I am certainly no cheerleader for democracy, but I think Robin Hanson goes to far here.

Longtime readers should not be surprised to hear my suggestion: even random pivotal voters tend to think in a far mental mode. When we make concrete choices about our own immediate lives, especially for our private consumption, we are in a pretty near mental mode.  Since near-far depends on distance in time, social distance, and unlikeliness, our mental mode becomes farther when our choices are about a more distant future, are about a wider scope of people, are seen by more people, are about more unlikely situations, or are unlikely to matter. So citizen votes in a democracy are pretty much a far fest

For those unfamiliar with Robin’s use of the terms near mode is a state of mind that among other things bases decision on a cold weighing of costs and benefits. Far mode on the other hand is a state of mind that tends to focus on expressing our values and making broad brush assumptions.

I think that Robin is right that most voters go into the booth and make choices based on either grossly simply models, appeals to their values or even their emotional relationship with a potential candidate.

Rarely do you hear people say: Congressman Joe is a connoisseur of dog meat, the author of several kiddy porn themed comic books, and once defecated on the American flag to win a three dollar bet. However, he really gets the details of health care  policy and we need some good ideas about Medicare reform, so I voted for him.

When the choices are over more abstract things we think in terms of values. If the candidate is a “bad person” then we don’t want to vote for him. Even if we knew we would be the pivotal vote most people still wouldn’t vote for Joe.

In contrast if your father was trapped in a chemical fire and Joe the fireman had all the same negative qualities but was the best rescue man around then many more people, I suspect most, would call Joe.

When we are faced with the immediate consequences of our actions we think in terms of results. Joe may be a horrible person but without him dad will be dead.

The fact that many more people’s lives might be at stake when Congress considers health policy does not induce the same concern over results because the act of voting is mentally further away from the consequences of your vote.

All that having been said, however, I think republican democracy works pretty well and one of the the things that helps it work is that swing voters don’t vote that far. Swing voters often vote on whether or not they see things getting better or worse.

Policy wonks might be dismayed that these folks aren’t even trying to think through policy but their failure to do so may actually stabilize the system. It gives representatives an incentive to try policies that actually make things better, rather than merely enacting our values.

To a small extent the swing voter effect also selects for politicians that are good at making things better. That is, those politicians who, for whatever reason, do in fact make things better are more likely to be elected than those who didn’t. Selection is different from incentives because the politician himself doesn’t have to have any idea what he is doing to be chosen by the selection effect but he does to be influenced by the incentive effect.

Unfortunately the selection effect is minimal because the link between good policies and good outcomes is incredibly noisy.

Now all of that being said, immigration is a much stronger selector than voting. More open immigration allows people to select the government they want and mitigates the harm of bad government by reducing the number of people living under it.