Democracy with Open Eyes

Ah, the dangers of having people learn what political scientists know.  About voters, that is: the Boston Globe ran an excellent story by Joe Keohane over the weekend about some of the various cognitive biases that researchers have discovered, and how they result in ill-informed citizens (but see too John Sides on the subject -- and Henry Farrell).  So, if voters aren't very rational, or very good at processing information, or are mostly ignorant about politics and public affairs -- what does that tell us about democracy?

Some, in fact, see such information as a threat to democracy.  In fact, it's only really a threat to one version of democracy, a Good Government, League-of-Women-Voters interpretation of democracy.  In that version, unattached individuals carefully study The Issues, decide what constitutes good policy, then spend time researching candidates to see what the candidates say about those Issues, and support those candidates with whom they agree.  Finding out that most voters do things in reverse order (they generally pick a party, support candidates because they belong to that party, then take issue positions from that party and even believe facts based on what party leaders tell them is true) pretty much destroys the viability of that version of democracy.  Fortunately (at least in my opinion), American democracy has never been based on any such fantasies -- although some Progressive era inspired political regulation, alas, has been. 

Matt Yglesias (naturally) takes the opportunity to plug his preferred version of democracy, which is far more majoritarian than what Madison intended.  Here's Yglesias:

This, however, is a reason our political institutions need to be reformed. Democratic accountability is based on the idea of holding incumbents responsible for their performance. But for that to work, election winners need to have the chance to implement their agenda. If the losers get to block a sound agenda, and then reap the rewards for having done so the system will be perennially off-kilter. 

Well, it's certainly better than goo goo democracy, but I (naturally) continue to not understand the logical jump that Yglesias makes.  Yglesias understands that elections feature (at least looked at from one perspective) three groups: loyal partisans on one side, loyal partisans on the other side, and swing voters who are moved mainly by the economy (although also by war, and by the popularity of the president independent of those other things).  Yglesias wants to make sure that those swing voters are acting rationally; if they're going to reward or punish the in-party for economic performance, then at least the in-party should, he argues, be able to implement the policies for which they will rewarded or punished.  So the problem with the American system is that, for example, the Democrats this November will be judged for policies that reflected not what Nancy Pelosi and Barack Obama think, but what Olympia Snowe and Susan Collins think.  Even if they're sincere (and there's going to be a large incentive for them not to be), where's the logic in punishing Obama for what the Benator did?

There are a few problems with that logic, however.  First of all, for better or worse, economic conditions are not direct consequences of government policy, especially in the short term.  Now, we definitely want the in-party to have a strong incentive to make people happy with the performance of the economy, but (for at least the president's party) that's going to be true regardless of whether the president's party gets to automatically implement its plans or not.  In other words, it's nice to think that democracy can't do much, but -- if only the in-party gets to implement its plans -- at least democracy can reward good economic policies and punish bad ones.  But that's not actually true.  Surely, Yglesias believes that George W. Bush's economic policies in the 2001-2004 were terrible, and yet he was rewarded for them because their long-term effects had not yet produced trouble.  For the most part, then, democracy can't really do what Yglesias wants it do, and therefore his concern about a "perennially off-kilter" system really doesn't hold up. 

On the other hand, a pure majoritarian system produces an odd dictatorship of the swing voters.  These voters, who we know are the least engaged and informed, will reward and punish based on current economic conditions (not on economic policies, or even something relatively sophisticated such as economic conditions over the last few years.  Just current economic conditions).  As a result, the winners will be able to implement whatever policies they choose, whether or not those policies are supported by voters at all.  That is, we know that pro-choice policies have resulted from the 2008 election even though it's highly unlikely that the election results had anything at all to do with voters changing their minds from 2004 to 2008 about abortion.

A Madisonian system, with checks and balances and separated institutions sharing powers, guards against that sort of dictatorship of the swing voters.  Elections still matter a lot, but other sorts of representation matter, too.  Elections matter not only because they reward good policy outcomes and punish bad ones, to the extent they do that, but also because they mobilize voters to join groups and make demands on the government, and they encourage politicians to take their representative relationships with constituents seriously.  And because elections don't determine everything, citizens are encouraged to continue that involvement between elections -- and they really do so!  Not all of them, of course, but quite a few, either as activists themselves, or as members of various groups.  All of that counts as "democracy," and much of it would be, at least in my view, a lot less successful in a true majoritarian system. 


At any rate, I think that the information we have on voters does nothing to undermine either Yglesias's preferred majoritarian democracy or my preferred Madisonian version.  It does severely undermine Goo Goo versions of democracy.


(Necessary disclaimers: 1.  "swing" voters include not just people who change their minds, but people who drop into and out of the electorate.  2.  Nothing about the theoretical case for Madisonian democracy demands the specific institutions of American democracy, and in particular there's nothing about Madisonian democracy that suggests the need for a supermajority Senate.  Madisonian democracy suggests that simple majority rule is a bad idea, but is agnostic about voting rules within one half of a bicameral legislature in a system with separated institutions sharing powers and with considerable real dispersion of authority via federalism).