Florida polling shows a small shift toward the Democrat. Most midterm voters have already made up their minds. And Dana Milbank is more certain than he should be about the problems with election forecasts. This is HuffPollster for Friday, October 10, 2014.
FLORIDA GOVERNOR: CRIST GAINS - Mark Caputo: "A University of North Florida poll of statewide likely voters shows Charlie Crist leading Gov. Rick Scott 43-38 percent -- marking the fourth survey this week that has the Democrat pulling slightly ahead. Crist's advantage (as with polls from 0ptimus, Public Policy Polling and SurveyUSA) is within the poll's error margin. So the race could be called a tie. But Crist's lead is almost outside that margin. And, as stated before: it ain't the topline, it's the trend. The trend is with Crist right now." The HuffPost Pollster tracking model, based on all public polls, shows a similar result. It shows a Scott advantage of just over two percentage points during the summer narrowing to dead even in recent weeks. [Tampa Bay Times, Pollster chart ]
MOST MIDTERM VOTERS HAVE MADE UP THEIR MINDS - Katherine Grace Carman and Michael Pollard: "For the next five weeks, RAND will be surveying members of the American Life Panel (ALP) leading up to the midterm elections in November….What sets apart the RAND surveys from other public opinion surveys is that we survey the same people each week. This allows us to observe changes in voting intentions over time as we draw closer to election time. Furthermore, many of these ALP members were participating in the panel during the 2012 presidential election. This allows us to link their responses today to responses during that time period and observe changes over a longer time period….most respondents are strongly leaning towards a particular candidate in both 2012 (74 percent) and 2014 (70 percent). But, those in the middle are not the same from year to year: For example, 15.8 percent are not strongly supporting any particular candidate this year but were strongly supporting one party in 2012….Comparing intentions in 2014 for the House and Senate Elections, we see much more agreement in who is in the middle." [RAND]
Democrats face attrition - Nate Cohn: "Democrats are struggling to maintain the support and enthusiasm of people who voted for President Obama, according to data from the RAND American Life Panel….Just 79 percent of Obama 2012 voters say they support a Democratic candidate for the House in 2014, compared with 82 percent of Mitt Romney voters who say they’ll support Republicans. A surprising 15 percent of Obama voters say they support Republicans, compared with 11 percent of Romney voters who prefer Democrats…..Obama's 2012 voters are also less likely to turn out this year….The loss of Democratic support and turnout might seem modest, but it is enough to erase President Obama’s 3.9-point win in 2012." [NYT]
A different story in battleground states? - Prof. Michael McDonald: "Imputing local voter enthusiasm from national surveys is silly; misses effect of local competitive races, e.g., Iowa will have high turnout...Pssst...don't tell Gallup and RAND that over 100K people have already voted in Iowa, almost 10% of the 2010 total vote." [@ElectProject here and here]
MILBANK: THE 'FOLLY' OF 'MATHEMATICAL CERTAINTY' - On Wednesday, the Washington Post's Dana Milbank's column took on election forecasting models: "In the old days, life had many hardships. Among these: the need to wait until Election Day to determine who had won. But now Big Data has saved us from this struggle. Even close races can now be predicted with mathematical precision. We know, for example, with 98 percent certainty that Sen. Kay Hagan, an embattled Democrat, will win reelection in North Carolina next month. We are even more certain — 99 percent — that Sen. Mitch McConnell, a vulnerable Republican, will keep his seat in Kentucky. And we are darn near sure — 91 percent, to be specific — that Sen. Mary Landrieu (D-La.) will lose. Throw all of these into our election model, add eye of newt and toe of frog, stir counterclockwise and — voila! — we can project with 84 percent confidence that Republicans will control the Senate next year. The above data are from The Post’s Election Lab, run by George Washington University professor John Sides, but his is just one of several election models that claim to predict results with finely tuned accuracy. As of Tuesday afternoon, Nate Silver’s FiveThirtyEight, which turned the academic discipline of computer models into a media game, gives Republicans a 57.6 percent chance of taking the Senate. (Decimal points are particularly compelling.) The New York Times’s model goes with 61 percent, DailyKos 66 percent, Huffington Post 54 percent and PredictWise 73 percent. The Princeton Election Consortium gives a 54 percent advantage to Democrats . Apparently they forgot to add the toe of frog." [WashPost]
The handicappers agree - Milbank: "Compare that to Stu Rothenberg, an old-school political forecaster supposed to have been made anachronistic by Big Data...Rothenberg told me the models 'hype' and are 'intellectually deceptive' because they 'convey a sense of mathematical certainty that is simply misleading.' Rothenberg uses polling and other data, but he also interviews candidates and is 'humble enough to know that every election cycle is different.' Charlie Cook, another old-school forecaster, thinks the Big Data models have a place, but they put too much science in politics. 'They err,' he told me, in assuming 'a precision and confidence that doesn’t exist in human behavior.'"
They have it backwards - The point of the probabilities is not about attaching "mathematical certainty to human election behavior," as Milbank put it on Twitter. It's about trying to quantify the uncertainty -- the potential for error -- in all the polling and other data used to forecast an election. In a column that coincidentally ran the same morning, Democratic pollster Mark Mellman explains: "How are people misreading [the Senate forecasts]? First, there is over-interpretation. When HuffPost Pollster says Republicans have a 51 percent chance of taking control of the Senate after this election, or fivethirtyeight.com says it has a 59 percent probability, many people interpret that as meaning, 'Republicans are going to take control of the Senate.' That is not at all what these forecasts say. Such comments turn a probabilistic forecast into a categorical one. It’s transforming a 51 percent or 59 percent chance into a 100 percent (or maybe a 98 percent) chance. If you bet it was going to rain every time there was a 51 percent chance of precipitation, you might come out ahead over time, but you would lose a lot of money...What the election forecasting models are really saying is that, in circumstances like these, we expect that 51 percent or 59 percent of the time, Republicans will win the Senate." [The Hill, @Milbank]
More reactions from Twitter:
-Kevin Collins (D): "If I could have everyone understand just one concept better, it might be quantified uncertainty" [@kwcollins]
-Nate Silver: Milbank's column produced this "actual newspaper headline: 'Election forecasts rely too much on science, data'... Contrast Mark Mellman's good column on Senate models...with Dana Milbank's crapola" [HeraldNet, @NateSilver538 here and here ]
-Drew Linzer: "If my model is wrong, I'll look at the data and tell you why. If Cook or Rothenberg are wrong, then what? Gut check and see you in 2016?" [@DrewLinzer]
HUFFPOLLSTER VIA EMAIL! - You can receive this daily update every weekday morning via email! Just click here, enter your email address, and and click "sign up." That's all there is to it (and you can unsubscribe anytime).
FRIDAY'S 'OUTLIERS' - Links to the best of news at the intersection of polling, politics and political data:
-David Uberti looks at how news media cover the polls -- and the outliers. [CJR]
-Nate Silver profiles the "Maroon 6," a set of crucial red-leaning states.