Review:
Culture’s Conquences: Comparing Values, Behaviors, Institutions and Organisations Across Nations by Geert Hofstede, 2001, Second Edition, (Sage, Thousand Oaks, CA) US$97.95, hardcover, pp. 596.
he publication of this cond edition of Culture’s Conquences marks an important moment in the field of cross-cultural studies. The first edition of this book, published in 1980, launched what some have called a revolution within the social sciences. Rearchers who previously had not questioned whether their theories of human behaviour were applicable to all people were suddenly confronted with a systematic framework of cultural dimensions suggesting that not only behaviour but the process and mechanisms governing behaviour could fundamentally vary bad on four cultural dimensions. The first edition of this book brought the field of cross-cultural studies to the forefront (or at least the front) of social science rearch. But Hofstede’s framework has come under inten scrutiny since its first publication, and, although this cond edition makes many needed changes and additions, it will be significant to e whether the intellectual community still finds simplistic frameworks, even as influential as Hofstede’s, to still be relevant to cross-cultural rearch. T
The cond edition is an appropriate place for business rearchers interested in doing cross-cultural
work to get an introduction to the field, as it includes a (limited) survey of much cross-cultural work within the past twenty-five years, as well as introducing the above mentioned framework. Yet, if new rearchers are inspired to engage in cross-cultural rearch, they must go beyond this framework to adequately reprent the dynamic and complex nature of culture on psychological process and behaviour. Hofstede’s framework is perhaps most appropriate as a teaching aid for introducing the notion of cultural differences to business students, as it provides an easy-to-understand framework with numerous examples of their practical applicability. Having studied under Hofstede during my undergraduate study-abroad days at Maastricht University, I have successfully followed his lead many times with my global marketing students in teaching such topics as intercultural marketing communications. Although Hofstede’s framework for understanding national differences has been one of the most influential and widely ud frameworks in cross-cultural business studies, in the past ten years or so it has also become one of the most widely criticized. Detractors contend that his dichotomid way of reprenting cultural differences leads to unjustifiable generalizations and ignores the subtleties and frequent contradictions inherent in many national cultures. Many social scientists contend that there is in fact no such thing as national culture, since subcultures within a country can vary so greatly in their values and beliefs. Moreover, the original data Hofstede prented in 1980 in the first edition of this book, from which he derives his framework, have been misunderstood and applied
in inappropriate ways, which has also led to criticisms of the framework from a methodological stance.
In this cond edition, Hofstede does not prent new data (the data prented are the same original data collected from IBM employees in the late 1960s and early 1970s), but includes in his discussions of each identified dimension of culture the findings from numerous studies which have subquently ud his framework, includes some additional countries, removes dated material, redoes calculations, and responds to his critics head on.
Impact of the First Edition of Culture’s Conquences in Business Rearch Although Hofstede’s work has largely fallen out of favour with many business scholars for a variety of reasons that will be mentioned shortly, it is still widely ud in business rearch. And there is no denying the huge impact his work has had on business thinking in the past twenty years. Hofstede himlf describes his impact as paradigm shifting (p. 73), and indeed his work was what made many business scholars first question the universal applicability of some of the most common business models—models that have since been shown to be culturally variable, such as the diffusion of innovation model and the advertising effects model.
The data published in the first edition as well as this cond edition of Culture’s Conquences were collected during a 53–country project to try to systematically understand cultural similarities and differences around the world, although in this edition 19 more countries are added, using data collected subquently by Hofstede, as well as by other rearchers. Using IBM employees in each of the 53 countries becau they are similar to each other on almost all dimensions except culture, Hofstede measures cultural values relating to spheres such as interpersonal relationships and hierarchies. Four dimensions of culture that helped explain the differences among the respondents emerge:
1. Uncertainty avoidance—a society’s tolerance of the unpredictable;
2. Power distance—a society’s acceptance of the unequal distribution of power;
3. Individualism/Collectivism—the extent to which the interests of the
individual prevail over the interests of the group within a society; and
4. Masculinity/Femininity—the relative strength of masculine vs. feminine
values in a society.
Since the initial publication of the four dimensions, numerous other rearchers have ‘confirmed’ the dimensions in various cultures around the world and utilized them to analy a number of business issues, such as variations in consumer behaviour, management practices, and portfolio management.
In 1987, a group of rearchers bad in Hong Kong and headed by Michael Bond published results from a Chine Value Survey they had administered in 23 countries (Chine Culture Connection 1987). This survey replicated three of the original four cultural dimensions that Hofstede had reported on, but their fourth dimension was different, as the survey was prepared bad on Eastern values rather than Western values. They termed this dimension long-term/short-term orientation, and when Hofstede re-published his results in a textbook version of the first edition
of Culture’s Conquences entitled Cultures and Organizations in 1991, he included this dimension as a fifth dimension of culture. It is also included as an added chapter in this cond edition of Culture’s Conquences.
Each country receives a score within a range of about 0–100 on each of the five dimensions. It is important if utilizing the scores in rearch to note that the scores place each country in a relative, not an absolute, position.
Hofstede notes that his cultural scores are indicative of a proclivity across the entire country that he has measured; the scores are not variables to be ud to predict individual behaviour. Of cour, within countries there is variation in levels of all the variables, and his measures do not account for this. This has been a primary area in which his data and framework have been misunderstood within business rearch—one cannot predict individual behaviour using his framework. Rather, the unit of analysis is at the countrywide level, and thus predictions can be made only at this generalized level.
This cond edition includes a chapter on all five of the above dimensions, explaining at length what they mean and reprent, and incorporates findings from the 1980s and 1990s into the discussions of each dimension.
Notable Additions to the Second Edition
The cond edition emphasizes the reporting of the many ‘validations’ of Hofstede’s framework that have occurred since the first edition was published. But this ems to be somewhat tautological, as he primarily includes only the studies that have confirmed his original framework. He neglects to include the many empirical studies that have occurred wherein results have challenged his framework, and he explains away the minor empirical variations that he does report on that have appeared in the literature.
Another marked addition to this cond edition is that Hofstede openly acknowledges to a large degree the Western bias of the entire rearch project, to the extent of outlining his own history, beliefs, values and scores on all dimensions, so they can be taken into account by the reader when evaluating and interpreting the results he prents.
In a widely expanded introductory chapter on the nature of values and culture, Hofstede flags the ecological and rever ecological fallacies early on. The ecological fallacy is when data collected at a countrywide level (e.g. Hofstede’s data) are ud to predict individual behaviour. Hofstede’s data have been ud in this manner numerous times in business rearch, and as mentioned earlier this is highly inappropriate. There will be a diversity of, for instance, people who score higher or lower on the masculinity scale in any given country. When a country scores high on the masculinity index, it implies there are more people in that country who subscribe to masculine values, but it does not say anything about how to determine whether any given individual will score high on the values. Without confirming at the individual level whether a person indeed subscribes to the values indicated at a countrywide level, a rearcher cannot assume an individual will act in the way ascribed to his/her country in general. Becau his data have been misud so often in this manner, Hofstede is careful to explain the units of analysis issues in detail.
The rever ecological fallacy occurs when variables correlated at the individual level are then ud to explain countrywide data. Although this error occurs less than the ecological fallacy, he cautions that this is also highly inappropriate (p. 16). Most importantly, he points out the issues are not merely measurement issues, but paradigm issues. Within psychology, a special issue of Journal of Cross-Cultural Psychology has recently been published in which just the paradigm issues of how to conceptuali and measure differences purportedly bad on national culture should be handled are explored. (See Singelis 2000 for a summary.)
In respon to tho critics who have contended that his data are stale, Hofstede notes (p. 66) that there is no one t of measures or items on a measure that will always account for national differences throughout time. He explains, for example, that the original questionnaire ud to measure masculinity in the 1970s has been modified in more recent studies, since the meaning of this construct in various societies has changed since then. Having acknowledged this, though, Hofstede still has a tendency to rationalize variations in the replication studies that he discuss rather than openly discussing whether his five cultural dimensions are still meaningful in the way they were envisioned.
With reference to the long-term/short-term orientation dimension, it is interesting to note that the othe
r four dimensions tend to divide in an East–West division, and this dimension does not. Hofstede, however, claims in this cond edition that the existence of this value ‘proves’ that there is a cultural basis for the economic growth of the so-called Asian dragon countries (p. 368), as there is a correlation between countries that scored high on this dimension and their unexpected growth during the 1990s. Of cour, growth of the dragon countries has dramatically slowed since the recent Asian economic crisis, and Hofstede admits the relationship between this dimension and economic growth could be temporary. Many scholars have pointed out, however, that attributing the dragon countries’ growth to culture is spurious, as the same cultural traits were often ud in the past to justify why they were economically ‘backwards’.
In the final applications chapter (intercultural encounters), Hofstede recognizes that his results suggest a relativism with respect to such macro-level concepts as democracy (he argues it is not necessarily the correct political system for all countries), capitalism (same as above) and human rights (a goal such as the UN has of coming up with universal human rights would not be supported by his data). While arguing for an incread relativism with respect to political, economic and moral systems may be a laudable goal, it also rves to illustrate one of the primary drawbacks of this cond edition: making sweeping statements where they are not justified by the data, and relating th
e framework to literally every aspect of human society and behaviour. For instance, Hofstede cites archaeological evidence that 4,000 years ago there were centralized governments in the Middle East and democracies in Scandinavia as evidence that the power distance variable has been at work for millennia (p. 117). It would appear that Hofstede subscribes to the view that all historical, political, economic, social (or any other) events that have ever happened throughout history, in the prent, or in the future are related to and can be explained by his national culture dimensions. While his framework may be appropriate for accounting for certain obrvable behaviours (i.e. the differences in rates of eating out in various countries), most business and cultural rearchers
would hardly buy into the argument that his framework of values can account for all world occurrences.
Summary
Hofstede outlines what he believes to be the five primary criticisms his original work has received (p. 73): that surveys are not a suitable way to measure cultural differences; that nations are not the best units for studying culture; that studying the subsidiaries of one company is not reprentative of national cultures; that the data are old and obsolete; and finally that five dimensions are not enough t
o reprent the complexity of culture. He answers each of the criticisms with a reply along the lines of, ‘things could be improved and changed, and I hope they are, but this rearch is a valid and important start’. He insists his work was a paradigm shift, and argues throughout this cond edition that the cultural tendencies he has tapped into are centuries old, and thus his data are not outdated. Hofstede firmly maintains throughout this edition that cultural tendencies and values are inherently stable, a stance many anthropologists would take umbrage at. Most social science rearchers, including in the business disciplines, now u a dynamic constructivist model of culture when engaging in cultural analys, and this development in cultural studies ems to have pasd Hofstede by.
Another weakness of this cond edition, which has been alluded to already, is his tendency to only review literature that supports his framework rather than a reprentative sample of the literature that has employed his framework. In discussing the marketing applications of his framework in chapter 9, for instance, he exhibits a particularly limited review, citing primarily the work of de Mooij (1998), a former associate of his, when in fact his work has been ud by multiple marketing rearchers, and a more balanced review would illustrate both the impact his work has had on the field as well as the inappropriate us of his framework within the field.
Overall, the main problem with Hofstede’s rearch program, including its reprentation in this cond edition, is that Hofstede holds onto a static vision of culture, a view no longer tenable in light of decades of social science rearch suggesting otherwi. He us historical occurrences to tautologically justify his five dimension scores, which implies that almost all behaviour and indeed historical conflicts are value-bad, another contention argued against in many business and social science disciplines.
In the end, the results of Hofstede’s rearch program can be uful to business rearchers if their goal is to compare actual behaviour (not process or motives) at the countrywide level. For example, Houston and Eckhardt (2001) u Hofstede’s categorizations to cross-culturally investigate a wide variety of obrvable behaviours relating to food consumption such as brand choice and brand loyalty in a wide variety of Asian countries. Such obrvable measures are descriptive accounts of phenomena that reveal themlves in a similar manner across cultures. Thus using a framework such as this is appropriate and can help to identify patterns of similarity in a global context. If one is going to u a framework like this, Hofstede’s is the most comprehensive and validated. Even then, the results of studies using his framework typically suggest new areas of inquiry: they are not conclusive. Hofstede himlf acknowledges that one of the
