Much has been written in recent years about the decline in area studies in American universities. Due in large measure to the dramatically increasing influence of (deductive) rational-choice and (quantitative) econometric models, imported wholesale into political science from economics graduate students studying comparative politics at the top U.S. research universities today appear to be investing less time and energy in gaining deep cultural and linguistic knowledge of their country or region of interest, and proportionally more time studying formal modeling and statistical techniques. Most leading political science departments, for example, now require a rigorous sequence of quantitative and formal theory courses for their first-year graduate students. Many departments now also offer students the alternative of taking an advanced sequence of mathematics/statistics courses as a substitute for a foreign language requirement. Indeed, students in many graduate departments (including my own) may now choose ‘methodology’ as a subfield of political science to replace one of the traditional, substantive fields. Increasingly, if almost imperceptibly, methodology has shifted from being a set of tools, i.e. means used to study politics, to being an object of study, namely an end in itself.
If I may be permitted a personal observation, in the field of Chinese politics I have noticed a significant shift in the research interests of my graduate students in recent years, away from consideration of qualitative research problems (driven by an interest in major theoretical/conceptual issues and guided by intensive exposure to the country, its institutions, culture and its language) toward quantitative research agendas, often driven by the availability of ‘large N data sets’ incorporating several standardized socio-economic variables which can be more-or-less mechanically regressed against one another to derive ‘statistically significant’ tests of hypotheses (which were, in many cases, suggested by the availability of the data sets themselves). Such research is reminiscent of the old anecdote about the drunk who looked for his lost car keys under a street lamp. When asked why he was concentrating his search under the lamp, he answered ‘Because that’s where the light is!’
The problem lies not with the techniques and methods of statistical and formal modeling themselves, but rather with their tendency, when used in isolation from other, more traditional research methods, to facilitate the displacement of analytical thinking by mere technical procedure.
From: Baum, Richard. 2007. Studies of Chinese Politics in the United States. In China watching. Perspectives from Europe, Japan and the United States, ed. Ash Robert, Shambaugh David, and Takagi Seiichiro, 147–168. London: Routledge. quote from 160, italics in original.
And I found it in: Alpermann, Björn. 2009. “Political Science Research on China: Making the Most of Diversity.” Journal of Chinese Political Science 14 (4): 343.