Like medicine, policing has been transformed in recent decades by the use of metrics. Here too the stakes are high: the fate of cities rests in no small part on the public’s perception of its safety, and mayors often stake their reelection on their ability to control crime or to bring down the crime rate. When the public and its politicians think of public safety, they think of the police, who are held responsible for the level of crime. However, like health and its relationship to the medical system, or education and its relationship to the school system, public safety is only partially dependent on the effectiveness of the police. It depends in part on other elements of the justice system: on the public prosecutors, the judiciary, and the penal and parole systems. It depends in good part on the propensity of the local population to engage in criminal activity, and that in turn depends on broader economic, ethnic, and cultural factors.1 And public safety also depends on the ease of committing crime. Some of the decline in crime in recent decades is a product of private actions by property owners. The opportunity for car theft, burglary, and other crimes has been radically reduced by defensive measures undertaken by millions of private individuals, whose acquisition of improved car alarms and home alarms has made these crimes more difficult. In addition, there are about one million people employed by private security firms in the United States.
Violent crime has fallen in the United States since the early 1990s. Rightly or wrongly, much of that decline is commonly attributed to changes in policing. And the major change in policing has involved the increased use of metrics, above all in the form of Compstat. Here is a case where diagnostic metrics for internal use have proved genuinely useful. But then again, the use of publicly released metrics to bolster the reputations of politicians and police chiefs has also created incentives for gaming and fudging the numbers, and for counterproductive diversion of effort.
Compstat (which originally stood for “computer statistics”) is a crime analysis and accountability system, first developed by the New York Police Department in 1994. Pioneered under Police Commissioner William J. Bratton, Compstat uses Geographical Information Systems (GIS) to track the incidence of crime. It involves the collection, analysis, and mapping of crime data in a rapid time-frame to discover crime patterns, as well as entailing weekly meetings at which police managers are held accountable for the results in their precinct. The data are used to pinpoint hot spots in which crime is concentrated, and to deploy police resources accordingly. In the decades since it was rolled out in New York, some variation of Compstat has been adopted in many large American cities.2 Compstat does seem to have contributed to the decline in reported crime—and indeed, to the decline of crime itself.
Yet in city after city, there have been questions about the accuracy and reliability of crime statistics. Insofar as Compstat is a system of informational and indicative metrics, it seems genuinely useful. But when the mayor pressures the top brass to show improvements in the overall numbers, and that pressure, in turn, is placed on the district commanders who are led to believe that their career advancement depends on a steady diminution in crime, the message sometimes heard by lower-level police officers is that they will be penalized for an increase in reported crime. And that creates pressures for fudging the numbers.
Such problems preceded the rise of Compstat and exist independent of it. In 1976 the social psychologist Donald T. Campbell (of Campbell’s Law, see chapter 1) noted that President Richard Nixon’s declared crackdown on crime “had as its main effect the corruption of crime-rate indicators, achieved through underrecording and downgrading the crimes to less serious classifications.”3 And that continues. The most widely reported metric of crime is the Federal Bureau of Investigation’s Uniform Crime Report. Based upon the reports from each city, the FBI compiles data on four major violent crimes (homicide, rape, aggravated assault, and robbery) and four major property crimes (burglary, theft, motor vehicle theft, and arson). Less significant crimes are not included in the index. The index is widely publicized and is regarded as a crime report card. When the crime rate goes down, elected officials tout their success. When the crime index goes up, the politicians are criticized by their rivals. The politicians, in turn, put pressure on their police chiefs to reduce the crime rate, who in turn put pressure on those below them in the police hierarchy.
All of this creates tremendous temptations to demonstrate progress in reducing crime by massaging the figures. As one Chicago detective explained,
“It’s so easy.” First, the responding officer can intentionally misclassify a case or alter the narrative to record a lesser charge. A house break-in becomes “trespassing”; a garage break-in becomes “criminal damage to property”; a theft becomes “lost property.”4
In each of these cases, what had been a major offense becomes a minor crime, not reflected in the FBI Uniform Crime Report. The temptations to understate crimes is sufficiently great that the New York Police Department devotes substantial resources to auditing the reports it receives, and to punishing officers found to have misreported.5 But not every police force has the resources—or the will—to create these countervailing forces.
Nor is the problem confined to the United States. In London, England, the Mayor’s Office for Policing and Crime set as a performance target a 20 percent reduction in crime. The target was passed down the chain of command, from the police commissioner to the constables on the beat, whose chances for promotion were linked to hitting the 20 percent target. In 2013, a whistle-blower from the London police force told a parliamentary committee that massaging statistics had become “an ingrained part of policing culture”: serious crimes such as robbery were downgraded to “theft snatch,” and rapes were often underreported so as to hit performance targets. As a retired detective chief superintendent put it, “When targets are set by offices such as the Mayor’s Office for Policing and Crime, what they think they are asking for are 20% fewer victims. That translates into ‘record 20% fewer crimes’ as far as … senior officers are concerned.” Such underreporting and downgrading of crimes “are common knowledge at every level in every police force within England and Wales,” he added. Other experts explained the various techniques for improving the performance metrics: choosing not to believe complainants; recording multiple incidents in the same area as a single crime; and downgrading incidents to less serious crimes.6
Another temptation is perhaps more troubling still. It involves an additional key metric of police success: arrest statistics as a purported measure of effectiveness. Ed Burns, a former Baltimore police detective for the homicide and narcotics divisions—best known as co-creator of the HBO series The Wire—has described the process of “juking the stats,” by which police officials could orient the activity of the department toward seemingly impressive outcomes. As a detective in the narcotics division, Burns sought to meticulously build a case against top drug lords. But his superiors were uninterested in that prospect, which was consuming manpower and would take years to produce an arrest. They were interested in enhancing the metrics, and since arresting five teenagers a day selling drugs on street corners yielded better statistics than arresting a drug king-pin after a multiyear investigation, they favored the course that quickly produced the higher numbers. From their point of view—and from the point of view of the politicians to whom they reported—every arrest was of the same value. The course of action that produced the best performance indicators did little to diminish the sale of narcotics.7 When every unit is of equal weight, the temptation is to go after the easiest cases.8 In Britain, this process of directing police resources at easier-to-solve crimes in order to boost detection rates is known as “skewing.”
Metrics, then, have played a useful role in policing. But the attempt to use metrics as a basis of reward and punishment can lead to metrics that are less reliable and even counterproductive.