What the Data Say About Police

Roland Fryer, professor of economics at Harvard, wrote this op-ed for the WSJ, published yesterday.

In 2015, after watching Walter Scott get gunned down, on video, by a North Charleston, S.C., police officer, I set out on a mission to quantify racial differences in police use of force. To my dismay, this work has been widely misrepresented and misused by people on both sides of the ideological aisle.

I have not yet reviewed the full research article published in the Journal of Political Economy. So the WSJ op-ed is essentially the Cliff’s Notes version. Even so, there are some important take-aways here.

As we see so often in studies of race and criminal justice, raw numbers of racial “disparity” in the use of nonlethal force are very large, but they come way down when controlled for relevant factors. About 2/3 of the disparity vanishes when the controls are applied. Fryer notes, quite correctly, that the fact that the bias cup is 1/3 full means that use of his research to assert that systemic police bias is a myth is an incorrect use. The fact that the cup is 2/3 empty, though, is also a reminder that raw, uncontrolled numbers are essentially worthless. The common use of such numbers in public debate is equally wrong. As for whether the remaining correlation is proof of bias or the result of factors the researchers did not control for, Fryer can’t say. “As economists, we don’t get to label unexplained racial disparities ‘racism.’ ”

The current raging controversies involve lethal force, and here Fryer reports a negative result. “No matter how we analyzed the data, we found no racial differences in shootings overall, in any city in particular, or in any subset of the data.” Two other studies based on different data produced similar results.

The popular narrative that black men are in grave danger of being shot by police in circumstances where white men would not be is refuted by the data.

One more point is less well known but just as important. Fryer also studied the correlation between investigations of police departments and subsequent crime rates.

We conducted the first empirical examination of pattern-or-practice investigations. We found that investigations not preceded by viral incidents of deadly force, on average, reduced homicides and total felony crime. But for the five investigations that were preceded by a viral incident of deadly force, there was a stark increase in crime—893 more homicides and 33,472 more felonies than would have been expected with no investigation.

Now why would that be? The simple explanation is the right one.

The increases in crime coincide with an abrupt change in the quantity of policing activity. In Chicago alone after the killing of Laquan McDonald, the number of police-civilian interactions decreased by 90% in the month the investigation was announced.

The calls to abolish police departments are music to thugs’ ears. To safeguard the residents of our most depressed areas, we need more policing, not less. Jason Riley has this column today, quoting Franklin Zimring for an important point. I don’t often agree with Zimring, but he nails this one.

In his 2007 book, “The Great American Crime Decline,” Franklin Zimring describes violent crime as a “regressive tax whereby the poor pay much more” and observes that “because both victims and offenders are concentrated among the same disadvantaged populations, a major crime decline might produce a double benefit—fewer victims as well as fewer offenders arrested and punished for serious crimes.”

More policing and more effective policing means less crime, with potentially the greatest gains in the areas now suffering the worst crime rates, the poor areas. Severity and certainty of punishment both affect would-be criminals’ decisions to commit crimes, but certainty has the greater effect. Improved deterrence and the resulting reduced crime rates will therefore ultimately result in reduced incarceration.

Abolishing police forces, or even substantially cutting them back, would be a disaster, with the greatest harm befalling our poorest citizens. Fryer’s research, and that of others, confirms what common sense tells us.