The progressive mobilization myth is alive and well

The progressive mobilization myth is alive and well

Bad data and shoddy analysis is not the way to win

As I wrote in “Progressives’ mobilization delusion,” progressive politics took a wrong turn roughly a decade ago by convincing itself in the wake of Barack Obama’s reelection that all Democrats had to do to win was boost turnout among young people, Black people, and Latino people, likely by emphasizing left-wing stances on cultural issues.

That delusion was based on a number of errors and misconceptions, perhaps most notably literally undercounting the number of white people, old people, and people with no college degree in the electorate.

Unfortunately, what I learned from Ron Brownstein’s October 7 article “What Democrats Need to Understand About the Changing Electorate” is that Mobilization Myth thinking is alive and well in the movement, even in the wake of a 2020 election that should have put the nail in the coffin of key aspects of that narrative.

The article is based around what he characterizes as “an exhaustive analysis of electoral landscape” by a group called Way to Win. The report argues, to quote Brownstein, that “the key to contesting Sun Belt states such as North Carolina, Georgia, Texas, and Arizona is to sustain engagement among the largely nonwhite infrequent voters who turned out in huge numbers in 2018 and 2020.” He also writes that in these conclusions, Way to Win “echoes the findings of other Democratic strategists such as Mike Podhorzer, the longtime political director of the AFL-CIO.”

I think Brownstein’s article is an important document. The report it’s based on is the result of collaboration with a ton of organizing-centric partner institutions, echoes the views of the AFL-CIO’s political director, and in other respects represents the dominant thinking of the progressive wing of the Democratic Party.

It’s also, unfortunately, based on data and analysis that are clearly incorrect — in some ways egregiously so.

Did Florida Hispanics swing left in 2020?

(Ethan Miller/Getty Images)

A huge red flag around this whole report is that in their discussion of the Hispanic (or “Latinx”) vote in 2020, the authors say this swathe of the electorate swung toward Democrats relative to 2016 — including in Florida.

In the original version of the report, this data was attributed to TargetSmart, but these are not conclusions that TargetSmart themselves reached. Way to Win has now updated to clarify that this is merely their analysis of the TargetSmart voter file data, which is helpful, but they are not addressing the underlying problem that the analysis is mistaken.

Instead, the beginning of the report now has this disclaimer:

The visualizations throughout this report rely on a relatively crude method for counting any individual voter with a score from 50-100 as a “likely Democrat” and coloring them with blue shades in the accompanying graphs, and any individual with a score from 0-49 as a “likely Republican” and coloring them with red shares in the accompanying graphs. These visualizations are not meant to serve as a proxy or precise predictor of actual vote choice. But they do provide a stable, cross-state tool for evaluating key trends in voter participation.

That helps explain the inaccurate numbers, but again, it doesn’t really address the underlying issue, which is that when you base your analysis on inaccurate numbers, you get bad answers.

It’s very clear that Florida Hispanics shifted right in 2020. We saw it in pre-election and post-election surveys. And in Florida especially, the places with the largest Hispanic populations had the biggest swings against Biden.

This data from Catalist is a much better look at what happened with the Latino vote than what Way to Win is selling.

The rightward swing of Florida Hispanics was an extremely unsubtle feature of the 2020 election. It was very large in scope, it was widely discussed on Election Day because Florida’s polls close early, and it also led to two Miami-area House seats flipping. That someone would run with such an obviously wrong analysis suggests you should be very skeptical of everything they say.

But what’s worse, when you understand Way to Win’s methodological proviso, you’ll see that their own analysis completely undermines the mobilization story they are trying to sell.

Florida Democrats turned out the base and lost the argument

Go back to the proviso, and you’ll see that Way to Win is not even attempting to look at actual vote choice.

What they are doing instead is using data that is supposed to help you guess, ex-ante, who a given person is likely to vote for. The idea is that before you do a get-out-the-vote drive, you need to know which people you actually want to come out and vote. So when they say in the chart that Florida Hispanic support for Democrats rose 1.9 percentage points, what they are saying is that the mix of likely Democrats to likely Republicans in the Florida Hispanic voting pool improved.

That’s great news for Florida Democrats, except that when we look at the election results, it turns out that they did way worse with this demographic group.

That’s because the incredible success Republicans had with Latino Floridians was driven overwhelmingly by persuasion. Lots and lots of people who seemed like Democrats based on their past track record decided to vote Republican instead.

In other words, Mobilization Theory completely failed here. It’s true that there were many sporadically-voting or non-voting Latino people in Florida. And it’s true that many of them have demographic characteristics that would have led you to believe they were likely to be Democrats. It’s even true that it’s possible to mobilize them and get them to vote. But that doesn’t guarantee you victory because persuasion matters a lot, and in this case, Republicans won the persuasion argument.

Now I feel like every time I make the case that swing voters are important and a reputation for moderation is helpful, people come out of the woodwork to say that’s all obvious and there’s nothing new here. But that, I think, is the significance of this Way to Win document. A meaningful faction in Democratic Party politics is so bought-in on Mobilization Theory that even when their own data shows it’s wrong, they just kind of fudge it with misleading charts rather than revisiting their assumptions.

More basic errors

It’s challenging and probably pointless to fact-check every single chart in a document that’s so intellectually lazy that it tries to make you think Florida Hispanics shifted left in 2020. But as an example of how bad the analysis is, check out this chart which purports to show that higher turnout helps Democrats win states.

But just squinting at this, it’s not obvious there’s any correlation here. Turnout went up in every state, and Democrats did better in most states, but there’s no clear relationship between turnout surges and improving Democratic vote share.

In fact, when Milan Singh looked at all 50 states, the correlation (which is weak) goes in the opposite direction.

Indeed, thanks to education polarization, it’s far from clear that we should expect high turnout to favor Democrats anymore. If you look at special elections held between the 2018 midterms and the 2020 presidential election, you see that Democrats ran 4.7 points stronger than Hillary Clinton in 2016. The 2020 general election results were quite a bit worse than that for Democrats, suggesting that the low salience of special elections was helping the party.

Here’s another slide that’s very odd, criticizing Democrats for spending more money on ads touting bipartisanship than on ads about the economy.

I agree it would be a mistake not to talk about the economy in your ads. But if you combine “the economy” with “jobs” (to say nothing of small business, special interests, lobbying, taxation, and debt), then clearly the party was in fact keeping economic issues front and center.

Indeed, my view is that Democrats’ paid media program is the party’s messaging strong suit. In the paid media space, Democrats try pretty hard to think about the actual electorate and pick messages that are likely to appeal to persuadable voters. The whole problem is that their “earned media” strategies (what the candidates and allied groups say and what kind of coverage they seek) tend not to match this.

But regardless of whether or not you agree with that, there’s something slippery about a group that goes for this “spent more on bipartisanship than the economy” laugh line when it’s clearly false.

The way to win is to take research seriously

I don’t think there is anything wrong with progressives reacting to their frustrations with the Obama Era by deciding they wanted to try out a different idea. But the flip side of experimentation is you need to be willing to be honest with yourself.

The Mobilization Theory was based, in part, on an erroneous assessment of the actual demographics of the country.

It was also based, in part, on underestimating the extent to which politics would become nationalized, and it was going to become way harder to have Democrats winning Senate races in Great Plains states that vote overwhelmingly Republican on the presidential level.

And it was based, in part, on some conjectures about non-voters and their likely behavior.

There’s nothing wrong with trying new things or with making mistakes. But what should really worry people is this tendency to try new things, see them fail, and then pretend they worked anyway. Doubling down on a failed strategy is a bad idea. Beyond that, though, life is unpredictable and we don’t know what will happen next. To respond to unforeseen events in an effective way, you need to be able to look at the world honestly. And I’m afraid significant swathes of progressive politics aren’t up to that challenge right now.