“One of the great mistakes is to judge policies and programs by their intentions rather than their results,” economist Milton Friedman said in 1975. And one of the main problems with contemporary criminal-justice reform efforts is their failure to pay attention to the facts of criminal offending.

Three such facts should guide effective reforms. First, crime is highly concentrated among active offenders—most thefts, robberies, burglaries, and shootings are perpetrated by a small fraction of chronic offenders. Second, individuals don’t specialize in specific offense patterns; rather, active criminals engage in a “cafeteria style” of offending, choosing offenses from a menu of options. Third, the most active offenders are embedded within dense criminal networks. These three facts mean that the intensity of crime is highly nonlinear and that changes in crime rates are largely driven by individuals ranking in the top 10 percent of the criminal-offending population.

Criminal offending and the probability of detection and facing consequences involve a repeated set of interactions. As a result, across-the-board changes in the consequences doled out for criminal actions can generate big shifts in the quantity of lawless behavior. This logic can be explained through the idea of gambler’s ruin, which Nassim Nicholas Taleb discusses in his 2007 book The Black Swan. Repeated plays of a hand of cards, or any other game, in which, with each sequence, one faces even a slightly larger probability of losing will eventually bring total financial ruin. If sanctions produce some deterrence or incapacitation of crime, then small changes in those consequences for active repeat offenders will yield changes in the crime rate. The networked nature of criminal behavior also means that changes in sanctions among repeat offenders will have knock-on effects. These can involve (if sanctions are relaxed) a spreading sense that consequences are no longer as certain or severe for illegal behaviors; increasing acts of retaliation among disputing criminal factions; and a general sense within a community that chronic offenders are not being held accountable and that self-protection is essential. Simply put, the behavior of the “power few” of criminal offenders—whether they’re left undeterred or subjected to strong social controls—will have a remarkably strong influence on the overall crime rate.

In the case of criminal-justice reform, policies that marginally lower the risk of sanctions will generate major effects when implemented across the board. Conversely, programs or policies that target low-risk or infrequent offenders will have little relevance to the problem of serious crime and crime rates. Reformers who seek to scale back the criminal-justice system need to understand the nonlinear distribution of criminal offending and focus on those at the upper end of the distribution.

Attempts to scale up programs that are inconsistent with these realities run the risk of generating harmful effects. They also waste an opportunity to learn how to prevent crime while minimizing the footprint of the criminal-justice system. When evidence emerges that reforms are not having the intended effect, they should be redesigned or abandoned.

We have good evidence for programs that are consistent with the facts of criminal offending, work effectively to prevent crime, and lessen the impositions of the criminal-justice system. Unfortunately, many reformers have ignored this evidence. The resulting rise in serious crime and violence in America would have been preventable had the policy community paid more attention to what works and avoided wishful thinking.

Photo: Yalonda M. James / San Francisco Chronicle via AP


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next