The Monkey Cage

Jim: I agree with much in this piece — e.g., lots of people don’t make choices based on ideology, and swing voters are more likely to think in terms of performance.

Two issues, though:

1) I do remember Zaller’s model in the Monica Lewinsky piece. He did find that the ideological extremism of presidential candidates was associated with lower vote share. Perhaps just Goldwater and McGovern are driving that result. But it would be interesting to see if that’s true in other kinds of races.

2) What about people whose partisan identities are not yet crystallized? These would include some young people as well as some immigrants. Based on the “critical years” idea, we would expect political events to be crucial at this stage of their political development. If the GOP is out-of-step with their attitudes — e.g., on gay rights or immigration or what have you — will they essentially lose these cohorts forever?

Another thought: The ANES self-report question is widely viewed as misleading. It surely captures something, but it’s not clear that it captures what we mean when we go on to say things like people do or do not make choices based on ideology. Or on policy preferences for that matter.

To just expand on what Hans’ said, I believe Jim Stimson points out this complication with ideological self-reports in the very same book that this author cites. Stimson’s finds that responses to specific policy questions tend to be consistently be more liberal than self-reported ideology.

What is the true “ideology” of a voter who says she is conservative but has mostly liberal policy preferences? That seems more like a philosophical than an empirical question.

But however we interpret it, significant numbers of American’s like the word conservative more than the policies currently associated with that term. It would have been nice if this author addressed that. But maybe we should be grateful for baby steps.

Sorry about the typos in that last comment. It should read “…what Hans said…” and “Stimson finds that responses…” and “…numbers of Americans like the word…” This is why I don’t blog :)

I agree that this piece is a very useful corrective, albeit more for the media than for politicians. It is good to show reporters and pundits the importance of valence issues (a term the piece maybe should have included) and stop credulously accepting (and glibly producing) narratives about “mandates” and overstating the extent to which voters think spatially. To the extent that they do this out of ignorance (not the only explanation, I’m afraid, see Marjorie Hershey’s JOP article, “The Constructed Explanation”) this piece is helpful.

But for party politicians it’s a different picture. The fact that valence or performance issues typically trump the effect of ideological or platform changes within their typical range (to say nothing of “campaign effects”) doesn’t give parties much direction. One might say it suggests that they should just stand pat and wait for the tide to bring them in, but if the changes in positioning they can execute only matter a little some elections are close enough for them to matter. So parties still want to know how best to maximize the limited influence they have on the outcome.

The other point is that saying not everyone thinks spatially doesn’t mean that there isn’t issue voting; it’s just a multi-dimensional world.

I appreciate the reaction and commentary on this post (as well as the others). I have already learned a lot from it. There’s much food for thought in your feedback.