Firedoglake Polls Take Some Heat

Firedoglake has been commissioning some polls on moderate Democrats, done by SurveyUSA, showing some grim results for the Democrats in question: Rep. Steve Driehaus (D) trailing former Rep. Steve Chabot (R) 56-39 in Ohio's first district, Rep. Baron Hill (D) trailing GOP challenger Mike Sodrel 49-41 in Indiana's ninth district, and Rep. Tim Bishop (D) leading his challenger by a thin two percentage points in New York's first district.

But the polls have taken some criticism for their sample sizes: Swing State Project and The Washington Post's Chris Cillizza have both raised questions.

The issue is the samples of 18-34 year-olds polled: in the Ohio poll, zero percent of respondents were from the 18-34 age group; in New York, 1 percent were 18-34; in Indiana, 5 percent were 18-34.

18-34 year-olds are generally tough to poll. At the younger end of the spectrum, college students and recent graduates tend to move around a lot, live in group houses, or live with their parents--and many of them only have cell phones. SurveyUSA does automated polling, which doesn't sample cell-phone-only respondents. Many automated election polls have proven to be extremely accurate, so that's not to knock the entire methodology; but it is a specific difficulty in a specific age group.

Firedoglake and SurveyUSA have responded to the criticism here, and Firedoglake makes note that it has nothing to do with the methodology of its commissioned polling.

Jay Leve of SurveyUSA responds to say that "It is a sample of likely mid-term voters. There is no one "right" way to draw a sample in a congressional district 10 months from a midterm election, but this sample was drawn carefully and defensibly. To be included in the sample (SurveyUSA's criteria, not the client's):  the voter had to be registered with the secretary of state; had to have a telephone; had to have voted in 2008 and had to have voted in either [2006 or 2002], and had to confirm that he/she resided in the district being surveyed.  By design, this was not a survey of registered voters (which would have resulted in a younger sample).

"Reasonable people can disagree about exactly what percentage of the electorate in 2010 will be age 18 to 34 , and I am not defending any specific turnout target. But most would agree that midterm voters are older."