Levitt's results -- as interesting as they are -- seem to me to be far too localized & limited to justify the kind of sweeping policy change that you're suggesting here, Zach. Besides, if he's right, then the people to convince are the _politicians_: get them to stop raising & spending money, and the campaign finance laws will simply be besides the point!

Phil, 1. I'm not advocating a sweeping policy change, and 2. Will most (or any) readers of Freakonomics share your objection that his results aren't generalizable?

As for how to end the money madness in campaigns, you sound an awful lot like Newt Gingrich circa 1993.

I thought you were advocating the complete removal of public financing, on the basis of Levitt's research. If that's not what you were doing, but instead were merely speculating on the off chance that _perhaps_ people who read the book will be led to press successfully for this policy change, then the short answer to your question is "surely not".

I didn't make an actual policy prescription in my post, so I don't know what you took to sound Gingrichian. (I don't remember Gingrich's line on these matters.) My point was that _if_ someone believed the claim that money is basically epiphenomenal in campaigns, _then_ the people they should try to get to accept it will be the campaigning politicians, with no need for an actual change in the law. Surely it would take just a small handful of candidates opting out of raising money on their campaigns altogether, and then winning despite getting outspent 50-to-1 (or more) by their opponent, to convince the rest of the political world to get off the fundraising bus. Politicians generally hate all the time & craziness involved in fundraising, and if they came to think they could get elected without it, then most of them would be happy to let it go.

My own line on campaign financing is (i) overturn Buckley v. Valeo (in the direction John Rawls would want, that is); and (ii) put in a system of very substantial public subsidies for elections, much more than what we currently have, including a 'matching' component that would give someone in the public system funds equal to those spent on the behalf of an opponent who opts out of the system (thereby disincentivizing anyone to go outside the system in the first place).

Let me be more specific. I'm arguing about a novel policy, not the existing setup. This new policy, as advocated by Roth, i.e., federal financing of Congressional campaigns, is probably a popular proposal, but it is not salient.

In order for any policy to be enacted, it must be supported by a winning advocacy coalition. My question is whether the skepticism introduced by Freakonomics would inhibit this coalition from ever reaching critical mass.

And in 1993, Gingrich was arguing that CFR would be unnecessary if we'd only elect politicians who didn't play the game.

Ah, I see now. (And, really, looking back at the original post, the unclarity was my fault, not Zach's.)

I would offer two main worries, then.

First, it's not quite right to say that it is a real premise of the arguments of Roth and the like that money is actually electorally potent. The main relevant premises they would need for their argument are (i) that politicians _think_ that money is electorally potent, and (ii) that this belief (right or wrong) is politically distortive. It seems to me that (i) is obviously true; there is some literature on (ii), which seems to me to be a bit mixed, but partly that is because it is hard to operationalize the question of the influence of money on the formulation of legislation itself. (Up-and-down votes, on the other hand, are easy to run statistics on.)

The second worry is that, even if you have correctly identified a main premise of Roth and others, there are many ways in which their proposals may be consistent with Levitt's results. Levitt's data is the efficacy of expenditures in repeat match-ups of the same candidates (which he focuses on, in his typical sort of data-seeking brilliance, to do an end-run around the empirically-unmanageable question of the intrinsic comparability of candidates). But it may turn out, consistent with his data, that candidates still have too many incentives to gather money. I can think of two relevant nonlinearities that I don't see how his data could get at, as well as one confound.

For starters, it may be that one typically does have to pass some sort of spending threshhold (different from district to district, surely) in order to be a real contender. You do flag this parenthetically in the main post, in terms of barriers-to-entry, but it also means that candidates still always need to get a general fundraising machine together. Once they've got such a machine, even the small margins suggested by Levitt may be more than enough of a reason to crank that machine to whatever max that a candidate can. I hasten to add that, if this turned out to be correct, then campaign finance laws are exactly the right kind of remedy for this sort of situation -- i.e., if monetary competition between candidates turns out to be basically a big waste of time & resources (and hence an instance of a market failure), but unavaoidable without significant government-imposed market constraints.

The second nonlinearity may be that, as the gap between candidates' spending grows, so too does the efficacy of the spending of the better financed candidate. Candidates at rough parity aren't getting much out of their expenditures, but when they aren't even at rough parity, then the more _unanswered_ ads, etc. that the leader can get off, the more bang that those ads will carry. (Up to some saturation threshhold, surely.)

Finally, as noted above Levitt's data is all from repeat match-ups. It may be that money is basically impotent _in such elections_, when the voting public has already had plenty of time in the past to hear about & form opinions about the two candidates; yet still expenditures can be worthwhile when there is a fight to be had over the public perceptions of one or the other or both persons.

Isn't "public funding" a misnomer, since what it is applied to is in fact government funding. Candidates have always raised the bulk of their money from the public, unless their name is Corzine, Murray, or McCaskill in which case a lot comes from their own deep pockets

It's a completely standard use of "public", as in "public university", or "public housing". Cf. definition 2.b. (and also 7) from
http://www.m-w.com/dictionary/public