No ordinary man could be such a fool

From Broken Windows Works by Matt DeLisi.

Broken Windows policing receives credit—rightly—for being part of the crime turnaround that saved New York and other cities. The theory, originating with George L. Kelling and James Q. Wilson, argued that tolerating too much local disorder created a climate in which criminal behavior, including serious crimes, would become more likely, since criminals would sense that public norms and vigilance were weak. In practice, this meant that police should crack down on so-called low-level offenses. When the NYPD started doing this during the 1990s, aided by the Compstat crime-mapping and accountability system, it became clear that the low-level offenders were also often wanted for more serious felonies. In short, there was a continuity of bad behavior, just as Broken Windows suggested, and order-maintenance policing, as the Broken Windows approach is also called, reduced disorder and led to the apprehension of many serious criminals, helping reinvigorate formerly-troubled neighborhoods. The last quarter-century in New York offers a powerful case for the theory’s accuracy.

Now, a professor says that Broken Windows theory is bunk. Writing in the New York Daily News, Northeastern University professor Dan O’Brien, who conducted a series of studies along with several colleagues, insists that Broken Windows “is inaccurate and that we as a nation need to move away from the policing strategies that it inspired.” O’Brien, et al. found a small but significant association suggesting that areas with greater social disorder had greater crime—but, they concluded, once other control variables were considered, that association became insignificant. Thus, the foundation on which Broken Windows is built “doesn’t hold up to scrutiny.”

Like too much scholarly work in academic criminology, however, the new study offers more sophistry than sense. The scholars’ meta-analysis—that is, a study of a collection of studies—purports to disprove Broken Windows, but in fact more often simply restates its core principles. Though “disorder might not elicit crime,” the authors claim, they agree that “there are certain types of disorder that can create ecological advantages for criminal activity.” The creation of such environments is exactly how Kelling and Wilson described the process of neighborhood disorder—absent community control—leading to the promotion of antisocial behavior. “Disorder does not encourage crime,” O’Brien and his coauthors go on, “but makes it easier to commit crimes.” This, too, merely paraphrases Kelling and Wilson, who wrote in 1982 of the “folk wisdom that happens to be a correct generalization—namely, that serious street crime flourishes in areas in which disorderly behavior goes unchecked.”

Broken Windows has always been disfavored on the left despite the correlation between the introduction of Broken Windows policing policies at the height of the crime wave in the 1990s and a subsequent sustained fall in crime to levels now not seen since the 1960s when the nation was smaller, more homogenous socially more equal, and in many ways less complex.

It is not only a fair statement that correlation is not causation. It is actually a rock solid truth. Correlation only supports the possibility of causation but has no predictive power as to actual causation.

I agree with DeLisi that there is much compelling evidence supporting the efficacy of Broken Windows policies and the net benefit to society of Broken Windows policing, especially for those in the most marginalized communities.

However, we can't just dismiss criticisms as without plausible merit. No one anticipated the rise of crime in the 1960s and no one accurately forecasted the sustained decline since the peak in the 1990s. There is much we do not know.

In addition, implementation of the philosophy of Broken Windows policing has taken many different forms in many different places. To get to some degree of confidence we need to see whether the same treatment always produces the same results. When the treatment is actually always a different combination of things, it becomes harder to reach the level of proof we are seeking.

Broken Windows policing is much more a philosophical way of looking at cause and consequences and so appropriately the treatment should always be specific to the local circumstances. And if it is always materially different, in the purest of technical terms, it becomes hard to "prove."

However, the fact that the implementation of the same philosophy seems to usually achieve similar types of beneficial outcomes is powerful evidence on its own that serves as its own critique of the cavilling and jesuitical nit-picking of the ideologically motivated.

I think one of the more powerful pieces of evidence supporting the power and efficacy of Broken Windows policing was what we saw with the Ferguson Effect. In 2014, in Ferguson, Missouri, Michael Brown was shot and killed while attacking a police officer. Through many activist organizations, his death was used to advocate in many cities and counties to abandon order-maintenance policing (Broken Windows).

Pretty consistently, wherever Broken Windows retrenched or was abandoned there was in the year following, a rise in crime. So while the overall decline in crime continued across the nation, in select areas, there were spikes upwards. This was the Ferguson Effect.

While the usual academic suspects who oppose Broken Windows did much research and found no evidence of a Ferguson Effect, people living in the real world could see it with reasonable clarity. And now that many city police departments are quietly reinstating order-maintenance policing, crime rates are beginning to fall again in those cities.

Heather Mac Donald has done much first class reporting on the phenomenon, including her book The War On Cops documenting some of the consequences of the Ferguson Effect.

I think the evidence for the benefits and effectiveness of Broken Windows policing in all its many manifestations is quite compelling. The fact that it is not easily provable because it has all the dynamism and variance of the real world does not change the fact that when you introduce well-designed Broken Windows policing crime goes down, when you take it away, crime goes up again, and when you reinstate it on the sly, crime goes down again.

In regard to the academic critics who claim Broken Window policing philosophy does not work, George Orwell's words come to mind.

One has to belong to the intelligentsia to believe things like that: no ordinary man could be such a fool.