Be Careful With Polling on Spending

In my post over at Plum Line today, I beat up on CNN because they claimed their new poll showed that people want deep domestic spending cuts, when in fact as usual all it showed is that people want spending cuts in the aggregate but dislike specific cuts. Guess what? Gallup has a poll, and it shows exactly the same thing. Gallup's version has the popular "cutting spending other than defense, Social Security, and Medicare." Matt Yglesias interprets that as indicating that "non-defense discretionary spending" is unpopular (technically, that independents agree with Republicans about it) and therefore programs for "both anti-poverty spending and 'win the future' stuff on infrastructure" could be in trouble.  But I'd bet heavily that it's just an artifact of poll design. If instead they had asked about cutting spending other than health care, transportation, and education, that would have been popular! If they had asked about cutting spending other than programs that keep people out of poverty...once again, popular!

We know this stuff. Spending is unpopular in the aggregate, but virtually every polled category of spending is popular with only one real exception, foreign aid (and that one is dubious, because it mostly reflects wildly inaccurate ideas of current spending levels). Mostly, though, public opinion on this is inconsistent, nothing more.