Monday, May 16, 2011

Public Health - Just Due It?

“Program Goal: To improve the safety of our community for vehicular, bicycle and pedestrian traffic” – The San Rafael Police Department (SRPD) on the Automated Red Light Photo Enforcement System

“Oh No!!” – This author after receiving a red light camera ticket in the mail.

According to publicly available data, the advent of red light violation enforcement by cameras in the city of San Rafael has been associated with a twelve percent decrease in accidents at camera-enabled intersections. Between November 2008 and April 2009 there were 48 accidents at the intersection of 3rd and Irwin in central San Rafael, compared to 43 between November 2009 and April 2010. The prevention of five accidents (none of the accidents were fatal by the way) is nothing to sniff at, although, from a scientific standpoint, it is far from proof that these cameras improve public safety. Consider the other possibilities. Have accidents decreased because yellow light times have been extended (from 3.0 to 3.5 seconds)? Or, have motorists intentionally avoided 3rd and Irwin and chosen other routes instead? Or is this finding a statistical fluke? One thing that is not at all flukish is the amount of money collected in fines from violators. The current fine is $479, of which the city of San Rafael receives 30%, and with the addition of court and traffic school fees, the sum approaches $600. If you consider that since its inception, the city has averaged roughly 467 violations/month, over a six-month period, the yield (in fines charged per prevented crash) works out to an estimated $268,000. Expensive. Perhaps too expensive?

But, is it possible that a city that utilizes red light camera enforcement increases safety not just at the enforced intersections, but across the entire city? This is consistent with James Q. Wilson’s “broken-window” theory (embraced by former New York police commissioner William Bratton) that visible disorder in a neighborhood leads to greater lawlessness citywide. So rather than look at San Rafael in isolation (I was unable to get city-wide data on crashes from the SRPD), it makes sense to take a larger view of the effect of red light camera enforcement. And I should stress that, nationwide, this is not a trivial problem – according to the Insurance Institute for Highway Safety (IIHS) there were over 2 million intersection related crashes in 2009, resulting in 7,538 deaths.

Just this past February, one of the largest studies to date on this topic was published by the IIHS. The report, lead authored by Wen Hu, compared cities with and without red light camera enforcement programs across two time periods (1992-1996 and 2004-2008). The investigators looked not just at the enforced intersections, but citywide, to discern differences between those with and without red light camera enforcement. The authors report a 21% greater decline (35% vs 14%) in fatal red light crashes in cities with camera enforcement versus those without it. The investigation also reports an estimated 17% decrease (versus expected rates) in fatal crashes at signalized intersections in cities with cameras. This study has a lot of strengths – it is big – looking at 62 U.S. cities (14 with camera programs and 48 without) and focusing on an important outcome (fatal crashes). It also does a reasonable job of creating a nationwide sample and of adjusting for population (crash rates per 100,00 in population) and accounting for outside influences (what researchers call confounders) such as population density and land area. In sum, it is reasonably convincing. But, there are important caveats and limitations.

One of these is the question of conflict of interest – the IIHS is funded by the insurance industry, an industry that benefits not only if there are fewer car accidents but also if more drivers accumulate points on their license (increased fees for these drivers). But, let’s give them the benefit of the doubt and assume that they are interested in studying road safety in an unbiased fashion. So, what then are the limitations of the study? First, those cities implementing cameras had a higher baseline rate of crashes (65% higher) – meaning they had larger room for improvement. Second, there were significant differences in rates across cities (crashes increased by 165% in Raleigh, NC while decreasing by 75% in Chandler, AZ). Third, the “control” group – those without cameras – has two major outliers whose crash rates more than tripled across the study period. Removing these two cities from the analysis might have led to much more modest results. Fourth, the analysis could not account for changes in yellow light times or other interventions that might have been the true cause of decreased fatal crashes. Finally, this data excluded crashes from illegal turns on red (the vast majority of red light tickets in most locales are for illegal turns). Thus, this study is unable to inform the question of whether policing failure to come to a complete stop on red benefits the public welfare. So, what does this all mean? Well, in my opinion it means that the jury is still out on the public health impact of red light camera enforcement.

On the other hand, the jury is no longer in deliberation for my red light camera enforcement citation. Well, to be clear, there was no jury – just an honorable Superior Court judge and a twelve second videotape. My judge was not sympathetic to the plea that a failure to come to a complete stop on red was not equivalent (penalty-wise) to that of busting through a straight-ahead intersection. Oh well. For myself and others perps, we can take heart that our fine money is not only supporting the San Rafael Police Department (and Redflex Systems of Phoenix AZ), but also a whole host of public services ($ amounts based on the previous penalty schedule of $445): $8.92 County General Fund, $17.15 Criminal Justice Facilities Fund, $ 13.72 Courthouse Construction Fund, $19.60 EMS, $9.80 DNA Identification Penalty Assessment, $13.72 Maddy EMS, $3.43 Automated Fingerprint ID System Fund…and the list goes on. Sometimes a state budgetary crisis and a public health initiative can become inextricably linked. Public health…just due it?