A member of the Baltimore Police Department removes crime scene tape from a corner where a victim of a shooting was discovered, on July 30, 2015. (Patrick Semansky/AP)
Opinion writer

Here at The Watch, we’ve pointed out a couple of times the interesting overlap between police abuse and violent crime. In many of the same cities where we’ve seen a surge in homicides over the past few years — such as Baltimore, Cleveland, Chicago and the St. Louis area — we’ve also see compelling evidence of widespread police harassment, bias, abuse and corruption. In Baltimore, for example, headlines about police corruption compete with headlines about the latest homicide.

This lends at least some anecdotal support to what we might call the alternative Ferguson effect — the idea that when police are seen as abusive, corrupt and antagonistic, they’re less trusted by the communities they serve. There is less cooperation between police and citizens. And perhaps more importantly, when people don’t trust the police to resolve disputes, they tend to take matters into their own hands.

Now, a new study by the National Institute for Justice has found some empirical heft to support this theory.

Much speculation, but little empirical research, exists regarding the mechanisms linking police legitimacy, as reflected in public attitudes and perceptions, to crime rates. One study found a relationship between Google searches for phrases related to police violence (e.g., Black Lives Matter, police shootings, police brutality) and increases in monthly violent and homicide rates in U.S. cities between July 2014 and June 2016 (Gross and Mann 2017). Although this is the first study to document variation across cities in public concern with police use of force after Ferguson, it does not specify how that concern provoked increased violence.

This jibes with polling showing that whites (who have lower homicide rates) have a lot more trust in police than do blacks (who have higher homicide rates). The Atlantic reported on such polling last year.

Only a third of blacks say police do an “excellent or good job” in treating racial or ethnic groups equally, compared with three quarters of whites. Just under a quarter of blacks say the local police does “only a fair job” and about forty percent say they “do a poor job.”

And it’s not just the initial incidents; blacks report a lack of confidence in the ability of law-enforcement agencies to police themselves. Only a third of blacks in the Pew report say police are doing a good or excellent job in “holding officers accountable when misconduct occurs.” But 70 percent of whites say the same. This lack of confidence seems to be exacerbating the historical mistrust experienced by black residents. About 80 percent of respondents in the Pew report believe “these deaths signal a larger problem between police and the black community.” Just 54 percent of whites and 66 percent of Latinos say the same.

Neither of the studies links this lack of confidence to upticks in crime in the areas in which it’s concentrated. But they do suggest that well-publicized incidents are reinforcing historic distrust, and that this deters citizens from contacting police to report crimes. And given the reliance of police on their relationship with local communities, they raise the possibility that the problem in cities grappling with rising crime may be less that officers subjected to scrutiny are backing off from proactive policing, than that communities subjected to repeated abuses are losing faith in those who are sworn to protect them.

The new NIJ study notes another study that looked at the rate at which residents of Milwaukee called police for help after a high-profile incident of police abuse. As the authors point out, this probably is a better indicator of public trust in police than a pollster simply asking, “Do you trust police?”

Desmond et al. (2016) found that calls for police service dropped significantly after the police beat and badly injured an unarmed black man in Milwaukee; the decline was especially large in predominantly black neighborhoods. Their study also produced some evidence that widely publicized violent encounters between police and African-Americans elsewhere in the nation also reduced calls for service in Milwaukee. The researchers concluded that the decrease in calls for police service undermined public safety. In line with this interpretation, they reported that homicides, which are not likely to be affected by a reduction in calls for service, rose sharply after the police beating in Milwaukee.

One advantage of the Desmond et al. (2016) study of the social response to perceived police misconduct is its reliance on behavioral indicators rather than opinion surveys (or internet searches) to measure police-community relations. … “When it comes to relying on and cooperating with the police, what one does might not resemble what one says one will do. … People’s attitudes toward the criminal justice system might be poor, even misleading, indicators of their real-life dealings with the police.”

Of course, it’s likely that there’s more than one cause for the surge in homicides in many big cities. So the study also looks at some other possibilities. One likely culprit is the rising demand for heroin and other opioids, which likely caused a shift in black markets and, as a result, the violence that comes with turf wars. This makes some sense. We also saw a major surge in violent crime in the late 1980s and early 1990s as the popularity of crack cocaine disrupted the established black markets.

The study’s authors also look at “de-policing,” more commonly known as the Ferguson effect. This is the idea that protests, viral cellphone videos and general public criticism of police has made officers less likely to engage with citizens, which results in higher rates of violent crime. There is definitely evidence that police have become less proactive and that they are making fewer arrests. There’s also evidence that de-policing is more pronounced in nonwhite communities. But the new NIJ study findslittle correlation between arrest rates and homicide rates. From the study:

In summary, the UCR arrest data for large cities provide ambiguous support, at best, for the de-policing version of the Ferguson effect. Arrests for both serious and minor offenses decreased in 2015 when homicides were increasing, but they also fell several years earlier, when homicides were decreasing.

Frankly, if de-policing means fewer arrests for petty crimes such as disorderly conduct and drug possession, with no discernible overall effect on public safety, maybe that’s a good thing. The study also finds that the rate at which police in big cities are closing homicide cases is falling. That is, the cops are solving a smaller percentage of murders. Perhaps less police harassment over the little stuff — which also means less opportunity for escalation and conflict — will result in more cooperation and trust in police when there are serious crimes that need to be solved.