A “horse-race” poll is one that asks people whether they plan to vote and, if so, for whom. These surveys are appearing at the rate of half a dozen per day right now. I don’t think the tide can be stemmed, because we have a constitutional right to ask other people about the election and print the results. And I acknowledge that I follow polls obsessively, checking them several times a day and reading all about the minutiae of party weighting, robocalls versus live calls, and “house effects.” CIRCLE will even ask horse-race questions on our own youth poll, soon in the field.

But what if they all just went away?

Although there would be pros and cons, I think we’d be much better off. The theoretical framework that helps explain why is Jürgen Habermas‘ distinction between instrumental and communicative reason.

With instrumental reason, you know what you want and you deploy resources, including speech, to get what you want. For that purpose, horse-race polls are very useful. For instance:

• If you are Karl Rove, you are now moving money to House races instead of supporting Mitt Romney, because the presidential race looks lost to the GOP, while the House is more “in play.”
• If you are a leftish critic of Barack Obama and you live in a blue state, you are probably contemplating not voting (in response to the Pakistan drone bombings and many other issues). You might feel differently if you thought that Romney was about to take your state and the national election was close.
• If you work for the president’s re-election campaign, you’re not paying any attention to uncompetitive states, like California, Texas, and New York (combined population = 83 million), but you’re suddenly very interested in Nevada (pop. 2.7 million).
• If you are Barack Obama, you think you’re ahead, and so you’re inclined to run out the clock, rather than, for example, explain what you propose to do about difficult public problems.
• It’s possible that if you are a Republican, you are losing interest in voting for president, which would make the polls a self-fulfilling prophesy.

One way to assess instrumental action is in terms of the outcomes. From my perspective, Karl Rove’s targeting his millions is bad because he is supporting the wrong people. But lefties’ boycotting an uncompetitive election is fine, even though I am not personally moved to protest this president. (For the record: I voted for Nader in 1996, when I was certain that Clinton would win. This election is closer, and I much prefer Obama to Clinton.)

But we shouldn’t simply assess instrumental action by its outcomes, because that damages other values. In politics, one of the fundamental values is equality: everyone should count for the same. Another value is some version of integrity: you should say what you believe. And deliberation is a value: we should exchange reasons with our fellow citizens and give everyone a hearing. We should deliberate out of respect for other people and also because we might be wrong: deliberation is an opportunity to learn.

If you are sure what you want to happen and you have a pretty accurate sense of how your fellow citizens are going to vote, reasons become relatively unimportant, and some citizens count more than others. Forget about reasoning with all those New Yorkers and Texans: the electoral college outcomes in their states are utterly predictable. Forget about what should count as the best argument for your core positions, because those may not be  tactically valuable things to say. Pick the issues and arguments (no matter how trivial) that seem most likely to win you electoral college votes.

That’s advice for candidates and PACs, but individual citizens may regard the election as a spectator sport and believe that it’s interesting and worth their attention just insofar as the outcome looks close. That is wrong because our job is to decide how to vote and then move on to making other decisions that improve the world. Horse-race polls just distract us.