By Marleen Bornat
Steinmo (2008), Bates (1998) and Tilly (1990) propose different approaches to the study of social science and history. Tilly (1990) categorises these approaches by distinguishing between small-scale versus large-scale analysis and humanism versus social science. I will use my capstone topic, the political representation of women, and show how the two methods advocated by Bates (1998) and Steinmo (2008) would change the type of research question one would investigate and the kind of conclusions one is able to draw.
Bates (1998) proposes a method he dubs “analytical narrative” which combines historical narratives (accounts and stories) with the analytical rigor of rational choice theory, particularly game theory, to investigate a problem/puzzle. If one wants to examine the underrepresentation of women in the political sphere, one may for example ask: “Does clientelism pose an obstacle to women’s political representation?” To research this question, one would conduct interviews and surveys with politicians while also designing a game that helps to uncover behavioural patterns. This method may (i) yield new predictions that can be examined and (ii) provide a model that can be tested in other cases.
Steinmo (2008) advocates a different method named “historical institutionalism” which looks at institutions to unravel the sequencing and change of social, economic and political behaviour over time. It views human beings as both norm-abiding and self-interested and studies individuals as well as institutions in a particular context (ibid.). A research question one would investigate as a historical institutionalist would be: “How do institutions influence women’s descriptive representation in the German legislature?” Consequently, one would holistically analyse how institutions structure political behaviour, by for instance empowering and/or constraining male and female politicians differently across time. Historical institutionalism would only allow for conclusions to be made about how institutions constrain women’s representation in this particular context (this context can range from several countries to individuals).
Although the aforementioned methods as well as categories proposed by Tilly (1990) appear to be relatively fixed, most research rather falls somewhere along the continuum and usually only identifies to some extent with a particular category.