Roll Call: Latest News on Capitol Hill, Congress, Politics and Elections
February 7, 2016

Florida’s 2nd District: When You Don’t Need to Poll

On Thursday, Democrats released a poll showing Democrat Gwen Graham running neck-and-neck with GOP Rep. Steve Southerland II. They could have saved themselves a few thousand dollars because you don’t need a poll to show that Florida’s 2nd District will be competitive. But that also doesn’t mean that Southerland is on the verge of defeat.

As my colleague Stu Rothenberg wrote in a Roll Call column last October, the Panhandle seat is one of the most polarized districts in the entire House. Democrats just haven’t been able to get over the top in the past two election cycles.

“Voters in this district are incredibly polarized. It’s unlikely that 51 percent of the voters in this district would vote for any liberal Democrat, while close to 47 percent of district voters will always vote for the Democrat, no matter who he or she is,” Rothenberg wrote in the column last fall.

Indeed, in the last three presidential elections, the GOP nominee has received 52.4 percent, 51.9 percent and 51.5 percent. The Democratic nominee received 46.5 percent, 47 percent and 47.8 percent. That’s remarkable stability across three very different election years of 2012, 2008 and 2004.

Southerland knocked off Democratic Rep. Allen Boyd in 2010, and Democrats targeted the congressman for defeat in 2012. Three pre-election polls showed Southerland in a dead heat with former state Sen. Al Lawson. But that gave a false impression that the seat was on the cusp of turning over.

“So, I don’t have trouble accepting the Lester poll or other private Democratic surveys showing Lawson running even with Southerland in the mid-40s. But Lawson’s problem is that he is starting to bump up against his ceiling. He will need to get half of the votes cast in the district, and he can’t get there next month,” Rothenberg wrote at the time. “Given that, while some observers look at Lester’s poll and see a possible Lawson victory, all I see is a candidate getting his base vote — a vote that, because of the district’s makeup, will fall a few points short of what he needs.”

“Southerland is likely to win re-election with somewhere from 51 percent to 53 percent of the vote, and that assessment is based on the makeup of the district, which, I believe, augurs well for both a close outcome and a Southerland victory next month. And that’s why I have resisted moving the race to a more competitive category.”

Southerland defeated Lawson 53 percent to 47 percent.

Democratic strategists will argue that Graham, the daughter of a popular former U.S. senator, will outperform the African-American former state legislator. It’s certainly possible, but observers on all sides should be aware that we have seen the beginning of this movie before. It’s up to Democrats to prove this will have a different ending.

Once again, Florida’s 2nd District is rated Lean Republican by Rothenberg Political Report/Roll Call.

  • terjeanderson

    Nathan doesn’t seem to grasp that knowing who is ahead is far from the only reason that campaigns poll.

    While the top line numbers may be interesting (and what is released publicly to get some attention and convince donors to invest), a smart campaign is polling for many different reasons.

    At this point a campaign like Graham’s wants to know how various demographic groups view the race, who is undecided or persuadable, what are the issues and messages that resonate with different groups, and to use that information to plan campaign strategy.

    Spending a few thousand dollars to do that kind of research a year before the election is very smart spending by an candidate (Democratic or Republican) planning to mount a serious campaign. It is the kind of political intelligence needed to run an effective campaign that targets the right voters – even if that beneath the radar aspect of it escapes the attention of folks who only look narrowly at the horserace top line numbers.

    • nathanlgonzales

      terjeanderson- You don’t seem to grasp that this poll was conducted for an outside group, EMILY’s List, and not the Graham campaign.

      • terjeanderson

        So Emily’s List won’t share their data with the Graham campaign, or will find no value from the data in shaping their own approach to the campaign?

        Anyone who runs effective campaigns will tell you that (whoever conducted the poll) there is far more valuable information for campaigns coming out of public opinion research than simply the top line numbers that you focused on. They aren’t simply spending a few thousand dollars to know the race will be competitive- they are spending a few thousand to understand the dynamics of the race.

        But hey, thanks for the snarky comeback…. obviously I hit a nerve.

  • John Allegro

    For a brief introduction to the primitive superstitions of socialism, I generally refer people to Igor Shafarevich’s brilliant essay Socialism in Our Past and Future:

  • John Ramos

    Collectivists claim that particular forms of collectivism, like “democratic socialism” or “social democracy” will eliminate the problem, identified by Lenin, of who decides for whom.

  • georgecolombo

    This writer appears to know very little about Florida and its politics… or about local politics, in general, as far as I can tell. Without writing an entire column here, allow me to simple cite the Clelland/Dorworth race in 2012 (a race in which I was involved) to make two points: One, Republican voters are certainly willing to bail on a Republican candidate if they’re given persuasive reasons to do so; and, two, the beginning of Clelland’s momentum in the race was a poll just like the one cited here. The rest of the upset doesn’t happen until polling demonstrates that a win is in the realm of the possible so the idea that Democrats could have saved money by not polling is absurd.

Sign In

Forgot password?



Receive daily coverage of the people, politics and personality of Capitol Hill.

Subscription | Free Trial

Logging you in. One moment, please...