Culture's Consequences: Comparing Values, Behaviors, Institutions, and Organizations Across Nations, 2d ed.
The second edition of Hofstede's book about culture dimensions, Culture's Consequences, covers a tremendous amount of new literature and adds a modest amount of new data to the first edition of 1980. It updates literature about the now ubiquitous dimensions of individualism/collectivism, power distance, uncertainty avoidance, and masculinity/femininity that he adapted from Inkeles and Levinson's (1969) review of the national character literature in the first edition. The book includes data on these dimensions for 13 nations or regions not included in the first edition that previously only had been available in separate articles. It also adds data for a long-term/short-term orientation measure, provides scores for linguistic regions of several multi-language nations, and offers "index score estimates" for 16 nations not covered in the original study.
Perhaps the first edition of Culture's Consequences did not create the field of comparative cross-cultural studies, but it certainly has shaped the field's basic themes, structure, and controversies for over 20 years. The second edition will have a different role. The field has moved too far to be comprehensively restructured by any one project, and several broad-scope culture analyses compete with Hofstede's. Instead, the second edition makes several contributions to issues the field continues to face: it defends the validity of the original measures against recent critiques, evaluates the evidence for global changes in culture that affect the original data set's utility, and provides a thorough review of cross-cultural research.
The original data set and the measures certainly continue to generate controversy. The pattern of the controversy has been long fixed. Critics continue to question things like the heavy reliance on data from a single company, the increasing passage of time since the data were collected in the late 1960s and early 1970s, and what many find to be a nonintuitive link between the questionnaire items and culture dimensions. Central to Hofstede's strategy for defending the original measures is a thorough process of correlating them with other data sets that include multiple nations. The correlations are summarized in a "validation" section of chapter 2, the main methods chapter, and are discussed in depth in the chapters covering each dimension. The most persuasive evidence for validity and continued utility comes from a few direct replications and some large-scale studies using values measures similar to Hofstede's--Schwartz's (1994) Value Survey and the World Value Survey (Inglehart, Basanez, and Moreno, 1998). Correlations for these studies and a number of other smaller-scale projects are provided in an extensive statistical appendix.
The validation strategy also includes showing correlations with various demographic characteristics of nations, a number of which are used to construct plausible "culture's antecedents" arguments. For example, wealth produces individualism, and "as long as the wealth of nations grows, the individualism of those nations' citizens will increase" (pp. 254-255). More generally, Hofstede provides evidence for moderate culture change in individualism/collectivism, less in power distance, less still in uncertainty avoidance, and least in masculinity/femininity. Despite change, the ranking of nations remains relatively stable.
Hofstede is not shy about defending the validity of his measures but he stops well short of advocating that these or any other set of dimensions can represent all that one needs to know about culture. He repeats and reinforces his view that culture dimensions alone are not up to the task of preparing any individual or company to operate successfully in a given culture, but are better used to provide a substantial yet preliminary orientation. He argues that ethnographic analysis and personal experience are needed to actually function in a given locale. Culture dimensions add some discipline and substance to more comprehensive but less structured analysis and experience.
Had Hofstede never reported any data from the IBM study, the second edition of Culture's Consequences would still provide an invaluable review of the culture literature. The chapters covering the five main culture dimensions are very thorough. The later chapters about organizational culture and intercultural relations are insightful, but less comprehensive. Some parts of the review are discomforting due to characteristics of the material being reviewed. One occasionally gets the feeling of a huge jigsaw puzzle with a lot of missing pieces, a few pieces thrown in from other puzzles, and a few pieces that have been badly damaged. The pieces do not always fit easily together, and Hofstede sometimes seems to be trying too hard to make them fit. Notably, when a statistical relationship is found between the original measures and data from another project, the rationale to justify the relationship is sometimes strained. For example, over the years I have become accustomed to the idea that large power distance is consistently associated with compliant behavior, an acceptance that power differences are a normal part of life, and attitudes supporting various specific personal adjustments to power differentials, like don't criticize the leader. I have also become accustomed to the idea that large power distance can be associated with either strong emotional or attitudinal support for those in leadership roles, a dependent reaction, or for emotional or attitudinal rejection of leaders, a counterdependent reaction. But these sorts of multiple implications make it difficult to state hypotheses and make it very easy to rationalize more or less any association found between power distance and attitudes about leaders. In the case of power distance, I find the application to be reasonably clear: leaders should expect followers in large power distance societies to do what the leader says without objecting, but leaders should also watch their backs and be prepared for unexpected rebellion. Similarly, leaders in low power distance societies should expect followers to argue with them, but once agreement is reached, they face less risk of subterfuge.
These sorts of paradoxes leave me wondering whether I unfairly want to "shoot the messenger" at times when he tells me about a world that is more complex than I would like. I prefer, for the present, to interpret mainly the most stable relationships, given sample sizes constrained by the relatively small number of nations in the world, perhaps the frequently replicated relationships or those with p < .01 significance levels. Are there many consistent and strong relationships? Yes. For the present, I also prefer that, when each of us gets in the messenger role to speak about the more difficult aspects of culture, we speak softly. In the power distance example, perhaps more research will help me sort out the circumstances in which the dependent or the counterdependent reaction will occur.
I have several concerns about the book. One is that in terms of evidence for good overall validity for the set of nations taken as a group, some particular data points are probably misleading. This is the same as recognizing that data about any given individual may be problematic even in a well-validated selection instrument. For international management scholars, it implies that one should take care when relying exclusively on Hofstede's database to provide a theoretical context for research about a single nation or about culture differences between a small number of nations. Several studies of culture distance subtract the scores for a host of nations in which a multinational corporation (MNC) has overseas operations from the culture scores for the nation where the MNC has its headquarters. If this is done, the reliability of the culture distance scores are disproportionately affected by the accuracy of a single data point--that of the nation where the MNC has its headquarters. For example, I have some trouble being confident about the modest level of power distance reported in Japan, unless IBM systematically attracted Japanese employees who would have been misfits in Japanese companies. I would contest that many analyses of Japanese culture fit my own observation of quite a large power distance. Is my subjective assessment on this point unsystematic and biased? Perhaps. The good news is that Hofstede's review actually helps to answer questions like mine about the validity of any single nation's score on a particular dimension by directing our attention to other multiple-country projects. It also helps identify multiple sources of information for pinning down the culture of any single nation or small set of nations that scholars wish to compare.
One direction for improvement in the value dimension literature lies in continuing to look for distinctions among elements within each of Hofstede's dimensions and considering important elements of culture apart from values. Uncertainty avoidance includes rule orientation and xenophobia, for example. I find the cultural justification for linking these to be plausible, but I expect that they may well be empirically distinguishable. Reducing the field's emphasis on values alone by linking value dimensions to other equally basic social phenomena, like use of various components of social structure and social institutions, also continues to seem important (Smith, Peterson, and Schwartz, 2002).
Another improvement would be to find ways to adjust more effectively for response bias than Hofstede's approach of standardizing across a large number of correlated items. The attempt to cure response bias by standardizing the data across items in each nation is a substantial improvement over comparing means in raw data, but it is a little bit like the attempt to cure a bacterial infection with antibiotics. The antibiotics may well cure the infection, but they also create a host of other problems, or side effects. In the case of standardizing, these side effects are well known in the literature questioning the reliability of difference, change, or discrepancy scores (Cronbach and Furby, 1970), as welt as in the literature about changing the data structure by ipsatizing continuous measures (Closs, 1996). When one derives scores by subtracting correlated variables (including an item and a composite of other items, as is done when standardizing across items), the relationship among the variables is changed, and the reliability of any composite subsequently made from them is changed. On the positive side, and the reason that the standardizing practice seems not to have killed the good bacteria along with the bad that infected Hofstede's raw data, is that the correlations among the items being standardized are modest.
Hofstede's work is thorough and careful. It is also, paradoxically, both underappreciated and overused. It is underappreciated when criticized for characteristics it does not have, like devaluing ethnographic research or lacking evidence of empirical validity. It is overused, as noted above, in many applications to single nations. One can only hope that the field of international management will use this work as thoroughly and carefully as it deserves to be used.
Closs, S. J.
1996 "On the factoring and interpretation of ipsative data." Journal of Occupational and Organizational Psychology, 69: 41-47.
Cronbach, L. J., and L. Furby
1970 "How should we measure change--Or should we?" Psychological Bulletin, 74: 68-80.
Inglehart, R., M. Basanez, and A. Moreno
1998 Human Values and Beliefs: A Cross-cultural Sourcebook. Ann Arbor, MI: University of Michigan Press.
Inkeles, A., and D. J. Levinson
1969 "National character: The study of modal personality and sociocultural systems." In G. Lindzey and E. Aronson (eds.), The Handbook of Social Psychology, 2d ed.: 418-506. Reading, MA: Addison-Wesley.
Schwartz, S. H.
1994 "Beyond individualism/collectivism: New cultural dimensions of values." In U. Kim, H. C. Triandis, C. Kagitcibasi, S. C. Choie, and G. Yoon (eds.), Individualism and Collectivism: Theory, Method, and Applications: 85-119. Thousand Oaks, CA: Sage.
Smith, P. B., M. F. Peterson, and S. H. Schwartz
2002 "Cultural values, sources of guidance and their relevance to managerial behavior: A 47 nation study." Journal of Cross Cultural Psychology, 33: 188-208.
Mark F. Peterson
Department of Management
College of Business
Florida Atlantic University
Boca Raton, FL 33431
|Printer friendly Cite/link Email Feedback|
|Author:||Peterson, Mark F.|
|Publication:||Administrative Science Quarterly|
|Article Type:||Book Review|
|Date:||Mar 1, 2003|
|Previous Article:||The Handbook of Organizational Culture and Climate.|
|Next Article:||Organizational Culture: Mapping the Terrain.|