Josh Dyck

Josh Dyck is assistant professor of political science.

• Print

• Comments

How is the current financial crisis affecting the election?

When the election context is a tanking economy, that’s bad for Republicans, particularly if they are the incumbent party. It is not the conversation the Republicans want to have. When the economy is in poor shape, people vote against incumbents. But this can cut both ways. The electorate votes for Democrats when they want job growth, but in 2004, when John Kerry tried to claim ownership of an issue that his party doesn’t own—national security—it was a debacle for him. This is because each party owns certain issues. The Republicans are trusted to cut taxes and be tough on foreign adversaries; the Democrats are trusted for providing a social safety net and promoting job growth. The financial crisis has fully shifted the campaign conversation to an issue on which voters tend to trust the Democratic Party more than the Republican Party to fix.

Your research has found that people who register to vote close to elections tend to vote, which is why campaigns should focus on registering voters very close to an election and work to boost turnout of these voters. Is this holding true for this year’s election?

The presidential race is shaping up to be close, which means that turnout could very well be the deciding factor again, and insofar as the parties can win the turnout battle, swinging a swing state is possible. The findings from our paper suggest that there is a fertile ground to be had in registering new voters in weeks leading up to the closing deadline, which is the first week of October in many states. Voters registered in the course of an election campaign are more regular voters than those who registered during the course of the year at the DMV, etc. Also, given the scores of new voters brought into the primary process through the competitive and historic Democratic race, it will be interesting to see if these folks turn out on election day. The fear for Obama has to be that many young voters who registered in the spring are college students who are notably transient. If they changed residences or moved after graduation, bringing them into the fold during the primary may be insufficient to get them to turn out in November.

The polls seem to change from week to week. Just how accurate is most polling?

Polling is accurate, but you have to understand what accurate means. A typical poll of 1,000 voters is accurate to within a margin of error of about plus or minus 3 percent, 95 percent of the time. This often means that when we think we are observing changes in polling, we may just be observing random error. The best way to track polls is by taking the running average as the Web sites pollster.com and fivethirtyeight.com do. This gives us a better handle on which poll movements are due to random error and which are due to actual campaign-specific effects.