the divergent paths of sociology and political science

Fabio

Here’s a question that came up the other day in discussion with a colleague: Up until 1975 or so, political science and sociology were in similar boats, but only political science adopted mathematical modeling as a core research method in the last 30 years. Sure, there are pockets of mathematical soc folks, but now major chunks of political science have been recast in formal models, like theories of political conflict and competition (game theory), policy formation (Euclidean policy space models), legislative politics (more game theory), etc.  And all serious researchers are expected to know these theories. There has been no equivalent transformation of sociology. Why the divergence? Here are some are easily falsified hypotheses:

• Sociology is more comparative/institutional. Wrong – poli sci has deep comparative and institutional traditions.
• Sociology has ethnographers and historical types. Wrong – so does poli sci, especially non-Americanists,  but I’ll grant that we have a bigger group of them.
• Sociology is dominated by progressives who hate math. Wrong – leftists in poli sci (and econ) just recast their theories in the new formalism (see Romer and the analytical Marxists for examples).
• Sociology doesn’t have research problems that are easily recast in math. Wrong – soc has demography, mobility processes, etc.
• Sociology studies dynamics, while econ inspired formalism studies equilibrium states. Wrong – there is a lot of dynamic sociological models, such as stochastic process models, vacancy chains models, differential equations, etc.
• Sociology didn’t have the data bases or statistical technique that was comparable to poli sci. Wrong – Up until the 1970s, sociologists were often ahead of poli sci folks on surveys like the Wisconsin studies or the GSS. Soc folks were also ahead on path models and multiple regression when poli sci folks were still wowed by punch cards.
• Sociology gets the math deprived students. Wrong – True to an extent, but poli sci before the 1980s was never known as math central. They were just as math deprived as we are.

There are a few other hypotheses I can’t immediately rule out: people who like math dislike the topics sociologists study (crime, family, etc.), which might have been true pre-Becker, but that begs the question; math intense people have tried to infiltrate sociology through grad admissions, faculty hiring, or publication, but have been rejected; unobserved structural inertia of the field that preserves the research styles of the post-Parsons era. Any other guesses?