Why I don't link to house district polls, congressional district polls, etc

I see people quoting polls from CD 18, or a state house seat in Dallas, and I think, "why?"

First, I don't report polls unless I know who the pollster is, preferably with the exact wording of their question provided as well. Polling isn't quite like indictments ("a prosecutor can indict a ham sandwich") but it's close. Unreputable pollsters have plenty of tricks at their disposal to get the numbers they want to get.

Second, in terms of predicting a statewide race, a poll from any given house seat is not helpful. It's not a representative sample. If I take a poll in Austin, and it shows Ronnie Earle leading in the Dem lt gov race, do I know anything more than before? Not really.

This is especially true when there is sufficient data. Right now, Rasmussen and PPP is polling regularly, and a few new polls appear every week. There's no lack of data. There are limited exceptions, of course. If a trend can be established, that might be useful ("Gee, candidate X was winning by 10 last month and now is losing by 25!") to know, especially if one candidate was attacking on issues specific to the area of the congressional district.

But in general I think such polls confuse more than they aid. So I don't mention them here.

Posted by Evan @ 02/13/10 04:32 PM


Previous Entry | Home | Next Entry


No comments yet

Add Comments

No flames or impolite behavior. HTML will be stripped. URLs will be transformed into hyperlinks.