Measuring values for cross-cultural research

This paper investigates the empirical relevance of the recent critique that values surveys, as they are, suffer from the problem of measuring marginal preferences rather than values. By surveying items from cross-cultural surveys by Hofstede, Inglehart and GLOBE, we show that the marginal preferences problem is pervasive, even in cases where improved survey items are used. As an alternative to survey items about the importance respondents attach to various objectives, we therefore propose the use of so-called happiness or experienced utility functions as method to assess more directly what people across cultures care about and how much. Several recent contributions show the feasibility of such methods. We conclude that value measures based on cultural variation in the structure of happiness appear viable and superior substitutes for survey-based value indicators. Robbert Maseland, Radboud University Nijmegen, Department of Political Science, PO Box 9108, 6500 HK, Nijmegen, the Netherlands, Tel: +31 243 611 564, Fax: +31 243 612 379, E­ mail: [email protected]. André van Hoorn (corresponding author), Radboud University Nijmegen, Department of Economics, PO Box 9108, 6500 HK, Nijmegen, the Netherlands, Tel: +31 243 612 344, Fax: +31 243 612 379, E-mail: [email protected].