Hall, Hofstede, Huntington, Trompenaars, GLOBE: Common Foundations, Common Flaws
The notion that “national culture” shapes the behavior of the populations of discrete national territories (countries) both within and outside of organizations (for example the decisions and actions of managers and consumers) has extensive support within both the academic and management consultancy communities (Breidenbach and Nyiri 2009). Research, teaching and training which attribute such causal power to national culture rely heavily on the conceptions and descriptions of such cultures by Geert Hofstede; the multi-authored Global Leadership and Organizational Behavior Effectiveness project (GLOBE); or Fons Trompenaars2
(hereafter, ‘the Trio’). An indication of the popularity of their work and of “national culture” as an explanatory variable within the academic (overwhelmingly management) arena is that Hofstede’s research is one of the most cited in the entire Social Science Citation Index (Parboteeah, Hoegel, and Cullen 2008). Those citations include critiques, but largely they are supportive. His magnum opus
—has become an almost emblematic citation in a number of management disciplines. Both GLOBE’s and Trompenaars’s research is also very widely cited (Tung and Verbeke 2010). The popularity in management of the Trio’s claims—despite the theoretical, empirical, and practical limitations of their research—is, perhaps, not surprising, given the increasing pressures on academics and others to be instantly and visibly “relevant” (March and Sutton 1997) during an era
of enormous acceleration of the inter/trans-nationalization of business and markets. Their views about the characteristics and consequences of “national culture” receive legitimacy from deeply entrenched belief in national primordiality and uniqueness (Willman 2014; McSweeney 2009).
Although the Trio have at times engaged in intense criticisms of each other’s research, they have much in common. Their differences are, as Earley states, only “minor variants on one another’s styles” (2006, 923). The postulates they share include—national cultures are: (1) values—defined as invariant transituational preferences; (2) universally shared by the population of a country; (3) coherent (contradiction-free/integrated); (4) the fundamental cause/source of behavior and artefacts; (5) stable; (6) identifiable from the mean scores of answers to self-response survey questions from a minute portion of a national population; (7) depictable in league (ranking) tables (indices) of “dimensions” (quantifiable comparators; see Taras and Steel 2009; McSweeney 2002a). In short, each national population location is conceived of as a container of an undifferentiated measurable culture which molds the social in its supposed geographical domain.
Each of the seven postulates (above), but 6 and 7 especially, have been critiqued (see Bock 1999, 2000; Breidenbach and Nyíri 2009; Brewer and Venaik 2012 Duncan 1980; Earley 2009; Fang 2005, 2012; Gerhart and Fang 2005; Harzing 2006; Johnson, et al. 2005; Kitayama 2002; Kuper 1999; Maseland and van Horn 2010; Magala 2005 McSweeney 2002a, b, 2009; Moore 2012; Schwartz 1994; Willman 2014). Space does not permit a review of the commentary on all of the postulates. Instead, the chapter considers some aspects which have received comparatively less attention in the management literature. These are: downward conflation (the belief or assumption that the macro (in this instance the national) is replicated at, indeed creates, lower hierarchical levels (organization, individuals, or whatever). More specifically it discusses a crucial methodological error (the ecological fallacy) which characterizes an extraordinarily large number of papers purporting to apply the findings of one or other aspect of the Trio’s work. This section also considers the cultural generalizations of Samuel P. Huntington who shares postulates 1 to 5 with the Trio and those of Edward T. Hall who shares postulates 2 to 5. However, his primary focus is on the populations in or from multinational/regional locations which he calls “civilizations.” Although the Hofstede, Trompenaars, and the GLOBE study also make general claims about such large populations, they do so to a lesser extent than Huntington. The chapter then considers the implied, and sometimes explicit assumption, in their work and the work of their followers that national cultures are coherent, that is, integrated and non-contradictory.
National culture cannot logically be said to have uniform and enduring national “consequences” without that invalid assumption. Following that, it discusses the postulate of the fixity of national boundaries within which unique and stable national cultures are said to be located. Finally, it considers a misleading representation of intra-national variation—the view that countries are composites of multiple mono-cultures. The discussion of these matters is preceded by a commentary on the notion of culture employed by the GLOBE, Hofstede, Huntington, and Trompenaars (but not by Hall).
Culture as Values
What is, or is meant by, ‘culture’—national or other? As early as the 1950s, Alfred Kroeber and Clyde Kluckhohn estimated—in a survey of English language sources only—that there were already over 160 definitions of culture (“and its near-synonym civilization”) in use (1952). One can distinguish between at least five types of cultural theories: psychological (culture as subjective values); mentalism (or cognition); textualism; intersubjectivism; and practise theory. On a very basic level these schools offer opposing locations and conceptions of culture. However, as Michael Hechter observes, few concepts of culture “are bandied about more liberally in popular, normative, and explanatory scholarly discourse than that of values” (1993, 1; emphasis in original)—defined as enduring transituational preferences. Taras, Rowney, and Steel’s analysis of 136 publically available instruments for measuring culture revealed that almost all existing instruments and their underlying models of culture are focused on values and overlook other attributes of culture (2009). Within many sub-fields of management, it is the bedrock definition of culture. It is this notion of culture that the Trio relies on. GLOBE might seem to have also compared countries on the basis of “practices” and thus produced two different sets of cultural indices. But both are espoused value indices, as what GLOBE labels “practices” are not practices in the sense of the actions of individuals or collectivities but merely respondents’ views about existing social values “in my society” (Earley 2006).
But in any event, the identification of culture defined by GLOBE, Hofstede, Trompenaars, and Huntington—as causal transituational preferences—confronts a variety of impediments. These include: their unobservability; the possible roles of a host of other psychological constructs (desires, goals, motives, needs, traits, aversions, tastes, valences, sentiments, and so forth); the multiplicity of definitions of values—as early as 1963 Campbell listed 76 uses of the term—and the opaque link between values and actions.
In contrast, Swidler (1986), for instance, states that the values “model used to understand culture’s effects on action is fundamentally misleading” (273). John Meyer and his colleagues state that “[a] notion of ‘abstract values internalized by individuals through socialization simply leaves out too much’ and is given ‘too much reified inevitability.’” They describe as “primitive” the notion of culture as “a cluster of consensual general values” (Meyer et al. 1994, 11–12, 17) and they explicitly reject the Parsonian idea (which the Trio heavily draw on) of a general value system into which individuals are socialized (Meyer et al. 1994, 12; see also Joas 2000; Rohan 2000; Ailon 2008; Bock 1999; Breidenbach and Nyíri 2009; Cooper 1982; Gerhart and Fang 2005; Maseland and van Hoorn 2010). Furthermore, the assumption that values are unaffected by context—that they are invariant transituational preferences—is also at odds with an immense amount of contrary evidence (Ewing 1990; Shweder 1999).
Culture as Determinate
“National culture” (hereafter “culture”)—if it is assumed to exist—can be theorized on a range from the scarcely significant to the dominant driver. The view that culture has “affects,” “effects,” “influence,” “consequences,” “manifestations,” “impacts,” or “outcomes,” whether deemed weak or strong, are distinguishable from claims that merely point to possible statistical relationships, associations, or correlations. At one causal pole, culture is represented as the foundation for just about everything social. Culture supposedly orchestrates behavior. An important, perhaps the most influential, attraction of the Trio’s depictions of culture is the breathtaking claim that it shapes the social action of defined populations enduringly and predictably. It “affect
[s] human thinking, feeling, and acting, as well as organizations and institutions,” Hofstede and Hofstede state, “in predictable
ways” (2005, 31; emphasis added). “[L]anguage, food, buildings, houses, monuments, agriculture, shrines, markets, fashions and art,” are Trompenaars states, “symbols of a deeper [subjective] level culture” (1997, 21). Subjective values are treated as the incontestable causal core. The ontological status of the “inner” is distinguished from the “outer” (institutions, practices, and so forth) but at the same time national culture is their cause. “Values,” Allport (1961) states, are “the dominating force in life” (543). “Culture,” Etounga-Manguelle states, “is the mother, institutions are the children” (2000). Johns (2006) describes culture as “a contextual imperative.” In the tradition of early to middle Parsons (1951), culture is conceived of as a “normative pattern structure of values” (37) which act as a hierarchically superordinated control system (Schmid 1992). Beneath, or behind, the “luxuriant variety,
even apparent randomness” (Ortner 1984, 136) of life is posited a causal psychoculture. The basic idea, as Clifford Geertz critically observes, is that culture is “a set of control mechanisms—plans, recipes, rules, instructions (what computer engineers call ‘programs’)—for governing behavior” (1973, 44). It is held to be as Hofstede described it, “the software of the mind. In contrast, Geertz states, “culture is not a power, something to which social events, behaviors, institutions, or processes can be causally attributed” (1973, 14).
GLOBE, Hofstede, and Trompenaars occasionally, and inconsistently, mention possible causal influences in addition to national culture. But mere acknowledgement of other cultures or non-cultural factors without incorporating them into their dimensional models of causally determinate values is an empty gesture (McSweeney 2009).
What evidence do the Trio provide of the power of culture? There is zero empirical evidence derived or derivable from their questionnaire and/or interview-based depictions of national cultures or statistical representations of those cultures, of an influence on individuals’ behaviour (Gerhard and Fang 2005). The asserted link between the descriptions of a national culture and national action is not extracted from, and is not extractable from, respondents’ answers. It is presupposed.
Societal-level models of all types, not just the cultural, often lack clarity about causality (Oyserman and Uskul 2008). A “cause” is described (well or badly) as is/are the outcome(s). But the causal process, the linkages between cause and outcome, is too often not unfolded for the reader at least. Instead of descriptions of situated causal mechanisms, the mere fact that two conditions exist, or are supposed to exist, in the same time and space is, together with a general causal theory, treated as sufficient evidence that one caused the other.
There is an inverse relationship between the compoundness of a concept and the number of cases attributable to, or covered by, it (Mahoney 2004; Sartori 1970). Sub-national analysis will often demonstrate the information poverty of national averages and reveal considerable heterogeneity within countries (Smith, McSweeney, and Fitzgerald 2008). Even if we suppose that (a) the Trio’s mean values scores are accurate national averages; and (b) that values are causal—both highly contested notions—deducing the subnational from anyone of the Trio’s averages and rankings is at best wholly speculative. As Starbuck states: “comparisons between averages may say nothing about
specific situations” (2004, 1245). National-level data obscures considerable within-country variation. Of course, there are some national uniformities; for instance most cars are driven on the right-hand side of the road in Brazil; in India many drive on the left-hand side—because of legal requirements, a legacy of British colonial rule. The claim that national uniformities are a consequence of “national culture” is a mere assertion that ignores other possible explanations. In any event, there is a vast body of empirical data depicting considerable behavioral variation within
countries (see, for example, Au 1999; Camelo et al. 2004; Goold and Campbell 1987; Huo and Randall 1991; Kondo 1990; Lenartowicz et al. 2003; O’Sullivan 2000; Streeck and Thelen 2005; Tsurumi 1988; Weiss and Delbecq 1987; Yanagisako 2002). In short, as Peterson, Arregle and Martin (2012) state, there is an increasingly documented variability in cultural, institutional, and economic characteristics within nations.
Examples of positive correlations between one or other of the Trio’s measurements and a practice (organizational or other) are sometimes held out to be evidence of a causal relationship between the former and the latter. But a correlation, of itself, is not evidence of causality and almost any causal theory will generate some correct predictions. Thus, identification of confirming examples is not proof that a theory is correct (Starbuck 2004). Positive examples can be found for almost any theory. For example, table salt dissolves in warm water every time someone utters a “magic word” before immersing the salt in the water. Looking only at positive examples fails to reveal a vital falsification. The salt is, of course, equally likely to dissolve without the “spell,” as the spell has no influence, but a positive test stra...