When elephants fight, the grass gets trampled: the GLOBE and Hofstede projects

Abstract

Four aspects of the debate between Hofstede and Javidan et al. are highlighted. First, it is shown that characterizing cultures either on the basis of aggregated self-perceptions or on the basis of aggregated perceptions of others in one's society are not equivalent procedures. Each has inherent errors, and neither can be considered as providing the one best way to denote national cultures. Furthermore, the number of dimensions of national culture that can be usefully studied must be proportional to the limited number of nations available for comparative analyses. Third, although Hofstede and Javidan et al. appear to differ on optimal ways of aggregating individual-level data to the nation level, both appear to have done so in a way that does not prevent detection of differing relations between items at different levels of analysis. Finally, we need greater clarity as to the ways in which national wealth relates to other aspects of culture. It is a major component of contemporary national cultures, and must be retained as an element within nation-level analyses.