Forecast: the race between Biden and Trump is within the true margin of error


But the election is not held today. It will take place in just under four months. That’s a lot of time.

And given the size of Biden’s lead (sure, but not a blowout), the race can definitely change enough to characterize this race to be within the true margin of error.

To be clear, I don’t mean that Biden and Trump have similar chances of winning, far from it. Biden is clearly the favorite.

What I mean is that people often don’t realize what the odds really mean. It is possible that something has no chance of happening, but it is quite conceivable that it will happen.

Every time you see a published survey, you will see that the result comes with a margin of error. The margin of error essentially means that for a given population, 95% of the time the result of a survey will be within the margin of error of the true value. There is a 2.5% probability (1 out of 40) that a result falls outside the margin of error on the low side, and a 2.5% probability (1 of 40) that a result falls outside the margin of error on the low side. tall.

It is not surprising to receive a survey result that is outside the margin of error. It is even less surprising when a survey falls within the margin of error of the other results, albeit at the upper or lower end of the published results.

Borrowing from this concept of margin of error, there is literally no reliable forecast known to suggest that Biden’s advantage is outside that 95% confidence interval when projected toward November.

You can see how a career can change from this moment on by looking at history.

Earlier this month, I examined the 13 elections with presidents in office since 1940. I observed that in two of the 13 presidential contests, the difference between where the national polls were at the moment and the eventual margin was greater than 10 points.

That is more than the current margin between Biden and Trump. Two of 13 times is more than 2.5%. One of the 13 times is much more than that.

If you were to expand to headline-less contests, you’ll find even more examples of big changes. I wrote about the 1988 elections a few weeks ago. The difference between the polls at this time and the election result (Republican George HW Bush for 8 points) and the polls in mid-July (Democrat Michael Dukakis for 5 points) was double digits.

This landmark study doesn’t even take into account that Trump probably has a better chance of winning the Electoral College than the popular vote.

Importantly, you can see many different forecasts and come to conclusions similar to mine.

The betting markets, which I think are overestimating Trump’s chances, give him a 2 out of 5 (40%) chance of winning a second term. (It should be noted that the betting markets gave Bob Dole a similar chance to win in 1996, when he faced a deficit similar to Trump’s now. Dole lost.)
Jack Kersting’s forecast, which I have quoted before, puts Trump’s chance of winning around 1 in 6 (just north of 15%).
There’s also The Economist’s forecast, which gives Trump a 1-in-10 (about 10%) chance of winning.

There are clearly differences between these probabilities. The similarity between them all is that Biden is ahead, though not enough to have an advantage that can be seen anywhere near being outside the true margin of error.

The worst result for Trump is having a probability of 1 in 10. His shot would need to be below 1 in 40 to be considered outside the true margin of error.

In fact, we still have many events and possible game changes ahead of us.

There are conventions, which have sometimes (though not recently) really shaken things up historically.
We have an economy that appears to be more prone to shocks than usual, which could change the opinions of voters.

Also, the coronavirus pandemic is bad right now, but there is no real way to know how things will be when voters cast their ballots. Perhaps the case rate is lower. Maybe there will be better treatments. We just don’t know.

Speaking of the coronavirus, it appears that this campaign is not even really underway. Usually, a presidential campaign is the main news right now. But with the pandemic and protests against police brutality, it has been story number three in the past month.

The campaign will eventually become story number one. With a compressed time frame, we may see bigger jumps later in the campaign than we are used to in modern campaigns.

The bottom line is, unless the economy totally collapses or Biden begins to lead closer to 20 points than 10 points, this election will never come close to being safe in his corner at this early juncture.

.