Robo-polls and Mitt Romney's "boost"

For the second time in as many weeks I have rushed down to WGBH Boston studios to do short segments on a local news interview show, Greater Boston. In both cases it was to comment on the New Hampshire Primary, and in this regard I am the surrogate for those more expert on the topic who are much busier than me -- and besides, I live in Massachusetts, so the trek is really a grand detour to heading home. The first segment -- taped on April 30 -- focused on the campaigning now taking place in NH, while yesterday's focused on the latest SurveyUSA poll (for WBZ-TV, released on Monday, the 7th) that gave Mitt Romney a boost to the lead in the contest among likely voters. There is much hype and spin associated with the Romney boost -- the local media quickly bought into the idea that he had won the GOP candidate debate held in California last week. The problem with that explanation is found in another state-based poll issued by SurveyUSA on the 7th -- this time for California and this time showing Romney in a very distant third position, in statistical tie with non-candidate Fred Thompson (who ranked at 11 percent in both polls). So why the so-called boost? Perhaps it is the fact that Romney -- an already familiar face to New Hampshirites who are not likely to be subject to the religion issue -- has launched his TV ad campaign ahead of the pack and is getting some numbers from that. If so, it is a flash-in-pan lift, for the real politics of NH primary season is retail and Romney's boost will quickly fade among those whose opinions are sensitive to such unique media hyping as "I am veto man". But it can also be a manifestation of the "robo-poll" methodology used by SurveyUSA -- the subject of considerable controversy among professional pollsters, rendering SurveyUSA as one of the "black sheep" among pollsters. In a politically sophisticated place like NH -- especially during primary time -- the use of "interactive voice technology" polls is not likely to get the same kind of reception or response (and cooperation) rates as it might in other states -- like California. The potential flaws in IVT would be magnified in NH -- thus making the poll results very suspect. The Boston media, however, was quick to buy into this spun poll -- although the Boston Globe's coverage did raise the issues of credibility well into their article. I think it would have been wiser for the rest of the media -- including Greater Boston -- to have held back the hyping of these results until a more credible outfit (e.g., the UNH Granite State Poll which uses more traditional methods for WMUR and the Globe) has its say. I get the sense that the results would be a bit different.... Tags: New Hampshire Primary, Greater Boston (WGBH), Mitt Romney, SurveyUSA, Boston Globe, UNH Granite State Poll, WMUR-TV, robo-polls, California primary, polling, GOP debate, accountabilitybloke, Mel Dubnick