New study shows how demos are on Dems' side for 2020s. But they can't assume it'll work out

I wanted to draw your attention to a study from a national data firm named Catalist that did a deep dive on who Americans voted for in 2020, and the demographic breakdown is fascinating.

First, let's look at who Catalist is, and how they performed their analysis.

Catalist is a progressive data utility that works with public interest organizations and Democratic campaigns. We have been collecting and maintaining a national voter registration database – usually called a voter file – for more than 15 years. Broadly speaking, voter files provide a detailed view of the electorate, because they include the full list of people who voted in an election, collected directly from each chief election official in the country – when it comes to understanding voter turnout, this is not a survey, and it’s not a sample.

Voter file databases also vary in their methods and approach. Many voter files emphasize individual-level accuracy, optimizing for political marketing in a given election cycle. While we certainly do this at Catalist, we also realize that special attention and methods are needed to provide an accurate, or even reasonable, aggregate view of the electorate. We’ve invested a lot of time in trying to tackle these specific problems, with an emphasis on accurate aggregate-level statistics over time. The Catalist voter file represents hundreds of thousands of hours of work collecting, processing, and verifying information over a decade-and-a-half of dedicated research. Our process supplements the voter file with large-scale survey data, precinct results collected as widely as possible, and statistical modeling meant to tackle these specific challenges. Our results are not going to be perfect – nothing is – but we feel these efforts pay serious dividends, and give us a unique perspective on what happened in this election.

So Catalist put together this information and looked at the demographics of people who voted for Biden and Trump in 2020, and voter turnout in general. In doing so, they cut across racial, geographic, age and educational lines, and it illustrates a lot as to why our political parties act the way they do these days.

An area where Dems continue to gain are in suburbs, where the Biden/Harris ticket continued the strength that Dems showed when they gained the House 2 years prior.

Moving to...the suburbs, we see a consistent gain for Democrats over the Trump era, going from 51% Democratic support in 2016, to 54% in 2018 and 2020. In other words, Biden was able to maintain the gains that Democrats enjoyed in the suburbs in 2018. In cities, on the other hand, Biden lost a little bit of ground compared to the midterm. While the densest urban areas remain deep blue, Democratic support peaked in 2016 and 2018 at 75%, and fell slightly to 72% in 2020.

As I've mentioned before, it was in the Milwaukee suburbs where Biden won Wisconsin in 2020, much like how Tony Evers did in 2018.

In Wisconsin, Waukesha County was unambiguously crucial to Biden's coalition. Specifically, Biden's biggest gains over Gov. Tony Evers' 2018 map were in the WOW counties, around Milwaukee. Biden improved 13% in Waukesha, and it helped cancel out Trump's gains in the west. pic.twitter.com/DmAF5TyZVo

— J. Miles Coleman (@JMilesColeman) November 5, 2020

By contrast, Biden did NOT win our state by driving up turnout in Black communities, and actually lost Latino votes compared to Clinton in 2016. Catalist notes that Wisconsin not seeing much change in the results among non-white voters was in stark contrast to the situation in states like Georgia and Arizona, where voters of color were decisive in Dems winning those states.

But let's also not forget the second numbers to consider with elections - turnout. And Catalist notes that America's electorate continues to include more voters of color - particularly in 2020, when some groups had major increases in turnout, resulting in a notably more diverse group of voters than the group that elected Barack Obama 12 years prior.

These changes resulted in the most racially diverse electorate on record. We estimate that 72% of the electorate was white, down 2 percentage points from 2016. This comes almost entirely from the decline in white voters without a college degree, who continue a steady decline (in percentage terms) over our entire dataset, and indeed even longer than that. As recently as 2008, white non-college voters were a (bare) majority of the electorate. While they remain the largest plurality voting bloc in the country, they are down to 44% of the electorate in 2020.

And that change in the electorate worked in Biden and Harris' favor. Even though Trump actually got a higher proportion of voters of color, the added turnout of these pro-Dem groups overrode the percentage changes in some key states.

Add it up, and look at how diverse the group of Biden voters were compared to the universe of Trump voters, of whom nearly 3/5 were white people who did not graduate from college.

So much cool stuff in the Catalist 2020 presidential vote survey but this graphic in particular does a great job of illustrating of who makes up the parties right now https://t.co/QvHegp8RWF pic.twitter.com/TqnpdF00Iy

— Aaron (@BobbyBigWheel) May 10, 2021

This means Dems have to listen to and include people of color, while Republicans are increasingly reliant on a group of low-educated white people that are becoming a smaller part of the electorate. This feeds upon itself, as the GOPs are increasingly turning to racist messaging and voter suppression to try to stir up white folks in the sticks instead of trying to grab more of the votes of growing groups of people of color in populated areas. Which is ironic because Trump gained with Latinos in 2020, particular in Florida and Texas, where those gains allowed Trump to hold onto those states in 2020.

Now that doesn't mean that demographics are destiny, and Dems continuing not to get the votes of rural, non-college white people is a significant barrier when it comes to winning seats at the state level, particularly in places like Wisconsin. And the fact that Trump made gains with certain groups of color in 2020 means that Dems cannot be smug and think they have locked up the votes of those groups, but instead must listen to and deliver for these people.

But it does explain a lot, and it illustrates why so many elections have become a sort of trench warfare where "outvoting the other guys" is more likely to win than having November candidates work for an overall consensus where most groups like the winner. And given that Dems have more of the support of growing demographics, it helps explain why GOPs increasingly do not seem to want to win races on ideas, but instead would rather scheme tactics to get power even if they don't have the support of the people.

Here's a one-file look at all of the stats that Catalist put together at the national level for elections since 2008. Dig in and see what you find!