Thursday, October 23, 2008

Looks like someone did the research to back up my article about polls.

I wrote an article yesterday saying we should stop looking at the polls. I said they were biased, but didn't have any research to back it up. I thank Wizbang and Scott Martin for doing the work for me.

A Look at Polling Bias
Wizbang has a post up that may explain the difference we repeatedly see between polls and election results, and which definitely sheds some light on the bizarre big leads many polls are showing both nationally and state-by-state:

2006 was a bad year for republicans, a year when republicans stayed home and democrats used the opportunity to win a number of close races and take over control of the House and Senate. In a number of states, therefore, it's not surprising that democratic party supporters gained a few points (usually 1 to 3 points) relative to 2004 in voter participation. So I went back and looked at voters by party affiliation, and compared those balances to this year's weighting by Survey USA. In thirty-six states, the party affiliation weights for democrats used by SUSA was five points or more higher than in 2006, a high-water mark for democrats. In twenty states, the party afiiliation weights for democrats used by SUSA was ten points or more higher than in 2006, and in eight states, the party affiliation weights used for democrats by SUSA was thirteen points or more higher than in 2006. Significant battleground states affected by this bias are as follows:

Pennsylvania: D+5 in 2006, SUSA using D+19, 15 point variance
Indiana: R+14 in 2006, SUSA using R+1, 13 point variance
Nevada: R+7 in 2006, SUSA using D+6, 13 point variance
Colorado: R+3 in 2006, SUSA using D+9, 12 point variance
Iowa: R+2 in 2006, SUSA using D+10, 12 point variance
Virginia: R+3 in 2006, SUSA using D+9, 12 point variance
Ohio: D+3 in 2006, SUSA using D+13, 10 point variance
Missouri: R+1 in 2006, SUSA using D+7, 8 point variance
North Carolina: R+1 in 2006, SUSA using D+5, 6 point variance

So in nine battleground states we are seeing major variances between the actual voter turnout during an election with a very depressed Republican electorate and this election, in which the electorate seems rather charged up. Where is the logic in that?

If it has to do with increased voter registrations favoring Democrats, you have to ask at least two questions:

1. What is the likelihood of these new registrations actually voting? Historically it is quite low, but with the dynamics of this election it could be higher than usual.

2. They are completely ignoring the tens of thousands of people in many of these states who were employing "Operation Chaos" during the primaries and switching their registration to vote for Hillary Clinton in order to keep her alive in the race. Why is this? The news of the large amounts of these Rush Limbaugh voters was certainly widely reported.

Additionally, as has been pointed out before, national polls focus largely on urban areas, which tend to be much more liberal-oriented. Even if they get the party percentages close, they are much more likely to attract "strong democrats" and "weak republicans."

I'm so sick of polls.

see more from Scott Martin here:

1 comment:

  1. Polls certainly are crazy this election. But in terms of this particular post, the research needs to be done again, because SUSA doesn't weight by party ID.

    So, their results could be wrong for any number of reasons, but it's not because they're weighting Democrats too heavily in their sample. They only weight by demographics (age, race, sex, etc.). Whatever ends up being the party ID in their polls flows from that. It's not something they select. Democrats could certainly be overrepresented in their samples, but it's not a decision that SUSA is making. Also, given that party ID weighting varies from poll to poll (and like in SUSA's case, is sometimes not used at all), a thorough discussion of this would need to include an analysis of several of the major polls.

    For a discussion on whether its appropriate to weight by party ID in the first place, read here.