Over at The Economist, Will Wilkinson depicts a murky future for gluttonous Americans who want to have their cake and eat it, too:

According to a new CNN poll, 63% of Americans want the imminent “super committee” to call for tax increases on high earners and 57% want large cuts in domestic spending. Sounds sensible! But don’t get your hopes up. Only 47% favour big cuts in defence spending. Worse, a mere 35% support “major changes” in Social Security and Medicare. I guess there’s a reason defence and old-age entitlements dominate the budget: voters like it that way.

The longtime contradiction between what voters would like to keep (everything) and what they’re willing to give up (very little) is a topic I’ve discussed before. But Wilkinson’s post is worth mentioning for another reason. It reveals some major shortcomings in the way polls are framed. I’m not talking about the kind of framing that leads respondents to certain conclusions; that’s a fairly common problem and people know about it. What I’m talking about are the kind of questions that are isolated such that respondents can actually get away with calling for “big cuts” in one question while not having to stand by that answer in later questions about what should get put on the table.

Take this CNN poll, for example. The survey effectively absolves respondents of the responsibility for making a choice. Asking people whether they’d accept cuts to defense spending in one question, and entitlements in another, tricks poll-takers into considering each of these sectors out of their proper context. Such questions are misleading because they presume there is nothing else on the table when respondents know that there are other options the pollster hasn’t mentioned. Choosing not to cut defense spending becomes easier when you know you can just cut from something else that’s not in the question.

Lawmakers don’t have that luxury. Neither should poll respondents. Wouldn’t polling be much more enlightening if the data told us that given a strict choice between cutting entitlements and cutting defense spending, respondents preferred the latter by such-and-such a margin? The snapshot of voter priorities would be much more informative than what we get now, which is a general sense that everything is important. This is also partly why I like the New York Times‘ You Fix the Budget interactive feature. It shows users they can’t get the country to where it needs to be with an uncompromising appetite for popular programs and an aversion to taxes. It also circumvents the don’t-cut-anything impulse by telling users that they can have some of what they like—just not all of it.

Image credit: Laenulfean