Sunday, November 1, 2020

Who Will Win?

Who will win?

This is the question everyone is asking right now.

I gave up making predictions eight years ago after I predicted that Mitt Romney would defeat Barack Obama in 2012.

I now confine myself to analyzing and assessing the data I look at without making a definitive prediction. That is probably especially prudent this year when there is so much going on that is unknown and without precedent (the effects of Covid, surge in early voting etc.) 

In any election analysis I start first by looking at history. That is especially true when one of the candidates was on the ballot last time. I look backwards to see what the votes looked like in the last election. Defeating an incumbent starts first with understanding how many votes they got the last time and where they were from. 

Defeating an incumbent begins with getting voters who supported the candidate the last time to change their vote. After that, you hope to recruit new voters to your candidate and cause.

Looking Backward

Donald Trump beat Hillary Clinton in 2016 by winning Pennsylvania, Michigan and Wisconsin by razor thin margins.  

These states provided Trump with 46 electoral votes to give him a 304-227 margin in the electoral college.

Had those 46 electoral votes been in Hillary's column Trump would have only had 258 electoral votes and would have been 12 short of the 270 necessary to win..

The total margin in those three states were as follows:

Pennsylvania     44,292

Wisconsin          22,748

Michigan           10, 704

Trump also won Florida with only a 1.2% vote margin and North Carolina with 2.8%.

When that is the baseline from the last election it is not difficult to make the case that Biden could easily turn the table on Trump.

Consider if you have a result in a state that was 51%-49% the last time. 51 out of 100 voters were for Trump. 49 out of 100 were for Biden. It would only take 2 of Trump's 51 voters to switch their vote to turn the state. 

Trump could do the same to Biden in some other states.

Trump only lost New Hampshire by .4%, Minnesota by 1.5%, Nevada by 2.4% and Maine by 2.9%.

Democrats have a big advantage going into any Presidential election in that they are certain to win large electoral rich states like California, New York and Illinois. Just these three states provide 104 electoral votes which is almost 40% of the 270 that is needed.

However, Democrats struggle for votes in most of the rest of the country.

Consider that Hillary Clinton was able to get more than 50% of the votes in only 14 states (including D.C) in 2016. Trump got more than 50% in 23 states.

This alone should show you why the Founders established the electoral college system. 

I like this chart which shows the percentage of vote for Hillary in 2016 by state. Many more states went big for Trump than they did for Hillary.


Trump got at least 55% of the vote in 18 states. Hillary only did that in 10 states.

Those 18 states gave Trump 112 electoral votes balancing out California, New York and Illinois.

There is little question you would like to be the Democrat in a Presidential race these days. Knowing you have three big states locked up before you start is a big advantage. 

Any Republican has a much narrower path to victory. However, in the past the Democrats have come up against the reality that most of the rest of the country is center right rather than liberal in its political views. That is why Democrats who have won recently (Bill Clinton, Barack Obama) did all they could when running for President to appear as centrists rather than liberals.

Has that changed? The Democrats have made a big bet this year that it has. Granted, they have tried to disguise their agenda by putting good old Joe up front. The big question is whether voters will see through it?

The Democrat establishment was clearly afraid of letting Bernie Sanders or Elizabeth Warren head the ticket after they saw what was happening in Iowa and New Hampshire. They closed ranks behind Biden in South Carolina despite Joe's woeful finishes in the first two primary contests in Iowa (4th) and New Hampshire (5th). 

In 2016, in my final blog post before election day, the following  states were all considered swing states and polls showed all of these within the margin of error. The odds were against Trump but I pointed out that he clearly had a path to victory within these states.

Trump won the first six states on this list. He also won Wisconsin which was considered totally out of reach. He lost New Hampshire by 2,700 votes, Nevada by 27,000 votes and Colorado by less than 5 points. 



North Carolina






New Hampshire

All of these states are still key battleground states in 2020 with the exception of Colorado. Georgia may be in play now when it was considered safe for Trump the last time. Wisconsin is definitely in play. Arizona is as well. Add Minnesota to this list.

These 12 states will determine the 2020 election.

Where We Are Now

You would like to be Joe Biden right now if you look at the polls. He is running ahead. Even better, he is looking better than Hillary did on the eve of the last election. 

According to the poll average, Trump is running 4.5 points worse today than he was against Hillary Clinton in nationwide popular vote polling four years ago. 

The President is doing better in the battleground states but he is still worse off by 2.5 points than he was when he beat Hillary Clinton in these states.

Those numbers suggest that Joe Biden will win.

However, it is difficult to not question a number of the polls when they are showing numbers like Biden +12 (CNN), Biden +11 (Economist/YouGov) and Biden +10 (NBC/WSJ).

Are they seriously suggesting that Joe Biden is going to have a bigger winning margin over Donald Trump than Barack Obama had over John McCain (+7), Bill Clinton over Bob Dole (+9) or FDR over Thomas Dewey in 1944 (+7.5)?

It makes me want to question the entire methodology in these polls.

This Gallup favorability survey would also suggest that Joe Biden is the favorite. Notice that the candidate with the higher favorability rating with likely voters won every election from 1992 to 2012. Romney and Obama were tied but Obama had a higher favorable rating with all adults.

That was not true in 2016 as Hillary had a higher favorability rating than Trump and still lost.

Interestingly, Trump actually has higher favorables in 2020 than he did in 2016. However, Biden is still 9 points higher than Trump in the Gallup survey.

Can Trump overcome this disadvantage? How much are the recent revelations about Hunter Biden's affecting Joe's favorables leading up to election day?

If I am in Biden's camp I would be feeling good about the polls and the difference in the favorable ratings.

What would concern me if I was Biden?

I have never seen a bigger enthusiasm gap between two candidates as there is with Trump vs. Biden. You would have to point to Obama vs. McCain to find anything similar. It is quite remarkable.

The momentum also clearly favors Trump. However, this is probably not as important as in past elections because of the large number of early voters this year. 

The early vote is huge but it does appear that Black, Hispanic and youth votes have not turned out in the same percentages in the early voting in key states as we have seen in past elections . This can be made up on election day but Democrats have generally indicated in polling that most were going to vote early. A lot of that seems to be driven over fears of Covid which surveys show is much higher with Democrats than Republicans.

If I were Biden I would also be concerned about the number of White working class voters who are turning out to vote. This demographic is what made Trump the winner in 2016. Trump won White non-college voters by 37 points last time. He won men in this group by 48 points.

Biden needs to cut into this margin. That is especially true in the Rust Belt states. He has to be particularly concerned about new voters turning out in this group. The odds are that any new voters in this group are not turning out to vote for Joe Biden as a candidate.

If I am Trump I would be most concerned about the votes of White, college-educated women. I would also be concerned about the polls indicating that I am losing support in the age 65+ demographic. This is generally the most reliable demographic to turnout to vote. It has also tended to be more conservative than other other groups. Losing voters here would hurt Trump badly.

In the end, this election will be determined by turnout. Candidates win who best turnout their voters.

There were a couple of great examples of this in 2018 and 2019.

In 2018, the Republicans lost a Senate seat in Alabama when their candidate was Roy Moore. This was in a state that Trump won by 28 points two years earlier. Democrat Doug Jones won by 1.5 points.

How did he win? Jones got 92% of the votes Hillary did two years earlier. Moore only got 50% of Trump's total---500,000 fewer votes than Trump got. They didn't vote for the Democrat. They simply stayed home.

The same thing occurred in Kentucky in 2019. Trump carried Kentucky 62.5%-32.7% in 2016. Incumbent Republican Governor Matt Bevin lost by 5,000 votes to Democrat Andy Beshear.

The Democrat got 89% of the votes that Hillary received in 2016. Bevin only got 59% of the votes Trump got. Another 500,000 Trump voters seemed to have stayed home.

The conventional wisdom in politics has been that high turnout helps Democrats. I am not sure that is true in the Trump era as he seems to have a unique ability to motivate people to vote who have sat on the sidelines before.

For example, look at these statistics from Trump's rallies in Pennsylvania on Saturday as reported by GOP Chairwoman Ronna McDaniel. Approximately 25% of those who signed up to attend those rallies were people who did not vote in 2016. This is consistent with the stats I have seen for almost every Trump rally this year.

The rally on Saturday in Butler, PA was a sight to behold. I have been in Butler a number of times and my jaw dropped when I saw the size of that crowd. Butler's population is about 14,000.

Take a look at this Reuters photograph. The Secret Service estimated there might have been 50,000 people there.

Who is going to turnout to vote?

Keep these stats in mind as well.


Trump won in 2016 despite only 55% of White, working class voters turning out to vote. That compares to 72% of white college graduates. They are also the largest voting bloc comprising 41% of total voters.

There is more upside in votes in this demographic, by far, than any other group if Trump's support is still at the same level as 2016. His objective is to make sure they vote.

There is untapped potential for Biden in the Black and Latino vote. However, the polls seem to indicate that Trump has made inroads here. What happens with this vote will also be key on Tuesday.

If you want an excellent in-depth analysis of the state of the race and a prediction of where it ends up on November 3 I recommend you read this piece by Philip Stutts.

Stutts is not a pollster but has access to the largest data collection, analytics and AI company in the United States. He sees Biden the winner of the popular vote but Trump winning the electoral vote with a smaller margin than 2016.

He bases that on voters being most concerned about safety and security issues (ending riots, funding police, rebuilding the economy) and breaking for Trump at the end. I might add that this is also a benefit of incumbency. It is always a risk to vote for the challenger. The devil you know is always safer than the devil you don't. That is especially true in times of uncertainty like we are in right now.

He also thinks that Trump wins with his brand over Biden--projecting strength and optimism over weakness and pessimism.

Finally, he thinks there are a lot of "shy" voters out there who will vote for Trump that pollsters have missed.

Read the entire article. It has a lot of good insights.

Looking Forward

What should you look for on election night to get a sense of where things are going?

I am going to look first at the returns coming out of Kentucky. Kentucky's polls close at 6pm and it usually has the earliest reported results. In 2016 when I saw the early counts coming in from Kentucky, and Trump pulling in over 60% of the vote, I anticipated a good night. It proved true.

I am going to be interested in the reports of how heavy the election day voting has been. Larger turnouts might favor the Republicans. Lighter turnouts the Democrats.

I am going to be interested in any exit poll data that comes out early. I am going to be cautious in relying on these polls too much since it is unclear to me if the models can be relied on as they might have been in the past with so much early voting this year. However, if those exit polls show a marked upward trend in turnout with White working class voters, or a poor turnout in Black precincts, or a shortfall in the youth vote it might give some hints on where things are going.

The results from Florida, North Carolina and Georgia are going to be particularly important early in the evening. If Trump holds these states with margins similar to 2016 he may be able to duplicate what he did the last time. This will show that Biden has not been able to strip away the support Trump had in 2016.

If Trump loses any of these it is going to be difficult to see how he can win.

If Trump were to be in a close race in Texas or Ohio he will be in serious trouble.

On the other hand, if Trump is doing better in states like New Hampshire or Virginia than he did the last time, Biden needs to worry even more about Pennsylvania, Michigan and Wisconsin.

It is also going to be critically important that the news source you watch for election returns is providing context for you as to where the votes are coming from. 

Are they early (more likely Democrats) or election day votes (Republicans)?

Are they coming out of urban (Democrats) or rural (Republican) areas?

What is going on in the suburbs? Is Trump holding his own compared to 2016? If he is not, Trump has to do even better in the rural areas or hope that Biden is underperforming in the big cities.

How do the vote percentages compare to 2016 in key counties in the swing states?

For example, in 2016 Trump lost Miami-Dade county by 30 points. If he keeps Biden's margin below that number it is going to be very positive for Trump in Florida.

On the other hand, if Trump lags in Sumter County in Florida (location of the large Villages retirement community,) which he won by 39 points the last time, it could indicate that he really is leaking support with the age 65+ demographic.

The so-called experts will start the night giving Donald Trump about a 10% chance of winning the election.

Nate Silver of is predicting that Joe Biden has a 89% chance of winning.

The Economist has Joe Biden with a 96% chance of winning.

Considering all of the analysis I have done I think they are vastly underestimating Trump's chances.

I expect to see a lot of close races in many of the swing states I listed above.

A lot will depend on turnout.

There is no question that Trump is the underdog going into election day. 

That is where he started when he first announced for President back in 2015 and nothing much has changed since then.

I don't know what is going to happen. I am not making any prediction or putting down any odds.

However, if I have learned anything over the last five years it is that it is not wise to underestimate or bet against Donald Trump.

I can't count the numbers of times he has been counted out by the media.

He is still standing.

Will he still be standing on November 4?

We will find out together.

It may prove to be one of the most consequential nights in American history when so much hangs in the balance.

May God bless the United States of America.

No comments:

Post a Comment