Why Are Primaries Hard to Predict?

I actually wrote this a few years ago, but it’s relevant to our discussions about House Majority Leader Eric Cantor’s recent primary election loss. Here’s what I wrote:

Presidential elections are predictable. At the national level, economic conditions — measures like the change in per-capita income in the year or so preceding the election — usually do a good job predicting the popular vote. At the state level, you can usually predict the outcome given the vote last time, after adjusting for the economy and the home states and regions of the candidates. . . . This has all been known for a while, at least since the work of political scientists Doug Hibbs in the 1970s and Steven Rosenstone and Robert Erikson in the 1980s. . . . [But] it certainly doesn’t mean that every election can be accurately predicted ahead of time.

Presidential general election campaigns have several distinct features that distinguish them from most other elections:

1. Two major candidates;
2. The candidates clearly differ in their political ideologies and in their positions on economic issues;
3. The two sides have roughly equal financial and organizational resources;
4. The current election is the latest in a long series of similar contests (every four years);
5. A long campaign, giving candidates a long time to present their case and giving voters a long time to make up their minds.

Other elections look different. . . .

Now let’s look at the primary election between Cantor and David Brat (recently considered in this space by Robert Boatright):

1. The incumbent was well known, but the challenger was not.
2. Candidates very similar in their ideologies. Not identical, but they were mostly competing to show how far to the right they were. In the larger context of American politics, there was very little daylight between them.
3. The campaign was highly asymmetrical, with the two candidates having different sorts of campaign resources.
4. Contested Republican primaries are not regular events in this district.
5. The campaign was short.

The point of all of this is not to say that I could’ve predicted Brat’s success; rather, it’s the opposite. This election lacked a lot of the features that give predictability to national general elections.

Andrew Gelman is a professor of statistics and political science at Columbia University. His books include Bayesian Data Analysis; Teaching Statistics: A Bag of Tricks; and Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do.



Success! Check your inbox for details. You might also like:

Please enter a valid email address

See all newsletters

Show Comments
Most Read Politics



Success! Check your inbox for details.

See all newsletters

Next Story
Sarah von Billerbeck · June 18, 2014