Traffic stop data won't prove anything about race

By James Alan Fox, 04/20/99

In response to the recent news reports of discriminatory police practices on the roadsides, a bill is speeding its way through the Massachusetts Legislature to collect data on each traffic stop made by police in the state.

If this measure survives challenges from the law enforcement community, police officers would not only be required to record the race of drivers (and passengers) for every car stopped, but also detail information about the reasons for the stop and what transpired as a result.

Proponents of the bill insist that such data are essential to document the prevalence of race-based police stops. Moreover, it is hoped that the data collection process itself will deter officers from making patently racist traffic stops. Ironically, this could be true, if only because this process might deter the cops from stopping anyone.

Of course, some confirmation is sorely needed of the existence and extent of race-based traffic stops. However, it is doubtful that the particular strategy proposed will provide anything more than faulty data, and thus will be anything more than a massive time-waster.

The first and most basic problem is how to interpret the race tallies that would arise from this approach. Suppose, for example, that 20 percent of the traffic stops made by the Brookline Police involve black drivers. What kind of inference will this permit?

Do we compare this race breakdown with the racial composition of the Town of Brookline, or with the race characteristics of Brookline car owners? Should we not consider the population of surrounding cities and towns? Let's not rule out the number of Newton residents who drive up Beacon Street on their way into the Back Bay or of Boston residents who use Harvard Street to reach Watertown.

More fundamentally, the reasons or ''probable cause'' for traffic stops are as critical as the race of the motorist. The United States Supreme Court has prohibited traffic stops based on race, but not based on reasonable suspicion. How can we tease out differential rates of stoppage from the possibility of differences in driver behavior?

Proponents of the new legislation expect that the police questionnaire will help to understand the reasons for traffic stops. But if one distrusts the racial fairness of police officers, how can the responses of these same officers to a roadside survey be anymore trustworthy?

Ultimately, the issue is whether or not black and white drivers are equally likely to be detained for the same outward behavior - that is, when driving on the same street at the same speed at the same time in the same manner. The only way to tell is to experiment.

Experimental studies allow us to isolate the effect of one variable, like race of driver, from all other factors - those that might legitimately raise police suspicion. Simply, one could direct white and black motorists of the same age, gender, and auto type to drive the same route in the same fashion at the same time. The number of police stops of each race would then be recorded, and any difference would necessarily be attributable to race effects. We might also focus such an experiment in predominately white communities where the potential for racist policing may be the greatest. Of course, this undercover experiment would require the cooperation of leadership, so that warnings and tickets to these experimental ''confederates'' would be excused.

This is hardly a far-fetched idea, and is similar to the enforcement strategy used to identify stores that sell cigarettes to minors. Criminal justice policy has truly benefited from a wide range of experimental studies, such as experimental evidence concerning the benefits of mandatory arrest for domestic assault.

This may be the best, if not the only, valid strategy for monitoring police behavior. However, the hesitancy of political and police leadership to experiment in such a way is understandable. Yet, any alternative that does not separate race from behavior could do more harm than good.

We can all be pleased that Attorney General Janet Reno and a number of police executives have taken the issue of race and policing head on. Let's respond with the sensible approaches for policing the police, rather than burdening them with meaningless and misleading paperwork.

James Alan Fox is the dean of the College of Criminal Justice at Northeastern University.

This story ran on page A19 of the Boston Globe on 04/20/99.