Primary Performance Does Not Predict General Election Outcomes

John J. Pitney, Jr.
Professor of government at Claremont McKenna College and coauthor of American Government and Politics: Deliberation

When candidates win primaries or caucuses, they often try to spin the outcome as a portent of the general election. So what do the nomination contests to date tell us about what will happen in the fall? Not much.

The primary and caucus electorates are different from the general electorate. For one thing, turnout is smaller. As of last night, the total number of votes in Michigan’s Republican and Democratic primaries seemed on track to approach a record 2 million. By contrast, with a so-so turnout for a general election, 4.8 million Michiganders voted in November 2012. Primaries attract strong party identifiers, and in many states, voters have to register with a party in order to take part. Almost by definition, then, primaries under-represent political independents.

Caucuses have even lower turnout than primaries because they are more time-consuming. The parties have tried to enable participation by active-duty military and disabled veterans, but many other groups (e.g., single parents, shift workers) find it hard to attend these meetings. In no sense do caucuses make up a representative sample of their states.

There is an even simpler reason why a primary or caucus victory does not foreshadow state results in the general election. The other party may just be a lot stronger in that state. In 2008, Barack Obama won the Wyoming caucuses and the Utah primary – even though he had practically no chance of taking either deep-red state in November. (McCain carried both with more than 60 percent.) This year, Hillary Clinton swept the South Carolina primary. Unless something catastrophic happens to the GOP, her November chances in South Carolina are roughly equal to her chances in the Powerball lottery.

Donald Trump would deny that this analysis applies to him. He claims that his primary victories are hugely significant because he is getting an amazing number of Democrats to cross party lines. This claim is questionable. Exit polls show that Democratic identifiers accounted for no more than six percent of the voters in early GOP primaries – about the same level as 2012. In Michigan yesterday, the figure was 7 percent, two points lower than it was four years ago.

Overall GOP turnout is up because of the excitement surrounding the race, but the makeup of the electorate is not much different. Emory University political scientist Alan Abramowitz tells The Christian Science Monitor that Trump is “not bringing in a whole bunch of working class voters as a proportion to the electorate, and when he keeps claiming that Democrats are coming over to vote for him, that’s not true, either.”

Trump has touted a poll showing him winning 45 percent of Hispanic voters in the Nevada caucuses. As with caucus-goers more generally, Hispanics who voted in the GOP caucus were unrepresentative of the larger population. (Hispanics made up only 8 percent of the caucus total, compared with 18 percent in the 2012 Nevada general election.)   A recent national survey shows that 80 percent of Hispanics have an unfavorable view of Trump and that only 16 percent would vote for him against Hillary Clinton. Indeed, the Hispanic electorate might grow as more and more non-citizens take out naturalization papers for the express purpose of voting against Trump.

The lesson: nomination politics is not the same as general-election politics.

Comment on this article


Please note, statements and opinions expressed on the Fox&Hounds Blog are solely those of their respective authors and may not represent the views of Fox&Hounds Daily or its employees thereof. Fox&Hounds Daily is not responsible for the accuracy of any of the information supplied by the site's bloggers.