Democrats didn’t lose because of ‘defund the police’ in 2020
There is a lot of chatter about Democrats needing to move right on crime — in part, the argument goes, because they got hurt by campaigning on defunding the police in 2020.
But the critical premise just isn’t true. “None of the candidates included in this analysis supported defunding the police,” concluded a post-election report by a trio of Democratic-leaning groups released in May 2021. “But nearly all were targeted with paid ads claiming they did.”
That analysis focused on congressional Democrats. But Joe Biden didn’t run on defunding the police, either. Sen. Bernie Sanders (I-Vt.), the most prominent figure in the left wing, disavowed the idea. A small number of progressive Democrats, most prominently Rep. Alexandria Ocasio-Cortez (N.Y.), did embrace the defund rhetoric, it’s true, but those Democrats largely won, in part because they ran in very liberal districts. It’s not clear that a single congressional Democratic candidate who could have won a race lost it while emphasizing defunding the police.
A second faulty narrative, closely related to the first, concedes that few 2020 Democratic candidates embraced “defund” but focuses instead on the role of activists. This version of the story goes like this: Progressive advocates pushed defund into the national discourse and a small number of Democratic officials embraced it, which in turn allowed the Republican Party to peel away voters by falsely casting the broader Democratic Party as pro-defund.
If this narrative were true, it’s not clear that Democrats could have done much about it. The party can’t tell activists what to say and shouldn’t try. The idea that liberals can’t tout ideas that are fine in the areas they represent but potentially off-putting elsewhere runs counter to the very notion of local representation. And of course, Democrats can’t stop Republicans from caricaturing their positions.
But here’s the thing — this narrative isn’t true, either. Numerous election analysts have concluded that defund, while an unpopular framing of police reform, didn’t specifically hurt Democrats electorally in 2020. Defund, these analysts conclude, was just one of a broader series of attacks from Republicans that played on frames long used against Democrats — namely that the Democrats are too left, socialist and overly supportive of Black causes. Many Republican ads in 2020 hammered the Democrats not on defund but on the notion that Democrats were allowing “rioting” in cities in the wake of George Floyd’s killing.
“The ‘defund’ narrative about 2020 solidified in many people’s minds before most evidence was even available,” said Robert Griffin, research director at the Democracy Fund Voter Study Group. “Since then, there’s been an increasingly stark contrast between the continued strength of that narrative in the media and the relative scarcity of evidence to support it.”
It’s complicated to isolate how a single issue motivates voters. But we do have polling from the summer of 2020, when defund became a major part of the national debate. Biden’s numbers didn’t change much then, nor did those for congressional Democrats. Compare that with August 2021, when Biden’s numbers plunged as the Afghanistan withdrawal went poorly.
When a political event shifts voter preferences, it usually shows up in polls. Defund never did.
So why did “defund cost the Democrats” become such a pervasive narrative? I think there are four primary reasons.
First, Biden’s victory was narrower than anticipated, leaving Democrats looking for some ready explanation for that. But they went looking in the wrong place — the polls were overstating Biden’s lead the whole time, in part because they missed first-time voters who turned out to support Donald Trump.
Second, Democratic leaders, including influential Black Democrats such as former president Barack Obama and Rep. James E. Clyburn (S.C.), suggested publicly that backlash to “defund the police” was a major factor in the outcome. Obama and Clyburn are great politicians, but they aren’t experts in political data. Clyburn, for instance, argued that Jaime Harrison, a Senate hopeful from South Carolina in 2020, was severely harmed by defund. But I wrote a piece a month ahead of the election predicting Harrison’s almost exact margin of defeat based on previous showings by South Carolina Democrats.
But defund was also a convenient story for Democratic leaders to embrace — they could blame activists for the party’s worse-than-expected performance instead of examining the decisions of Biden, House Speaker Nancy Pelosi (Calif.) and other party officials who barely managed to defeat one of the worst presidents in American history.
Third, Democrats lost a net 12 seats in the House even as Biden won, driving a narrative that Biden did “well” because he successfully navigated issues like defund while congressional Democrats did not. But Biden didn’t actually do much better than House Democrats. He defeated Trump by 4.4 percentage points, while House Democrats ran 3.1 percentage points ahead of Republicans. Put another way: Biden won 224 House districts, and House Democrats won 222.
Fourth, voters in focus groups said they didn’t like defund and suggested it was one reason that they voted Republican. I’m not arguing that defund helped the Democrats. But what focus-group participants say motivated their votes isn’t a perfect indicator of what actually motivated them, particularly on a fraught issue such as defund.
In the summer of 2020, a host of things were happening at once: There were huge racial protests; some of them involved rioting; some activists invoked defund and other anti-police rhetoric; and Republicans argued that the protests were a violent rampage destroying American cities. It would be virtually impossible to isolate the slogan “defund the police” from the broader protests — and of course a Republican voter might rationalize their vote by talking about defund, even if they also opposed the broader protests against systemic racism.
Since the 2020 election, there has been a real electoral battle happening on crime policy — between more centrist Democrats and more left-wing ones. Some cities did reduce police funding in the aftermath of the protests, and a surge in crime has led to reversals and, in some cases, even increases in police budgets. In the New York Democratic primary for mayor, Eric Adams defeated a progressive opponent who favored reducing police funding. Voters in San Francisco and Los Angeles, both heavily Democratic, might vote to recall progressive prosecutors.
At the same time, some advocates of major criminal justice reform are winning their races, including district attorneys in New York and Philadelphia, and some of the 2020 police reforms remain in place.
If most Democratic officials, particularly those in swing districts or states, had been running on defunding the police or much more lenient crime policies, that would likely pose a big risk for the party. But that has not happened. The narrative that issues around crime are clearly hurting Democrats is just not accurate.
‘Wokeness’ has not pushed voters without college degrees away
This narrative is related to the one around defund. Policing is part of a broader series of issues on which the Democrats are said to be too “woke” and too prone to speaking in language that appeals to progressive White college graduates but turns off less educated voters, including many voters of color who usually back Democrats.
There are two shortcomings to this narrative. The first is that “wokeness” is so undefined that it’s hard to specifically attribute anything to it. While “woke” was once a term used by Black people to signify awareness of racial issues, it is now largely employed as a catchall pejorative against liberals.
If being woke means taking issues of racial inequality seriously, that in theory could be a problem for Democrats — but it’s not. In the last cycle, they won control of the House, Senate and presidency after talking about racial disparities more bluntly than ever.
But wokeness critiques are often about, say, liberals using the term “Latinx” instead of “Latino” or being too focused on transgender rights. Since it is hard to quantify or specify this kind of wokeness, we can’t credibly attribute many of the Democrats’ challenges to this. That said, I find it unlikely that there is a single voter in America who was going to back a given candidate but then didn’t after hearing the candidate use the term “Latinx.”
The deeper problem with this narrative is that the Democratic Party doesn’t have some new problem best understood as trouble with voters without college degrees.
Over the past decade in particular, there has been a decline in Democratic Party support among White Americans without degrees, while White Americans with four-year degrees have become more Democratic. Biden won about 55 percent of the vote among White Americans with degrees in 2020, compared with 37 percent among those without degrees, according to data from Catalist.
But that huge difference in education background isn’t as apparent among voters of color. Democrats won about 90 percent of Black college graduates, about 65 percent of Asian college graduates and about 60 percent of Latino college graduates in 2020 — and a similar number of those in those groups without degrees.
The big problem is that Democrats lost ground among Latinos in particular in 2020. They do seem to have lost more ground with Latinos without degrees than those with degrees, although we have limited data on this question.
The better, more nuanced narrative is that Democrats are slipping backward with Latinos, and should do more to address that, but their biggest problem is the same as ever: White people.
The focus on degrees and so-called education polarization is particularly wrongheaded. What is happening is that voters are increasingly aligning their broader philosophical views to their votes. So voters, particularly people of color, who view themselves as conservative but might have voted Democratic in the past are now aligning with the GOP. Democrats aren’t specifically losing voters without college degrees — they are losing voters, with or without degrees, White and non-White, who consider themselves ideologically to the right of the current Democratic Party.
“Trump did better in 2020 among conservatives who identified as White, Black, Hispanic and Asian American,” said John Sides, a Vanderbilt University political scientist and co-author of an upcoming book about the 2020 elections.
And these voters probably see the Democratic Party as too left of them not strictly on racial issues but on other issues as well. It’s not about “Latinx.”
Education issues are not crushing Democrats
The narrative here is rooted in the argument that some combination of school closures, Democratic candidate Terry McAuliffe’s comments playing down the role of parents in education policy, and Republican candidate Glenn Youngkin’s attacks on critical race theory led to the Democrats’ defeat last fall in Virginia. The recall a couple of weeks ago of three San Francisco school board members further bolstered the case.
San Francisco voters did vote out three members of their city school board in part because they changed the admissions processes at an elite public school in an effort to increase the number of Black and Hispanic students there.
But that wasn’t a case of Democrats being defeated by Republicans. In San Francisco, Democratic voters, based on the recommendation of the Democratic mayor, voted out extremely-left-leaning school board members, who will almost certainly be replaced by very-left-leaning people. I suppose Biden and congressional Democrats should not run on changing the admissions requirements at elite high schools — but I don’t think they ever had much intention of doing that.
In Virginia, an analysis by Michael Hartney, a political science professor at Boston College, found that places where school closures lasted longer were more likely to back Republican nominee Youngkin in the 2021 Virginia gubernatorial race. But Hartney found that the closure effect explained only about 0.8 percentage points of Youngkin’s margin. Youngkin ran almost seven points better than Trump did in Virginia a year earlier, so school closings were far from the whole story.
Saying “I don’t think parents should be telling schools what they should teach” no doubt didn’t help McAuliffe. But in New Jersey, where education issues weren’t as prominent, Democrats also did much worse than in 2020. And Democrats remain more trusted than Republicans on education. The Democrats’ weak showings in New Jersey and Virginia are much better explained by the general anti-incumbent-party sentiment that usually arises after a presidential election and the very high turnout by Republicans in those races than anything else.
Why are these narratives so pervasive? For two reasons.
First, the Democratic Party is increasingly divided between a left wing and a more centrist wing. In city politics, where Democrats are dominant, these tensions are clearest, and they matter greatly. Police funding is likely to be increased more by center-left Democrats, while more left-wing Democrats are more willing to push policies to make schools more racially equitable. And each side works to elevate their local-level victories into the national conversation as part of this broader and larger intra-party battle.
But mayoral elections in New York and school board elections in San Francisco tell us little about the broader Republican vs. Democratic conflict. These localized fights are getting elevated nationally right now for another reason — the two wings of the Democratic Party are battling over who is to blame for the party not doing as well as Democrats hoped.
“Defund/wokeness/equity/Latinx is killing us,” goes the centrists’ argument, “The Biden-Pelosi ‘be moderate plan’ isn’t working” is the left’s answer.
I think neither side is to blame. The portrayal of Democrats as wide-eyed hyper-liberals policing everyone’s speech is coming from Fox News and other conservative outlets — and that can’t really be stopped. Biden’s governing approach basically had no chance of succeeding after the emergence of high inflation and more covid variants, neither of which he had much control over. And the party that controls the presidency basically always loses the midterms.
Most important: It’s the voters, stupid.
Lots of Americans prefer the conservative party to the more liberal one. It’s not clear that any tactical changes by Democrats could prevent the GOP from winning at least 45 percent of the vote nationally and a majority of the White vote, which means Republicans always have a fighting chance to win presidential and most Senate elections. Republicans successfully cast Democrats as too left for many voters before anyone had ever heard of Ocasio-Cortez, “defund the police” or wokeness.
Of course, candidates and tactics matter. It is obvious Biden should not promise to defund the police. But the clearest evidence we have is that messages, candidates and tactics don’t matter a lot — and we can’t be sure which ones will work or how much they will work ahead of time.
Democrats are muddling through, but they can’t change the fundamental dynamics. We have two parties about equal in size in a zero-sum fight for power.