In the run-up to the Nov. 3 U.S. election, several U.S. pollsters are showing Biden’s support leading Trump by a wide margin. However, the 2016 election still leaves many half-hearted about the polling results: with the last pollsters taking a collective beating, are the results still trustworthy this time?
Four years ago, with polls mostly showing Hillary with a bigger margin of victory, Trump had a surprise general election victory. That urged pollsters to reassess and adjust their methodology. At the same time, there are far fewer hesitant voters in this election than last time, which significantly reduces the uncertainty of the race.
This means that the polls may be more credible than in the last election, and Biden does have a greater chance of winning.
Adjusting the weight of opinion among low-education white voters
From the last loss, many pollsters have found that a major factor in skewing the results is that pollsters are ignoring an important demographic: the white population without a higher education.
While it’s unclear exactly why, it turns out that whites without a college degree, who are typically Trump supporters and also make up a large percentage of the American electorate, are more likely to reject polls. In the 2016 election, 44 percent of voters were white without a college degree, according to the Pew Research Center.
The opinions of this demographic were not well reflected in the results of the last poll.
So before this election, many pollsters adjusted their survey methodology. They adjusted the calibration data by weighting education levels in the survey to better reflect the electorate.
For example, when a poll collects data with a smaller percentage of adults without a college education than the agency expects for the size of the population, it automatically gives a higher weight to the responses of this group, allowing their opinions to be more accurately reflected in the survey results.
Fewer voters are hesitant to vote
In the 2016 election, many voters are on the fence when it comes to the two candidates, Hillary and Trump. As a result, polls in the final weeks before the election often show that more than 20% of voters are still undecided.
This adds a lot of uncertainty to the outcome of the general election. It also turns out that as more and more voters make up their minds, Hillary’s once large lead in early October almost disappears on the eve of the election.
And this year, there are clearly a lot fewer voters who are hesitant to decide. Reuters polling data shows that just under 7 percent of voters are still hesitant, almost half as many as four years ago.
As a result, many political analysts say they are more confident in this year’s polls because the American people have shown they are more certain of their choices in this election.
“This election is more reminiscent of 2012, when there was also an incumbent president in the race and a much higher level of voter decision,” Virginia Elections political analyst Kyle Kondik said.
“And it’s because fewer voters are hesitant this time around that Biden’s approval rating has been able to stay above 50 percent.” Kondik said, “In my view, a support advantage like Biden’s is more qualitative than Hillary’s because it shows that Biden has the majority.”
According to a Reuters national poll, Biden currently has 52 percent support nationally, compared to Trump’s 42 percent. This means that even if all of the current hesitant voters vote for Trump, Biden still has the support of a majority of voters. However, it should be noted that the U.S. elections implement the electoral college system, which means that Biden needs to win more electoral votes to win.
Investing more resources in key state polls
In the 2016 election, despite national polls showing higher support for Hillary and final results that did show Hillary getting more popular votes, Trump was ultimately defeated as he won more electoral college votes. This has led pollsters to devote more resources to state-by-state polling this year, especially in some of the key swing states.
In the Reuters survey, for example, they conducted 36 polls in six key battleground states between Sept. 7 and Nov. 3.
More polls don’t necessarily enhance accuracy, but they can help analysts more closely track changes in voter opinion over long periods of time, making it more likely that they’ll spot flaws in the survey data.
So far, the Reuters poll shows Biden and Trump neck-and-neck in Arizona, Florida and North Carolina, while Biden has a clearer lead in Wisconsin, Pennsylvania and Michigan.
Recent Comments