I’m about to do something you rarely see a pollster (or someone who commissioned a poll) do, especially before an election.
I am going to attempt to walk back the results of the two polls we commissioned, specifically in the races for Senate District 34 (Republican Ellyn Bogdanoff vs. Democrat incumbent Maria Sachs) and House District 69 (Republican incumbent Kathleen Peters and Democrat Scott Orsini.
This is not to say we wrong or even that we goofed. It’s just that, upon taking a second survey of those districts with a tighter screen on who is likely to vote this November, the results are substantially different.
I did not want to present these polls as “new polls” because then it may be construed that one candidate had closed a gap or another candidate had extended their lead. These polls are more accurate than the first surveys and will be considered the benchmark as we continue to track these races.
In the first round of polling, our samples mirrored an enhanced measure of the voter registration numbers in the district. This second round of polling uses a sample of voters that have a history of voting in both 2010 and 2012. Only those voters who said they were likely to vote in the upcoming election were included.
There is increasing evidence, based primarily on national trends in midterm elections which have already occurred, that suggests there will be record-low turnout this fall.
Specifically, of the nearly 123 million voters who were eligible to cast ballots in 25 statewide primaries for governor and U.S. Senate, only 18 million did so, reports Mark Barabak of the Los Angeles Times.
That’s a 14.8% turnout.
If precedent holds, a study from the Center for the Study of the American Electorate says the nation will record the lowest midterm primary turnout in its history; already more than half the states holding primary elections this year have had record-low turnouts, including California.
With this in mind — and at the suggestion of some of the “brightest minds in Florida politics” with whom I rely on for guidance and advice — we re-ran the polls in SD 34 and HD 69. Like I said, the results are significantly different.
Previously in SD 34, we had Maria Sachs with a 14-point lead over Ellyn Bogdanoff. The second-look poll, which still has a heavier Democratic sample, pegs the race at Sachs 48 percent, Bogdanoff 45 percent.
It’s important to note that these numbers are more consistent with some of the internal polling to which I’ve been made privy.
Previously in HD 69, we had Kathleen Peters and Scott Orsini in a virtual tie. And while those numbers still hold true in a poll of all voters, once a screen for participation in the 2010 midterm election is added, Peters lead increases significantly, as she leads Orsini 50 percent to 41 percent.
It’s too late now to go back and put the first set of poll numbers back in the bottle. Nor do I really want to, because conducting those polls prompted us to do more research than normal about the state of the electorate. But now that we have new information, we knew we had to share it as soon as possible with our readers.