Let's take a deeper dive into the differences between the 2016 and 2020 elections.
One Year he won, the other he lost. He got more votes in 2020 than he did in 2016. So how could he possibly lose? I'm glad you asked that.
“The biggest impact the third-party voters had in 2020 is perversely not in voting for the third party in 2020. By which I mean, it's what happened to Gary Johnson voters and Jill Stein voters from 2016.”
Got it yet? The easy answer is not non-partisan voters, but it is, voters who chose to vote for a third-party in 2016. In fact, a record number of people voted for someone other than one of the major two political parties in 2016, in some cases two to three times as many. Let's take a look.
Historical Partisan Voting
While I will make a larger argument for this theory later, to illustrate I want to take a look at the five states that Trump won in 2016 but lost in 2020. Those states were Arizona, Georgia, Michigan, Pennsylvania, and Wisconsin. Before looking at the numbers specifically, let's note the following.
Three of those states Trump won by less than 1% in 2016 (Michigan, Pennsylvania, and Wisconsin). All three of those states had not been won by a Republican in this century.
In the other two states, Arizona and Georgia, Trump and Republicans had already been losing ground since 2000. In 2016 Trump's margin was more than 2% less than Romney's from 2012, and in Arizona, his margin was less than half what Romney's was in 2016.
In short, we have Trump winning those five states in 2016, but three of those he won by less than a point, and the other two he won, but by a significantly smaller margin than the Republican nominee did four years before.
Now, let's look into the numbers from both President Bush and President Obama and the differences between their election, and re-election campaigns.
In 2004, President Bush improved upon his 2000 numbers in four of those five states (The exception was Wisconsin, where the numbers were virtually identical). However, the numbers didn't change the outcome. He still won two of those states (Arizona and Georgia) and lost three others (Michigan, Pennsylvania, and Wisconsin) just as he did in 2000.
In 2008, President Obama improved upon Democrats 2004 numbers in each of those five states, but again, this did not change the overall outcome. Obama won three of those states and lost two, the exact same ones Democrats won and lost in 2000 and 2004.
Unlike President Bush in 2004, President Obama did not improve on his numbers between 2012 and 2008. Republican nominee Mitt Romney actually performed better in each of those states than John McCain did in 2008, with the one exception being Arizona, McCain's home state. However, all five states again had the same overall outcome as they did in every single year this century. Three of them went for Democrats, and two Republican.
So, to review, prior to 2016, those five states had the same outcome in each of the previous four elections. President Bush improved his numbers between elections, and President Obama went down.
What Happened in 2016?
I'm glad you asked. It's the same question I asked myself after Trump's historical victory that year. How did Trump win three states that Obama won by double digits in his first election? In order to answer this question, we need to separate two ways to look at winning and losing in any campaign. The first number to look at is the percentage of a vote the candidate receives. The second number is the margin of victory, or the difference in percentage between the two candidates. First, let's break down the winning number in each state compared to President Bush.
Trump's percentage of the vote in three of those five states (Arizona, Michigan, and Wisconsin) was lower than Bush in 2000. His percentage was better in two, Michigan (+1.2%) and Pennsylvania (+1.80%).
If we compare Trump's 2016 election to Bush's 2004 re-election campaign, Trump actually got a smaller percentage of the vote in all five of these states.
Despite having a similar percentage to Bush in 2000, and a lower percentage than Bush in 2004, President Trump was able to win all five of those states, three of which President Bush was never able to win.
So how did Trump get a relatively same percentage of the vote, or lower than Bush, and win so convincingly? Now let's look at that second statistic, the margins of victory.
Trump's margin of victory was lower than Bush in two states, Arizona (-2.70%), and Georgia (-6.60%). Both of those states were won by Bush in 2000 and Trump in 2016.
Trump's margin of victory in the other three states (Michigan, Pennsylvania, and Wisconsin) was an impressive average of 3.77% more than Bush. All three of those states Bush lost, and all three of those states Trump won.
By comparison to Bush's 2004 re-election campaign, the Republican margin in Arizona and Georgia has been cut significantly, and the margins in Michigan, Pennsylvania, and Wisconsin were the best for Republicans in over two decades.
So HOW and WHY did Trump win in 2016 when he performed relatively the same, or worse, than President Bush?
The answer to the "how" and the "why" are different. The how? Hillary Clinton. I could make the argument she was the most unpopular candidate in over 50 years, but that is an article for another day. For the purposes of this blog, let's stick to the "Why," which has a fairly simply answer. More voters cast a ballot for a third-party candidate in any election since Ross Perot appeared on the ballot in 1996.
Almost seven million voters cast a ballot for a third party in 2016. This was more than the combined total of votes received for third-party candidates in the previous three elections combined.
In the five states we have broken down, the third party candidate received 5.32% of the vote on average. The average percentage of the vote third-party candidates received in 2012 was 1.34% and in 2008 it was 1.24%. This means third-party candidates received nearly four times as many votes than third-party candidates did in either of the previous two elections.
So who did those candidates take away votes from? Hillary Clinton of course. Clinton received an average of -3.52% less than Obama while Trump's was relatively the same as Mitt Romney (-0.46%) in those five states.
OK, so what happened in 2020 then?
There is another simple answer here, which is that voting for third-party candidates went back to normal levels in 2020. The voter who couldn't vote for Hillary Clinton in 2016 voted for Joe Biden. Consider this:
In the five states we've been looking at that all flipped, Trump got, on average, +0.39% more than he did in 2016. He actually got a higher percentage than he did in four of the five states. Georgia was the only difference.
In those same five states, Joe Biden received, on average, +3.49% more than Hillary Clinton did in 2016. His margin in those states, on average, was only one point higher (+1.02%) than Hillary Clinton in 2016.
Joe Biden received, on average in those five states, almost the identical percentage that Obama did in 2012. The average difference? It was less than a twentieth of a percent (-0.03%).
One Bonus Stat: While Biden was nearly identical on average to Obama's re-election campaign, he actually underperformed Obama's first campaign from 2008. In those five states, Biden's average in 2020 was -2.29% worse than Obama's was in 2008.
The 2020 election was statistically similar to election history over the past two decades where 2016 was very different. In terms of percentage of vote, Joe Biden performed the same as Barak Obama did during his re-election in the five states that flipped. In terms of percentage of the vote, Trump's average percentage was almost the same as 2016, a difference of less than half a point (+0.39%). The simple math tells us that Trump performed equal to 2016 in terms of percentage where Biden performed much the same as traditional Democratic nominees, where Hillary Clinton significantly underperformed compared to other Democrat nominees (-3.49%).
The Biden did just enough to win the election, he didn't win in a blowout. He simply performed like a traditional Democratic nominee should, and comparted to some, actually underperformed. Remember, in those five states, Biden's average percentage was actually worse than Obama's from 2008 (-2.29%).
The difference this time around was much the same it was in any election since 2000. Swing voters went for the winner and against the loser, and not to a third-party candidate. Where third-party votes hurt Hillary Clinton in 2016, those same votes went back to Joe Biden in 2020.
I would never take anything away from Trump, he was one of the most impressive candidates in modern election history, and the way he campaigned became a game changer. That being said, it's worth asking one big question about the 2016 election. Did Trump WIN that election or did Clinton LOSE it? I can tell you by the numbers, at the very least, it's a combination of the two.
Notes and Things to Pick At
In an attempt to be as honest and forthright as possible, this is the area where I mention certain facts that can influence an article and how you look at it. So here are some things you should know
1. In this article, I discussed "averages" in the five flipped states. The "average" I refer to takes into account negative numbers on one side and positives on the other side. So there could be arguments made against some of the math here, but I use the math specifically to point to a larger point.
2. Although I could have, I did not dive deep into voter registration and how states are specifically trending. The obvious caveat to take into account in these numbers is that the Republicans have been losing votes in two outlying states (Arizona and Georgia) have been gaining them in the three Midwestern states. I could most likely do a separate post splitting the outlying states from the Midwestern states, but at least for me, the conclusion would still be similar.
3. In the number comparisons, I feel the need to remind you that while Bush performed better in his re-election, Obama performed worse. So one of those was more popular a the time of his re-elect than the other, and some comparisons might reflect this.
That is all for now folks, thanks for reading, and feel free to shoot me a note on why you agree or disagree, or with a follow up question you might have.