[ad_1]
In the aftermath of the party conventions and riots in Kenosha, Wisconsin, there has been some evidence that the US presidential race between Joe Biden and Donald Trump is beginning to harden in the major decisive states.
A Monmouth University of Pennsylvania poll released this week gave Biden a four-point lead among the state’s registered voters, down from 13 points in mid-July. Polls from other swing states, including Wisconsin and Arizona, have been better for the Democratic challenger.
While some state polls show a tighter race, almost none have Trump leading the way. But the betting markets have moved dramatically in the president’s favor since the conventions, now giving him a nearly 50-50 chance of winning re-election.
Voters are also skeptical. A recent Pew Research Center poll of American voters gave Biden an eight point lead, but found that only 82 percent of the former vice president’s supporters thought he would win. Trump supporters are much more confident: Nine out of 10 think he is on his way to reelection.
So why is there such a disconnect between national polls, where Mr. Biden enjoys a stable lead, and the perceptions of voters and those who bet on the prediction markets?
One explanation is that a national advantage does not equate to an electoral college victory. Hillary Clinton won the popular vote in 2016, but still lost the race. The same thing happened in 2000 when Al Gore lost to George W Bush.
So some pundits prefer to focus on statewide poll averages showing Biden has a narrower lead at this point in the race than Clinton did, especially in those states that helped propel Trump to the race. victory.
However, simply comparing the state poll numbers in 2020 with those in 2016 can be misleading and could suggest that the race is closer than it actually is, for two important reasons.
1. In states where it mattered most, 2016 polls underestimated support for Donald Trump. Since then, pollsters have tried to correct their mistakes.
National polls were mostly accurate in 2016, but state polls, especially on the battlefields, tended to underestimate support for Trump. Polls underestimated their support in Wisconsin by 6.5 percent, and in Pennsylvania and Michigan by four to five points.
One of the main problems, according to an autopsy by the American Association for Public Opinion Research (AAPOR), was that many state pollsters did not adjust their samples for education, meaning they had too many highly educated voters among their respondents.
This mattered less in the elections prior to 2016. In 2012, education was not as correlated with the choice of voters. However, in 2016 education was arguably the most important demographic factor in determining which candidate voters (especially whites) would back.
Many state pollsters have tried to rectify the problem by weighting their samples by education, along with more traditional demographics such as age, gender, and race.
To illustrate the difference this can make, the FT asked three state pollsters to remove the education weights from their samples and to recalculate their latest surveys. To varying degrees, all three would have estimated a higher advantage for Biden. More dramatically, Biden’s lead in a Florida voter poll conducted by the University of North Florida jumped from 6 to 11 points after the education weighting was removed.
2. There are fewer voters at stake than in 2016, and some polls suggest they favor Joe Biden.
The weighting of education alone does not explain the deficiencies of the surveys in the run-up to the last elections. The Marquette Law School of Wisconsin voter poll was one of the few that made a weight for education in 2016, but it was still out by six points.
Charles Franklin, director of the survey, attributed the error to “opinion dynamics at the end of 2016.” There were many undecided voters who did not like both candidates, but they opted for Trump in the last week before the election.
There isn’t much pollsters can do about the “late breakers,” who made up 13 percent of the 2016 electorate in Wisconsin, Florida and Pennsylvania, according to the AAPOR report.
A related theory is that there were many “shy” Trump voters in 2016 who were unwilling to admit their preference for pollsters. But the AAPOR report and subsequent investigation have found little evidence for this.
This time around, there is less chance that latecomers will disrupt the race. For example, the latest Marquette poll of Wisconsin has Biden at 50 percent and Trump at 46 percent among likely voters. That leaves only the 4 percent who say they are undecided or vote for a third party.
By comparison, in the latest Marquette poll of 2016, Clinton stood at 46 percent and Trump at 40 percent, leaving 14 percent undecided or supporting a third. So Biden’s lead is smaller than Clinton’s, but he’s arguably in a better position.
Second, there is some evidence that undecided voters are more likely to elect Biden. In Marquette’s August poll, 58 percent of Wisconsin voters who didn’t like both candidates said they would vote for Biden, according to Franklin.
That said, Election Day is still two months away and external factors could change the race.
Trump vs Biden: who is leading the 2020 election polls?
Use FT’s interactive calculator to see which states are most important to winning the presidency
Americans widely disapprove of Trump’s handling of the coronavirus pandemic. However, it is unclear to what extent this hurt Trump in November and the net effect of his law and order rhetoric. Voter turnout is another big unknown.
Finally, many of the problems that affected electoral polls in 2016 remain. Big-name pollsters in key states are grappling with lower budgets and higher costs. That means high-quality state surveys, which can cost anywhere from $ 10,000 to more than $ 20,000 per survey, are very rare.
“The structural challenges that the survey field had in 2016 are still in place in 2020,” said Courtney Kennedy, director of survey research at the Pew Research Center.
