I suppose that the pollsters, wanting to think that they are doing something scientific, just want to measure what people think. But time and again, the results they give us are that Americans want to reduce foreign aid, believing that it is something like a quarter of the budget, whereas it is under one percent, reduce taxes, and increase spending on pretty much everything else.
Okay, so we know that.
How about asking questions that begin, "Given that the budget is finite," and go on to pose a choice: defense spending or education. Social Security or defense. Foreign aid or defense. (Sorry, I'm getting repetitive.) Housing aid or education. Scientific research or education. (That's starting to get harder.)
The objection will be, I suspect, that the pairing of the alternatives will influence the answers. Well, that might be interesting. If you set up the pairs correctly, you might get some priorities out of it. Or you might find that the American public prefers spending on education to spending on research to spending on defense to spending on education, a circular and again illogical set of preferences.
But wouldn't it be more helpful to see answers like that?
Inspired by this article.