In a recent article we wrote for THE WEEKLY STANDARD (“The Canary In The Coal Mine,” Sept. 23, 2013), we reported on the findings of six surveys conducted to test the strength of the congressional exemption of Obamacare issue.
Included in that analysis was a report of a ballot test that showed Sen. Mary Landrieu trailing her GOP challenger, U.S. Rep. Bill Cassidy, by 30-57 percent. This number is well outside the norms of current Louisiana polling, in which Landrieu leads by a considerable margin – so we double-checked the weighting, adding in party affiliation and race as variables to see if that accounted for the discrepancy. Yet even after re-weighting, Landrieu was still losing to Cassidy 39/51, or -12pts, in our test group.
Why such a discrepancy from other polls? It is important to understand that the polling we reported, as we said, was done among a subset of 7,500 likely Louisianan voters who had served as a test group, receiving a week’s worth of light but apparently still quite effective messaging on the congressional exemption issue.
Because our aim is to assess the efficacy of not only the message but also the message delivery vehicles, we also polled a control group made up of likely voters who had received no information about the Congressional exemption, also now weighted similarly for party affiliation and race. Among the control group, made up of people who didn’t know about the congressional exemption, Landrieu was in fact leading Cassidy by 46/34, or +12 points – in line with other current Louisiana polling. Contrast that with the test group, where before we even mentioned the exemption in the survey questionnaire, one week of light messaging turned the ballot test to 39/51, combining a -7 drop for Landrieu with a +17 gain for Cassidy, yielding an overall net -24 point difference in Landrieu’s standing on the ballot.
Among Republicans, the ballot test moved from 82/8 for Cassidy in the control group to 91/2 for Cassidy in the test group, reflecting a significant +15 point hardening of his support among Republicans. Among Democrats, the ballot moved from 68/11 for Landrieu in the control group to 63/26 for Landrieu in the test group, reflecting an even more significant -20 point shift away from Landrieu in her own base. And among Independents? The biggest move of all – from 29/23 for Cassidy in the control group to 57/30 for Cassidy in the test group, a +21 point increase for Cassidy among independents. In other words, whereas roughly half the independents in the control group were undecided, only 13 percent were undecided in the test group – and the vast majority who chose a side broke to Cassidy.
The strength of the message is further seen in noting how quickly and notably it affects sentiment when people learn of it. (Which is why Sen. Reid has tried so hard to obfuscate what Sen. Vitter is trying to do by attempting to block the exemption.) The format of the poll asked the initial ballot test question, then tested how well the issue of and facts about the congressional exemption were known, then assessed again, “if it were true”, what that would do to people’s ballot test. In the group that had already received our messaging, there was essentially no difference in their before and after ballot test, indicating they had already fully absorbed and formed an opinion about the congressional exemption; the information contained in the survey questionnaire was not new to them. But in the control group, where the information was new, the messages contained in the survey itself moved the ballot from 46/34 for Landrieu to 34/29 for Landrieu, from +12 to just +5, a -7 point drop.
(You’ll note that Cassidy’s score dropped on the ballot, too. What happened? About a quarter of Landrieu’s supporters and a quarter of Cassidy’s supporters responded by moving to the “Undecided” column – most likely because of the anti-incumbent information contained in the questions themselves, which raised doubts about the integrity of all incumbent politicians, and drove disgusted supporters of each of them into the “Undecided” column in equal proportions.)
The bottom line? Incumbents ought to be very, very scared of some of their colleagues imagining this will blow over.