Police forces in America have undergone heightened criticism and scrutiny in the last year or two following widespread protests of high-profile police shootings of black men. It has become fashionable in some conservative circles to claim that this wave of criticism has scared police from doing their jobs and to blame this criticism for increases in crime and perhaps even the unraveling of society.
This piece by Michael Barone is typical of the genre. Barone manages to blame “Black Lives Matter” for everything from recent ambushes of police officers to increases in homicides in major U.S. cities. This viewpoint believes there is nothing, or almost nothing, wrong with policing in America today, that Black Lives Matter was founded on false premises involving disproportionate police shootings, with a criminal poster child to boot, and that the critical response is leading police to pull back from their good crime-fighting work, ironically leading to an increase in the real problem of black-on-black crime. Black lives matter, indeed.
I am skeptical of many elements and implications of this narrative, though I am willing to follow data wherever it leads. There is enough suggestive data, and the implications, if true, are serious enough, that it is worth exploring. Overall, however, I believe the narrative involves cherry-picked examples and oversimplified correlations that spring from what are nonetheless valid concerns about the current discussion around policing.
The Facts About Killings of Police
After describing some of the worst things people associated with Black Lives Matter have said or chanted, many calling for the murder of police officers, Barone implicates the movement: “some people seem to be acting on that advice,” followed by a list of recent police killings, including several from August alone.
Such killings are terrible – but are activists responsible? Barone’s timeline is certainly questionable. The cited Minnesota State Fair chant occurred after all four of his examples of police deaths. Even if it caused nothing, that arguably makes such chanting even more indefensible – it certainly shows little concern for such deaths. Besides, it is only representative of a general tenor that has been prevalent for quite some time.
What evidence we should expect to see if the link is genuine? An increase in police being killed? The evidence suggests that so far this year such numbers remain among the lowest in decades.
Bill O’Reilly tried to implicate the movement on his show and was told that “there are fewer cops shot this year than last year. Are you willing to give Black Lives Matter credit for that?… Seventeen percent…. August was a bad month. In July there were none. Overall they are down. I don’t see an epidemic there.”
Any officer death is a tragedy, and if police protests were leading to an increase in people shooting police, that would be very concerning, both for the officers and their families, and for society at large. To strengthen the argument, one could present evidence that more police are being ambushed, as opposed to being shot by suspects with whom they are already engaged. To really strengthen the argument, one could present evidence that any of these recent killings were directly inspired by BLM, such that they likely would not have occurred without BLM, as similar killings did before them.
Thus far, however, I have seen no evidence in either case, and the only evidence suggests that killings of police officers remain among historical lows. (There may be those on the ‘police side’ who are focusing now on every new killing of an officer and committing some of the same logical fallacies of those who focus on every officer killing of an unarmed black man.)
It is true that the better parts of Black Lives Matter may not be doing enough to disassociate themselves from the worst things said by those sharing their slogans. While I saw nearly universal condemnation of the NYPD slaying last December from pretty much every person I had ever seen question anything about policing post-Ferguson, I have seen less more recently. Two recent high-profile defenses of BLM make many good points but say nothing about the frequency of excessive death-wishers in their ranks, even as they insist they are not anti-police.
I admit I find it difficult to have fair expectations; I have previously explored the economics that encourage “denunciation deficits.” While the situation is different enough to avoid too many parallels, the hastiness to link political rivals to the death of political allies reminds me not a little of the reverse attempt a few years ago to link Sarah Palin to the shooting of Congresswoman Giffords, which the right rightly derided. It would be just as wrong, though no one seems to be attempting it, to link any part of conservatism to Dylan Roof’s killing spree, which from what I understand was likely motivated by radical right-wing racialized websites. All of this makes me think of Tim Wises’s comments in White Like Me about the “white privilege” that allows whites to avoid having to answer for the bad things that white people do, while always expecting blacks to answer for any bad thing another black person does.
Still, I think there is a definite deficit here. But I’m not convinced any of it is to blame for recent officers being killed.
The Facts About Urban Homicides
Even if more officers aren’t being killed, more citizens are – at least in some places. Are protesters responsible for reversing two decades of crime decline? At first glance, there’s a strongly suggestive correlation. A New York Times article on recent trends includes this graphic: It’s hard not to miss that two of the top three cities – St. Louis and Baltimore – were also the two cities that saw the strongest reactions to killings by officers in the last year, as defined by any number of objective measures such as numbers of buildings burned down. The correlation seems too strong to be a coincidence, something to take very seriously.
But what exactly does it mean? Milwaukee’s numbers are worse than St. Louis and Baltimore, but no state of emergency was ever declared there. They did have large protests over a police killing – but that simply speaks to the sheer number of cities that have police killings. So did Cleveland – and they’re only up 16% (from 50 to 58). So did New York – and they’re only up 9%; cities jump up and down by those margins all the time.
There may be a case that riots cause temporary homicide increases – see also Cincinnati in 2001 – but it doesn’t seem clear that protests cause either. To strengthen the argument, one could, for example, compare major cities that experienced a given level of protest to ones that didn’t, but I haven’t seen any attempt to make correlations deeper than “hey look homicides are up in some places!” FiveThirtyEight crunched the numbers and found a 16% homicide increase in the top 60 US cities so far in 2015; slightly concerning, but too small and too early to conclude too much.
There’s also an inherent weakness in focusing on homicides. True, the data tends to be the most clearly and consistently defined and the hardest to manipulate, compared to other crimes. But homicides are relatively rare – occurring within an order of magnitude of 100 in most major US cities. They tend to occur between people who know each other, and as such are harder for police officers to prevent (not that innovative gang interventions and improved clearance rates can’t help a whole lot). It’s harder to tell if other more common crimes are increasing at similar rates.
There’s something to be said for the denunciation deficit here, too. It’s true that many activists seem more focused on blacks killed by police than the many more blacks killed by other blacks. There are reasons that make this imbalance more understandable. But the imbalance is often highlighted with severe criticism, in classic speck-in-your-eye style, from conservatives asking why activists don’t care about both when they themselves don’t seem to care much about either.
I am more optimistic. I do see imbalances, but I also see the goodness and sincerity of many of the people involved. I believe that sincerity will lead them to a better balance. I see signs that it is beginning to happen. And I think that grace and encouragement, rather than antagonistic criticism, is more likely to make that better.
The Facts About Police Response
I believe I have demonstrated that statistics about recent crimes are more complicated than they may appear. I think the same is true of the claims that protests and criticisms have caused police to “pull back” for fear of being punished for simply doing their jobs. Barone quotes a scholar about a “reluctance to act,” but quotes are hard to rely on. I found it amusing that the NYT quoted Ferguson mayor James Knowles saying “we barely pull anybody over anymore” while in the same article saying “People here say the police still treat residents suspiciously, still bark questions, still make arrests for what they consider trivial charges.” “Barely pulling anybody over” apparently means different things to different people.
The only statistic Barone notes is that Baltimore arrests plunged 60% after Freddie Gray. An extremely similar article in this genre by David French links to two WSJ articles with text that “arrests are down” in both Baltimore and St. Louis. Interestingly, the first notes that arrests had been trending down 22% from 2014 in Baltimore pre-Gray, but the three following weeks were down 40% from the past two years. The second article says St. Louis year-to-year arrests were down by a third between last August and November, though no comparison of the pre-Brown trend for 2014; the May 2015 article curiously contains no updated stats; it also says “arrests in Baltimore were down 56% in May compared with 2014,” – which is probably where Barone rounded up to his 60% number.
These statistics – a drop in arrests in both St. Louis and Baltimore immediately following their respective police killings – seem to be the only quantifiable statistics offered to support the “pull back” theory. Commentators across various sites are repeating those same statistics several months later without any indication of whether or not the trends have continued.
It’s also not clear what kinds of arrests are down. Arrests for homicides must account for a small percentage of overall arrests – especially considering dismal clearance rates. What percent of arrests generally come from traffic stops uncovering outstanding warrants related to previous traffic stops? Are those down by a third? Is this thought to be driving the homicide increase? What actual activity was reduced to cause the drop in arrests? Fewer arrests for outstanding warrants during traffic stops? Fewer traffic stops to begin with? Fewer attempts to find suspects identified by detectives? Fewer suspects identified by detectives to begin with? Fewer drug busts?
Some of those kinds of arrests might be less immediately linked to stopping violent crime than others. Baltimore has open data on arrests, so one could strengthen the argument for the “pull back” theory if they provided evidence that the arrest slowdown has continued, or if the kinds of arrests that are down can be argued to support the theory.
It seems reasonable to suppose that under the current atmosphere, some officers may be marginally more hesitant to stop a suspicious vehicle, marginally more hesitant to follow a potentially dangerous lead, marginally more hesitant to chase down a suspect who runs away. These outcomes could be troubling, and if more evidence is provided that these things are actually occurring, I will accept it. But so far I do not think a very strong case has been made.
Furthermore, there is one piece of evidence that I believe goes against the theory. If officers were marginally more hesitant to do various things, I would think that one of those things, if not the primary thing, given that it was in fact the catalyst for all of this, would be finding officers marginally more hesitant to shoot people. The data on police killings is complicated, but the best numbers I’ve seen showed roughly 1000 deaths per year pre-Brown, or roughly three per day. There are more official outlets tracking this year, with widely varying numbers, but they all still seem to be coalescing around the same average: three per day.
Let’s presume the vast majority of these are justified. In a post-Brown, post-Gray America, officers seem no less hesitant to pull triggers when they feel their lives are on the line, to say nothing of the continuing drip of questionable incidents. It may be possible to construct a theory that accounts for this – perhaps officers are less likely to shoot per dangerous situation, but emboldened criminals are creating more dangerous situations, perfectly canceling out any statistical change. But so far I have not seen a proponent account for this any more than I have seen them detail plausible relationships from arrest numbers. Instead, commentators just seem to be trading around the same old single statistics, implying their representation of a large picture that has not yet been demonstrated to actually exist.
A Valid Concern: Emotional Toll
If the evidence for protest-led violence is weak, why are many conservatives so adamant about it? I believe that it springs from a valid concern that is being filtered through a distorted understanding about policing and crime. The valid concern is that a lot of the current police criticism is coming in broad strokes full of ignorance and hate. This is inherently unhelpful for genuine police reform, but it also has had the effect, unappreciated by the larger population, of emotionally harming thousands of good police officers who feel that they haven’t done anything wrong yet are being attacked by people they’ve never met who have no idea what their jobs are actually like.
If this mental toll becomes too great, it may sap the motivation of many to continue policing altogether, and these are not the officers you want to lose, especially if you want genuine police reform. Unlike the rhetoric about a physical “war on police,” I believe the evidence is clear that this emotional war is very real, and while I haven’t seen strong evidence that this toll is causing widespread rank thinning, I’ve seen enough ancedotal evidence that the risk of it genuinely concerns me. I think activists would do well to consider these emotional effects of the overreach of the general rhetoric.
My theory is that those who are most concerned about this overreach are most likely to believe that it must be leading to bad things, and are primed to quickly accept any evidence that it’s already happening. These conclusions are heightened when viewed through an oversimplified, distorted understanding.
A Distorted Understanding: One-Dimension Policing
When judging the effects of the critical overreach, commentators tend to mirror the denial of their opponents. Just as the defenders of the best of Black Lives Matter make no mention of death wishes made under their banner, police defenders make no mention of even the possibility that some of these protests sprung from questionable deaths. They deny that the policing status quo had any problems that offered any legitimacy to the criticism at all. They depict a false choice that the existing way was better than this new way that is alleged to be resulting in the deaths of police and citizens alike.
It all seems to stem from a view of policing as a One-Dimensional Lever Of Activity. “Let’s say it used to be set at… LEVEL 8…. and even though police weren’t doing anything wrong, you don’t seem to like it set there… or, hey, even if a few police were doing something wrong at that level, well, this thing only goes up or down, so, OK, you wanna make our jobs harder, we can pull it back here to… say… LEVEL 5…. now you can’t complain about harassment or whatever… but just don’t come complaining to us when the crime gets out of control! See, look at these stats! It’s already happening!”
Barone says “the ‘broken windows’ policing BLM decries has saved thousands of black lives.” He thinks the lever went up to save black lives, and now it’s coming down, no longer saving them. But the link between crime and no-tolerance policing has been disputed for years. Criminologists have noted that cities that didn’t try New York City’s “tough policing” saw crime drops as big as the ones that did, suggesting that other nationwide factors may have been involved in the trends of the last few decades.
Even if no-tolerance policing does reduce some crime, it may be limited, and unsustainable, due to the way it tends to foster community resentment. Turn the lever too high, and you can lose effectiveness – if no one in the neighborhood likes you, they’re not helping you catch criminals, either. Barone seems unaware that the original “champion of broken windows policing,” former NYPD commissioner Bill Bratton himself, is now trying to replace that with “a new approach to community policing that attempts to repair the breach between police and citizens.”
The Good News
This means there is good news. Even if police are pulling back, and even if that is leading to some increases in crime, we don’t have to return to the status quo that led to widespread police resentment among urban minorities. Policing is not just a one-dimensional lever. You can fix overreach without hampering good police work and restoring community relations that help you work with the community to reduce crime, all at the same time. I believe it because I’ve read about it happening – in Cincinnati. In Los Angeles. It may be starting to happen in St. Louis.
It always happens when angry sides come together to listen to each other, to see good faith in the best of the other’s intentions, to admit the imperfections of our own, and to look for ways to work together. It won’t happen as long as we’re chasing every cherry-picked statistic to prove that the other side is totally responsible for the problem.