Polls could again underestimate voting intention for Donald Trump this year

Polls could again underestimate voting intention for Donald Trump this year

There is one week left until the United States presidential elections, on Tuesday, November 5, Americans will go to the polls to choose between the alternative of the Democratic Party and the Republican.

The latest polls, conducted by Reuters and Ipsos, showed that Vice President Kamala Harris has a marginal advantage of 46% over the 43% intention to vote for former President Donald Trump.But if the scenario of the previous presidential elections is repeated, in which the vote for Trump was underestimated by national polls, he would actually be the one who would have the advantage and the greatest chance of winning.

LR Chart

When Trump faced Joe Biden to govern in the current presidential term, the current president achieved 306 votes in the Electoral College out of the minimum 270 to retain the presidency. so that he obtained 51.38% of the votes. While Trump reached 232 votes (46.91%), although he did not win, the average of the pollsters prior to the vote indicated that Biden was going to keep 52% ​​of the votes but Trump was only going to reach 42%, which which reflected that support for Biden was exaggerated by 3.9 percentage points and support for Trump was underestimated by five percentage points.

With this precedent, the current doubt among analysts is whether the results that the polls are showing, in favor of Harris in several cases such as that of Bloomberg News Morning Consult, They are true or there is an under-registration in the votes that will reach Trump in the official race. If so, the leadership that Harris apparently has would be threatened by having a lower advantage than Biden had four years ago over the same opponent.

In the previous vote, it was found that pollsters failed to underestimate the proportion of white and older voters who finally cast their vote and support for Biden in urban areas was overestimated compared to the actual results, so it identified the problem and Several biases were eliminated when choosing the sample, but Courtney Kennedy, supervisor of survey design and data science at the Pew Research Center told The Wall Street Journal, “I wish I could say that the polling industry has done things to give us great confidence that polls will not systematically underestimate Trump’s support again, but that is not the case. Pollsters have made a great effort to correct the error, but there is no miracle solution.”

Manuel Camilo González, professor of international relations at the Javeriana University, pointed out that even if the polls indicate a greater intention to vote among Americans for a candidatethere will always be the risk that whoever wins the popular vote will not become the president of the United States by failing to achieve the necessary 270 delegates in the Electoral College.

In the history of the country it has happened five times, three times in the 19th century and twice in the 21st century, one of them in which Donald Trump surpassed Hillary Clinton in the electoral college but not in the popular vote; This is a scenario that can also occur this year, where a candidate wins by popular but not electoral votes.

“The pace that the Trump and Harris campaigns have adopted lately in the key states and with the most delegates is due to the need to ensure a sufficient number of voters in each state. according to the number of population. There is a marked inequality in the weight of the vote from one state to another,” he added.