Polls for the U.S.A. 2020 elections

  • Oct 24,2020

The team which stays behind the poll for the 2020 elections, created their own working model, based on many factors: polls, economic conditions, presidential approval ratings, political polarization, the presence of an incumbent. 

The only thing that they didn’t get factored in was the probability of a meteor strike, this says Andrew Gelman, Professor of Statistics and political science of Columbia University. The meteor strike is not a likelihood event, so it’s pretty slim to disrupting the electoral landscape between July and November. But for example, only a few people would have guessed that this election campaign was going to be hit by a global pandemic. 

The Former Vice President Joe Biden is leading at the polls nationally. So for now is more likely that he will be the next American president. After the upset victory of Donald Trump in 2016, a lot of professional pollsters are spending time thinking if something will change at the last minute. 

Four years ago, a lot of mistakes and political circumstances led a lot of analysts to predict that the Hilary Clintons Presidency was only Trump to win. This mistake came after an election cycle that seemed unique for that time. Trump as a political outsider and host of the former reality show and Clinton as the first female nominee of a major political party. 

Gelman says that his model which currently gives Biden 9 of 10 probability to win. People have said that 2020 is a special year, but when you look back at every election, for example, 1948 was the first election when Franklin Roosevelt hadn’t participate in a long time. 1952 was the year when Dwight Eisenhower was very unlikely to figure. 1960 John F. Kennedy – the first Catholic. Every American election has something special and unusual. 

With this view, the pandemic doesn’t look so unusual for the American elections. 

According to all of that, until the day of the elections, there is a lot of time for things to change. Courtney Kennedy is a director of survey research at Pew Research Center. She says that we can expect the coronavirus to effect on the polls and the forecast for the outcome of the election. 

Beyond the uncertainty which is added by the pandemic, polls still have the heard job to predict who is going to vote. According to experts, they said that a lot of the problems in 2016 were a result of state polls overrepresenting college graduates, who were officially pro-Clinton. 

These days if the virus makes a gap between people who are likely to vote and which actually vote it cloud have the accuracy of the poll. Nathaniel Rakich, an election analyst for prominent forecasting outlet, says that the virus may disrupt the election process in other ways as missing bulletins, or long queues. 

The polling could also be complicated because of the economic impact of the pandemic, If the people lose their homes and they need to move that can involve re-registering to vote. 

On the other hand, the pandemic yielded benefits for polls, because more people agree to participate. More people are at home when the phone rings. The polls are enjoying a bump inaccuracy as the pandemic conditions make by mail increased. 

This year when Trump is not so famous commodity the Americans had a reason to weather another one like him is going to occur. But the early voters could be different from the election-day voters. 

Franklin says that he doesn’t think America has such good empirical tests about how much the early voting matters. Because this is the first time the situation is with such a big increase. Pollsters will watch if the one who already voted is different than the others who haven’t make up their minds or haven’t yet vote in-person.

The last question is if the mistakes in 2016 have permanently shifted toward less accurate polling. 

In 2018 the polling has been very accurate. All the polls which are made carefully could give good results.