Photographer: Scott Eells/Bloomberg

Question Day: How Do Issues Work in Elections?

Jonathan Bernstein is a Bloomberg View columnist. He taught political science at the University of Texas at San Antonio and DePauw University and wrote A Plain Blog About Politics.
Read More.
a | A

To kick it off today, Noah Berlatsky asked:

Often you write about policy positions or issues that won't affect voters choices in Congressional or Presidential elections. What issues or policy positions do actually have an impact in those elections with voters rather than with elites or interest groups? Are there any?

Excellent question.

I'd start with the idea that party identification is a huge factor in vote choice, which tends to strongly limit the additional direct effects of any particular policy issue. Then on top of that is a basic "How are things going?" effect, which is largely (although not necessarily exclusively) about the economy, but less in terms of policy positions than just whether it's perceived to be good (helping the in-party) or bad (helping the out-party). And then add to that perceptions of the president, which in turn are largely but not exclusively determined by party and the economy.

So what breaks through that, making a specific issue have a direct effect on vote choice? Generally, it's going to be something that's even stronger than party -- so something that has a major real, tangible effect on some group of voters, and separate from their normal party leanings. That does happen, but it's usually going to pretty limited. But I'd think of something like immigration and Latino voters; it's certainly possible that for those who are strongly in favor of a comprehensive bill (that is, for a path to citizenship), it might trump party or be a major factor for those with weak party ties.

Note that it has to cut differently than normal party ties to have any effect. A lot of people of course feel very strongly about abortion ... but virtually all of them are already sorted by party, so anything new coming up about abortion isn't going to change anything about their votes. At best, it could affect turnout if "their" issue is suddenly more salient, but if that really is their issue, then it's probably already pretty important for them even if it's not currently in the news. Otherwise, you're actually getting reverse causation: People who always vote for Democrats, say, and broadly agree with Democrats on most issues, feel as if the cause of this particular vote is abortion if it's been featured in the campaign or has been in the news -- but they're going to be voting for Democrats either way, and all that changes is what they're thinking about when they do that.

That's direct effects. For indirect, there's some evidence that there can be effects from what's in the news. So for example with George H.W. Bush, who was perceived as a good foreign policy president but a bad domestic policy president, his approval ratings would go up if foreign affairs dominated the news, but down when the big story was the economy -- even though voters' perceptions of Bush on the economy and Bush on national security would stay the same. Presumably "issue ownership" works the same way; if issues that people trust Republicans on are in the news, it will tend to help Republicans. I'm somewhat skeptical of that concept ... I actually have a new book on my desk, by political scientist Patrick Egan, which looks into the question, so when I've looked through it I'll perhaps post something more.

Even more indirect would be any effect that specific issues might have on partisanship in the first place. I assume (but don't know the literature, so I can't say for sure) that it's hard to disentangle effects of "issues" as opposed to just group membership. Say, that is, that you are a union member ... to what extent does just being a union member make you likely to become a Democrat, as opposed to a specific issue about unions?

So mostly, issues are going to have only minor direct effects, and the indirect effects are either small or complicated. Of course, in a close election, even very small effects can make a huge difference, so even if it's small, it could be extremely important.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Jonathan Bernstein at

To contact the editor on this story:
Frank Wilkinson at