Tuesday, October 10, 2006

A Little Political Science: Polls

Today, USA Today displayed a poll proclaiming "Dems Gain Big Lead." Now I'm not saying I'm quite as confident now as I was before that I would win my bet with Publius, but it is interesting how un-useful generic "party control preference" polls really are. Take a look.

The pollers only actually polled 622 people who were self-proclaimed "likely voters" nation wide over the weekend. I'm assuming they cold called people, which is the standard method. There are 435 House races this poll purports to encompass. That means that they only polled an average of 1.43 people per Congressional district. It is likely many districts aren't represented at all. And there is no mention of how people see their OWN Representatives, who even now have approval ratings which average over 60%.

Also, publishing this poll during a major national crisis is just bad journalism. The North Korean situation is sure to re-focus voters away from the Foley situation, and as such, the data was already obsolete by the time they went to press.

I still think the GOP will retain the House, although it will be a much closer call. The Democrats overreached on Foley, and cannot answer the question, "What will YOU do about North Korea - try to bribe them again and just trust that they'll finally act in good faith?"

1 comment:

Orrin Johnson said...

People vote on a lot of things that the legislature has less control over than foreign policy - abortion springs immediately to mind.

But the Congress actually has quite a bit of control over foreign policy, especially with regard to funding. And even so, it matters to me - and I think to a lot of other people - whether or not the Congress will support the President in times of crisis (even if they disagree on aspects of it), or if they will actively seek to cut him off at the knees.