The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Attorney General Sessions Properly Links an "ACLU Effect" and the Chicago Homicide Spike
The Attorney General approvingly discussed my article linking the dramatic decline in street stops in Chicago under an ACLU agreement and the subsequent homicide spike. While Professor John Rappaport has a different take on this "ACLU Effect," his unsupported analysis does not fit the data.
I was pleased to learn recently that Attorney General Jeff Sessions found convincing my soon-to-be-published article analyzing the 2016 Chicago homicide spike. In a speech to a law enforcement training conference last week, the Attorney General discussed the shocking 58% year-over-year increase in homicides in the Second City in 2016. As the Attorney General explained, an article Professor Richard Fowles and I co-authored collected data on the subject, concluding that the best explanation for the spike was a dramatic decline in street stops by the Chicago Police Department (CPD) at around the end of 2015, which in turn triggered the spike at around the beginning of 2016. Using multiple regression analysis and other statistical tools, we estimated that the decline in stops led to approximately an additional 200 victims killed and over 1,100 victims shot. We also provided our reasons for believing that the decline in stops was attributable to an agreement between the City of Chicago and the ACLU regarding street stops implemented at the end of 2015. We also noted that contemporaneous reports from the streets of Chicago during the spike had also identified the causal factor as the "ACLU effect"—and our findings supported those reports. The Attorney General cited these findings as a cautionary tale about well-intentioned but ultimately-harmful regulations on law enforcement.
One supporter of the ACLU—Professor John Rappaport—quickly attacked the Attorney General's speech, writing on Slate that Sessions was "scapegoating" the ACLU for the homicide spike. To criticize the Attorney General, Rappaport criticizes our study, calling the ACLU effect a "mirage."
It is, of course, much easier to be a supporter of ACLU police "reforms" if they unerringly produce nothing but benefits. But our study suggests that the real world is much more complex and that ACLU-style consent decrees can produce deadly trade-offs.
Professor Rappaport begins by observing that as CPD stops were substantially declining, Chicago suffered an increase in gun crimes, but not in other crimes. Rappaport writes: "So Cassell and Fowles need some theory for why street stops, when conducted en masse, were depressing gun violence but not other crimes—such that, when stop and frisk dropped, gun crimes alone shot up. They don't have one."
Actually, our article articulates a very specific, empirically-supported theory. As explained in our article (p. 24), the available empirical research suggests that some forms of proactive policing—such as targeted gun patrols or stop, question, and frisk tactics—are uniquely effective as suppressing gun crimes:
Interestingly for our purposes—i.e., for trying to explain why Chicago's gun crime rates increased dramatically but not rates for other crimes—one of the conclusions suggested by such studies is that stop and frisk (at least in some forms) will reduce gun crimes, but not other violent crimes. For example, a study in Indianapolis found that targeted offender gun patrols reduced gun-related crimes substantially but did not appear to affect violent crimes without firearms. See E.F. McGarrell et al., Reducing Gun Violence: Evaluation of the Indianapolis Police Department's Direct Patrol Project (NCJ-188740 Nat'l Inst. of Just. 2002). Another study, in New York, found few effects of SQF on robbery and burglary rates. See Richard Rosenfeld & Robert Fornango, The Impact of Police Stops on Precinct Robbery and Burglary Rates in New York City, 2003-2010, 31 Just. Q. 96 (2014) (finding no statistically significant correlations between SQF and burglary or robbery using yearly rates of crime and SQF stops across all 75 precincts in the New York City Police Department).
Our article also collects research showing that the typical mechanism by which aggressive policing is expected to prevent gun crimes is not by physically removing guns on the streets but rather through deterring the carrying of guns in the first instance—something which could uniquely reduce firearms violence. Rappaport may not agree with our theory—or have some as-of-yet-unpresented data to contradict it—but to write that we "don't have a theory" is simply false.
Professor Rappaport goes on to suggest that other research suggests that when proactive policing slows down, we should expect a higher increase in property crimes than in violent crimes. But the study he cites to prove this point—an important recent paper by Professors Rushin and Edwards—looked at federal consent decrees orchestrated by the U.S. Justice Department on all forms of policing, not just stop and frisk. Accordingly, it is perhaps unsurprising that broad decrees had broad effects on all types of crime. But what we were examining in our paper was a specific ALCU agreement focused narrowly on street stops.
Rappaport next wonders why we "ignore" other cities, noting that street stops declined in New York City without a subsequent homicide spike like Chicago's. But as we explained in detail in our article—and have previously discussed here at Reason—good reasons exist for concluding that a reducing street stops in Chicago would produce much different effects than in New York:
In 2016, New York's homicide rate was only 3.9 per 100,000 population, while Chicago's was 27.8—a rate more than 600% higher. But the relevant differences between the two cities may be even higher than this already staggering difference suggests. Looking at homicides committed by firearms, in 2016 New York's rate was 2.3 compared to Chicago's rate of 25.1—a rate almost 1000% higher. This is important because, as discussed earlier, gun crimes may be particularly sensitive to stop and frisk policies. In addition, because New York has such a small number of guns and gun crimes (relative to Chicago and many other cities), it can concentrate resources on preventing gun crimes in a way that other cities cannot….
Another problem in equating New York's circumstances with Chicago's is that the level of police power is different. Famously, New York has high levels of law enforcement. . . New York had about 153 law enforcement employees for every homicide committed in the city, while Chicago had only about 17 employees for every homicide committed—about a 800% difference. The difference is even greater if one combines both the gun homicide and police force numbers. Per gun homicide, New York has roughly 260 employees, while Chicago has only 19—well over a 1000% difference. To this point it might be objected that a homicide is a homicide, so it makes no sense to break out gun homicides separately. But homicides are not all alike. To the contrary, in general, homicides committed by firearms are more difficult to solve than other kinds of homicides, only adding to the relative difficulties for the Chicago Police Department. Moreover, in 2016, about 23% of New York's homicides were gang-related, while roughly 67% (or more) of Chicago's homicides and shootings appear to have been gang-related. Here again, gang-related homicides may be more difficult to solve than are other homicides, particularly in Chicago.
Rapport also notes that in 2016, other cities had homicide spikes, suggesting that homicides can increase for reasons apart from changes in police tactics. But our research tried to explain what was going on in one important city—Chicago—where our research tools and data could explore what was happening. Every city has its own story to tell, and we have tried to tell one chapter of Chicago's.
Rappaport even wonders whether Chicago's 58% homicide increase was unusual, linking to an intriguing paper by Professors Wheeler and Kovandizic. They explain that before jumping to the conclusion that some sort of homicide "spike" needing explanation exists, this typical volatility should be considered. While this caution is well-taken, what happened in Chicago was an exceptional spike outside of normal variability. Chicago's 2016 increase in homicides was the largest single-year homicide increase of the last 25 years among the five most populous United States cities. Rappaport fails to mention that Wheeler and Kovandzic's own "normal" volatility model shows that Chicago 58% 2016-over-2015 increase in homicides was "near the upper limit of the . . . 80% prediction interval"—i.e., near the very top of the 80% confidence band that encompasses normal year-to-year volatility. Indeed, after having surveyed homicide changes all across the country in recent years, Wheeler and Kovandzic specifically concluded that "[s]ome recent increases in homicides are clearly noteworthy, such as Chicago in 2016 . . . ."
While our paper focused on the time period for which we had full data (2012-16), Rappaport points to a homicide decline the next year—2017—in which Chicago's homicides fell from about 750 to about 640—a figure still well about where homicides were in 2015 before the decline in street stops. As we explain in our article, the most likely causes for the 2017 decline appear to be a combination of countermeasures that the Chicago took in response to the homicide spike. But Rappaport draws the remarkably sanguine conclusion that this decline proves "we can fight crime" without streets stops. Yes, it is true that crime can be fought in other ways—simply not as effectively. If our modeling is correct, relying on these other ways comes at the costs of the deaths and shootings of hundreds of additional victims, predominantly African-Americans and Hispanics located in some of Chicago's most impoverished neighborhoods.
Rappaport is seemingly unwilling to stipulate that the disappearance of hundreds of thousands of street stops by CPD in 2016 led to even a single additional homicide. Instead, Rappaport tries to pin the blame on other factors. For example, Rappaport suggests that a reduction in funding for a social program known as "Ceasefire" might be the cause. An analysis by the Chicago Crime Lab found that this explanation doesn't fit, because the ceasation in funding was well before the spike took place.
Rappaport also cites the release of the Laquan McDonald video depicting a CPD officer shooting a young African-American. The causal mechanism that Rappaport advances is that the video's release "affected how and when citizens called the police." That is a speculative hypothesis for what happened in Chicago in 2016. We investigated that very subject at length in our article (pp. 37-43), explaining that any reduction in police-citizen cooperation would be expected to result in an increase in all crimes, not just gun crimes. Of course, as discussed above, Rappaport himself has to admit that this is not the pattern seen in Chicago crime data in 2016. In any event, we collected monthly 9-1-1 call data from Chicago during the relevant time period. Directly contradicting Rappaport's speculative hypothesis, the call data show a longterm downward trend—no sharp break around the time of the McDonald video release. If Rappaport has any data to support his speculative theory, he has yet to provide it.
As a final fallback position, Rappaport argues that, even if declining street stops caused the homicide spike, the ACLU should not be blamed. Rappaport, who previously interned in the ACLU's National Legal Department, tries to invent an incongrous timeline to deflect attention away from the ACLU. Stop and frisks clearly began to fall in November 2015, but he asserts (without support) that was "six months before CPD finished preparing officers" to implement the ACLU agreement. Rappaport's chronology is wrong. As we document in our article, the ACLU's agreement in August 2015 envisioned revised general orders relating to street stops "by December 31, 2015." CPD promulgated the new general orders on this schedule shortly before December 31—precisely when the decline in stop and frisks occurred.
In addition, the link between the ALCU agreement and the decline in stops has been specifically acknowledged by … the ACLU. In February 2016, when it was clear that the number of CPD stops had fallen dramatically—but before the deadly consequences had been fully revealed—the ACLU took credit, saying the decline was actually "a good thing" produced by the agreement.
In writing our paper, Professor Fowles and I had no intention of trying to "scapegoat" anyone. Our goal was to spark a discussion about how to save lives in Chicago, particularly on the city's south and west sides where a disproportionate number of deaths result. We appreciate the fact that the Attorney General is now discussing this question as well. The best answer remains to be determined. But too many lives are at stake to try to simply sweep under the rug the clear possibility that police street stops may play an important role in assuring public safety in Chicago and elsewhere.
