Does funding police better serve public safety than defunding them?
Core argument: Data from 2013 to 2020 indicate that Black Americans were disproportionately affected by police violence, with higher rates of being killed by police, shot by police, and unarmed when shot than other groups.
From 2013 to 2020, data indicates that Black Americans were disproportionately affected by police violence, with higher rates of being murdered, shot, and unarmed when shot by police compared to other groups. This underscores the importance of addressing public safety concerns through strategic police funding rather than defunding. Economic implications suggest that well-funded police departments can implement better training and community engagement programs, potentially reducing incidents of violence and improving public trust. This approach aligns with Biden's stance on enhancing public safety by investing in law enforcement, which could lead to more effective crime prevention and a more stable social environment, ultimately benefiting economic growth and community well-being.
Ed Comment:Skip this. there is much better data on this than this article. I don’t want to be the person calculating the data. but here’s what can be gleaned from this data (given that I calculated it, put it in the database. 332mm total US population x (13.4% black alone +2.8% two race) = 54mm black (assuming all 2-race are black, which would be the upper bound) Murders per capita: Black: 61k/54mm = 1,130 murders per million; Not black = (155.3k-60.1k)/278mm = 340 murders per million; blacks are 1130/340 = 3.3x approximately more violent than not-blacks Shot by police per murdered: Black 2,498/60,980 = 4%; not black (9,942-2,498)/(155,253-60,980) = 7,444/94,273 = 8%; shot per murdered assumes murders is a surrogate for serious engagement with the police, and that murders is predominately by the same race Unarmed per shot: Black: 413/2498 = 16.5%; not black (1315-413)/(9942-2498) = 902/7,444 = 12.1% 16.5%/12.1% = 1.36 police are about a third more like to be shooting at an unarmed black suspect than an unarmed not-black suspect. 12 to 16% of those shot by police are unarmed seems surprisingly high to me. Given that blacks are approximately 3.3x more likely to be murderers than non-blacks, one would expect the police to respond more aggressively to blacks than whites. Yet shot by the police per approximate engagement with the police (as approximated by murders per capita) is half as much. although, it would appear that police assume black suspects are more likely to be armed—presumably approximately 3.3 times more likely based on the murders per capita. That’s taking the data further than it ought to be taken. Especially if there is better data.
"...Part of the generally higher level of mayhem in the United States is that we have relatively trigger-happy police departments that kill a lot of people. Even so, combining FBI Uniform Crime Reports data with information from Mapping Police Violence shows that good old-fashioned murder is even more common than police officer shootings and a lot more common than cops shooting an unarmed suspect. It’s also worth underscoring that while police violence against civilians disproportionately impacts African Americans, so does murder. Indeed, the racial gap in murder victimization rates is larger than the racial gap in police violence...."
The problem here is not the wrangling over whether or not it’s true that Sunrise tweets hurt moderate Democrats at the polls; the problem is that mainstream institutions like CAP and CBPP are playing footsie with an idea that’s substantively bad, while progressive activist groups like Sunrise adopt hardline anti-policing stances that have nothing to do with their nominal missions.
On a personal note, last July there was a scary gun battle right near my house that got a lot of media attention because it happened in a hotspot dining corridor, and Jim Acosta happened to be there to take a video of it. That same week, Amazon opened its brand new “just walk out” supermarket of the future, literally across the street from the shooting. Earlier in July there had been several shootings nearby, but for the past six months, there’s been a squad car parked right by the Amazon Fresh store. And no more shootings! I can’t say for sure why that particular block has been blessed with 24/7 police presence. Is it because of Jeff Bezos’ clout? Is it because the 14th Street restaurant owners generate a lot of tax revenue for the city, so MPD caters to their needs? Is it because Logan Circle homeowners like me have a lot of juice? I don’t know. But someone got the city to make a substantial investment in getting people to not shoot up that stretch of 14th Street, and I really appreciate it.
The fact that most people living in high-crime neighborhoods do not benefit from that level of responsiveness from their local government is a huge substantive problem.
Murder is a big deal in America
The United States is a very rich country that enjoys a very high standard of living. Indeed, I often think Americans who walk around the Nice-Looking Old Buildings districts of major European capitals end up underrating how wealthy the United States is compared to France or Italy.
But there are a few areas in which America falls short, and the greatly elevated level of violent death is one of the major ones.
Part of the generally higher level of mayhem in the United States is that we have relatively trigger-happy police departments that kill a lot of people. Even so, combining FBI Uniform Crime Reports data with information from Mapping Police Violence shows that good old-fashioned murder is even more common than police officer shootings and a lot more common than cops shooting an unarmed suspect.
It’s also worth underscoring that while police violence against civilians disproportionately impacts African Americans, so does murder. Indeed, the racial gap in murder victimization rates is larger than the racial gap in police violence.
Given American history, it’s not hard to see why “the police should act with more restraint” is coded as a racial justice cause whereas “violent crime is a really big problem” is not. But that is the outcome of a number of contingent historical processes and does not reflect an objective reality whereby crime is a problem facing white people and police misconduct is a problem facing Black people. Among Democrats, white people are less enthusiastic than Black or Latin people about the idea of having more cops in their neighborhood, probably for the very good reason that most white Democrats have less to fear from crime and less to gain from law enforcement.
Misconduct on the part of armed officers of the state has major societal implications beyond the incidents of misconduct themselves, which explains their prominence in our political discourse.
But the absolute numbers matter, too. A policy change that led to a 99.9 percent reduction in the rate at which unarmed people are killed by the police but also generated a 0.75 percent increase in the murder rate would generate a net loss of Black lives (you would, however, save white lives because police violence is less racially disproportionate than murder). Now is there really such a tradeoff in policymaking? I’m not sure. But it illustrates why the big gap in the quantities is such a big deal. To save large numbers of people by reducing police violence requires huge reductions in the rate at which it occurs. By contrast, small changes in the murder rate make a big difference, simply because murder is much more common.
Policing is a well-established deterrent to crime
Sherrilyn Ifill, president of the NAACP Legal Defense Fund, responded to Biden’s state of the union remarks by asking, “Where is the data to support this is ‘the answer’ to current public safety challenges? Do the highest funded depts have the lowest crime rates? Are those depts more effective in solving violent crimes?”
The cross-sectional question “do well-funded departments have more or less crime?” comes with a serious endogeneity problem. Hiring firefighters is a way to control wildfires. But we have more wildfires in states with high levels of spending on firefighters than we do in states with low levels of spending. That’s because wet states on the east coast don’t have wildfires, so we don’t invest in creating firefighting services that could fight them. The more prone your geography is to wildfires, the more money you spend on firefighting.
So scholars need to try to find more sophisticated causal strategies. Steven Mello looked at the inclusion of a program giving federal grants to cities to hire police officers in the 2009 American Recovery and Reinvestment Act and compared trends in cities whose applications were approved to those whose applications were denied. He writes that “police and crime follow similar trends in high and low scoring cities prior to 2009, but trends diverge as high scoring cities receive hiring grant funding.” Each additional officer generates about $350,000 worth of reduced crime victimization so “expanding the police force is easily cost-effective for the average city in my sample.”
One issue with Mello’s study is that since murder is reasonably rare (and was considerably rarer in 2009 than it is today), generating evidence for statistically significant impacts on the murder rate is challenging. He finds statistically significant declines in violent crime in general and a directional decline in murder, but one that doesn’t reach the statistical significance threshold.
Aaron Chalfin and Justin McCrary try to get around that problem by abandoning the natural experiment approach to look at a much bigger dataset of all variations in crime and police spending across cities. Basically, they are asking Ifill’s question but using advanced math to try to address the endogeneity. They can’t guarantee that they’ve eliminated all endogeneity bias with their efforts, but they find $1.67 cents in crime reduction benefits for every $1 invested in policing — in this case largely driven by a reduction in murders.
More recently, Chalfin (this time with Benjamin Hansen, Emily Weisburst, and Morgan Williams Jr.) finds that “each additional police officer abates approximately 0.1 homicides” and improves on prior scholarship by documenting the race-specific effect, which is “twice as large for Black versus white victims.”
This seems persuasive to me — having more police officers around reduces crime. Most people’s intuition that hiring more cops is a good way to improve public safety is correct, as is their intuition that people who are very averse to hiring more cops are probably just not that fired up about crime reduction as a policy goal.
That said, there are certainly other investments that work, too.
There are other things that reduce crime
Summer jobs programs for kids are a well-studied intervention that, as Charles Lehman explains in this Manhattan Institute policy brief, have a very solid track record of reducing crime.
Interestingly, these programs don’t seem to generate long-term benefits for the kids enrolled in terms of superior educational outcomes and long-term earnings. If a kid is enrolled in a summer job program, they are less likely to commit a crime while enrolled, and that’s about it. As Lehman argues, we should probably invest not only in summer job programs but in more research as to whether we can tweak the programs to deliver a wider range of benefits and/or to target them more narrowly where they do the most good. But it’s absolutely promising.
As previously discussed in a column on the opioid epidemic, Medicaid expansion appears to reduce crime, seemingly because it increases the availability of substance abuse disorder treatment.
Annalisa Packham and Jillian Carr find that distributing SNAP benefits more frequently reduces theft at grocery stores, but in a separate paper, they find that it also increases domestic abuse, so I’m not as enthusiastic about that one.
Reducing lead exposure very clearly reduces crime, but it does so with a large lead time. If we invest today in lead abatement, then kids born a year or two from now will commit fewer crimes when they are teenagers.
The main thing that I would note about all of this is that increasing police staffing in cities with average to above-average levels of violence clears an absolute cost-benefit hurdle and does so relatively easily. So if “let’s spend more on the police” is something that can be done on a bipartisan basis or that helps win lots of votes, then there’s simply no reason not to do it. Medicaid expansion, of course, is great. But the blue states have already done it, so there are limited returns.
Meanwhile, when we look at CBPP’s proposal for non-police approaches to public safety, I see a lot of ambiguity. They argue that we don’t necessarily need to be using armed, uniformed police officers to be doing so much traffic enforcement. I absolutely agree (see “Automate as Much Traffic Enforcement as Possible”), but in practice this is a form of increasing police manpower. The idea is that you are taking work off police departments’ plates, which frees up more officers to do policing. The same is true for the idea that we should have more civilian mental health crisis responders. Within a normal-person framework where cops patrolling the streets play an important role in reducing crime, hiring non-cops to do non-cop work is a way of enhancing police manpower. It’s only within the weird worldview of anti-police ideology that “we don’t necessarily need cops to be spending their time writing speeding tickets” becomes “we don’t necessarily need cops.” We need cops to stop rival gangs from shooting at each other and terrifying or maiming bystanders in the process!
But the only real tradeoff I see is around the lead issue, where there is simply a difference between investing in controlling crime now and investments in crime control that pay off over the long term. I am a huge proponent of lead cleanup and always encourage people to think more long-term. But I also challenge progressives to do some self-scrutiny here. The Bipartisan Infrastructure Bill that Biden signed features by far the most significant investment in lead abatement in American history, and progressives mostly downplayed it. In the end, I’m glad they came around and passed the bill. But there’s a difference between fighting enthusiastically for lead abatement (what Joe Biden did) and bringing lead up in a whataboutist way to argue against police funding and then not actually being passionate about lead abatement when a lead abatement bill is on the table.
You should fire bad cops
Someone is reading all this and banging their head against the wall saying, “GODAMMIT YGLESIAS IS TOTALLY IGNORING POLICE MISCONDUCT!!!!!”
But it’s crazy that we managed to get a place in the discourse where the conventional way to demonstrate your concern about police misconduct is to argue for arbitrary budget cuts or to suggest that normal, workaday non-misconduct is somehow unnecessary or abusive. If a bus driver got drunk, crashed the bus, and killed an innocent person, you wouldn’t say, “that goes to show we shouldn’t have buses.” You’d punish the bus driver. And if the Bus Driver Union had rules that made it impossible to hold abusive bus drivers accountable, you’d try to change the rules. You wouldn’t say, “these drivers are reckless and out of control so let’s run the buses less frequently.” If the bus is useful, you run the bus. If the driver is engaged in misconduct, you punish the driver. Derek Chauvin is in prison because he murdered George Floyd, which seems extremely appropriate to me.
In general, though, it is too difficult to discipline or fire cops for misconduct and too easy for bad cops to circulate from department to department. These are absolutely things that should be changed.
But if a manager decides to start being tougher on staff — asking more of them and firing a larger share of the worst performers — and also decides that prior recruiting practices haven’t been rigorous enough, are they somehow also going to promise payroll savings? Of course not. They’d need a larger budget.
My guess, though, is that the deep roots of progressive dysfunction on the topic of policing stem from the fact that there’s solid reason to think that police unions and police collective bargaining meaningfully increase misconduct. Adam Serwer did a great piece titled “Bust The Police Unions,” but unlike “defund the police,” that never became a viral meme on the left. Tackling police union power would, obviously, be difficult coalition politics for progressives since the idea has negative implications for other public sector unions. By contrast, “cut the police budget” is comfortable.
But that’s a sloppy way to think about policy. And while I’m glad that Biden and congressional Democrats haven’t embraced it, it’s a real problem that the institutions that are supposed to do policy analysis for progressives seem so twisted around on this.
Republicans are the party of austerity
Mainstream Democrats should not be approaching this issue from a position of defensiveness or as a question of sloganeering.
It was Bill Clinton and then-Senate Judiciary Committee chair Joe Biden who created the original federal police funding initiative as part of the 1994 crime bill. Republicans opposed that bill, and Donald Trump spent much of the 2020 campaign trolling Biden about it. Obama boosted federal police funding as part of ARRA in the face of near-unanimous Republican Party opposition. Every single Trump administration budget request proposed cutting police funding. All throughout 2020, Democrats and Republicans argued about fiscal support to state and local governments (i.e., police departments) with Republicans in opposition. Then in 2021, Democrats got the chance to deliver big checks to support the hiring of teachers, cops, and front-line public servants, and they did so in the face of unanimous opposition.
Rick Scott and the NRSC recently released an 11 Point Plan for America that calls for eliminating all federal programs that duplicate local government functions — i.e., they once again want to cut the federal police spending.
This all reflects the basic logic of the larger political dispute in the United States. Democrats believe that rich people are undertaxed and that this creates an opportunity to fund more generous public services. Republicans believe that rich people are overtaxed and that in order to reduce their tax burden we should cut back public services. Police departments are a public service that costs money. Republicans are willing to do things like have state legislatures pass symbolic bills telling cities not to cut police funding. But are Republicans willing to send money to low-income cities to help them build larger and more effective police departments? Of course not. Are they willing to have the federal government step up and fill the need? Again, obviously no. They will salute a Blue Lives Matter flag, but it’s a low-tax party first and foremost.
Democrats need to remember that and to say it — they are the party of public services, the party that will fund the police and expand Medicaid so addicts can get treatment and provide subsidized summer jobs for teens and clean up the lead. Democrats will do all the things.