Can Detective Work End Black-on-Black Crime?
There is no debating that, as in so many other areas, Americans are racially segregated when it comes to homicide. There is also no debating -- at least among people who are actually familiar with the numbers -- that homicide is especially concentrated in the black population. The FBI reports that in 2010, a year in which whites outnumbered blacks six-to-one in the population at large, there were nearly as many black-on-black homicides as white-on-white homicides: 2,459 vs. 2,777. Interracial killings paled in comparison; there were 447 black-on-white homicides and 218 white-on-black homicides.
By contrast, what to do about this problem is hotly debated. Some say aggressive "Broken Windows" policing and increased incarceration have helped to bring homicide down over the past two decades. Others, even some who concede these strategies can work up to a point, see cops and prisons as a threat to black lives rather than as a protector of them, saying the only acceptable way to reduce homicide is to address its root causes or rehabilitate criminals.
LA Times journalist Jill Leovy, in her highly impressive and exhaustively reported new book Ghettoside, presents a very different idea. She notes that, through most of American history, blacks have been denied the protection of the law -- killing a black person has come with little official consequence, signaling that black lives do not matter and leaving vengeance in the hands of the victims' allies. Even today, she writes, "like the schoolyard bully, our criminal justice system harasses people on small pretexts but is exposed as a coward before murder." The solution is more police activity -- focused on solving homicides, not preventing them. Only then will these communities view the state as having a legitimate monopoly on the use of force, cede the task of retaliation to the law, and cooperate with police investigations.
Leovy carefully infuses a true-crime page-turner with the numbers and historical backdrop that are required to understand it. She tells the tales of Bryant Tennelle, a murdered black teenager and the son an LAPD detective, and of John Skaggs, the white Republican super-investigator who doggedly tracks down the victim's killers. Her storytelling is superb, and her command of homicide statistics and the history of African-American violence is downright intimidating. Those of us focused on policy, however, will wish she spent more space giving her theory the modern-day empirical support it deserves.
Certainly, the theory has much to recommend it. The idea that Leviathan's justice system can supplant humans' natural tendency toward cycles of revenge killings can be traced all the way from Thomas Hobbes's thinking in the 17th century to Steven Pinker's stunning 2011 book The Better Angels of Our Nature. Pinker convincingly argued that the rise of government contributed to a centuries-long decline of violence, and he even mentioned the possibility that a lack of legal protection incubated black crime in America. Scholars Leovy cites -- including William J. Stuntz, Eric H. Monkkonen, and Randall Kennedy -- have similarly posited that various forms of "underenforcement" have plagued black communities since the days of slavery and Jim Crow and remain a major problem today. It is eminently plausible that if every homicide were investigated by someone as determined as John Skaggs, poor black neighborhoods would be safer and less distrustful of the police.
But many readers will finish the book wanting direct answers to two basic questions: In the context of today's America, do improved homicide solve rates actually reduce future homicides? And if so, is this really because communities come to rely on the law, or is it simply because solving homicides locks up some murderers and deters others? Leovy doesn't address these issues head-on, so I set out in search of answers elsewhere.
Testing the theory involves more than just raw clearance rates, of course. Leovy writes that segregation is a key factor too -- it's how, she explains, "relatively modest differences in homicide clearance rates by race produce such disparate outcomes." This is important, because the racial gap in clearance rates is indeed modest: An extensive investigation several years ago by the news organization Scripps found that 78 percent of cases with white victims were cleared, compared with 67 percent of cases with black or Hispanic victims. Scripps also found evidence of abysmal performance in cities with poor black enclaves, though -- rates went as low as the 20s and 30s in Chicago, Detroit, and New Orleans -- and Leovy says clearance rates are around 40 percent in the areas of L.A. she covered.
One would think, in our increasingly data-driven crime debate, that the stats wizards would be running numbers left and right to see whether, all else being equal, low clearance or arrest rates -- especially very low rates in poor, segregated areas -- are associated with higher future crime. Unfortunately, there are relatively few studies taking this approach, many of which are decades old, and I have not found any that explicitly take segregation into account. To make matters worse, few have tried to disentangle the effects of deterrence, incapacitation, and greater deference to the rule of law.
A late-1970s study of Houston did find that clearance reduced homicide, though it also found boosting clearance rates to be very labor-intensive. A study of eight cities in the 1990s found that higher clearance rates in one year predicted lower homicides in the next, both in the overall data and in three of the cities studied individually, all heavily black -- Atlanta, D.C., and New Orleans. (There was also a statistically significant effect in Richmond and Miami, but not Atlanta, when clearance rates and homicide rates in the same year were compared.) A 2001 Heritage Foundation analysis focused on violent crime in general found an effect from clearances as well.
However, a 1982 study found no persistent link between clearance rates and murder. Results were also mixed in a Steven Levitt paper from the late 1990s; murder was the only crime for which arrests weren't consistently tied to lower rates (see Table IV). And a late-2000s study of 20 cities from the St. Louis Fed found an inconsistent relationship between murder arrests and murder rates too -- though the one city where there did seem to be a strong relationship was New Orleans.
In addition, the Scripps report singled out a few cities for improving their clearance rates or watching them deteriorate between the 1990s and the 2000s. The Justice Department has murder-rate data for Durham, N.C., and Santa Ana, Calif., places where rates improved dramatically, as well as for Flint, Mich., and Dayton, Ohio, where the opposite happened. Here are those cities' murder rates divided by the nationwide rate since 1990 (it should be noted that Santa Ana, which is overwhelmingly Hispanic, has a very low black population):
One shouldn't make too much of this, given how weak some of the trends are and the host of other problems Flint is dealing with, not to mention all the more mundane questions of correlation and causation. Combined with some of the evidence above, however, it bolsters the case that this topic deserves a lot more study. One promising approach would be to fund a pilot program in a few segregated, low-clearance cities and carefully monitor the effect it had on crime and community relations -- not just on the clearance rates themselves.
In the meantime, though, readers should be wary of Leovy's dismissive attitude toward preventive policing. She never offers a compelling case against it, and it's entirely possible to pursue preventive and reactive strategies simultaneously. Frankly, the evidence for Broken Windows reducing crime, while disputed, is much more compelling than the evidence for improved clearance rates' doing so. Both New York City and Leovy's L.A. experienced notable crime drops during the tenure of police chief Bill Bratton, the world's leading Broken Windows practitioner, and even the decidedly left-leaning researchers at the Brennan Center say that some preventive practices (such as putting more officers on the street and using the computer program CompStat to deploy them where they're needed most) can reduce crime.
With Ghettoside, Jill Leovy has given readers not just a gripping story, but also an important theory with a solid historical grounding. The nation's number-crunchers should give Leovy's ideas the thorough vetting they deserve.
Robert VerBruggen is editor of RealClearPolicy. Twitter: @RAVerBruggen