Matt Yglesias is lying about Defund the Police

While many centrists have spent time derailing the “Defund the Police” slogan as bad messaging, in a recent blog post Matt Yglesias claimed that it wasn’t the slogan that was bad, it was the idea itself. I welcome his new focus on the actual policy, instead of trite lecturing at activists about how to best do activism. But I do not believe Yglesias seriously engaged with the topic, but rather cherry picked evidence to support his claim. I read his piece in detail, and below I explain what he gets wrong.

First, let’s be clear on what it means to “Defund the Police.” Rayshawn Ray, a sociology professor from the University of Maryland and fellow at the Brookings Institute, says that “‘Defund the police’ means reallocating or redirecting funding away from the police department to other government agencies funded by the local municipality.” This seems to be the widely accepted definition.

To reiterate, defunding the police simply means taking money and resources away from today’s police departments and moving those same resources towards other public institutions. It is certainly not the “austerity” that Yglesias originally asserted.

Thus, if one accepts this definition of “Defunding the Police” then to prove that it is a bad idea, you would need evidence showing that money allocated to the police is better spent than on alternative public services. I followed every citation in the Yglesias piece and none of them provided this sort of evidence.

Instead Yglesias simply tries to prove that more police = less crime, he writes:

“And the evidence across a large range of studies is that having more officers on patrol leads to less crime, including violent crime.”

The citation he uses for this claim links to another article he wrote for Vox in June where he cites a few primary sources. These studies are convincing and show a relatively strong trend that more police patrol means fewer reported crimes.

Yet, “reported” is carrying a lot of weight here. First, because of their power, crimes committed by police are rarely reported and thus unaccounted. Second, evidence shows that people who witness violence committed by police are less likely to report crimes to the police. Thus, if we increase police presence, and the likelihood of police violence, then we are actually increasing the chances that a person is a victim of police violence, while decreasing the chance that crimes are reported.

These facts together point at the flaws of studies evaluating police presence by measuring the amount of crime reported. Still, after examining the evidence, I think Yglesias is probably correct that there is a causal relationship between police patrol and crime in the short run.

Yet, this is not a concession that we should not defund the police. When assessing the overall impact of a reform we must evaluate it on both the pros and the cons. For the reform to be “good” then the pros need to outweigh the cons. Yglesias never mentions the downsides of increasing police. More police almost certainly means more victims of police brutality (one-third of Americans killed by strangers are killed by police) and perhaps infringement on people’s personal freedoms. Do we want to live in a society where armed agents of the state are constantly patrolling our neighborhoods? Do the marginal reductions in crime justify potential increases in police violence? A person actually interested in doing a cost benefit analysis would have included an analysis of the costs.

Speaking of costs, Yglesias also completely ignores the opportunity cost of money spent on the police. Every dollar spent on the police is money that cannot be spent on other public services. None of the studies he cites compare policing to other alternatives to reducing crime. In fact, his main focus is on a 2018 study that compares localities that received free federal money to spend on policing and those that did not. Free federal money also assists the local economy and a healthy economy reduces crime, and thus could be a confounding variable.

Of course, crime isn’t the only measure that matters in policy evaluation, we should measure policies on a range of measures that summarize people’s overall utility gains. Since measuring happiness can be tricky, one way economists examine the overall impact of a policy is the social returns — for every $1 you invest, how much you save later. While Yglesias does not mention social returns in this blog post, he did mention in a Vox article earlier this year that the social returns on policing were $1.63.

This sounds great, if you ignore areas where there are higher returns. Studies have shown that investing in early childhood education yields social returns between $4 and $12. A 2016 study led by the World Health Organization found the social returns on investments in mental health to be $4. Sure, I’m cherry picking results, but that’s exactly what policymakers should be doing, finding the social programs with the highest returns and maximizing their investment. There are other areas of potential investment where research is less clear, but we should not default to status quo bias and spend that money on police simply because that is where it is already being spent.

Local governments cannot control federal or state budgets and thus smart legislators should spend their funds on the institutions with the highest social returns, the evidence is pretty clear that that is not policing. Thus, true policy wonks would advocate for local legislatures to shift money from policing to policies with higher social returns….sound familiar?

And I think that is the crux of the argument. The evidence we do have shows that money is better spent in other places than on policing. But of course the $1.63 Yglesias cites is not linear. Money spent on policing, like most things, has diminishing returns. So the first dollars spent on police likely have much higher returns than a $1.63.

So if you’re trying to maximize returns, where would the first dollars on policing go? Almost certainly to crime investigations and emergency responders, not patrol. And that brings me to the next point Yglesias gets wrong about “Defunding the Police,” it’s not about completely stripping funds from all law enforcement, rather the current police institution. I think Yglesias knows this but he still chooses to make defunding the police synonymous with just letting people commit crimes without consequence.

The Defund the Police’ movement has been pretty clear that defunding the police does not equate to eliminating all the public services they currently provide. Sure, we should end armed agents of the state patrolling our communities, but for many current police responsibilities it means moving the job and resources to a more specialized force. Despite this, Yglesias takes time to show that more people trying to solve crimes means more crimes being solved as if that is evidence that defunding the police is bad. This really shows how deeply he misunderstands the demands of Defund the Police activists.

In other areas, Yglesias further demonstrates his misunderstanding. I think this full quote is worth including.

“But the abolitionists were very successful in owning the brand of “people who care a lot about racial justice.” And most liberals care a lot about racial justice. So the thing to do if you don’t want to be a contentious internet figure is to take some sensible liberal idea and say that’s a form of defunding police.

Many people have, for example, been advocating for creating a civilian corps of mental health crisis responders who could take over certain activities from the police.

That’s a perfectly reasonable idea, especially for larger cities that can logistically maintain rapid dispatch capabilities for multiple overlapping services.”

He’s unclear here, but it seems as if Yglesias is suggesting that moving money from police to a mental health team is not defunding the police. That’s exactly what defunding the police is, moving resources from the police to other public institutions. This can happen in other ways such as shifting resources from police to an automated traffic system, decriminalizing drugs, and removing police from our schools. Yglesias does not discuss these ideas.

Eventually Yglesias pivots to discussing his own preferred policy, increasing the amount of officers, paying them more, and a series of reforms that he claims would reduce police brutality. He offers some vague proposals:

“We need policy reforms that address the actual problem here. That means operational changes to the rules about how cops are supposed to behave. It means structural changes to the job security of officers with spotty records. And it means a sustained ongoing effort to dislodge the warrior cop ideology and replace it with an ethic of public service.”

Ironically, these ideas are more slogans than actual policy recommendations. How do policymakers “dislodge the warrior cop ideology?” When it comes to more concrete reforms, Defund the Police activists agree that these changes are better than nothing, but they certainly are not sufficient and are often used as distractors to avoid addressing the core of the problem.

Later, Yglesias claims that defunding the police would not have prevented George Floyd’s death but suggests that his proposed reforms would have, he writes:

What drives me a little bit nuts about the liberal sanewashing of “defund police” into this idea about mental health services, is that while it’s true cities should provide better mental health services, that would not have helped George Floyd at all.……The fact that defund police activists managed to draw attention away from the idea of changes that would have averted Floyd’s death to their own project is an impressive triumph on their part, not some kind of tactical failure.

First, Yglesias is wrong to assume that any measure of police reform would have prevented Floyd’s death. Eric Garner was killed in a chokehold in New York after they were banned. Just because you have rules doesn’t mean that police will follow them. After all, as Yglesias himself notes, accountability for police is nearly non-existent. Second, it seems much more likely that George Floyd would be alive today if we had defunded the police. Derek Chauvin was Floyds judge, jury, and executioner. A more human system would not have people detained by armed agents of the state before there is any proof of a crime.

Finally, while Yglesias primarily stays focused on his perceived merits of the policy, he does stray into polling in an attempt to assure us that Black Americans agree with him. He cites a Gallup poll that shows that just 22% of Black Americans support police abolition. But in that exact same poll there is a separate question that phrases the policy behind the Defund the Police movement, much better. The question asks respondents if they support “reducing the budgets of police departments and shifting the money to social programs.” 70% percent of Black Americans said yes.

In fact, 47% of all Americans agreed with the statement. Just a few percentage points away from a majority. The activism seems to be working and working very well.

But people like Yglesias could do real damage to the cause, just 5% of Republicans currently support Defunding the Police. It is going to take nearly all Democrats to get the majority of Americans on board, and Yglesias, along with most of his readers, are Democrats.

Yglesias’s readers should be aware that he created a hypothetical scenario and then he argued with himself. Of course, he won. He either does not understand that defunding the police means moving money to other, more efficient, public institutions, or chooses to purposefully mischaracterize it.

But for those of us who care about having honest policy discussions more than winning Twitter fights, we should engage with the Defund the Police movement as it is, not whatever Matt Yglesias claims it to be.