Is There Still Hope for Trump?

• Published on July 6, 2020 on Linkedin

Latest polls show that Biden is leading between 4-12 points in terms of winning chances. (not the popular vote, the difference there is bigger, Hillary Clinton even won the popular vote, but no use). Odds on betting are implying a similar picture: 2.64:1 for Trump and 1.62:1 for Biden. If the 2016 memories were not fresh, I ‘d say "it is over for Trump now. This is not something he can turnaround." (possibly he won’t, either)

Therefore I checked how the polls were in Hillary Clinton’s favor. Was the difference similar, or bigger? How did the trend change? Most importantly, what was the reason that polls missed the results that far?

In 2016, last week of the elections, polls were implying that the Hillary Clinton was still leading but not as far as in that April, where the difference was over 10 points. Trump closed the gap end even in August there were polls saying that Trump is leading.

At this time of the year, 4 months towards elections, Clinton seemed still 5-6 points ahead. Since the difference is now bigger for Biden, I can say Trumps chances are relatively lower. Besides, in 2016 we saw different trends among months even weeks, either in favor of Clinton or Trump, whereas during 2020 the trend has been all the time pro Biden, with only small breaks.

When we look at excuses of pollsters, it is possible that we might face same problems also this year. The main argument was that undecided voters did not turn out to vote proportionally. That was a strong bias towards Trump among them. Another argument was the Shy-Trump effect. Some claim that Trump voters did not cry out loud pro Trump, especially in polls. I am sure there is ground for the theory but I am not sure the effect would be significant enough. Another argument is education level of those participated in polls vs. reality. Although pollsters make related adjustments, we see that higher educated people tend more to participate in polls, and they also tend more to vote for Democrats. Especially the better predictions in national polls compared to state polls are attributed to this education factor. Experts say adjustments in national polls were done better than the state polls.

The question is, did pollsters take lessons from 2016 ? Possibly yes, but is it enough the predict the results this time better? We’ll see. There are many factors and we’ll have live debates that will effect people’s views, voting decisions etc. So this brief post is not about analysing any electoral outcomes, rather just a remainder how was it in 2016.

Here is a detailed analysis of 2016 polls.