McCain, the Media, Money, and Montesinos (and Obama Too)

So Barack Obama continues to raise millions upon millions of dollars, and if he wins the election a lot of people will certainly attribute his victory, at least in significant part, to this money.

But should they?

We addressed this topic in Freakonomics. Our argument was based on a clever piece of research by Steve Levitt (scroll down to “Using Repeat Challengers …”) in which he analyzed legislative races in which two opponents ran against each other more than once. Here’s why it was clever:

If Candidate A wins by 20 points and outspends Candidate B by 50 percent, it might be natural to assume that it was the money that made the difference. But how do you really know? It is hard to separate a candidate’s natural appeal from the appeal that is created by spending money on organization, ads, etc. So by measuring repeat challengers — i.e., races in which the candidates’ natural appeal stayed more or less constant — Levitt was able to isolate the impact of the money.

Here’s how we wrote up the results:

[T]he amount of money spent by the candidates hardly matters at all. A winning candidate can cut his spending in half and lose only 1 percent of the vote. Meanwhile, a losing candidate who doubles his spending can expect to shift the vote in his favor by only that same 1 percent.

What really matters for a political candidate is not how much you spend; what matters is who you are.

Now of course you could argue that money can help change voters’ views of who a candidate is. Isn’t that the purpose of the standard campaign TV ad? (Actually, most of the ads these days seem to want to change voters’ views on the opponent, but that’s just the flip side of the same coin.) And a recent tally of Obama’s spending shows that he has spent $160 million on TV ads, easily dwarfing every other expenditure. Staff salaries, for instance, were $44 million; campaign events cost $16 million.

Now comes word that he spent $21 million on TV in the first week of October alone, and is buying up prime-time network space at the end of October to run a 30-minute infomercial.

As I wrote above, this spending will probably be seen as central to Obama’s victory if he wins. But our argument is that money is more a symptom of a winning campaign than a cause.

In other words: it’s not that raising a lot of money helps a candidate become more appealing and therefore do better; it’s that better candidates raise a lot of money because they are so appealing. Just remember: about a year ago, Mitt Romney was loaded and John McCain was just about broke. If money is so central to elections, why couldn’t Romney put McCain away? And how on earth did McCain end up winning the G.O.P. nomination?

It’s also interesting to note that Obama is using the media — well, buying the media, in the case of the infomercial — to get his message across, while part of McCain’s campaign message is that the media itself is the enemy. The antagonism between the McCain campaign and The Times in particular has been operatic. According to, the McCain campaign views The Times as “a partisan rag.” Here’s what McCain senior adviser Steve Schmidt had to say:

Whatever The New York Times once was, it is today not by any standard a journalistic organization … It is a pro-Obama organization that every day attacks the McCain campaign, attacks Sen. McCain, attacks Gov. Palin, and excuses Sen. Obama.

The McCain/Palin anti-media attitude that was given full voice at the Republican convention continues to play out on the campaign trail. Consider this Washington Post report from a recent Palin campaign event:

Palin’s routine attacks on the media have begun to spill into ugliness. In Clearwater, arriving reporters were greeted with shouts and taunts by the crowd of about 3,000. Palin then went on to blame Katie Couric‘s questions for her “less-than-successful interview with kinda mainstream media.”

At that, Palin supporters turned on reporters in the press area, waving thunder sticks and shouting abuse. Others hurled obscenities at a camera crew. One Palin supporter shouted a racial epithet at an African-American sound man for a network and told him, “Sit down, boy.”

But this doesn’t mean the McCain-Palin campaign has nothing to gain from the “partisan rag” media. McCain and Palin have both made good use of a report in The Times exploring the relationship between Obama and onetime domestic terrorist William Ayres.

So what is the right way to think about the relationship between money, the media, and campaign outcomes? Is it wise for Obama to spend so much on media? Is it wise for McCain to risk alienation of the media? Would all that money and energy be better spent on something else?

There may be some wisdom to be gleaned from a strange incident in the not-too-distant past in Peru. A few years ago, the economists John McMillan and Pablo Zoido wrote a fantastically interesting paper about Vladimiro Montesinos, who ran the Peruvian secret service under President Alberto Fujimori.

Montesinos was extremely corrupt and brazen. Not only did he routinely bribe anyone who could help Fujimori maintain power — more than $3 million a month went to judges, police officials, opposition politicians, and TV station owners — but Montesinos also kept ledgers of these bribes and even videotaped the transactions. Sadly for him (but good for McMillan and Zoido), Montesinos was busted, and the economists were able to analyze the bribe data.

Of the four main categories of bribe beneficiaries — police, judges, politicians, media owners — whom do you think Montesinos paid off the most?

Here’s the answer, as summarized by Richard Morin in The Washington Post:

It wasn’t even close. “One single television channel’s bribe was four times larger than the total of the opposition politicians’ bribes,” [the economists] found. “By revealed preference, the strongest check on the government’s power was the news media.”

So while Obama may be wasting millions of dollars in general, at least it seems he is wasting them in the right direction. As for McCain: well, it’s not too late to start sending out a little something to your friends in the media.

[Note: I’ll be discussing this topic early tomorrow morning on The Takeaway.]


Along the lines of Nathan's comment, I think that 1% seems like it is way too low.

As with most situations involving money, I believe the law of diminishing returns applies. I have a sneaking suspicion that the first $100,000 in any campaign goes quite far compared to dollars $99.1M-$99.2M.

This would hold especially true when the candidate has little household name recognition.

John Jay

I think it's like Formula 1 auto racing. A certain dollar amount gets to you to, say, a lap time of 90 seconds at a certain track with a certain quality of driver. Another 50% increase in budget gets you to 88 seconds. Another 50% may get you to 87.5 seconds. The value is at the margins - like with most businesses or elections - and is disproportionate.

Francesco Lovecchio

It is true that Montesinos bribes to newspapers were much higher than to politicians and judges, in particular bribes to television channel owner were the highest of all. However, what was illuminating was the fact that Montesinos cared more about the papers read by the masses than elite newspaper read by the most educated people.

Joe Smith

A politician can either "feed" the media (treat them well) or have the media feed on him.


Two problems with your logic: first is incumbency, as another user has pointed out. Someone who is running for reelection stands a much better chance to win than someone who stayed out of public view for four years. Many studies show that getting reelected is much easier.

But even if we compare two candidates who lost, situations change from one year to another. You'd have to control for indirect media exposure - if one of the candidates is a celebrity or holds another office -, partisanship in a particular state or country at that time (obama won't ever get more than 60% of votes in texas even if he spends 700 billion dollars campaigning there) and how much money was collected through popular or corporate donations.


Given how presidential elections -- particularly this one -- work, I'm not sure that expenditures are as trivial as Levitt has shown.

Even if it's true that massive expenditures of money can only move the election by 1% or so, this still might be significant, if the election is close enough that a 1% shift matters. Even though polls are showing Obama ahead by several percent, there are still "undecideds" in the polls, plus if you throw in margins of error, we cannot really say that the 2008 election isn't running that close.

So a 1% push may, in fact, be significant.

Then there's the electoral college, which means the election is not a single country-wide vote but rather 50 separate elections of various sizes. In some states the candidates are much closer than are shown nationally; targeting expenditures to those states in which a 1% move is likely to matter, might in turn make all the difference in the electoral college vote and be the difference between victory and defeat.



So, Steve Schmidt thinks the NYT "is a pro-Obama organization"? Then I wonder why Governor Palin called it "generally very accurate"?

They can't have it both ways.

Eric Gunnerson

You're missing the obvious about the McCain relationship to the media. Their complaints are merely a case of working the refs - they complain that the media is too hard on them, and then the media reflexively moderates their tone.

Despite the fact that the corporate-owned media cannot honestly be defined as "liberal".

lee schipper

As a transport freak=onomist I recognize the spending for political adds as something like jockeying for space on a road - if I don't get there, you will.

If the spending is not so effective, why have the amounts skyrocketed? The expenditures employ people, to be sure, but the same $$ could have built bridges, fixed oral bridges, bridged knowledge gaps, etc etc. In short, the moneys are just thrown away. Having various spending doesn't really help, either.

What kind of psychology would change this pattern? I for one stop watching network TV during campaigns because I just can't bear the commercials.

lee schipper

berkeley CA


For this campaign the deciding factor will not be how much money was spent but the fund raising process. Obama's strategy of bringing in a huge number of small donors rather than going after big donors speaks volumes as to whom he will be listening. The fact that he is raising record breaking amounts of money will be viewed positively. These small donors rightly feel that their contribution is making a difference and giving them a voice - So when Palin was nominated, there was a huge surge in first time small donors leading to a record breaking month of fund raising.


I like this article, because it elucidates a few of the dents in a flawed notion I keep running into: Marketing Products = Campaigning for People.

The notion seems simple, and there are plenty of professionals in politics and marketing who will tell you that they can sell anything -- a bar of soap, tampax, or a person -- with the same set of principals.

I understand that consultants and other "experts" in both politics and marketing(and especially those who travel back and forth between the two fields) often tout the great persuasive powers of their "unique" methodology (usually on tv as a qualified correspondent).

This is, however, simply a result of their need to advertise their own wares (Being on tv as an "expert" increases their asking price. To get back on tv or even in print, you must first arrive with a simplistic info-bite that people can easily understand and react to). It also arises from a cacophony of bragging and chest-beating that vibrates amidst a chorus of highly competitive, incredibly self-righteous people.

Sure, some methods used in either marketing or political campaigning (Mailers, tv spots, testimonials, etc.) can help increase the exposure of a product or a person. Yet, as Levitt's findings seem to suggest, more exposure does not equate to more votes in the way that it can produce more sales.

Ross Perot and Michael Dukakis had a ton of exposure... How many of you voted for them?

Sure, Ross Perot could be considered a "bad product". Perhaps we really can call a bar of soap or a candidate a "product" on equal terms. This, however, begs a question: Are there or are there not a different set of criteria for judging a person than a bar of soap?

I could go on...



Money is the mother’s milk of politics.” --Jesse Unruh


How does this study account for the fact that the country winds up electing (and re-electing) the likes of Bush and Cheney, i.e. the most unappealing candidates ON THE ENTIRE SURFACE OF THE EARTH?


With regard to the initial claim that campaign spending has little effect, it's probably worth noting that Larry Bartel's, in his recent book, came to the opposite conclusion: that campaign spending is meaningfully predictive of the ultimate winner.

So while Levitt's method is clever, and the conclusion is appealingly counter-intuitive, it should not be taken as certain. Certainly one can waste money running a hopelessly flawed candidate (it could be the mission statement for the Democratic Party!), but that's not the same "as money has little effect".

Bartel's research focuses on presidential elections, and it might be that congressional and presidential elections have different dynamics, maybe there is differences in the variability of appeal in the two arenas. So both findings could conceivably be correct, if further specified.

Levitt's method controls for innate appeal, but it seem to me, as others have alluded, that there may be other important uncontrolled factors e.g. the effect of one candidate having lost previously, what another writer called "indirect" publicity, &c.



The trouble with the Romney-McCain example, of course, is that we have no counterfactual. We can't go back in time and assign the two men equal amounts or money or, better yet, give McCain a large spending edge. One could make a compelling argument that the only reason Romney (who certainly does fit the traditional mold for a Republican nominee) was even competitive was that he held a major financial advantage. There are any other number of variables to consider as well in this equation, not the least of which is that there were other viable candidates (Huckabee) seeking the position.


I firmly believe that money finds its way to appealing candidates. No one can buy appeal.


While there is certainly something to the seleciton argument (popular candidates raise a lot of money while unpopular ones raise less), the problem with the repeat challengers analysis described above is that it makes a faulty counterfactual argument. It tries to estimate the impact of dollars spent by looking at the same candidates over time (which assumes the natural appeal is fixed) with differing levels of money spent. However, the results of the first trial are known to voters and influence their behavior in the second trial. For example, if candidate A spends millions smearing candidate B and wins in election 1, he has created the impression in voters minds about candidate B that carries over in election 2. In that case candidate A could spend less the second time around and simply have to remind voters of their previous impressions of candidate B.


Econometrics is a mathemetical way of looking at economics. If you plot the stock history you can get a time variable which is random in behaviour. Using mathematical theory for random vaiables (or stochastic processes),which is well developed, we can use a time filter which automatically decides buy/sell decisions for a group or category of stocks. You dont need any expertise to do this. You can also use optimization techniques with boundary conditions to help you monitor or analyze the phenomenon. This approach is good for any process in nature.


I would have to disagree with the "hardly matters

at all" result. Prima faci 1% seems like a small

number, it would have quite probably changed the result in at least the 2000 presidential election, so i would think that it mattered quite a bit in close elections.


Money is necessary but not sufficient to win. (Having more money is not necessary, but could help. Better safe than sorry.) The fact that money doesn't decide candidate A over candidate B doesn't mean money is unimportant. On the contrary. It means that because both A and B need money they are limited in what they can say and promise so that the money keeps flowing. Money constrains and skews the political agenda; it limits what can be said in one direction, and pushes what can be said in the other. It spins the discourse and thus how people think about politics, economics, and what is "efficient", just and fair.