How Much Does Campaign Spending Influence the Election? A Freakonomics Quorum

We have long argued (most recently in this Marketplace podcast) that campaign spending isn’t nearly as influential in elections as the conventional wisdom holds.

This week, with the G.O.P. presidential hopefuls in South Carolina spending lots of money (and time and effort) and everyone’s talking about “super PAC” spending, we thought it was a good occasion to air this question out further. We’ve convened a Freakonomics Quorum on the topic, soliciting replies from a few folks with expertise in the realm. Thanks to all of them for participating.


Douglas Holtz-Eakin, the director of domestic and economic policy for the 2008 John McCain presidential campaign, has been chief economist of the President’s Council of Economic Advisers and was the sixth director of the non-partisan Congressional Budget Office.

Beware the Super PAC? 

The political season is upon us (candidates are in South Carolina as I write this) and with it the POMP (Punditry Outrage about Money in Politics) season.  This year POMPous breast-beating focuses on those pesky “Super PACs” that can collect unlimited amounts of cash and weigh in on behalf (but not plan or coordinate with) a candidate. With certitude and POMPosity, Americans are told their electoral system is being sullied by monetary incentives and activities.

I will save my defense of systems driven by economic incentives and activities for another occasion.  Let’s talk about POMP.  To begin, make no mistake about it, there is a lot of money in modern American politics.  And if my first exposure to the fundraising demands of modern, national electoral politics during the McCain presidential campaign is at all representative, the mechanics can be eye-opening. 

I shall never shake the scene at the back of the Straight Talk Express: McCain holding court with the press while simultaneously fielding continuous handoffs of cell phones from aides dialing donors for dollars.  Still, and for the record, the McCain loss was not the result of the vaunted Obama small-dollar internet donor campaign, his cynical “yes-I-will, oops no-I-won’t” policy on public financing, or any other aspect of the financial tsunami that washed over the McCain campaign.  It had more to do with the financial tsunami that hit the country.

Still the POMP crowd insists money is killing the system.  If so, how does money matter?  There is now a reasonably robust empirical finding that pure campaign spending in head-to-head Congressional elections does not tilt the playing field in any significant way.  And there is a corollary that traditional PAC spending (which got funneled largely to incumbents) had comparably small effects.  These results suggest that if money matters, the impact must derive from other channels.  So, in the interest of promoting a more nuanced research agenda (hopefully to be financed by POMPs!), let me suggest a few channels that introspection suggests merit investigation.

Modern politics is about a lot more than the ultimate head-to-head to matchups.  Serious consideration is given to (a) success at the top of the ticket, whether it be the presidential, gubernatorial, or other race; (b) factors that contribute to getting out the vote; (c) endogenous determination of the quality of candidates (a euphemism that encompasses (i) candidate recruitment, (ii) the incumbent/challenger getting the bejeezus beaten out of her/him in the primary, and (iii) and the systematic generation of uncomfortable votes prior to an election); (d) the ability to legally defend and/or challenge the legitimacy of votes; (e) the time, focus and energy cost to the candidate of fundraising activity. 

All of these involve resources; that is, money.  And they are far tougher to link to specific electoral outcomes.  I encourage the Freakonomics community to get to  work. 

My bet is that when all is said and done, it will be POMP versus circumstance.  And the latter will win.  At any moment the country has real challenges, real needs, and real demands for leadership.   And those trump money.

Robert Shrum, a senior fellow at New York University’s Robert F. Wagner Graduate School of Public Service, has been a senior adviser on many Democratic campaigns, including Dick Gephardt (1988), Al Gore (2000), and John Kerry (2004).

In politics there is certainly no linear relationship between amount of money and degree of success. Just ask the well-heeled Republican losers of presidential primaries past — former Texas Governor John Connally, former Texas Senator Phil Gramm, and former Mayor and front-runner Rudolph Giuliani. Or how about Howard Dean, who raised and spent nearly $40 million before crashing and burning in the 2004 Iowa caucuses?

Big money without the right message can become a penny waiting for change. Thus Dean misread the Iowa landscape. While voters were focusing in on one overwhelming question — which candidate had the best chance to beat George W. Bush? — Dean and Dick Gephardt were engaged in a well-financed exchange of petty negative ads. Maybe Dean never could have been a plausible answer to the determinative question in Iowa in any event. But he never seemed to try — and left the field almost entirely open to John Kerry who, Iowans rightly judged, could give Bush a real run for the White House. Similarly, this year no amount of cash could have rescued the malaprop Rick Perry; GOP caucus goers decided, once again rightly, that he couldn’t face up to Barack Obama — or measure up to the requirements of the Oval Office.

Kerry’s come-from-behind win in Iowa also illustrates the other side of the coin: You don’t need the most money, but you do need enough. Kerry took out a mortgage on his home to keep his campaign afloat when he was written off in the fall.

In 1960, Hubert Humphrey’s resources in the West Virginia primary weren’t remotely equal to John Kennedy’s — and Kennedy’s victory there all but sealed the nomination for him. Gephardt, in his first run in 1988, carried Iowa but then couldn’t raise enough money fast enough to compete on Super Tuesday. In 2012 the disparity in super-PAC money let pro-Romney forces dismember Newt Gingrich during December. Gingrich is probably fatally flawed, but it didn’t help that he couldn’t defend himself or go after Romney early on.

Finally, money doesn’t make all the difference — unless it does. Much as Obama did in 2008, Kerry raised prodigious sums, a lot on the internet, during the primaries four years before. The campaign then made a mistake of accepting federal funding in the general election; this meant that Kerry had the same amount of money for a thirteen-week campaign that Bush had for eight — because the Republican convention came more than a month after the Democrats, and Bush could keep spending primary dollars in the meantime. The result was a form of financial disarmament which deterred a swift response to the Swift Boat ads — because that would have drained limited end-of-the-race funds in mid-August. More broadly, as the 2008 Obama experience suggests, s
taying outside federal funding could have let Kerry broaden the list of target states and potentially prevail in a tight contest where a football stadium’s worth of voters in Ohio decided the outcome.

So gold doesn’t always glitter in politics — but you better have some of it, and sometimes, sometimes, having the most can matter the most. 

Jeff Milyo is an economics professor at University of Missouri at Columbia. His research includes campaign finance, state and local health policy, elections, and the media.

The misperception that political spending drives electoral outcomes is reinforced every campaign season by sensational media coverage, post-election debriefs from losing candidates and the exaggerated rhetoric of professional reform advocates.  And this first presidential election cycle post-Citizens United promises to bolster that errant view as sanctimonious posturing by pundits on the evils of money in politics will likely crescendo to a spectacle rivaling only a North Korean grief orgy.

It is true that winning candidates typically spend more on their campaigns than do their opponents, but it is also true that successful candidates possess attributes that are useful for both raising money and winning votes (e.g., charisma, popular policy positions, etc.).  This “reverse causality” means that campaign spending is potentially as much a symptom of electoral success as its cause.

In order to identify the treatment effect of campaign spending on electoral success, researchers exploit natural experiments. For example, imagine re-running a race between two candidates but varying the campaign spending of each; repeat that exercise enough times and you have an experiment that will allow you to observe the causal effect of campaign spending, all else constant.  That’s basically the approach taken by Steve Levitt in his seminal study of repeat meetings of the same Congressional candidates over time.

Levitt finds that changes in campaign spending produce negligible changes in electoral outcomes when candidate characteristics are held constant.  Now that doesn’t mean that candidates don’t need to get their message out to voters.  We’re talking about marginal changes in campaign spending.  Given you are already spending a million dollars running for a House seat, another hundred grand or so won’t make any appreciable difference.

Of course, repeat meetings of candidates don’t happen by chance, so Levitt’s study is susceptible to the criticism that it isn’t the cleanest of experiments.  However, I have poked at those results without being able to overturn them, even though I was highly motivated to do so (Steve is a great friend, but the professional acclaim I might have had from reversing his finding far outweighs the value of one friendship at the margin).

I have examined several other natural experiments and found similar results. For example, large shocks to campaign spending from changes in campaign finance regulations do not produce concomitant impacts on electoral success, nor do candidates with vast personal wealth to spend on their campaigns fare better than other candidates.

These findings may be surprising at first blush, but the intuition isn’t that hard to grasp.  After all, how many people do you know who ever change their minds on something important like their political beliefs (well, other than liberal Republicans who find themselves running for national office)?  People just aren’t that malleable; and for that reason, campaign spending is far less important in determining election outcomes than many people believe (or fear).


So money doesn't effect election outcome? Oh ... Well in that case I am going to start a new political party which will bring into the national dialogue a contrarian viewpoint to the NeoLiberial dogma of the two presiding parties. I'm sure I will be effective with my, oh, 10 million?

The problem with all of these rants about how money has no effect is that they are tragically myopic in viewing how politics should function. Sure, if your either/or it doesn't matter because corporations play both sides; but if your an other, well you don't have a political hope in hell. So essentially this article supports the one party system in America: the money/capital party. And as such all other voices and ideas are suppressed so that the status quo can be maintained.

Bu$y B

@ ecocollectivist - neither the article nor any of the comments put forth an opinion on how things"should" be, but refer to how things actually are. Its fairly obvious that a) one needs a certain (large) amount of money to even get in the game, and b) campaign donations reap rewards after the candidate is elected, hence the strategy of donating to both candidates. Correctly analyzing reality does not equal endorsing the status quo.


@Bu$Y B - Who said anything about "how things should be"? I am analyzing how things are. The fact is money wins elections. Their analysis is myopic in that it only legitimizes two players. This is not a true analysis but rather continues the propaganda and illusion of a democratic system. My opinion about "how things should be" would be to ban all political advertising, have publicly funded elections with seriously constringent donation rules, a more robust dialogue system, multiple parties in order to progress a diversity of views, funding for town hall meetings both in real life a in the virtual world, etc. I could go on. These are my opinions on how it should work. But my analysis of the system is that it marginalizes non-corporate grassroots democracy and only legitimizes a singular orientation; this is primarily done through campaign funding. This "analysis" is completely inside the box and this is nothing more than propaganda because it neglects to analyze all the various factors at play. They fail to correctly analyze the status quo by leaving out a very real aspect/result of the funding mechanism. Or are you telling me all the disenfranchised views and people are not real? The people on the streets just don't exists?


Nicholas Bowman

Seriously? You (correctly) criticize the correlation implies causation fallacy, and then go on to do something equally bad. A lack of correlation doesn't imply that there isn't an effect.

Specifically, if someone spends more on advertising than someone else and does worse in the election that doesn't mean that money didn't buy any votes. Candidates spend more money because they know they need those extra votes. As far as the comparison of spending in congressional races goes, it wasn't clear if this was comparing differences in spending during a campaign or in different election cycles. If it was in different election cycles, there are SO many factors that aren't the same, even if they are running against the same candidate. Once again, they probably spend more because they think they NEED to. The same would apply to different time points in the same campaign. The amount of spending almost certainly inversely correlates with how much the voters know of or are likely to vote for the candidate pre-advertising. As such, you can't very easily tease out what the TRUE effect of the spending is.

One way you could better address this issue would be by looking at races with two candidates and use their poll numbers going into a race and their poll numbers right before the election and then compare how much they spent and how much their poll numbers changed.


David D

Guided by this posting, I withheld all campaign contributions for this election cycle. I'm now wondering if the grassroots targeted Democratic/Obama campaign spending was more effective at garnering votes than past efforts? Even if so, when the Republicans update their game plan next election cycle, any effectiveness edge might just disappear...

Gene Silvers

There's a certain size that a campaign has to reach in order to give every one of the voters a full look at its candidate. The further beyond that size that the campaign gets, the more unfairly-large it gets. The spending is what creates the bigger and bigger size. In other words, these guys are just plain wrong. They should become political candidates for some local offices and give their opponents enough money to outspend them 5-1 and see how far they get. LOL

Ranganath Nayak

Yes, but does funding from big money go for feet on the ground or for TV and radio and newspaper advertising?

Ranganath Nayak

If campaign spending has a relatively modest effect, why is there so much of it, all provided by politically smart people?


The author fails to understand that he and no one else has provided any evidence that all of the money pouring into Republican campaigns does not elevate GOP chances as the GOP is increasingly the minority party politically and demographically, without the massive amounts of money pouring into GOP coffers, there would be no GOP at all.


Phil Gramm didn't run to win but to get some money he could spread around in turn for favors. Come on! Phil probably looked clammy and sweaty reading the morning paper!


This is an interesting forum. I found it because I looked directly for the correlation between $ and success. It seems that every journalist I read just takes the assumption that those who have the dough will win. Period. I especially liked the notion of reverse causality which I never thought of before -- I entered the wrong search phrase but was corrected. Thanks


The idea of democracy is that we elect a president to lead America with the popular shared vision. When corporations buy presidents, these presidents owe favors in exchange (there is a high correlation), and our shared vision also changes to favor the corporate supporters. This is why during election years Obama said "make this county great, and all other years he says "make this county a little bit better", so if becoming president is a function of your charisma to attract sponsors, and you believe this “reverse causality” is fine, Jeff Milyo you are making a terrible assumption, companies by definition need to generate profits, so then how does this greed equate with our shared vision? Do you vote on who leads the companies that buy our presidents? If you truly don't believe that people are that "malleable" then I think you need to open your eyes.

Ryan Baker

So.. money can matter, but only if it's used correctly? Sounds like that leaves the case for limiting money in politics on very firm footing. If you don't have enough to match your opponent, then you'll always be worried they'll exploit that by intentionally creating a financial quagmire for you. So being proactive is going to be a part of a candidates strategy, and by allowing large donations, this encourages them to seek out and bend their will to the potential sources of those donations.

It would also seem somewhat relevant to see how money affects long term electability. It would seem somewhat reasonable to guess that if one candidate is "on the ropes" so to say in terms of political finance, they are more likely to need to give bigger promises for smaller amounts. While those may have no impact on the current election cycle, they're more likely to result in a future scandal that will harm the next cycle.

Now that might at first seem uninteresting, because of course we'd prefer politicians who don't offer rent seeking advantages to donors, and thus we should prefer not voting for those with scandals on their record. But if you look at it a different way, say that there is a probability that a candidate is going to be more or less resistant to such corruption all other things being equal, we should prefer the one who is going to be better before the stresses of fundraising force him of course, assuming of course that we can minimize or eliminate those stresses.