Generic polls defectuve

[CARLHAYDEN]

As usual, there are a bunch of generic polls which purport to show that the Democrats are preferred to take control of the House of Representatives.

Well, these polls are defective for several reasons:

First, turnout.  Most of these polls are based on registered voters and not 'likely voters' (or a poor screen is used). Using a reasonable (historical) voting screen, the Republicans typically run much better in most off year elections than generic polls suggest.

Second, all Congressional districts are not created equal (at least in partisan distribution).  Lets look at the results of the 2004 Congressional elections in California (which represents approximately one eighth of the House).  While 22 of the 33 districts won by Demcrats were won by at least a two to one margin (two thirds of them), only 7 of the twenty districts won by Republicans were by a two to one margin (or more).

Third, Congressional elections (for House members) are NOT generic.  People may prefer a generic candidate of one party BUT 'their' representative (of a different party from the generic selection) is an exception.

Now, events can radically impact election results, and nobody knows what may happen between now and election day.

However, at this time there is no credible trend either way.

As an aside, I have learned of focus group interviews with Democrats voters, where considerable unease has been acknowledged about the purging of Senator Lieberman.