Tom Pepinsky

Category: Politics

  • Interests, Ideas, and Identity

    Nothing is more productive for stale academic debates than a momentous yet unexpected political event. 2016 has seen at least two such events: Brexit and Trump. As a sometime-participant in a stale political debate on ideas versus interests in political economy, these two momentous events have been nothing if not intellectually productive.

    My hunch is that the next big conceptual move for political economy is to take identity seriously. The task for political economists now is to integrate identity into our theoretical architectures as a conceptual primitive rather than as a nuisance, a behavioral distraction, or as merely a consequence of something else.

    [If you are reading this and pounding the table, saying “I’ve been doing this for years,” please bear with me. I recognize that scholarship on identity and political economy has a long history. I am proposing that we need to do it more, and rather differently.]

    What would this look like? Take, for example, the literature about preferences for economic integration. One view holds that people’s preferences are a function of their economic interests: in the simplest Stolper-Samuelson world, low skill workers in advanced economies should oppose trade, and high skill workers should favor it. In a world in which ideas are foundational, individuals favor trade because they have learned (or come to believe in) a cosmopolitan worldview in which trade is good—they possess a causal belief about what trade does. In a world in which identity matters, people oppose trade because they are part of a community that opposes trade.

    The first thing to observe about this identity-based explanation is how vacuous it seems. “One does X because one is an X-doer” is infinitely generative claim, but that is because it is nearly tautological. Here is a more pointed way to think about identity in political economy. When people search for information about what they should do, how do they go about it? One view is that they consider costs and benefits, although not necessarily in a materialist or egoistic way. Another view is that they consider their existing beliefs about how the world works, and then try to apply them by analogy. The third is that they look for cues among the beliefs and actions of people whom they consider to be like them. That third possibility is an argument for identity in political economy.

    Another example: Why does Spain remain in the Euro? An argument about interests would suspect that it is in the interests of Spain’s leading sectors or a majority of citizens to remain within a monetary union with (most of) the rest of Western Europe. An argument about ideas would imply that the Spanish public and/or Spanish policymakers holds a belief about what leaving the Euro would mean. An argument about identity—one that seems consistent with my own impressions—would hold that being part of the Euro means identifying with a particular European project and aspiring to take part in that. My point is not that this argument is novel; far from it. Rather, my point is that explanations such as this one must be brought in to the “hard core” of political economy, whether it be from the perspective of open economy politics or critical political economy.

    The good news is that there are many places to start for integrating identity into political economy. At the individual level, social identity theory is (in my view) among the most productive literatures in the social sciences. At the aggregate level, so-called “constructivist international political economy” has tended to be more interested in ideas than in identity (there are exceptions) but can otherwise offer useful insights. There is a rapidly growing literature on identity and political economy in comparative politics and development economics, although this literature has tended to be more interested in either explaining why identity comes to matter (why nationalism? are ethnic groups “real”?) or on the political and economic consequences of ethnic diversity or polarization. In American politics, racial orders are an obvious place to start, but even something as simple as partisan identity ought to taken seriously by political economists as not just an uninteresting control variable but as a theoretically meaningful explanatory concept.

    So what should the road ahead look like? No one seriously believes that identity doesn’t matter at all in politics; the debate should be about when, under what conditions, how much, and in which domains? The first step will be to establish some empirical domain in which identity explains some political economy outcome that neither interests nor ideas can explain. The second and more interesting step will be to integrate ideas, interests, and identity into a coherent theoretical framework. Fortunately, there is good new theoretical work to build on. George Akerlof and Rachel Kranton and Moses Shayo and Sharun Mukand and Dani Rodrik have made essential contributions. The smart money is to follow these leads.

    There are some interesting conceptual challenges that political economists working on identity will need to ponder. In a political economy of identity, how one comes to understand one’s own identity is itself interesting. I suspect that one of major conclusions that we will draw from Brexit and Trump is that modern political economies and communication technology are more communally isolating than ever before, making it easier and easier to construct everyday social worlds in which we interact meaningful only with people who are “like us.” Sociology’s perspective on this is homophily in social networks. How do these networks facilitate identification itself, or shape how one understands new information about an in-group? As an exercise, pause for just a moment and ponder just what the concept of a “basket of deplorables” signified in this light.

    Another challenge is how to put identity into a causal model? Identities are not fixed or static, but emergent and situational—they can be shaped, they can evolve, and they can themselves be the consequence of identities and interests. One reason why identity has struggled in much political economy is that it is plausibly epiphenomenal on some other deeper thing. I outlined one version of this debate several weeks back, and my own thinking has evolved even from there.

    A related issue is the distinctiveness of identity as an explanation. At a recent meeting of the International Political Economy Society, one argument introduced by a colleague is that standard mainstream IPE fared relatively well in explaining the outcome of the 2016 election. Those who stand to lose the most from globalization and deindustrialization voted against the incumbent party. So do we conclude that our existing theoretical architecture “worked”? Or do we include that we must ask different questions, why a particular class of distributional losers understood their interests to be represented in one particular way by a party campaigning using a particular discursive strategy? There are no easy answers. But these are essential questions for knowing whether a research community, in this case IPE, must adapt in response to a momentous political event.

    I expect that this issue of identity in political economy will be a major theme in my own work over the coming years (see a preliminary output here [PDF]). I hope to see others take up this task as well.

  • Inferring Whether the Polls Were Correct

    Let’s say we want to estimate a quantity \theta. We form an estimate of that quantity \hat{\theta}_{A} = 51, with a 95% confidence interval of (49,53). Let’s say we form another estimate \hat{\theta}_{B} = 49, with a confidence interval of (47,51). And then it is revealed to us that \theta = 50. Which estimate, \hat{\theta}_{A} or \hat{\theta}_{B}, is the correct one? Can we infer that \hat{\theta}_{A} is correct?

    It is easy to see that the answer to the first question is “we can’t tell, the data are equally consistent with both estimates.” The second question is more subtle, but the existence of \hat{\theta}_{B} suggests to us that we ought to be cautious about inferring that \hat{\theta}_{A} is somehow “correct.”

    This toy example reveals something fundamentally rotten in the election polling postmortem.

    Many polling pundits are arguing this week that the polls were “correct” in some sense because polling results produced estimates with confidence intervals (or credible intervals) that captured the final two-party vote share, either nationally or by state. Here is one example but there are many others to find. The above example makes clear that such inference should not be drawn. If we call the election poll aggregates \hat{\theta}_{A}, the results from Tuesday’s election (call them \theta) are equally consistent with the hypothesis that the aggregated estimates were perfectly unbiased and that those estimates were biased upward by four points.

    The general point is this. The confidence interval around the two-party votes share estimate from polls reflects the standard error of the…estimate from the polls. It is not a confidence interval that captures the actual two-party vote share except under the hypothesis that the data generating process that produces the polls is the same data generating process as generates the vote. The same point holds for polling aggregates. We may not infer anything about the accuracy of the polls or the quality of the poll aggregates from the relationship between the election result and some confidence interval except for by maintaining that hypothesis. If we maintain an alternative hypothesis that the polls were systematically subject to substantial error in modeling turnout and/or voter intentions, these results are also consistent with many such hypotheses about the size of that error.

    This point has momentous implications for public opinion polling and for American democracy. If one makes inferences about the quality or correctness of polls from C.I. coverage, then one might conclude that there is no need to reevaluate the polls themselves. Estimates of uncertainty are necessary in public opinion polling, but they also make it hard to diagnose fundamental, systematic error. The more informative way to proceed is to identify errors, as Sam Wang has done (“The business about 65%, 91%, 93%, 99% probability is not the main point”), and going forward, to learn how to minimize that error.

    There is no way to avoid the secondary conclusion that this will be hard. As I wrote two days ago,

    Future aggregates for future elections by sites like 538 are going to use historical performance (i.e., prediction error today) to weight or “adjust” future polls. It is possible that some polls were more accurate than others because they had better models of turnout and voter intentions. It is also possible that all polls were just off (“correlated errors,” in the lingo), and some of these randomly happened to be less off than others. If the latter is true, then adjustments in the future will be worse than useless—they will be chasing noise.

    The future of election polling is not “whose polling aggregation method had the greatest uncertainty?” The future is “whose polls are the most accurate, and how do we know?” Anyone who suggests otherwise is either confused, or trying to sell you something.