A few months ago, the Lincoln Police Department implemented a new policy that encourages officers to take advantage more often of a state law that allows vehicles driven by suspended drivers to be impounded for 30 days. The policy change went into effect in late February, and provides an opportunity for a bit of evaluation research.
Since the policy is intended to deter suspended drivers, if it's working we would expect a decrease after implementation. Here's the problem: you can't simply count up the number of suspended driving tickets, because the overall level of traffic enforcement would influence that independent of the policy change. When officers are writing a lot of tickets, they are more likely to encounter suspended drivers then when they are writing fewer tickets.
Traffic enforcement activity by police officers varies considerably over time for several reasons, such as weather, service demands, and staffing levels. As an example, the peak month for tickets in 2015 was March, with 8,046 tickets (both warnings and officials), while the low month, December, produced 5,718. In order to account for these fluctuations, a good measure would be to calculate a percentage of tickets that yielded a suspended driver: suspended driving tickets divided by total tickets.
As the chart shows, that percentage takes a significant drop in the three months since the policy change. In fact, March, April, and May are the lowest months during the entire comparison period, which begins in January, 2013. If the percentage during the past three months had been the same as the average over the preceding 38 months, there would have been 159 additional suspended driving arrests from March through May.
I think this is pretty strong evidence that the policy is having the intended effect, although I'd like to watch it over a longer time period. Sometimes the impact of a crackdown initiative decays quickly over time. Officer Luke Bonkiewicz, whose research credentials are better than mine, is doing some more sophisticated work with these data. In particular, he is also looking at repeat suspended drivers. His work may shed even more light on the efficacy of impounding vehicles. Police practitioners ought to do this stuff more often: simple evaluations that, despite some methodological warts, provide evidence about what might be working (or not.)