Poll reading

I don't know why, but I still read Newsweek even though I basically spend my time just getting pissed off at what I'm reading. Some of their columnists just make things up and have become unreadable (Samuelson, Will), the editor is a pompous ass (Meachem), their science writer gets things wrong pretty much every column where I'm in a position to know (Begley), and as a whole, they focus way too much on meta-narrative and assuming that the truth must lie somewhere between what people with a (D) or (R) are saying, with no attempt to actually discover and report facts. All politics, no policy.

Anyway, one theme of this week's issue is that America is a bunch of fickle mush-heads. Now, I happen to generally agree with this, but they way their writers get to this conclusion is wrong. For example, I've read and heard (from many places I should note, not just Newsweek) that the American citizenry doesn't want to make tough choices, that we are babies who want contradictory things. A poll with then be cited, whereby a majority wants things that are diametrically opposed. But this is not proof that any individual in America holds these opposing ideas! For example, here's a hypothetical poll asking people about the deficit and taxes and spending:



If you were to only read the "top-line" results, in bold on the bottom, you'd laugh and say ha, stupid people, they want to reduce the deficit with no tax hikes and no spending cuts! Ha ha ha, they are so dumb! But that isn't true -- there are three types of people in this "poll" each of whom holds an entirely self-consistent view of what should be done. It is just that, in aggregate, the results make no sense. I see this all the time in media reports. Guh.