Some recent pieces by Ezra Klein (here and here) and Jonathan Bernstein (here) made the very important point that 2012 was supposed to be the year that donors bought the election, but that basically didn't happen. As Ezra wrote:
[I]t’s hard to look at the 2012 election, with its record fundraising and the flood of super PACs, and all the rest of it, and come away really persuaded that money was a decisive player. And yet the way we talked about money in the run-up to the 2012 election, we really suggested it would be a decisive player. In fact, we suggested, quite often, that it wouldn’t just decide the election, but that it would imperil democracy itself.
I think Klein's description is accurate. A lot of political commentary, particularly that emanating from the combination of journalists, politicians, and activists he calls the "campaign finance community," claimed that the growth of Super PACs and other well-heeled political organizations in the wake of the Citizens United case would overwhelm the voters' voice. Yet despite unprecedented sums spent by these groups on behalf of Mitt Romney and many Republican congressional candidates, the votes ended up going almost exactly as we would have otherwise expected.
This isn't a new pattern. It has generally been very difficult to pinpoint any specific effect of campaign spending on a general election outcome. Of course, it's very difficult to do this in part because the research itself is so challenging to perform. That's because spending levels tend to correlate with other politically important things, such as the quality of the candidate (prodigious fundraisers tend to be prodigious vote-getters, incumbents tend to have better access to donors, etc.). So if we see Candidate A outspending Candidate B, and Candidate A wins the election, did she win because of the money, or because she had certain qualities as a candidate that made her good at raising money and securing votes, even though one didn't cause the other?
Economist Steven Levitt tried to get around this problem two decades ago with an innovative study. He looked solely at congressional re-matches: cases in which the same two candidates ran against each other multiple times. This allows us to factor out aspects of candidate quality and just isolate the impact of spending. He found that spending had detectable, but very small, effects. Each additional $100,000 spent by a candidate correlated with roughly 0.3 additional percentage points of the vote. (That was only for challengers, by the way. He found no effect of spending by incumbents, with whom voters already had some familiarity.) That's not a huge effect. You could raise an extra $1 million (that was a big number for a congressional race back in the 1990s) and only hope to get about three extra points for that, and the vast majority of congressional races are decided by far larger margins.
John Sides and Lynn Vavreck found similarly-sized effects in their study of advertising in the 2012 presidential race. If one candidate doubled his amount of campaign advertisements, they found, his standing in the polls could go up by about a point. This proved to be a very ephemeral effect, though, disappearing after about a day.
Now, none of this means that campaign spending doesn't matter at all, of course. Even with very small effects, an absurdly huge expenditure of money could be critical. But there is a saturation point where you just can't get any more ads in front of people's eyes. And we really don't have a sense of whether the ads put together by Super PACs are as effective as those aired by the campaigns themselves.
Bernstein is right to note one particularly insidious aspect of campaign money: the amount of time spent by incumbents on fundraising. This keeps them late into the night talking or meeting with well-heeled donors instead of, say, their constituents, and may well distort their views about the sorts of issues confronting the nation. Michael Miller studied the states of Arizona and Maine, which adopted optional public financing for state legislative candidates a little over a decade ago, and found that those candidates who took the public money ended up spending their extra time doing the things we generally want candidates to do: knocking on doors, speaking with local voters, and so forth. This use of candidate time is definitely something to consider as we think about the thorny issue of campaign finance.
But the idea that voters can be bought with enough money just doesn't hold water. And given that recent efforts to make campaign finance more fair just seem to result in more byzantine rules and less transparency, maybe we should stop trying to do that for a while.