Thursday, April 28, 2011

Economic Voting is Rational

John Sides wrote another excellent post about (actual) independent voters, noting that they tend to vote based on whether they believe times are good or not; Jonathan Chait complains that this is irrational behavior:
Of course voting on the basis of economic growth in the two or three quarters leading up to an election plus military casualties is irrational. That's Kazin's point. The point that it can be predicted doesn't make it rational. Short-term economic growth and military casualties are related to good governance, but only very, very vaguely.
Would Sides vote the way independents vote? Would he even want to be friends with anybody who did? I suspect the answer to both is no.
Chait is correct that there's only a loose relationship between what politicians do and short-term results. But that doesn't make it irrational to reward good results and punish bad ones! Voters who do that provide extremely strong incentives for politicians to pursue good policy (that is, policy that tends to result in solid economic growth and peace, and avoids disasters such as Katrina).

Strong partisans are also rational, even though they will support their party regardless of performance. How can that be? Because strong partisans (usually) want something that independents don't want: specific policies, and they're willing to basically swap their vote for party loyalty on those policies.

I'm not saying all voting is rational, but I certainly don't think that providing politicians with incentives to produce peace and prosperity would be high on my list of suspected irrational behaviors.

Update: Don't miss John's response.


  1. Chait is correct that there's only a loose relationship between what politicians do and short-term results. But that doesn't make it irrational to reward good results and punish bad ones!

    Similarly, there's only a loose relationship between what my cat does and whether it rains outside. But that doesn't make it irrational for me to give Smokey a good belt-whipping every time we get a nasty thunderstorm!


  2. Our system makes it pretty damn hard to vote rationally.

    First off, thanks to our dual-party dysfunction, we now only have a choice between the hyper-right conservatives or the center-right "liberals," meaning that, as an overeducated, overindebted, pissed off atheist progressive manchild, I'm left (if I feel sufficiently motivated on election day) casting a ballot for the lesser of two evils, rather than anything close to my ideal candidate, who doesn't have a snowball's chance in hell of winning.

    So basically, as an American, I'm voting DEFENSIVELY - trying to PREVENT the worst candidate from taking office by installing a candidate who may be only marginally less sucky.

    And can some political scientist tell me how, despite the fact that our officials are spreading out to the ends of the ideological spectrum, their actual policy prescriptions seem to be coming together - e.g. the Reps say cut $60 billion; the Dems say cut $30 billion? Where's the socialist/FDR option, if all the Democrats are supposedly are way out in lefty ideology land?

    Unless we get wider representation through more parties (which would ideally require a switch to proportional representation) I think political participation will remain unsatisfactory and disillusioning to many Americans. And when Americans think government is broken, they return to their (Boston) tea party roots and start taking to the streets...

  3. "First off, thanks to our dual-party dysfunction, we now only have a choice between the hyper-right conservatives or the center-right 'liberals,'"
    You're just playing a word game here, defining the mainstream left as "center-right" and defining the mainstream right as "hyper-right". A conservative could just as easily play this word game and say voters can only choose between the hyper-left and the center-left. The fact is that the two options cover American political opinion quite well, and the reason you're unrepresented isn't a problem with the system but because your beliefs are a tiny fringe.

    "[W]hen Americans think government is broken, they return to their (Boston) tea party roots and start taking to the streets... "
    I've seen a lot of silly claims about why the Tea Partiers REALLY mean what the claimant wants them to mean (many of them documented on this blog!) but this is the silliest. The Tea Partiers are conservatives upset that there's a Democrat president. They aren't upset about a lack of fringe left candidates; indeed they think (or claim to think) that Obama is actually fringe left.

  4. There's a lot of talking-past-each-other going on, mostly about what constitutes rational behavior. The argument for independent voters' rationality could be either that:

    (1a) given that x doesn't pay attention to politics but
    (2a) does have some relevant information in the form of general impressions of how things are going, if we assume that
    (3a) voting is the rational or otherwise preferable option, then
    (4a) voting on the basis of the relevant information at her disposal is the rational course of action for x to take; and
    (5a) independents are acting in accordance with (3a); hence
    (C) independent voters are acting rationally. That is, they're acting on decisions they've made using sound reasoning and the best evidence available to them --


    (1b) given that x doesn't pay attention to politics but
    (2b) does have some relevant information in the form of a general impression of how things are going, then
    (3b) if there are enough people like x, politicians will have substantial incentive to produce in x et al. a general impression that things are going well; so
    (4b) voting like x will tend to lead to good systemic consequences, whether x's reasons for voting so are good or not; and
    (5b) independent voters as a class and politicians as a class are such as to satisfy the conditional in (3b) and therefore to make (4b) true; so
    (6b) independent voters are rational insofar as their behavior brings about the consequences they intend it to bring about (politicians try to make things better for them) through the means by which they intended to bring it about (viz. voting for the incumbent party when things seem to be going well, and not otherwise).

    Of course both could be true.

    Chait and Sides are both doubting that the A-version holds -- that's the point of agreement Sides notes between them, though he also mentions that some respectable people defend A-version. Sides argues however that B-version (1) holds and (2) is strong enough for us to be able to call independent voters rational.

    Kazin and commenter Andrew are doubting that B-version is strong enough: the fact that independent voting is systemically rational isn't good enough for believers in Civic Virtue as Kazin at least seems to be. Of course JB is a known foe of civic virtue conceptions of democracy, so he's defending the claim that B-version offers a legitimate way to understand the practice of independent voting (if not necessarily independent voters as individuals) as rational.

    The question at hand is really whether voters must have this or that intention, and this or that level of information, for their voting to be good for democracy. Right?

    (If you've made it this far, you might be an analytic philosopher...)

  5. To Anonymous:

    It's amazing how easily words get misinterpreted when just written down.

    A) I'd argue there are many, many serious academic and media observers (Ezra Klein is the latest with his column about Obama being a centrist Republican) who would agree with me that there really is no liberal policy representation, at least not any effective/significant representation, in American government right now. There are token liberals like Bernie Sanders, but even his moment of glory was a futile one. When both parties (the party of big goverment AND the party of fiscal discipline) refuse to seriously entertain the idea of raising taxes and when the basic right to unionize is being rolled back and endangered across the country, I feel pretty confident in saying that I'm not just playing "a word game" when I say that there really is no liberal policy representation.

    B) I specifically inserted (Boston) in my description to refer to the historical tradition of popular protest in this country, NOT the recent incarnation of that on the fringes of the far right. I would include them as one manifestation, but a less important and less serious one than, say, the recent protests in Wisconsin by liberal activists.

  6. As the classicist explains, the issue here is semantics -- what do we count as "rational." As John Sides says about his response to the original writer, Kazin: "To me, his concerns implied that there was nothing in particular underlying their choices in politics." I think, in other words, he constructed a kind of straw man: To call voters "irrational" is to suggest that they're voting on the basis of nothing related to politics or governance. But Kazin and Chait, I think, were saying that in normal conversation (as opposed to the poli-sci literature, apparently), it's not really rational to vote on the basis of some immediate reaction to current conditions, if in the process you're electing people whose policies would likely create the kinds of conditions you don't want.

    In other words, "rational" means factoring in causal relationships, having some theory of what kinds of policies produce what kinds of outcomes, etc. Granted, average voters may not have sophisticated ideas about this; they may be relying on elites or on selection processes (as partisans do when they vote for party nominees) that they believe have already taken account of such things. But the Kazin/Chait critique is that lots of so-called independent voters apparently don't even have that kind of indirect thought process in play. That's what they're calling "irrational."

  7. Here's an analogy: If I'm just choosing a product -- say, which laundry detergent to buy -- it makes perfect sense to prefer one over the other because of immediate effects I'm able to see: This one gets my whites whiter and my colors brighter, or whatever. But if I'm voting on who should be CEO of Proctor & Gamble, it would be irrational to prefer one candidate over the other on the basis of what I like in laundry detergents, unless I knew at least a little something (or were relying on some plausible "signaling" system to inform me) about where the different candidates stood on laundry detergents. Otherwise I might be helping elect a person pledged to discontinue the very type of detergent I prefer.

    So, we're talking about two different kinds or levels of choice, and what's rational in the one case (choosing a product) isn't rational in the other (choosing a CEO). I get the feeling that some professional social scientists have gotten carried away with market-derived models of choice, and so are not clearly distinguishing these levels. Hence they're prepared to it's "rational" to choose a president in the way one chooses a product, instead of in the way one would choose how to run the company that makes the product.

  8. I have to say I side with irrationality here. I don't have the time right now to quote chapter and verse on the definition of "rational," but it often involves preferences that are stable and don't change when irrelevant alternatives are introduced (among other things).
    I don't think either of these things is true of independent voters.

  9. @Jeff: your product purchase/CEO selection analogy is really helpful actually. I was saying before that the dispute was over whether independent voters were lacking information because they don't follow politics closely, but that's too simple. The questions of (1) whether using a method that generally promotes positive results suffices for rationality, and (2) whether one's general impression as to how things are going does in fact provide any or sufficient or sufficiently relevant information to base a decision on -- those two are quite distinct.

    I'm inclined to think that if it's irrational to rely on a feel for things in deliberation, then the most important issue becomes how to choose the cues and experts to rely on in deliberation, since we can't possibly have enough information to deliberate in a way that would satisfy us if we were deliberating about something in our private lives. Can't escape epistemology ...

    (Btw, I never wished you happy Easter back, and it's passed now ... but happy decapitated chocolate bunnies, or half-finished boxes of matzah and bottles of Slivovitz, or whatever it may be!)

  10. classicist, thanks. It's malted milk eggs -- that's my weakness. :-)


Note: Only a member of this blog may post a comment.

Who links to my website?