What Drives Racial Differences in Speeding Tickets?

The Issue:

Blacks and Latinos are significantly more likely to come into contact with the police and the broader criminal justice system and generally experience worse outcomes than whites once sanctioned. Significant research has focused on how to determine whether these observed disparities are the result of discrimination on the basis of race. But there is little empirical research on how to determine whether any observed police discrimination is the product of widespread officer behavior or driven by the actions of a few "bad apples." Our analysis of traffic policing by the Florida Highway Patrol provides us a window to observe the individual ticketing practices of agents and whether it varies significantly by the race of the driver stopped. From the perspective of police departments and policy makers, identifying the concentration of misbehavior across officers is paramount for designing the appropriate policy response. If misbehavior is widespread, targeting the worst offenders may not lead to significant direct benefits. If behavior is concentrated among a small set of officers, interventions that apply agency-wide may be relatively ineffective and might needlessly sacrifice public safety effectiveness relative to a more targeted policy.

Minority drivers are more likely to encounter officers that offer no leniency to drivers of any race.

The Facts:

• In many states, the fine schedule for speeding jumps at particular speeds and officers have some latitude to grant leniency when writing tickets. For example, a citation for driving ten miles per hour (MPH) over the posted limit may carry a significantly higher fine than a citation for driving nine MPH over the limit. To avoid issuing a higher fine, officers regularly engage in the practice of "discounting" a driver's ticket by writing down a lower charged speed. Using administrative traffic records from Florida, where drivers face a roughly $75 higher ticket at 10 MPH over, we document that over 30 percent of drivers stopped by the Florida Highway Patrol are issued tickets for driving exactly 9 MPH over the posted speed limit, while less than one percent of tickets are written for driving 8 or 10 MPH over the limit. The distribution of ticketed speeds strongly suggests that officers systematically exercise discretion, commonly discounting tickets as a form of lenience. 
• We find that a significantly higher share of tickets for white drivers are issued at a speed that leads to lower fines compared to minority drivers. While drivers of all races benefit from leniency to some degree, the share of white drivers who benefit is higher: 34 percent of tickets for white drivers are issued for nine MPH over the speed limit, and this figure is only 25 percent for minority drivers (see chart). As in all studies where the goal is to test for the presence of discrimination, an important challenge is disentangling the extent to which the observed differences in treatment can be explained by racial differences in offending. In our setting, it may be the case that minority drivers are being stopped at faster speeds and are therefore less likely to receive a break from the ticketing officer. 
• There is some evidence that minority drivers who are stopped and ticketed are driving faster than white drivers who are stopped and ticketed. But the speed difference is small relative to the difference in lenience across groups. In our analysis of all traffic citations issued in Florida for the years 2005-2015 we have information on the speed that the officer charged on the citation but we do not see the original speed of the stop. To address this concern, we leverage the fact that a significant share of officers never give drivers breaks and effectively use these officers as speedometers for the underlying distributions of stopped speeds. While Black and Hispanic drivers are stopped at slightly faster speeds than white drivers (roughly one MPH greater), this gap is not sufficient to explain the disparity in lenience. This fact leads us to conclude that a significant share of the gap in treatment is due to officer discrimination in lenience.
• Ticketing behavior varies greatly among officers. Because officers write hundreds of tickets for speeding, we are able to disaggregate our estimate of discrimination to identify the degree of discrimination practiced by each individual officer. We compute an officer's lenience toward minorities relative to the officer's own treatment of white drivers (after taking into account other features of the traffic stop). Doing so, we find that 42 percent of officers explain the aggregate disparity. While this figure indicates that over half of officers are engaging in no discrimination in their practice of lenience — including a third who practice no lenience, writing fewer than one percent of tickets at 9 MPH above the limit — 42 percent of officers is a greater share than the "few bad apples" story would indicate. We also find that discrimination is more prevalent among white officers and male officers.
• In this setting, which policies would be most effective in mitigating the racial disparity in ticketing? Consistent with the 42 percent estimate, a policy of removing the 5 percent of officers with the most severe measures of discrimination leads to very small reductions in the overall disparity. Similarly, hiring policies that increase the share of officers who are minorities or women reduce the overall disparity in lenience but only modestly. The most effective means for narrowing the gap in lenience comes from reassigning officers across locations. In our setting, the least lenient officers disproportionately work in counties with large minority driver populations. This means that minority drivers are more likely to encounter officers that offer no leniency to drivers of any race. When we (counterfactually) resort officers to expose minorities to the most lenient officers, the aggregate disparity in discounting almost completely disappears. This highlights the important role that geography plays in generating disparities in policing. As has been documented in other studies, the intensity of policing varies substantially across locations, even within the same municipality, and Blacks and Hispanics are more likely to reside in neighborhoods with greater police presence.