Main map

Presidential polls: (None)
Dem pickups: (None)
GOP pickups: IN NC

Previous | Next


Downloadable data

News from the Votemaster

Romney Concedes Florida

Even before all the votes have been counted, Mitt Romney has conceded that he lost Florida. Obama is ahead by over 55,000 votes but nearly all the remaining absentee ballots to be counted come from Miami-Dade, Broward, and Palm Beach counties in Southern Florida, all of which are very heavily Democratic. These counties have nearly 100,000 ballots yet to be counted and Obama could easily pick up another 50,000 net votes in them. The provisional ballots must also be counted, but history shows that most provisional ballots are from Democrats who voted in the wrong precinct.

How Did Each Pollster Do?

The problem with polling is that there comes a moment of truth when everyone can see how good you are at your craft. That moment is now. We have done the following analysis. For each of the ten swing states, Colorado, Florida, Iowa, North Carolina, New Hampshire, Nevada, Ohio, Pennsylvania, Virginia, and Wisconsin, we took the final poll of each major nonpartisan pollster and called that its prediction. "Major pollster" was arbitrarily defined as one that had at least 10 published polls in 2012. Not every pollster polled every state, though. Here are their predictions.

ARG Romney Romney Romney Romney Tie Obama Obama   Romney  
IPSOS Tie Tie         Obama   Obama  
Marist College Tie Obama Obama Obama Obama Obama Obama   Obama Obama
Marquette Law School                   Obama
Mason-Dixon   Romney             Obama Obama
Opinion Research Obama Romney   Romney   Obama Obama     Obama
Pharos Research   Tie         Obama Obama    
PPP Obama Obama Obama Tie Obama Obama Obama Obama Obama Obama
Pulse Opinion Res.   Romney         Obama Obama Obama Obama
Purple Strategies Obama Romney   Obama     Obama   Tie  
Quinnipiac Univ. Romney Obama         Obama Obama Obama Obama
Rasmussen Romney Romney Romney Romney Obama Obama Tie Obama Romney Tie
Siena College               Obama    
Suffolk Univ.   Obama     Tie Obama Tie   Obama  
SurveyUSA Obama Tie   Romney   Obama Obama      
Univ, of New Hampshire         Obama          
Zogby   Romney         Obama   Obama Obama

The color coding is as follows. Light green is a correct prediction and gray is an incorrect prediction. Everybody who tried got Nevada right That was easy. Nobody thought Romney would win Ohio, although Rasmussen and Suffolk University rated it as a tie. Florida was tough, most pollsters got it wrong, but it was very close.

Now comes the moment of truth: how well did everybody do? Only those states where an actual prediction was made count, so ARG, for example, predicted 7 states and got 3 right and 4 wrong. The predicted tie for New Hampshire doesn't count, so the accuracy here is 3/7 or 43%. Here are the results.

Pollster Score States polled
PPP 100% 9
SurveyUSA 100% 4
Suffolk Univ. 100% 3
Pharos Research 100% 2
IPSOS 100% 2
Univ of New Hampshire 100% 1
Siena College 100% 1
Marquette Law School 100% 1
Marist College 88% 8
Quinnipiac Univ. 83% 6
Opinion Research 83% 6
Pulse Opinion Res. 80% 5
Zogby 75% 4
Mason-Dixon 67% 3
Rasmussen 50% 8
Purple Strategies 50% 4
ARG 43% 7

There are several things to note. To get in the list, a pollster had to run at least 10 polls this year, but they might well have all been in the same state. The number in the third column above is the number of states polled, not the total number of polls. Siena College, for example, is located in New York state and mostly polled New York, with one poll each of Connecticut and Pennsylvania. It did all right, but it is more of a regional pollster. Of the more adventuresome pollsters, with 5 or more states polled, the gold medal goes to PPP. Marist College gets the silver, and Quinnipiac University gets the bronze.

Many people have suspected Rasmussen of having his thumb on the scale for the Republicans. Scott Rasmussen appears on Fox News often, where he is called an "independent pollster." Now the question is: "How good is he?" Some people thought he was biasing the numbers toward the Republicans during the year and as we got closer to the election, would remove the bias so his final polls would be very accurate and everyone would say how great he was. It turns out not to be the case. He got only 50% right. You could flip a coin and get half of them right.

Most likely, Rasmussen has a flawed methodology but is reporting his numbers honestly. After all, he is an independent businessman and having a poor track record isn't good for future business. Two criticisms of Rasmussen stand out. First, all his polls are done in 4 hours--a big selling point to customers. The downside is that if he calls and no one is home, he just picks a different random number and calls it. The consequence of this approach is that he tends to miss the 18-29 crowd, who are out a lot and oversamples seniors, who tend to be home a lot. Since young people are largely Democrats and old people are largely Republicans, a bias is introduced here. Of course, he knows this and tries to correct for it, but it is better to call the original number 10 times if need be. The second criticism is that he doesn't call cell phones. Robopollsters are not allowed to. This also introduces a bias towards the Republicans. Nevertheless, PPP, SurveyUSA, and some of the other robopollsters manage to correct for it. Apparently it can be done.

Technically, if a pollster predicts that Romney will win a state by 1% with a 3% margin of error and Obama wins by 1%, the pollster got it right. Nevertheless, many people have more faith in pollsters that name the winners correctly. A more technical analysis can only be done when the final vote totals are in, which will happen when all the absentee and provisional ballots are counted.

Long Lines Mar the Voting Process

Many people had to wait in line for 3, 4, even 7 hours to vote. That is simply disgraceful. America is the richest and most technologically advanced country in the history of the world. It is not Iraq. In some cases, the problem was intentionally caused by state legislatures that wanted to make it hard to vote. In other cases it was mere incompetence. The Washington Post has a story about how to reduce the lines. Here are the main suggestions.

Election-law specialist Rick Hasen has said that when it comes to federal elections, Congress has broad powers to overrule the states. It could also declare election day a national holiday or move elections from a Tuesday to two days, the first Saturday and Sunday in November.

What Can the Republicans Do To Recover?

Pundits of all stripes are starting to announce how to fix what ails the Republican Party. Here is a small sample of their advice.

Welcome to the 2016 Campaign

Political junkies who are despondent that the campaign is over can cheer up. Sen. Marco Rubio (R-FL), is heading out to Iowa to start his 2016 campaign for the Republican nomination. Rubio is well aware that Romney did miserably with Latinos, and hopes that Mexican Americans will vote for a Cuban American. Maybe that will happen, but he first has to win the Republican nomination in primaries ruled by angry, old, white men and not many Latinos.

Getting Latinos to switch parties won't be so easy if President Obama has his way. It is likely he will make immigration reform a key item on his second term agenda. He could easily revive George W. Bush's immigration reform plan, coauthored by Sen. John McCain (R-AZ), and barnstorm the country trying to pass it. If he succeeds, the Democrats will get the credit. If Senate Republicans filibuster it or the House votes it down, Republicans will get the blame. There is little Rubio can do to change this if Obama decides this is a good way to lock Latinos into the Democratic Party for a generation.

Gerrymandering Gives Republicans the Edge in the House

Although not all House races have been settled yet, it is likely that the number of Republicans in the House will be at least 40 more than the number of Democrats. Yet about as many people voted for a Democrat for Congress as voted for a Republican. How can this be? The answer in a word is "gerrymandering." The 2010 elections put the Republicans in charge of redistricting in many states and they did an excellent job of drawing the district lines to favor themselves so that even in the event of an evenly split vote they would come out ahead. The only way to make the process fair is to have a nonpartisan commission draw the district lines as happens in California and some other states. But, of course, states where one party has all the power are not likely to do that unless the voters mandate it via a referendum, which is what happened in California.