What I learned
    at the Executive Sesssion on Gang Violence

1. The leading cause of death for African-American and Latino males 15-30 in big cities is homicide.

2. Most of those murders are committed in connection with the activities of persistent multi-offending groups: call them “gangs” if you like. They vary enormously in size, structure, and activity; a gang that wears blue and calls itself “Crips” in Indianapolis may have almost nothing in common with a gang that wears blue and calls itself “Crips” in Los Angeles. Even in LA, “Crips” refers to a set of culturally similar groups, not to a unified organization with a central command structure.

3. Most of the victims of gang homicide are themselves gang-involved. But the damage spreads much more widely. Gang violence is a neighborhood-killer.

4. The combination of witness/juror intimidation and the “code of the streets” that prevents testimony even against gang rivals means that even first-class homicide investigation isn’t a good solution to this particular homicide problem.

5. A huge problem does not imply a large number of problem people. Not all persistent multi-offending groups of young minority men use deadly violence, and most members of even violent groups aren’t themselves killers. The actual number of shooters, even in a big city, is likely to be in scores or hundreds, not the thousands.

6. The situation, though out of control in many places, is not uncontrollable.

7. Effective law enforcement against gangs rests on two principles: direct communication of credible deterrent threats, and group-level accountability for violence. Those are the principles that David Kennedy and his colleagues employed to such good effect in Boston’s Operation Cease-Fire, and subsequently elsewhere. Tell the gangs that if one of their members shoots someone, the whole gang will be subjected to the law enforcement equivalent of a full-court press, and be ready to carry out that threat if need be. The multiple offenses committed by gang members — from selling drugs to driving unregistered cars to playing hooky from school — represent multiple vulnerabilities to enforcement action if a given gang is chosen as a target. Direct communication is a force multiplier. The key is understanding that arrests and prosecutions are costs of law enforcement, not benefits.

8. The companion to the threat is the offer of services: drug treatment, education, tattoo removal, job placement.

9. Sometimes indirect pressure is effective: in Lowell, Massachusetts, a wave of violence by Asian youth gangs was controlled by telling older figures linked with, but no longer active in, the gangs that if a gang engaged in deadly violence the gambling interests of that gang’s OGs (“older guys”) would be shut down.

10. The techniques of community outreach in the service of behavioral change developed in Third World public health campaigns can be effective in changing neighborhood norms about violence. Gary Slutkin of the University of Illinois School of Public Health claims reductions of 35-60% in homicide rates based on outreach alone. (Outreach plus smart enforcement does even better.) The cost per life saved is on the order of $100,000, not counting any benefits from preventing non-fatal shootings.

11. Starving big-city police forces of resources in the face of gang violence (and the terrorist threat) is a truly dumb idea, but smart enforcement can be relatively cheap, and almost any big-city force is wasting more bodies on unproductive routine activity (traffic stops, unsystematic low-level drug busts) to do the job. Federal resources are often useful, and ought to be always available. That’s the responsibility both of the agency heads and of the United States Attorney in each district.

12. This is less a criminology problem than it is a public-management problem. Changing the behavior of officials is harder than changing the behavior of offenders. (I now have permission to source that quote to Scott Decker of the University of Missouri-St. Louis, though Adele Harrell of the Urban Institute, in another context, expressed the same insight in similar words years ago.

13. This is not a good problem on which to use “ready-fire-aim” “don’t-just-stand-there-do something” approaches. You have to know what you’re doing, which means knowing what’s going on, which means a combination of data analysis and human intelligence. The Boston group spent almost two years planning an intervention that was almost immediately successful once it started.

14. On the other hand, gang violence accounts for more deaths each year than were killed on 9-11. Thinking about getting ready to think about it isn’t really a satisfactory response. Doing what we now know how to do, with resources currently available, we could cut the number of youth homicides nationwide by at least a 25% within a couple of years. Anything less is unacceptable.

Update Matt Yglesias points out that the primary meaning of “O.G” is “original gangsta.” I stand corrected.