2020 Election Prediction

by Jonathan Hobratsch

I present my prediction for the 2020 Presidential Election. In short, I predict a victory for former vice president Joe Biden against the incumbent president Donald Trump.

A victory by Biden will make Trump the 10th defeated incumbent president in US history and the first since 1992 when Bill Clinton defeated Pres. George H. W. Bush.

A clear Trump defeat will likely look like a rebuke of Trumpism on the map itself. There is a chance the entire Southern border–where Trump has been frantically building a wall–votes against Trump, although Texas flipping probably won’t happen. Nevertheless, the Southern border is less supportive of Trump than they were in 2016. Additionally, the same can be said of the Rust Belt–Trump’s focus area in 2016. Democrats are poised to flip back three of the states won by Trump in 2016. Finally, Biden is likely to win back many of the Obama-to-Trump counties. Considering former president Barack Obama has been a special target for President Trump, losing these counties will be especially insulting for the incumbent president.

Why do I feel confident in a Biden victory? I see the following reasons for confidence:

The average polling margins favor Biden in most battleground states. Unlike in 2016, there is no clear poll tightening in the states or nationally. Biden is holding noticeable leads. We have many more polls than we had in 2016. Overall, we have more data. That is, more evidence.
Biden’s margins over Trump, nationally and in the states, are noticeably larger than Clinton’s margins over Trump in 2016. In many cases, the polls will have to be off twice as much as they were in 2016 for Trump to win–very unlikely. Besides, who knows if polls being off would necessarily favor Trump even if they were off? Nate Silver has a recent post showing Biden with 320+ electoral votes, even if the polls were off as much as they were in 2016.
Biden’s favorability is about +20 higher than Trump’s, according to Real Clear Politics. This is more than double Clinton’s favorability over Trump in 2016, according to Real Clear Politics.
Biden is polling higher support among women (the largest demographic), white men with no college education (Trump’s main demographic), and seniors (another Trump demographic) than Clinton did in 2016.
Third Party voting is likely going to hurt Trump more than Biden. The Green Party is nearly invisible in this election. The Libertarian Party, which is less visible in this election than in 2016, has a more conservative ticket than it did in 2016. It might be more appealing for conservatives that voted Trump, don’t like Biden, but now no longer want to vote Trump. Lastly, Kanye West’s write-in campaign probably won’t do too much, but he’s likely going to take chaos voters that voted Trump in 2016, rather than hurt Biden’s support.
All data that I’ve seen so far points to undecided voters positioning to vote for Biden over Trump. The Atlantic, New York Times, and the Washington Post all have recent articles arguing that these voters won’t swing to Trump in the last moment or won’t be as numerous as they were in 2016. Much of this has to do with favorability and the fact that voters now know how Trump operates as president, which wasn’t the case before. In 2016, voters had a choice between two negative favorability candidates. They had an idea of what a Clinton presidency would be. Trump was a mystery, so they went with him. These voters are now faced with Trump’s record vs. Biden’s likability and empathy.
Finally, early voting polls, high youth turn out, generic party polling, and high turnout in general, all points to voters opting Democrats in 2020. We know that early voting polls are heavily favoring Biden. Voting on election day would have to be extremely high for Trump, and I don’t see that if the polls aren’t tightening.

Not all signs favor Biden, however.

The Direction of the Country poll has been trending Trump’s direction (about -31 instead of -45 as was the case a month or so ago). I’ve heard some pundits/pollsters consider this poll to be the most important poll. However, it’s hard to know if the poll is trending this way because people feel optimistic that Trump might soon be out of office.
The recent economic report showing big gains in the GDP might favor Trump. However, those gains would have been made if they GDP hadn’t fallen as low as they had. Gains are also not up to pre-Covid levels.
Trump’s average approval and disapproval polls are trending in his favor. However, he still has a 44% approval and a 53% disapproval. I’m not sure how this can be a positive.
Latinos and, some argue, black voters as well support Trump more in 2020 than they did in 2016. However, I’ve seen data going both ways in regards to black voters– 90% who are polling for Biden. Some of this has to do with white voter overrepresentation in early voting. I think evidence for Latinos supporting Trump more than they had in 2016 does exist, however. Nevertheless, Biden will have more Latino voters than Biden will.
North Carolina, Georgia, Ohio, Florida, Texas, are all lean Conservative states that have used voter suppression in the recent past–generally aimed at black voters. There’s no reason to believe that votes won’t be suppressed in these states, especially considering how close the contests will be.
Some polls and pollsters — all conservative bias — are showing Trump winning states like Pennsylvania (a must win for Biden), Michigan, and Wisconsin. Avg poll margins aren’t tightening as in 2016. These pollster generally have a low pollster grade, according to FiveThirtyEight. It’s hard to really justify their results other than their being ideological wishful thinking.

I’ve created an algorithm weighing the Pro-Biden and Pro-Trump factors above in an attempt to predict the 2020 election. For average polling, if used Fivethirtyeight’s average polling as the samples are not as restrictive as they are at realclearpolitics.

All said, evidence is clearly stronger that the polls will be more accurate than last time, which means Biden will win. As stated, even if the polls are as wrong as last time, Biden still wins. Barring flagrant cheating, Biden will win.

See my prediction map below. The different colors stand for safe support, likely support, leans support, and tilt support. A state with a “safe” darker blue or red will probably see a 10+ margin for the winner. A state with a “likely” medium blue or red will see a +6 to +9.9 margin. A state with a “leans” light blue or red will see a +2 to +5.9 margin. A state with a “tilt” gray or brown color will see a +0.1 to +1.9 margin.

How confident am I in this map? Let me answer this by saying, if a state is at “Leans” (the light red or blue), I’m about 75% sure that will be accurate on election day. If it is a “tilt” state, then I’m only about 60% sure I’ll be accurate. I should note that Arizona, while leans on my map, came out as a +2.1 margin for Biden, which means it is almost a title. I support I’m about 67.5% confident in that state.