To read the full version of this content please select one of the options below:

A test of Hofstede's model of culture following his own approach

Michael Minkov (Sofia Campus, Varna University of Management-Sofia Local Center, Sofia, Bulgaria) (University of Tartu, Tartu, Estonia)
Anneli Kaasa (University of Tartu, Tartu, Estonia)

Cross Cultural & Strategic Management

ISSN: 2059-5794

Article publication date: 27 October 2020

Issue publication date: 29 April 2021




Recent studies exposed serious issues with Hofstede's popular model of culture, especially his uncertainty avoidance (UA) and masculinity-femininity (MAS–FEM) dimensions. However those studies did not focus on work-related issues as in Hofstede’s research.


We followed Hofstede’s approach to his dimensions more closely than anyone before in a large cross-cultural study. We used data from the nationally representative International Social Survey Program (over 50,000 respondents from 47 countries), measuring work goals and work-related stress in a way similar to Hofstede's.


UA and MAS–FEM, as measured and described by Hofstede, did not replicate. They lack internal consistency and the items that target them are not associated with Hofstede's UA and MAS–FEM indices. Instead, some of those items follow a very different and sound logic, invalidating Hofstede's UA and MAS–FEM theories. Our study provides additional evidence that UA and MAS–FEM are misleading artifacts of Hofstede's IBM database, with no analogues outside IBM. An improved, recently reported version of individualism-collectivism (IDV-COLL) replicated nearly perfectly, solidifying the validity of that dimension of national culture. A revised version of long-term orientation, called flexibility–monumentalism (FLX–MON) also replicated well.

Research limitations/implications

We discuss lessons for the cross-cultural field, including cross-cultural management, as well as policy-making by national governments, to be drawn from the controversial story of Hofstede's model. We advise a stronger focus on empirical confirmation and replication rather than excessive faith in fascinating, yet unproven theory.

Practical implications

To avoid further confusion, we advise researchers, consultants and managers to reconsider the use of Hofstede's UA and MAS–FEM and focus on the valid dimensions in the revised Minkov-Hofstede model.

Social implications

A number of national governments recently launched large-scale studies of their national cultures, based on Hofstede's model. The goal of those studies was to involve culture in the design of social and economic development policies. Studies of this kind should be founded on empirically sound models or else they can result in the formulation of flawed policies.


This is the first study of large samples from many nations showing that even when Hofstede's method is followed closely by focusing on work-related issues, UA and MAS–FEM do not emerge from the data, and this is not because of data deficiencies but because the logic of UA and MAS–FEM is demonstrably flawed. Our study also demonstrates new methods for the replication of IDV-COLL and FLX–MON, though without claiming that they are superior to existing ones.



This study was supported by the Estonian Research Council grant PRG380.Conflict of interest statement: The authors declare no conflict of interest.Competing interest statement: The authors declare no competing interests.


Minkov, M. and Kaasa, A. (2021), "A test of Hofstede's model of culture following his own approach", Cross Cultural & Strategic Management, Vol. 28 No. 2, pp. 384-406.



Emerald Publishing Limited

Copyright © 2020, Emerald Publishing Limited

Related articles