Florida was a breath of fresh air for those of us who spent days pouring over past races and deciding on a likely turnout model.
The Florida GOP primary is closed, with only registered Republicans are allowed to vote, so there is much less guessing on who will wind up in the voting booth. The one wrinkle in this year’s contest were the high numbers of early voters, which was about a third of the total turnout and must be factored in to ensure the most accurate numbers possible.
Looking at the polls released in the few days before the election, there was little disagreement that Mitt Romney would be the eventual winner. Every public poll since Jan. 25 has had Romney leading, and the last few days this turned his lead into a comfortable margin in most surveys.
Indeed, in the last three days Romney enjoyed a lead ranging from 20 points in the final Suffolk poll to 8 points in the Public Policy Polling numbers. Looking at these two polls, Suffolk was just about dead on when it came to their call on Romney’s support (47 percent in the poll versus 46 percent on Election Day) but was short on Gingrich’s total (27 percent in the poll versus 32 percent on Election Day).
For PPP, it was the reverse. The poll was right on the mark with Gingrich’s number (31 percent in the poll), but short on Romney (39 percent in the poll). Interestingly, the public polls continue to have some big issues with the third and fourth place finishers.
Just about every poll had Ron Paul and Rick Santorum tied or very close. With the final results in, Santorum almost doubled up Paul (13 percent to 7 percent). Since Florida is a winner-take-all state in terms of delegates and both Paul and Santorum skipped the state, the differential will be generally ignored. However, as this trend continues it speaks to a problem with the sampling that is being used in public polling.
With the exception of a few firms (PPP and Suffolk to name two) most public polls do not release their data or share their demographics so it becomes next to impossible to figure out exactly why Quinnipiac, for instance, had Paul running in third place two points ahead of Santorum. Did they over sample younger voters? Did they under sample evangelicals? Unfortunately, we’ll just have to guess.
With Gingrich and Santorum announcing that they will continue on for the foreseeable future, coupled with the fact that there are four caucuses and a primary followed by almost a month off before Super Tuesday, missing the boat on Santorum and Paul could have a large effect as delegates become divvied out on a proportional basis.
Granted, most people tend to focus on the winner and in the case of Florida, the polling got it right. But when you’re looking for quality from top to bottom, most of the polls came up a bit short.
Stefan Hankin is founder and president of Lincoln Park Strategies, a Washington D.C.-based public opinion firm. Follow him on Twitter at @LPStrategies.