John Hinderaker looks at some poll results and writes,
Democrats lead Republicans in the generic Congressional preference poll by 40%-37%. It’s a paradox: voters prefer Republicans on the issues, but still lean toward voting for Democrats. One could speculate about why that is true; I think it is obvious that the press’s ceaseless attacks on Republicans are part of the explanation.
The Gamble, a new book by John Sides and Lynn Vavreck, warns against any attempt to over-interpret polls or elections. (if you search for “The Gamble Vavreck” the book has many different subtitles, which strikes me as weird. Is PUpress micro-targeting?) They argue that the press is an opinion follower rather than an opinion leader.
Their view is that many voters are partisan and unpersuadable. Swing voters tend to move with the state of the economy.
I am a bit skeptical of the economic variable. They say that economic conditions in 2012 were good enough to re-elect Obama. They look at GDP growth in the first half of the year, which looks fair. But other measures, such as the employment to population ratio, look pretty terrible. Why should voters focus on the one but not the other? My hypothesis is that, given that we have only about a dozen observations (Presidential elections) to go by, with sufficient specification searching and information from prior research, you can get a good fit mostly on luck. The fact that the fit held in 2012 does relatively little to lead me to dismiss the luck factor.
Nonetheless, one of my take-aways from the book is that issues matter much less than pundits believe. The idea that voters are sending clear messages on issues is tempting to claim and rather difficult to defend.