Why empirical political scientists and political theorists talk past each other

Why do empirical political scientists and political theorists define democratic quality so differently? A diagnosis and prescription.

The latest issue of Perspectives on Politics, which just went up, includes my article “The Two Cultures of Democratic Theory: Responsiveness, Democratic Quality, and the Empirical-Normative Divide.” I’m always behind the times when it comes to paywalls, but here’s my best shot at a link for “blogging”:

The Two Cultures of Democratic Theory: Responsiveness, Democratic Quality, and the Empirical-Normative Divide

This is for professional political scientists, and admittedly harder going than the book review I blogged about a couple of days ago. The basic idea is that empirical political scientists very often use as their assumed measure of democratic quality “responsiveness”: the extent to which changes in public policy reflect changes in the preferences of the public, or the median voter. Political theorists, on the other hand, almost never define democratic quality this way: there are a host of other things that democracy is supposed to be about. I try to hash out where the disjunction came from; why it matters; what each side can learn from the other; and why there’s still room for legitimate differences and a division of labor. (Teaser version: we should expect people who measure political phenomena for a living to seek rough agreement on how to define what they’re measuring. We should also expect people who study political concepts and political values for a living to be legitimately dissatisfied with the inevitable simplification this entails.)

Academic readers should be able to download the full version easily as an .html or .pdf. Anyone else whose interest is piqued by the abstract should email me at my academic email address (not hard to find) and I’ll send you a version within a few days.

Update: The article, along with the rest of what looks like a great issue (I’ve been on vacation and have lacked the time to read it), is now ungated through the end of July, courtesy of Cambridge Journals. The link is: http://journals.cambridge.org/action/displayIssue?jid=PPS&tab=currentissue

Author: Andrew Sabl

I'm a political theorist and Visiting Professor (through 2017) in the Program on Ethics, Politics and Economics at Yale. My interests include the history of political thought, toleration, democratic theory, political ethics, problems of coordination and convention, the realist movement in political theory, and the thought of David Hume. My first book, Ruling Passions: Political Offices and Democratic Ethics (Princeton, 2002) covered many of these topics, with a special focus on the varieties of democratic politics and the disparate qualities of mind and character appropriate to those who practice each of them. My second book Hume's Politics: Coordination and Crisis in the History of England was published in 2012; I am currently finishing a book on toleration, with the working title The Virtues of Hypocrisy, under contract with Harvard University Press. A Los Angeles native, I got my B.A. and Ph.D. from Harvard. Before coming to Yale I taught at Vanderbilt and at UCLA, where I was an Assistant, Associate, and Full Professor; and held visiting positions at Williams, Harvard, and Princeton. I am married to Miriam Laugesen, who teaches health policy and the politics of health care at the Mailman School of public health at Columbia, and we have a twelve-year-old son.

4 thoughts on “Why empirical political scientists and political theorists talk past each other”

  1. Count me on the side of the political theorists here, though I'm not a member of the guild. The trouble with "responsiveness" is not so much that the plebes don't understand what is going on (the élites don't either: see the Greek crisis). It's also that no photograph of citizens' preferences can be guaranteed against circular Condorcet paradoxes, as Arrow proved. I argue here that such paradoxes are not a theoretical curiosity but in fact extremely likely, in mismatches between means and ends. Ergo, any workable politics must include horse-trading. An ideal democracy is a learning, Socratic polity.

    1. James–Relating to your article on cyclical majorities, I see a physical analogy in familar events that have had very bad consequences and proved very important to understand.

      I happened to be watching the US Open golf tournament yesterday, and they showed the Tacoma Narrows Bridge. This brought to mind the original Tacoma Narrows Bridge, known as Galloping Gertie, that destroyed itself in 1940, a few months after it was opened. That article in Wikipedia is interesting: https://en.wikipedia.org/wiki/Tacoma_Narrows_Brid

      In the article it states that the disintegration of the bridge was long attributed to forced resonance, but later was understood to be a manifestation of aeroelastic flutter, a much more complex phenomenon.

      Following the trail to aeroelastic flutter, I found the article on the disastrous problems in Lockheed Electras, in which the in-flight disintegrations were originally attributed incorrectly to metal fatigue with inadequate QC. https://en.wikipedia.org/wiki/Braniff_Flight_542

      Only later was the phenomenon of aeroelastic flutter understood, which led to significant changes in the field of aeronautical engineering design.

      Why these divergences from your original topic? Because I think the phenomenon is a very close analog of the mechanism of popular voting in a two party system. Whichever direction the majority moves the middle, the forces that tend to move it back in the other direction become stronger than the forces keeping it at its then-current location. And the farther the middle moves "off center," the stronger the forces become, while the longer the periodic frequency becomes.

      1. Democracies can certainly self-destruct, like autocracies and oligarchies. That of Athens just made bad decisions about Philip of Macedon. The Roman Republic is a better example, as warlord entrepreneurs like Pompey and Caesar bought elections to consular office that gave a legal cover to their essentially private wars and expanding armies. But the analogy to hunting in physical systems seems far-fetched. There are also two-party political systems that have managed to stay quite stable over long periods: eg Britain ever since 1700. That includes major, life-and-death wars: 1792-1815, 1914-1918. The suspension of party politics in 1940-45 was quite exceptional.

        Americans have had a distrust of parties ever since Madison. Environmental columnist John Roberts (formerly of Grist, now at Vox) nails it: American progressives hanker after bipartisanship, convincing their adversaries by argument. The alternative and much more realistic strategy is simply to go for the win.

        Talking about Roman elections. Why is the Pope the Supreme Pontiff, a priestly office in pagan Ancient Rome? Julius Caesar got himself elected to the office in 63 BC, because it gave him immunity from prosecution. Augustus kept it up, for different political reasons, as part of his project to wrap autocracy in a republican toga, and it became one of the imperial titles. Constantine Christianized the job and used it to assert control over the Popes, who therefore manoeuvred to get it back and succeeded under the pious Gratian. It's a defensive title.

    2. Hi James, Haven't had a chance to respond yet, nor to read your whole essay (reason for both: I'm on vacation). But in general, I'm very skeptical that Arrow-type considerations–perhaps best known to political scientists through Riker's work–are sufficient to invalidate the idea of rough responsiveness. For one thing, one of the big things parties and ideologies do is group citizens together in ways that make sense of what would–indeed–otherwise be pretty chaotic sets of preferences. That's why a Left-Right spectrum makes a rough sense, as does identification with a mass party. There's nothing logically incompatible with wanting to both ban abortion and raise taxes on the wealthy, but it's empirically uncommon. (In some ways, a multiparty system enabled by proportional representation makes responsiveness easier by making the parties a little less "mass": someone who wants high old-age pensions and fewer immigrants is out of luck in the U.S. but not in Denmark: good for Danish responsiveness, though bad for justice.) Ken Schepsle was one of the first to write about so-called "structure-induced" equilibria, mostly with regard to party leadership in legislatures; more recently UCLA's own Kathy Bawn and others (Hans Noel, UCLA grad now at Georgetown) have made a similar case with respect to the "long coalitions" enabled by ideologies. Gerry Mackie's book Democracy Defended, an intemperate but quite convincing attack on Riker and his followers, is in some ways a long meditation on how such equilibria work, and how consistently. In particular, Mackie searches hard for evidence that Arrovian cycles have ever occurred in a significant case–as opposed to a hypothetical or constructed one–and can't find any. Parties and ideologies do their work pretty well.

      One can also get constructed but not ridiculous majorities through public opinion polls, as Page and Shapiro did in The Rational Public with techniques that have been followed, in parallel or series with them, in other work. That is, most famously, if you asked people whether U.S. defense spending should go up, a big majority said yes in 1979. But that had plummeted to a small minority by 1984, after Reagan's buildup had been put in motion. Pretty interesting: suggests that there was something called a public on the question and that it was, in the aggregate, paying attention.

Comments are closed.