Wednesday, April 11, 2007

The Great American Crime Decline

In the past I've used the example of drought and efforts to relieve it to talk about whether or not government action has the impact on crime that policymakers frequently believe or assert. If you have four areas, all suffering from drought, and one brings in rain dancers, another cloud-seeders, another fervently praying ministers, and one does nothing, and they all end up getting enough rain the next month to break the drought, what happens? The first says the rain dancers did it, the second the seeders, the third the prayer, and the fourth, . . . well, the fourth says nothing because it did nothing. The first three will engage you at least emotionally in how their "solution" turned the trick when Mr. Spock, looking at the four, would rightly conclude that something else was at work. In the case of crime and public safety, it's not that what government does has no effect at all. It's that the magnitude of the effects claimed do not bear the scrutiny, much less the hyperbole, and maybe we ought to be making more effort to find out what that "something else" was and use our resources on maximizing it (them).

One of the best proofs of this example is Franklin Zimring's recent The Great American Crime Decline. In this quick and very readable work, you will find examination of all the proposed "solutions" for the crime drop of the late 80s and early 90s to today and the reasons why they look so much like rain dancing, cloud seeding, and prayer. In fact, you'll find that the "fourth" is Canada, a nation with much of the same foundation as we have (just p.o'ed a bunch of Canadians, eh?) which saw the same basic pattern of crime rise and fall without engaging in the massive incarceration, "broken windows" policing, increased abortion, or any of the other "solutions" that have been offered.

Zimring goes through the backstory of the "great crime decline" quickly and well. He doesn't cherry-pick the time frames to make one case look better than another. After an explanation of why homicide stats are the best long-term crime numbers (even today) for use in analysis, he's able to go back several decades to compare the ebb and flow of crime and prison populations and how those ebbs and flows don't always correlate. The range for impact of greater imprisonment on crime is 10%-40%, depending on the study, so a best guess of 25% (which several good studies come up with) is reasonable. Keep in mind, this does not prove that this was the best reduction of crime rates that could have been gotten from the resources used. So 75%, give or take, of the reduction came from something else.

What? Decline of the Baby Boom and high numbers of crime-prone young people aging out of criminality. This explanation got pooh-poohed some a decade or so ago, but it's not looking that bad now, in both Zimring's analysis and in the renewed growth of the "aging" population entering prison for the first time that we're seeing in OK and other states. But even demography can't explain all the decline. The economy? Well, historically, the correlations aren't much better there than for imprisonment. Again, like prisons and age groups, there's probably a connection, but not enough to point a finger definitively. Nor were improved and expanded rehab programs, decline of drugs, or gun control nationally widespread enough to account for national decreases.

But there are other candidates put out there in recent years—abortion, the fall of the crack market, and more cops. Again, however, these did not cut across the nation equally with similar enough impact everywhere to have brought about similar crime declines. And Zimring raises enough questions about the research that has brought these "solutions" forward to increase doubt about their impact. Not that he disproves them, or even raises doubts among their advocates, but he makes clear that resting the case for the "solution" on any of them is hazardous.

His most important chapter deals with what was happening in Canada at the same time, the same rise and fall of most crimes despite that nation's general lack of action. Now the crime patterns weren't identical and Canada, to its relief and ours, is not us, but, if a fourth person sits and does nothing while we're practicing our special solutions to a problem and the result for them is very close to ours, then we should at least be skeptical about whether our actions had anything to do with anything. His next most important chapter is an analysis of the extraordinary decline of New York City crime (which was accomplished without massive rises in incarceration) and its attribution to "broken windows" or "order-maintenance" policy pursued there more exhaustively than elsewhere. Trying to find other places to place the success, Zimring leaves unsuccessful. He concludes that the "natural experiment" NYC conducted ended up having as much impact as that attributed by the research nationally to incarceration. From that, Zimring tells us that the evidence does indicate that major improvements in public safety can be achieved without drastic changes in a community.

With reports of crime rates beginning to increase again, particularly violent crime, Zimring examines whether there might be a natural kind of floor to crime drops and whether the public's anxiety about crime has relaxed enough to allow more flexible responses than we've had in the last couple of decades. His answer is pretty much, well, let's keep studying and see. Still, he does give us these seven "lessons from the 1990s" to keep in mind as we approach more flux in our crime situation and how we address it:

1. The crime decline was real, was national in scope, and was larger and longer than any documented decline in the twentieth century.

2. The crime decline of the 1990s was a classic example of multiple causation, with
none of the many contributing causes playing a dominant role.

3. It will not be possible to comprehend what caused declining crime in the United States until more is known about the parallel crime decline in Canada.

4. New York, the nation's largest city, had a crime decline during the 1990s almost twice the national average, and the city's downtrend has continued through 2004, making it a natural laboratory for studying the effects of a lower crime environment on urban life.

5. Two kinds of parochialism have hampered the effort to understand the effects that social and criminal justice factors have on crime rates: (1) the failure by many investigators to consider data and insights outside their narrow disciplinary perspective, and (2) the failure to consider events outside the United States.

6. Since the national crime decline ended, almost as if on cue, in 2000, rates have stayed near the lowest levels of the 1990s. But there are indications that crime rates could drop further, perhaps much further, without major changes in the American social framework.

7. Whatever else is now known about crime in America, the most important lesson of the 1990s was that major changes in rates of crime can happen without major changes in the social fabric.


So Zimring doesn't give us "answers," just "lessons," which might be disappointing to those wanting more definite hope. But, as someone who went through this in the 90s and saw what we knew and didn't and what was claimed that has now been proven at best limited in its "truth," these seven lessons give us much better ground for facing the future, whatever it holds, than we had then. It's up to us, and frankly for grabs, whether we use them wisely. Or at all.

No comments: