A renowned political scientist (…and me) discuss quantification in soc sci

On the latest (March 1, 2016) podcast episode of The Ezra Klein Show, Ezra from Vox chats with political scientist Theda Skocpol (perhaps best known in the Philippines for her 1985 work Bringing the State Back In along with Peter B. Evans and Dietrich Rueschemeyer) about “how political scientists think differently about politics.”

The United States is a bastion of quantitative social science, and poli sci is no exception to this trend, but there’s no shortage of researchers here who feel a bit uncomfortable with this. I’m in a master’s program called “quantitative methods in the social sciences”, so most of the people teaching me things are clearly on one side of the fence, not just in terms of how they study things but also in their epistemological approach to social research. There are people here, for example, who believe that qualitative research is pretty much worthless as long as it cannot come up with a formal system of defining and testing the plausibility of its propositions. Researchers from Columbia University and the University of British Columbia have recently proposed a framework for the systematic integration of quantitative and qualitative research.

For Theda Skocpol, however, the quantitative vs. qualitative debate is subordinate to what kinds of investigations are most appropriate to answer the particular research question one is interested in. Theda Skocpol elaborates below:

[38:59] Ezra Klein: You and I, in past conversation, have talked a little bit about frustrations with the way that political science has evolved in a much more quantitative direction, and that you think that there are blind spots this gives the discipline. I’d love for you to talk about some of what those are – what happens when you focus so much on what you can measure and what you can model?

Theda Skocpol: Well, I’m a big believer in quantification when you can do it. So, for example when I told you that I sat down and put together databases that squeezed everything I could out of news accounts, that’s so that I could create trend graphs, and in the case of our work on Americans for Prosperity we’ve actually come up with measures of the strength of right networks in different states that we plug into statistical models and show that they have a significant impact above and beyond partisanship.

So I’m not against quantification. What I worry about is getting fixated on certain kinds of data and ignoring other kinds. For example, because data on opinion surveys are online, because data on congressional votes are online, and you don’t even have to leave your armchair to manipulate those data, there’s a heavy emphasis on more and more arcane investigations of exactly the same questions asked using exactly the same kinds of data. I’m a big believer in systematically gathering new kinds of datasets, and where possible, using them in interaction with others in quantitative as well as qualitative accounts of political developments.

The other thing that makes me different is that I’m interested in change over time, not simply snapshots. And I think there’s a lot of incentive in standard political science to just look at a snapshot, and not look at, say, the timing of how things are unfolding. [41:04]

[45:35] Theda Skocpol: I wanna point to one other book that I recommend people pay attention to. It’s by Kathy Cramer. She’s a political scientist who goes about studying public opinion very differently than sitting in front of a computer and calling up the latest surveys. She actually went into every part of the state of Wisconsin, repeatedly, found groups of people who were sitting around cafes, or stores, or libraries, talking about affairs in their town, and just listened to the way they talked about politics, for several years. Obviously, took a lot of work. And then, she wrote about how people really construct their political world. And she found the answer to a puzzle that I think helps us understand what’s happened in Wisconsin, that even though a lot of people outside of Madison and Milwaukee are actually benefiting from government programs, they think government serves only the people in the cities and the big university centers. They have an us-versus-them understanding of government and politics, they feel they’ve been betrayed by the political process. [46:51]

So I posted this because I agree with Skocpol. Later on she talks about how the best way to approach academic research is not to ask what the next question in the literature is, but to ask what kind of puzzle needs to be solved, and whether or not anyone outside academia has a stake in it.

I want to point out something else, however. A lot of people doing social science in the Philippines think they agree with Skocpol, but I don’t think they do. They will hear or read something like the above from Skocpol or from many other sensible people, and they will nod their heads in self-righteousness at silly quantitative researchers with their software programs thinking they can understand society with a few lines of code and fancy models. In this they will reveal that they have no idea what Skocpol is talking about.

There is a dearth, a scarcity, a downright famine of quantitative research in the Philippines in disciplines that don’t start with “e” and end with “conomics”. And there are many people who have an ideological stance against this kind of research, with pithy soundbites like “you can’t reduce people to numbers” or “you’re falling into the trap of logical positivism” or “society is too complex to be modeled”. Those are correct only to some extent – all models are wrong, but some are useful, especially when we can study a few people and be able to generalize to populations, or if we can uncover trends from messy-looking sources of data without having to manually sift through everything, or if we can describe the relationship between multiple measurements of aspects of society, or if we can narrow down the causes of a social phenomenon or the effects of a policy through careful experimental design.

There are many critiques of quantitative research and its overreach. These critiques make sense in the United States where many people do indeed think that methods such as ethnographic fieldwork and unstructured interviews can result in nothing more than opinion pieces that contribute nothing to a larger understanding of society.

These critiques make no sense in the Philippines when almost no one is doing the kind of quantitative work I describe above. The best way forward is not to sort ourselves into one camp or the other, but to be skilled enough to collect whatever data and undertake whatever methods of investigation are needed. And for that, we need future social scientists to be as capable with numbers and variables as they are with words and images.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s