HOFSTEDE'S MODEL OF NATIONAL CULTURAL DIFFERENCES AND THEIR CONSEQUENCES: A TRIUMPH OF FAITH - A FAILURE OF ANALYSIS
University of Essex
ABSTRACT Geert Hofstede's legendary national culture research is critiqued. Crucial assumptions which underlie his claim to have uncovered the secrets of entire national cultures are described and challenged. The plausibility of systematically causal national cultures is questioned.
Do nations have cultures? Within each of the ‘management disciplines’ there is a significant literature which assumes that each nation has a distinctive, influential, and describable ‘culture’ As Hickson and Pugh declare: '[i]t ‘shapes everything’ (1995: 90).
Other than a priori belief, what is the basis of claims that influential national cultures exist? What is the quality of the evidence appealed to? Frequently, within the management disciplines, the causal-national-culture accepting literature justifies its reliance on the notion of national culture by citing approvingly the work of Geert Hofstede who claims to have successfully 'uncover[ed] the secrets of entire national cultures' (1980b: 44). Whilst Anderson has vividly described nations as ‘imagined communities’ (1991) and Wallerstein states that he is ‘skeptical that we can operationalise the concept of culture ... in any way that enables us to use it for statements that are more than trivial’ (1990: 34), Hofstede claims to have identified the four (later five) 'main dimensions' of national culture along which countries can be hierarchically ordered (1980a, 1984, 1991). By 1998 he could confidently claim that the scale of acceptance of his notion of distinctive-identifiable-influential national cultures was such that ‘a true paradigm shift’ had occurred (480) (see also Sondergaard 1994: 453).
Hofstede's model could be evaluated in a number of ways. It could be compared with alternative depictions of national cultures, especially with those which have emerged more recently (for example, Schwartz, 1992). His notion of culture and values could be contrasted with arguably richer conceptions of culture (for example, Geertz, 1973). His project could be dismissed as a misguided attempt to measure the unmeasurable (MacIntyre, 1971; Smelser, 1992). His findings could be judged solely on the basis of their predictive value by reviewing the many smaller-scale replications. The approach of this paper, however, is an evaluation of his research methodology. Are Hofstede's identification claims warranted; what is the quality of his evidence; what presuppositions does he rely on and are they justified?
The paper proceeds as follows. First, it briefly outlines Hofstede's identification 'model'. Secondly, it explores the implications of theorizing culture as national. Thirdly, it unpacks and critiques Hofstede’s claims to have ‘empirically’ (1980a) identified multiple national cultures or differences between such cultures by challenging five crucial methodological assumptions he makes. Finally, it considers the plausibility of a determinate relationship between national culture and uniform national actions/institutions.
Hofstede's main research on national culture is principally described in Culture's Consequences (1980a; 1984). On a few occasions he has added to his model, but he has never acknowledged any significant errors or weaknesses in that research. Indeed many of his subsequent publications are robust, at times aggressive, defenses of his 1980 methods and findings. As most readers will already be familiar with Hofstede's national culture research, a very brief outline only is given here. Where necessary greater detail is provided in the critique of his research methodology.
Hofstede's primary data was extracted from a pre-existing bank of employee attitude surveys undertaken around...