Mitt Romney holds a commanding position in the GOP primary, despite acerbic opposition among Republican activists to his centrist record and near-universal distaste for his character among anyone who’s ever seen him or heard him speak. It often seems that Romney – who, with $250 million in personal worth, is easily the richest in the field – is attempting to buy the American presidency outright. So far he’s succeeding. This strategy can flop, as the Romney of 2008 can confirm, but his current advantage and the relaxation of campaign funding rules via Citizens United lead to the question: Does spending more money confer a significant electoral advantage?
A wide-angled view of the data suggests it does. In the past three national election cycles, the candidate who spent the most won 81% of the time in Senate races and 91% of the time in House races.
However, this doesn’t cement the case. In a famous argument, the Freakonomics team takes the correlation between money and electoral victory and flips it. “People who win political campaigns usually outspend their opponents by a lot,” concedes Stephen Dubner, the dark-haired, bespectacled half of the Freakonomics duo. “It’s not that money wins the elections. It’s that the candidate who is much more attractive from the outset raises more money.”
This argument appealed to my contrarian instincts, but, simultaneously, it threatened to invert my political understanding in a displeasing way. I was keen to poke holes in their case and pleased when I found hollow spots in key areas.
The foundation of their theory, a paper by the other half, Steven Levitt, examines thousands of elections where the same candidates competed against each multiple times. He chose this data set to control for candidate quality and other confounding factors (region, impact of presidential popularity, etc.). He found that, in these cases, election spending had very little effect on the outcomes. “According to my estimates, an extra $100,000 (in 1990 dollars) in campaign spending garners a candidate less than 0.33 percent of the vote.”
Early on, Levitt reports that previous studies overwhelmingly show campaign spending is less effective for incumbents than for challengers. Incumbents have previous campaigns and previous records that are difficult to alter through messaging. Having established this, Levitt then curiously chooses to study elections where the candidates are repeat candidates – and often incumbents. The problem, naturally, is that repeat candidates are known to the public, too, and they face the same diminishing returns on campaign spending. Therefore, money having little effect in these elections is a result that’s expected. And of questionable relevance.
A second serious flaw comes from trying to define the ‘intrinsic vote-getting ability’ of a candidate. Most pressingly, a candidate’s appeal may well result from a successful investment in public relations at the outset. (Tangentially, it’s important to note that in many races a candidate can’t even viably compete without reaching a prohibitively-high funding threshold. In this sense, money has already bounded the playing field.) This seems especially true of larger races (senatorial, gubernatorial, presidential), where candidates with shoestring budgets struggle to achieve exposure or even ballot access and therefore appear unworthy of the donors’ money. I don’t doubt that a victorious campaign has a self-fulfilling momentum, as the Freakonomics guys claim, but I think money is required to give it forward motion.
Lastly, even if we can establish a sound definition for the attractiveness of a candidate – one that doesn’t depend on superior initial funding – there’s no reason to accept Levitt’s assumption that “an individual candidate’s quality is constant over time.” Contradicting examples abound, from Nixon (219 electoral votes in 1960, 301 in 1968) to Carter (297 electoral votes in 1976, 44 in 1980) and beyond.
A realistic picture of election funding is not simply one of a straight line from spending to victory. Dubner and Levitt explore important complicating factors concerning incumbents/challengers, electoral momentum and candidate quality. But, while the common conception of campaign spending is both simplified and exaggerated, there’s still strong evidence for a major effect on electoral outcomes. I think this election will show, in contradiction to Levitt’s claims, that the influence of PACs is insidious and that public financing of elections is desirable.