The latest ABC/WaPo poll is pretty much a wall-to-wall bummer for Democrats, showing the president’s job approval rating dropping to 41%, the generic congressional ballot tied (this measure appears to overstate Democratic support by around 5%), and all sorts of indications that Republicans are more likely to turn out than Democrats this November.
Let’s untangle all this woe for progressives. There being no obvious reason for a sudden plunge in the president’s approval rating, I suspect this particular plunge is a polling outlier. It’s troubling for Democrats mainly because we’re at a point in the 2014 cycle where Obama’s approval ratings ought to be trending slightly up rather than down.
The generic ballot is mainly important late in the cycle when the shape of the electorate is more or less locked in. And all the indicators of superior Republican willingness to vote in a sample composed of registered voters are entirely predictable, and would almost certainly exist if the president’s approval ratings were ten points higher.
The trouble with the way this poll and others like it are going to be interpreted is that likelihood to vote, largely a function of age and ethnicity, is being attributed to variables like support for the president and for his party. So we get the illusion of a midterm election conducted as a referendum on Obama, and probably giving Republicans a better-than-even chance of victory in 2016, too.
The reality is that in a relatively stable environment (i.e., no dramatically changing economic indicators, no immediate threat of a major war), given the balance between the two parties and the distribution of presidential-year and every-year voters, you’d expect a Republican “win” in 2014 and another Democratic “win” in 2016. Confusing those expectations with gyrating public opinion on Obama–for him in 2008 and 2012, against him in 2010 and 2014–kind of ignores everything we know about turnout patterns that are going to exist no matter how the president or for that matter the opposition is “performing.”
