Does More Campaign Money Actually Buy More Votes: An Investigation
Every American knows that money buys elections, that votes come with a price tag. Given new reports of one millionaire's investment in the Virginia gubernatorial race, we decided to test that assumption.
Every American knows that money buys elections, that votes come with a price tag. Or, you know, we assume. Given new reports of one millionaire's investment in the Virginia gubernatorial race, it's worth testing that assumption. So, using data from the Center for Responsive Politics, we did.
The link between the two makes intuitive sense. More money means more television ads or mailings, which likely means more support on Election Day. But there are a ton of variables at play, so it can be hard to differentiate between the role of money (did more TV ads spur more votes) and the communications themselves (were some ads better than others)? There are turnout issues, questions of candidate viability, scandals. So many things go into campaigns, but few are as trackable as contributions.
Last Tuesday, Terry McAuliffe was elected governor of Virginia by a very small margin. As Politico reported on Monday, McAuliffe had a strong ally in that fight: California businessman Tom Steyer. Steyer is a very wealthy man. He poured $8 million into McAuliffe's race, according to Politico, funding TV spots, online ads, and door-to-door canvassing. Steyer's main priority is the environment, as The New Yorker's Ryan Lizza noted in a look at Obama's second term priorities in September. In Virginia, the Steyer-funded NextGen focused on areas of the state vulnerable to sea level rise form climate change to oppose Republican Ken Cuccinelli, a climate change denier.
Erin Lehane’s turnout operation was going door to door, particularly in Virginia’s coastal Hampton Roads area and on six college campuses throughout the state, hammering away at the message that Cuccinelli is an extremist on the environment and, well, everything else. … [T]he field component of NextGen had collected 10,000 pledge cards on college campuses — asking voters to commit to participating on Nov. 5 — and hit 62,000 doors in the days immediately before the vote.
Impressive. But effective? Field programs, the conventional wisdom goes, only account for a few percentage points in a race. A look at the county-by-county results from the Huffington Post shows that in the Hampton Roads area, in the southeastern part of the state, McAuliffe didn't do as well as President Obama in 2012 — though Obama outperformed McAuliffe in most of the state. So what role did that money play?
Last year, the country ran controlled experiments in 435 different places, putting a number of candidates up against one another and tracking how much each raised and spent. By combining data from the Center for Reponsive Politics (the people behind campaign money site Open Secrets) and the actual 2012 results, we're able to get a sense for how money and vote totals correlate.
More money was spent on votes in closer races.
First, we took a look at how the margin of victory in each race (the raw vote total separating winner from loser) correlated to the cost-per-vote. That figure is how much each winning member of Congress spent divided by how many votes he or she got. If a candidate spent $100 campaigning and got 20 votes, each vote cost $5. (On average, the 435 winners spent $9.84 per vote.)
Below, the relationship between votes and vote-cost. Each blue dot is one winning campaign. The vertical axis is the number of votes in the margin of victory; horizatonal, cost per vote. At the top left, Rep. Chaka Fattah of Pennsylvania's 2nd district, who won by a heavy margin at a low cost. At bottom right, Minnesota's 6th, home of Rep. Michele Bachmann, who won by a small margin at a high cost. The red dotted line is the most important part of the graph. It shows the trend in all of the data. (This graph, unlike the others, uses a logarithmic scale on the x-axis.)
Notice that the red line drops sharply down to the right. In other words, as races grew tighter (few votes in the margin of victory), each vote grew more expensive. More important votes cost more.
The more you outspent your opponents, the more you won by.
More interesting is a look at how the size of the winning candidate's victory corresponds to how much more that person spent. (The first three graphs in this post correspond to spending, not what was raised by the campaign.) Here are all of the races, with the vertical axis showing the point spread of the victory (e.g., a 60 percent to 40 percent win is a 20 percentage point spread). The horizontal axis shows the percentage more or less that the winning candidate spent.
This, too, looks the way we'd expect. The greater the difference between how much the winner outspent his or her opponent, the more of a spread in the end result. The red line goes from lower left to upper right showing that trend.
But notice all of the dots at the far right. That's largely heavy incumbents, who won against opponents that weren't well-funded (like Rep. Fattah). What if we filter them out?
In close races, that effect was heightened.
If we only look at races that ended up within a 20-point spread — not close races at that margin, but not completely lopsided — the pattern shifts.
The most significant thing on this graph is the slope of that red line. It moves up and to the right much more steeply than for the chart above, showing all races. In other words, there's more of a correlation between spending differential and vote differential. This supports the idea that in closer races, money makes more of a difference.
Most incumbent losses saw the incumbents outraised.
We were also curious whether or not incumbency affected the relationship. So we took raw vote total and compared it to the margin by which the winner outspent the loser in races where the incumbent lost — only a fraction of all of the races.
The more a challenger outraised his opponent, the more likely he was to defeat the incumbent. But nearly all of the insurgents actually did so, outraising the incumbents.
What does all of this tell us? That our shorthand for political success — more money, more votes — was validated in 2012. But what it doesn't tell us is the role of money in any particular race, nor does it tell us how outside parties, like NextGen, might have affected results (CRP only includes money raised by candidate committees) — particularly in an off-year election.
In general, elected officials are warranted in raising and spending as much money as possible. A lesson that Tom Steyer has clearly already taken to heart.