As the first wave of early exit polls is released, you might be tempted to find some signs of which way the political winds are blowing. But a few words of caution: Exit polls are not very helpful in gauging turnout. And because so many people vote early, they are incomplete.
The early exit polls are most useful in figuring out which campaign messages resonated with voters. So pay attention early in the night to issues. Ahead of the midterms, health care has consistently been the most important issue for Democrats; while immigration (and the economy) have been prominent issues for Republicans. The degree to which we see voters emphasize any of these issues in their exit poll responses could be a sign of Republican versus Democratic enthusiasm.
The early exit polls, according to CNN, indicated that 4 in 10 voters say health care is the most important problem facing the country, whereas 2 in 10 say the economy and immigration are their top issues.
One other interesting nugget from this first batch of exit poll results, according to CNN, is that 1 in 6 voters say this is the first time he or she has cast a ballot in a midterm election and a majority feels the country is on the wrong track.
Later, as we get more complete exit poll surveys, you'll want to pay attention to voter preferences for a particular candidate. History tells us that women, minorities and young people are all more likely to support Democrats, while men, evangelicals and blue-collar voters are more likely to vote Republican. Tonight, pay attention to whether there's a substantial uptick in the Republican or Democratic preference for any group.
Many of the most competitive races for the House of Representatives will be fought in the suburbs, where anecdotally we have been hearing that white married women could be moving away from the GOP. Whether that shift is real or merely anecdotal, we'll see tonight. A couple of key demographic groups to watch: white women as a whole, white college-educated women and married women.
In any group, small shifts from past years could be nothing more than statistical noise. But big swings could suggest real shifts that will help us understand the results.