Welcome to the the first of 51 (D.C. has electoral votes too) reviews of the states that are choosing the next president. At various times in American history, the popular vote and electoral vote line up almost precisely. This isn’t one of those occasions. The drama is all on the state by state tabulation side.
We’re starting in Michigan because it was a key surprise component in Trump’s 2016 upset victory, but I’m hoping you find the journey to places like Hawaii and Rhode Island interesting too. No state in the Union has managed to stay only red or only blue throughout it’s history. Today’s sure thing could be up for grabs as soon as the next cycle.
2016 Results: Trump 47.50%, Clinton 47.27%
This was close as hell, approximately 10,500 votes out of about 4.5 million, with Jill Stein taking 51,463 voters who must have preferred Clinton to Trump. Going back through the numbers is a great reminder of exactly how narrowly Trump threaded the needle. Or, if you prefer, all the many ways the Clinton campaign snatched defeat from the jaws of victory.
This was the first GOP win since 1988, and Michigan the reddest (compared to the rest of the country) since 1976.
2020 Polls: As of today, FiveThirtyEight is showing Biden with a 6.5 point lead. His national advantage is about one point more. We can look at this two ways. First, Biden has a nice lead, outside the margin of error, but not insurmountable. He’s consistently led non-partisan polling for the past several months, though some show him with a double-digit lead, while others have this as a 3 to 4 point race. It appears Michigan is moving the same direction as the country as a whole. Biden’s numbers here tend to improve or decline in the same direction as his national results.
The second point is Michigan remains slightly more pro-Trump than the country. In 2016, he ran a little over 2 points better here. Now it’s a single point. Very similar. Given both of these factors, I’d guess it will be tough for Trump to lose the popular vote by 4 or 5 points and still win Michigan. But we can also figure if he improves his national numbers, this will get closer too.
Key Historical Shifts: Dating from 1876 (which is as far back as I built my table lol), Michigan was more Republican than the nation in each election through 1952. This doesn’t mean Republicans won every presidential election here. FDR won 4 times. But in each instance, his margin was narrower than the national popular vote win.
For this whole stretch, you would have considered Michigan a red state and assumed a Republican couldn’t possibly win without it. Then something happened. I’ve been meaning to attempt to figure out what for years.
The same thing happened at around the same time in Minnesota and Pennsylvania. Don’t know if Michigan just didn’t like Ike (in 1952 Eisenhower ran just barely ahead here, but by his 1956 re-election gave up 4 points, though not enough to make it close), or if the increasing power of the United Auto Workers made Michigan blue. Either way, Michigan was a key part of JFK’s narrow victory in 1960. By 1964, Democrat Lyndon Johnson ran well ahead of his national landslide.
With one exception, 1976, when Michigander Gerald Ford was on the ballot, and did 7 points better than his overall result, the state has remained blue, with occasional bits of purple (1984, 1988, 2016.) If a Democrat is doing poorly in Michigan, it means there’s a problem nationally. Their candidate is failing with white working class voters, failing to inspire African American voters, or both (2016.)
Where Biden Can Improve on 2016: In his 2012 re-election, Barack Obama got 595,000 votes in (predominantly Black) Wayne County. Four years ago, Hillary Clinton only managed 519,000 votes. This gap was almost 7x her margin of defeat. Even if Biden can’t match his former boss, just splitting the difference should flip the state.
Another path is taking some Obama-Trump voters and making them Obama-Trump-Biden people. Macomb County, north of Detroit, and once home to a ton of auto industry jobs, is the spiritual home of the Reagan Democrat. When people talk about the white working class, those voters (and/or their ancestors) who picked Nixon in 1972, Reagan in 1980/1984, and then Trump in 2016, this is what they mean.
Obama won Macomb by 4 points in 2012. Trump won by 11.5 in 2016. Biden doesn’t need to win Macomb, but if he limits the deficit to a few points, he’ll win Michigan, even if the specter of Trump doesn’t help him in Wayne. Voters haven’t forgotten the 2009 auto bailout. And Biden’s ads are reminding them anyway.
Then there’s the upscale, traditionally GOP leaning suburbs, where Democrats won key victories in the 2018 midterms. Two congressional districts, CD-8 and CD-11, are good examples. Each sent Democrats to D.C. in November 2018. Both supported Democratic Governor Gretchen Whitmer and Senator Debbie Stabenow (narrowly) in the same election.
Trump won CD-8 by 7 points and CD-11 by 5. If Biden can get results halfway between 2016 and 2018, he’s got a third path to win Michigan, even if the other two options are blocked.
Where Trump Can Improve on 2016: I think it’s a certainty he loses Wayne County by more than in 2016. Beyond Clinton’s political malpractice and the possible benefit of having Kamala Harris on the ticket, the county was clobbered by COVID, with some of the highest death rates in the country in March, April, and May, impacting African Americans in particular. Trump is losing ground here for sure.
I also wouldn’t want to bet on a retired auto worker who supported Hillary last time suddenly opting for Trump now. He may not lose what he picked up last time, but how does someone who voted for Obama twice, plus Hillary, now suddenly decide it’s time to support Trump? Scranton Joe is the most Macomb County-friendly Democrat since Bill Clinton.
If Bernie Sanders were the nominee, Trump might be able to grab a few upscale suburban voters who actively dislike him but fear socialism. He’s trying to turn Biden in to Red Joe, and it may work with some voters, but not the same suburbanites who already held their noses and voted for Hillary last time. Trump has spent the last four years making them feel better about their decision.
So where can Trump turn to give himself a chance? The same rural counties where he racked up big margins in 2016. Michigan has a couple million voters who don’t live in metro areas, and while they were very pro-Trump last time, there were still some voters who stayed home.
If we can say Wayne County voters who weren’t excited about Hillary and stayed home because they didn’t think their vote was needed aren’t going to make the same decision this time (and I’m saying it), then it’s equally possible that rural voters who didn’t think Trump could win in 2016 will make sure to support him this time.
There was a jump in participation among less educated rural voters from 2012 to 2016, but it was still far less than those in more populated areas with 4 year or graduate degrees. A further narrowing of the gap in 2020 would give Trump a legit chance.
Forecast: This is a must-win for Biden. It’s not for Trump. Had Hillary won Michigan in 2016, Trump would still have won the election. The state is looking more relatively favorable to Biden than it was to Clinton.
It’s fun to talk about Biden flipping Texas, Arizona, or Georgia. And it’s very possible he wins all three. But it’s very hard to see him winning any if he loses Michigan. To get to 270, he needs to win here, plus Wisconsin, plus at least one more Trump state. There are all kinds of possibilities. Pennsylvania, Florida, etc. Those are more of a reach than Michigan too.
A Biden victory will not only include Michigan, it will have a margin here of at least 100,000 votes.