It’s unfortunate when party propaganda overrides objectivity when it comes to analyzing polling numbers, because it starts to color your impressions of someone’s future reliability. In this case, Steve Singiser:
OH-01: Right wing poll claims GOP lead over Dem incumbent
If you bought the FDL/SUSA polls from a few months back, then I suppose we should call this progress. At the time, you might recall, FDL/SUSA had Democratic freshman Steve Driehaus on life support, with former Congressman Steve Chabot (R) up by seventeen points. Now, at the end of May, even Chabot’s own internals are not that rosy about his chances. That said, the poll still gives Chabot a robust lead, with a 53-39 advantage over the incumbent.
The DCCC worked overtime to get party loyalists to spread the word that somehow the polling done by Survey USA for FDL was flawed in its methodology. The DCCC refused to release their own polls, however, saying that those were private. Most honest brokers would’ve refused to print such attacks without demanding that the DCCC provide numbers to the contrary, since they most certainly do have their own polling on every district across the country.
Markos himself weighed in:
But others were just happy to be tapped to whistle the party tune by the powerful and the important (which I suppose only works if you consider anyone at the DCCC to be a member of that group).
Let’s review: FDL’s Survey USA poll, taken between 1-12 and 1-14, found Chabot to be up over Driehaus, 56-39, with a margin of error of ±4.1% (crosstabs). They polled 600 likely voters, made their methodology available, and even went back and addressed the DCCC peddled-critique that the poll had undersampled young voter. After Survey USA re-weighting the poll, they found that the results did not change.
We challenged the DCCC at the time to release their own polling on OH-01, and even offered to poll the district together. They refused.
Five months after we released our poll of OH-01, the GOP firm Public Opinion Strategies poll finds Chabot still up over Driehaus, 53-39. They sampled 400 likely voters. What’s the margin of error? Where are the crosstabs? Nobody asks. Once again, in the CQ article, the DCCC “questioned the validity of the poll’s findings.” Does anyone ask them to provide their own numbers before printing this? Well, no.
Did Singiser mention that the polling about Colleen Hanabusa that was dead wrong about her chances in the Hawaii congressional election belonged to the DCCC? The poll that the DCCC was pushing to justify their support of Steve Case showed Hanabusa trailing him by 5 points. She beat him by 3.2%. The seat went to a Republican.
As SurveyUSA’s Jay Levy said at the time they released the Chabot-Driehaus poll, “the highway to high office is littered with the road kill of political operatives who find it easier to campaign against a pollster than an opponent. If the DCCC has polls showing congressional Democrats doing better, release them.”
But what could have happened in five intervening months? According to Pollster.com, most polls (both Democratic and Republican) show that the Democrats have gained between 4-6 points on the generic ballot since then:
SurveyUSA showed Chabot ahead of Driehaus by 17 point in January. Five months later, a second poll shows that there are 14 points separating the two. The logical conclusion would be, in the absence of any evidence to the contrary, that the second poll actually validates the first and the results roughly track the closing gap in the generic ballot.
Both FDL and SurveyUSA have said that FDL had nothing to do with the methodology of the polling. And Singiser has had no trouble relying on SurveyUSA results since then. So what is his problem with the SurveyUSA polling, specifically? If he doesn’t trust their methodology, he shouldn’t be quoting it regularly without noting that he finds their results suspect. Is it the fact that we were the ones choosing the district that somehow influenced the results? That’s a novel claim, and if so, he should certainly develop it further.
How Singiser arrived at “even Chabot’s own internals are not that rosy about his chances” is a complete mystery. Without further documentation on his part, there’s nothing positive to say about the intellectual leaps it took to get there.