Illustration by Kelsey Dake

Why Data Wonks Are Wrong About Presidential Elections: Ron Klain

a | A

Nov. 15 (Bloomberg) -- A little less than a year before the 2012 election, it’s a good time to ask whether presidential campaigns matter.

Do all the ads, speeches, mailings, debates, online activity and rallies really change minds? Or is the outcome of the election the product of underlying fundamentals that are scarcely affected by such efforts?

For many years, a group of political scientists, mathematicians and scholars have argued that a handful of factors determine of presidential elections, irrespective of the campaigns.

Most famous among those thinkers is Allan Lichtman, whose “13 Keys to the White House” model (which looks at factors such as incumbency, the outcome of the previous midterm election and per capita economic growth) has forecast the popular-vote winner in each of the last seven elections.

More recently, the brilliant data analyst Nate Silver has employed a three-factor model (presidential approval rating, economic outlook and opponent’s ideology) to forecast the 2012 outcome under a variety of scenarios. At least implicitly, he, too, is suggesting that the campaign itself is irrelevant to the result of the election.

One immediate reason to be skeptical of the models’ forecasting prowess is that they point in opposite directions: Lichtman has interpreted his keys to forecast that President Barack Obama will be re-elected in 2012, while Silver rates Obama’s chances at less than 50 percent.

Extrinsic Factors

In any case, we still face the bigger question: Are presidential elections decided by extrinsic factors that models can analyze without regard to the performance of the candidates or their campaigns, or do the campaigns determine the outcome?

It probably won’t surprise anyone that my view -- given a lifetime of work on presidential campaigns -- is that campaigns do matter, and often make a great difference. That’s not to say that underlying factors aren’t crucial. Of course, the state of the economy, the success of an incumbent president’s foreign policy and the ideology of his opponent matter a great deal.

But, to use just one example, if cold, hard economic data were decisive in elections, then President Ronald Reagan, seeking re-election in 1984 when the economy was beset by a 7.5 percent unemployment rate, wouldn’t have won 49 states. After all, his successor, President George H.W. Bush lost by more than 200 electoral votes when he ran for re-election in 1992, with the jobless rate at 7.4 percent.

The relative skills of Reagan and his Democratic opponent Walter Mondale as candidates (and the prowess of their campaigns) had a huge effect on the outcome of that contest. The same is true of the matchup between Bush and Bill Clinton in 1992.

More broadly, there are three reasons I put limited faith in the models and their presumption that the campaigns are irrelevant.

First, the models’ seemingly objective factors are loaded with ambiguity and interpretation that inject political handicapping. Lichtman’s keys, for example, include questions such as “Has the president achieved a ‘major success’ in foreign affairs?” or “Is the challenger ‘charismatic’?” These aren’t clear cut.

Even Silver’s data-driven model leans heavily on the subjective question of the “ideology” of the challenger. Silver tries to make this more scientific by assigning an ideological score to the 2012 Republican hopefuls; former Massachusetts Governor Mitt Romney gets a 49, and Texas Governor Rick Perry is rated a 67 (a higher score means a candidate is more conservative).

Immigration Debate

But Perry’s first major stumble in the primary race occurred when Romney attacked him from the right on the question of immigration: A generic “ideology score” can’t capture the fact that, for millions of voters (especially Hispanics) who see immigration as a make-or-break issue, it is Romney, not Perry, who is furthest to the right.

Second, the models cheat by capturing the performance of the campaigns and the candidates via backdoor measures. A major factor in Silver’s model is the presidential approval rating, which is offered as a measure of how the president is doing in his job. But as presidents get closer to Election Day, approval ratings rise and fall with the vagaries of the campaign, and are more likely to reflect the president’s standing in the campaign than his performance in the White House.

Voters who make up their minds to vote for the incumbent president begin to raise their grade of his performance, and voters who decide to vote for his opponent lower the approval score. Building a model that rests on an approval rating is a bit like building one that relies on the results of the presidential preference poll: If it’s accurate, it’s because it’s measuring the campaign’s impact, not in spite of it.

For example, between Sept. 26 and Oct. 10, 2004, President George W. Bush’s job-approval rating in the Gallup poll dropped from 54 percent to 47 percent, below the critical 50 percent divide. Did this reflect a change in his job performance as president, or news that some policy or program had gone awry?

No. The significant events in that period were the candidate debates on Sept. 30 and Oct. 8, in which the Democratic nominee, Senator John Kerry of Massachusetts, was perceived to have defeated Bush. If Kerry were a less skilled candidate, and Bush had won those debates, the president’s job-approval rating would have remained higher, and he would have coasted -- not squeaked -- to victory.

Presidential job-approval numbers after Labor Day of an election year are a barometer of campaign events, not a measure of the president’s performance. As a result, they aren’t an independent variable that can be determined outside of the campaign’s dynamic.

Third, and perhaps most important, the U.S. doesn’t hold presidential elections often enough in a given time frame to provide sufficient data to model the drivers of an election result. There have been just three presidential elections in the Internet era, and only one in the age of social media. Since 1976, only one campaign -- in 2008 -- has been run outside campaign-finance limits; 2012 will be the second such contest.

Campaign Finance

Just because the poorly funded, poorly organized, rudimentary campaigns of 1944 or 1964 had little ability to overcome extrinsic factors and affect the outcome those years doesn’t mean that the sorts of campaigns we saw in 2008 and will see in 2012 will be similarly ineffective.

Saying that a model captures the dynamics of presidential elections based on data from 1944 to the present (as Silver does) is like saying you’ve modeled the effectiveness of cancer treatments or the computing power of microchips, from 1944 to the present. Data from 1944 have no predictive power to determine how cancer drugs or computer chips work today.

The same is true of elections. It is ridiculous to compare President Gerald Ford’s 1976 re-election effort -- which spent about $80 million in 2011 dollars -- with Obama’s 2012 effort, which will spend more than 10 times as much.

In five-card draw poker, the hand each player is dealt at the beginning has an impact on the outcome. But how they play those cards -- the bets they place, whether they keep or discard, draw or fold -- decides the final result.

Likewise, in 2012, factors such as the state of the economy and the ideology of the Republican candidate will certainly affect the president’s chances of re-election. But in the end, how the campaign unfolds -- the messages the candidates offer, the campaigns they run, their performance on the stump, their get-out-the-vote efforts and their debate appearances -- will make the difference. Candidates and their campaigns will dictate the outcome, not calculators.

(Ron Klain, a former chief of staff to Vice President Joe Biden and a senior adviser to President Barack Obama on the Recovery Act, is a Bloomberg View columnist. He is a senior executive with a private investment firm. The opinions expressed are his own.)

To contact the writer of this article: Ron Klain at

To contact the editor responsible for this article: Max Berley at