiso, et al., remind us on how religious people have differential trust-levels (at the individual level, in a given country, for instance the U.S). In aggregate, and in particular at the level of regions or nations, they show that there are lasting differences, in the way individuals relate to their community, and they identify, and measure lasting differences in the way societal trust/ civic culture structures socio-economic and political interactions. Lastly, Statman relates social capital (societal trust), individualism, culture and religion, to, and contrasts them with various socio-economic categories (such as revenue, education, age, gender).

152Therefore, is argued that: one, complementarities rather than divergences should be emphasized across these approaches, and, two, that there is room for a variety of references in financial research, provided appropriate cross-references are made from adjacent disciplines.

153Furthermore, authors within the dimensionalist literature have extensively investigated the importance of serial correlations of their indices with many country-level measures, across a range of domains. There is evidence of serial correlation of national culture indices with varieties of capitalism (Schwartz, 2007^ΔΔΔ), with the rule of law, corruption and democratic accountability (Licht, et al., 2007^ΔΔΔ), with economic, demographic and political structures (Schwartz, 2004^ΔΔΔ), as well as with many other cultural variables, mentioned earlier. De Jong (2009^ΔΔΔ: p. 49.) summarizes similar, but somewhat less extensive, results for Hofstede’s and the Globe’s indices.

154Overall, the issue of national embededdness has been looming large in the social sciences, including economics, and many venues exist for financial scholars. For instance generic efforts to describe nations, going beyond the idea of national culture indices exist. So Hall and Soskice (2001) propose an index based on a dichotomy between contractual vs. relational economies. This distinction echoes historical evidence provided by Greif (2006) who studies the development of trade by Genovese and Maghribi traders comparatively, from the 10th to 13th centuries. He shows how reputation and family ties (among Maghribi traders) can be substituted for analytical accounting, contracting, and legal enforcement (with the Genovese).

155The definitional issue of what culture is, remains, as does how financial scholars should tackle it. This is what is considered next. In particular, it is highlighted that over-simplified views of culture are likely to fail, while, in conclusion, it is suggested to backtrack definitions of culture in approach tentatively labeled backward definition of cultures.

Conservatism/ autonomy

227Conservatism includes values that are important in close-knit harmonious relationships, in which the interests of the individual are not viewed as distinct from those of the group. These values are primarily concerned with security, conformity, and tradition.

228Intellectual and Affective Autonomy are values likely to be important in a society that views the individual as an autonomous entity entitled to pursue his or her own interests.

229Intellectual autonomy places an emphasis on self-direction and affective autonomy emphasizes stimulation and hedonism.

Mastery/Harmony

230Mastery accentuates active mastery of the social environment through self-assertion. Such values promote the active efforts of people to change their surroundings and get ahead of others.

231Harmony lays emphasis on harmony with nature.

Egalitarianism/ hierarchy

232Egalitarian Commitment emphasizes the transcendence of selfish interests. This group of values concerns voluntary commitment to help improve the welfare of other people.

233Hierarchy stresses the legitimacy of hierarchical roles and resource allocation.