Stuart Rothenberg:
On Sunday, a Huffington Post headline screamed what most Democrats were hoping: “GOP In Grave Danger Of Losing House In 2014, PPP Polls Show.” Of course, anything coming from Democratic pollster Public Policy Polling and MoveOn.org Political Action, which paid for the surveys, must be taken with at least a grain of salt.
PPP isn’t your typical polling firm. Its surveys often are intended to boost Democratic recruiting, fundraising or prospects. In this case, the “polls” were almost certainly commissioned to create a narrative about the political repercussions of the shutdown and the nature of the midterms.
It’s no coincidence, then, that the PPP memo accompanying the results, written by Jim Williams, observes, “The surveys challenge the conventional wisdom that gerrymandering has put the House out of reach for Democrats.”
Not surprisingly, the Democratic Congressional Campaign Committee sent out multiple fundraising emails in the hours after reports of the PPP polls surfaced, and dozens of Democratic candidates and liberal groups did the same.
That’s the standard modus operandi these days on both the right and the left: have a sympathetic media organization or polling firm assert some alleged finding, and then have fellow travelers cite the initial report to try to raise cash or create momentum. It is becoming (yawn — excuse me) a little trite.
Anyway, the Huffington Post reported that of the 24 House districts polled, “Republican incumbents are behind in 17 of the districts analyzed.” Of course, the “polls” did not include head-to-head ballot tests of likely nominees (even though the surveys could have included candidate names in many contests), but instead relied on a messy question that was part “re-elect” and part “generic ballot.” The results are of little or no use because that is not the choice voters will face on Election Day.
Moreover, at least five of the 17 Republicans who are “losing” either have no serious opposition or have less-than-top-tier opponents at this point: Steve King (Iowa’s 4th District), Andy Barr (Kentucky’s 6th), Kerry Bentivolio (Michigan’s 11th), Patrick Meehan (Pennsylvania’s 7th) and Sean P. Duffy (Wisconsin’s 7th). Bentivolio may not survive a GOP primary.
Each PPP survey asked seven substantive questions and four demographic ones. Some of the questions were loaded, and as I have noted previously in dissecting PPP polls, the “more likely/less likely” question is a horrible one to use in surveys because it tends to measure the underlying attitude rather than gather useful information about an issue’s eventual importance as a vote cue.
PPP conducted a “second ballot” after additional survey questions gave more information to respondents and allegedly found four more Republicans “losing.” But as David Nir of Daily Kos Elections wisely noted in writing up the results:
