urban institute nonprofit social and economic policy research

The Decline in Crime: Why and What Next? (Transcript)

Share:
Share on Facebook Share on Twitter Share on LinkedIn Share on Digg Share on Reddit
| Email this pageE-mail
Document date: October 26, 2000
Released online: October 26, 2000

Mr. Travis: For nearly half a century, the issue of crime has occupied a prominent place in our public discourse. Our policy discussions have been dominated by shorthand phrases such as "law and order," "three strikes," "gun control," "victims' rights," "truth in sentencing," "community policing," "graduated sanctions," "zero tolerance," "drug courts," "boot camps," "harm reduction," "coerced abstinence," "broken windows," just to name a few. And behind this everchanging policy potpourri has been the seemingly inexorable fact that crime has been rising in America. Our crime conditions seem particularly acute in the late '80s, when an outbreak of violence added new phrases to our lexicon: "drive-by shootings," "crack wars," and "super predators."

Now the headlines reflect a different message: Crime rates are at the lowest levels in a generation. Beginning in the early '90s, rates of violent crime, particularly juvenile violence, started to drop and have been dropping steadily ever since. And, with much less fanfare, property crime rates have been drifting downward, and adult homicide rates have also been on the decline.

So, why has this happened? Will this good news continue? And how can we create safer communities? These are the questions that we hope to address at our forum this morning. We hope to shed some empirical light on these questions, and to try in a very rough sense to unpack the various social forces that may have contributed to this good news. So we have assembled a truly stellar panel to help identify the social trends and the policies behind our declining crime rates, and to offer some predictions and policy choices for the future.

[The panelists] reflect a broad spectrum of researchers and practitioners. I'm very grateful that they have agreed to join us this morning. And I just want to say at the outset that our wonderful weather conditions have caused fog in airports, but hopefully not fog in thinking, and some of our panelists are on their way and have phoned in from various stranded airlines that they're eager to be here.

And I want to say that particularly as many friends and admirers are very pleased to see Mark Kleimann here back in action after a valiant and successful medical battle.

So, we have a lot of ground to cover in a few short hours. So, here's the format. Our first presenter will be Al Blumstein, known to people in this community; he is the Eric Johnson Professor at the Hines School of Public Policy and Management at Carnegie-Mellon, and also director of the National Consortium on Violence Research. He will present the macro trends in crime rates in the United States, and we will then hear from a series of brief presentations by researchers who have examined the contributions of gun control policies, prison expansion, changes in adult and intimate partner homicide and violence, changing drug markets, policing practices, labor market force, and demographics, and the contributions of each of these to the overall shift in crime rates.

These presenters have recently collaborated on a book that came out last month, entitled The Crime Drop in America. That book was co-edited by Professor Blumstein and Joel Wallman of the Guggenheim Foundation.

Before our break, we will then ask the remaining panelists to reflect on what they've heard and to help us understand whether there are one or more policy arenas or social developments that have not been accounted for or not accounted for adequately in the first set of presentations.

This is a lot to pack into a short time, and Al and I have had lengthy discussions with each of the presenters about keeping to time limits, and trained as we are in behavioral research, we've conducted a literature review and decided that the best way to encourage researchers to good behavior is to offer them application-free grants from the National Consortium of Violence Research if they keep within their time limits. We also understand that occasionally enforcement of rules requires swift punitive action, and that's my role.

Then we will return after a break and shift our focus to the future. We will focus on three broad domains that cut across the first set of presentations: first, communities and young people; second, police and other community-based responses to crime; and third, policies on sentencing and offender management, and policies on guns and drugs.

And we will ask first a mini-panel that I will create, and then the entire panel, to help us understand whether these downward crime trends will hold, what policy choices seem to represent the best investments for producing greater levels of safety, and what early warning signs should be heeded over the next few years.

And we hope, of course, that our discussions of the future will reflect the research on our recent past. If so, our policy choices will be better informed by the lessons we have learned, and the public can be offered something more than slogans.

Before turning the podium over to Professor Blumstein, I would like to thank our supporters—the Open Society Institute, National Consortium on Violence Research, the Guggenheim Foundation—that made the initial research and book possible, and the people who actually made the forum possible—the Justice Policy Center staff: Amy Solomon and Michelle Waul, Nichole Chigoy and Laura Johnson; and the Public Affairs staff of Karen McKenzie and Renu Shukla, who really put it together with great style. And I owe particular thanks to Al Blumstein, who, when I called, said he would love to be a partner in this venture, as we've done many others together, and whose energy and curiosity and analytical rigor are simply inspirational. So, Al Blumstein, let's get going.

Mr. Blumstein:—program funded by the National Science Foundation, a really innovative step for the foundation to create a consortium of collaborative researchers scattered around the country and around the world to try to deal with the whole issue of violence in a basic as well as applicable way, and cosponsoring the forum.

Jeremy talked about the crime trend, and I think it's useful to look at that crime trend, which from the '70s through the early '90s was relatively flat, oscillating up and down, but relatively flat, and we see this clear strong downward trend—red is homicide and the green is robbery. And Jeremy wondered how long this was going to continue. I can tell you that I've made a projection of this impressive straight line, and by 2007, homicide rates will go negative, and you can anticipate that that's going to happen on Easter Sunday. All of which is intended to say that this trend obviously can't continue in the same direction it has. It's going to have to start leveling off in various ways over the next several years, and perhaps we're starting to see the beginning of that because in the '99 Uniform Crime Reports, for example, the biggest cities—which were the first ones up and the first ones down—the biggest cities had a growth rate that was smaller than the rest of the country. Cities of half-a-million to a million had a decline rate of 2 percent, and cities of a million and above had a decline of 2 percent, while the nation as a whole had a decline of 8 percent. So we may be starting to see that leveling off, and we don't know what in the future might turn it up. But at least we know that it can't continue indefinitely, and changes have got to occur before 2007.

The theme of much of the conference, and much of the volume that Jeremy mentioned, was the theme of disaggregating the variety of forces that are contributing to the decline. I think you can't really understand the decline without understanding what happened before, because so much of the decline was undoing the rise that came before it. And the most important variable to disaggregate is age. What's gone on has been very different across different age groups. This is the arrest rate for homicide over time, which has been fairly stable from 1970 through 1985, fairly stable, after a rise in the '50s and '60s as the baby boomers came in and distorted so much of what's going on in the society.

In '85, we saw this dramatic bifurcation between young people, represented by this red graph of 18-year-olds, and if you look at the younger people, younger than 18, they all follow the same very sharp rise pattern, and very sharp drop pattern. These (indicating slide) are the 24-year-olds, and their pattern looks very much like the older folks, folks over 24. So young people went wild in the late '80s, older people were largely unaffected, and you'll see that in a variety of ways.

What I've done here is summarize that information, which is the ratio of the homicide arrest rates in 1985, which was the departure point by age, so that 15-year-olds, for example, tripled by 1993, which was the peak year of the young people's offending, and that growth declined with age, so that people over 30 at that point were not only not growing, but declining.

By 1999, these young folks stopped doing what they had been doing, and in this age range of the late teens/early 20s, they're still about 20 percent above where they were in 1985 at the takeoff point. The older folks have been declining steadily, about 20 percent below '85 in '93, which is when the young people peaked, and by now they're about 40 to 50 percent below. So one has to think quite differently about the young folks compared to the older folks.

The second dimension I want to talk about is the weaponry, because the weaponry has been a particularly serious aspect of the growth and of the decline. If you look at the weapons used in homicide by adults, [which] I'm calling 25- to 45-year-olds, there's not much change, somewhat of a decline, but fairly flat. [This is] in marked contrast to the group I'm calling youth, which are the people 18 through 24; compared to a base rate of 100, an index, they more than doubled in handgun homicides with no change in other weapons. The notion of the "super predators" is contradicted by this. It's not that the kids were worse; it's that the kids were doing what kids always do, which is fight. But they've been fighting with far more lethal instruments than the fists that they normally use, so that we see a major change in the weaponry. And when we go to the juveniles, people under 18, we see a quadrupling from a base of 100— almost 400 times as high a rate of use of handguns in homicides, and again, no particular effects in the other weapons.

So, if it were worse kids, they'd be doing bad things with all kinds of weapons. The theme here is kids and the theme is weapons, and handguns in particular. And that's been a major part of the growth.

It's also the case that there's been an important race difference in the offending and the victims. And, again, if we start with a base of 100, for African-American youth we see, again, two and a half times that by the peak in 1993. A major shift all in handguns, some shift in whites and Hispanics—and those have been difficult to sort out, but nothing like the shift in the African Americans. And part of the theme that really comes in here relates to the emergence of crack markets. If you look at adult arrest rates for drugs, [there's] not much going on with whites, but a major growth starting in about 1980 for the adults, reaching a peak in '89.

The juveniles didn't get recruited into this until 1985. So part of what we're seeing is the emergence of recruitment of sellers in the drug market, at least in part because of growth in demand for crack, and at least in part because the older sellers were being put in prison, and so it's a resilient market [that] recruited replacements, and we saw this major growth in the recruitment of African-American juveniles into drug markets. And if you're in the drug market, you're very vulnerable to street robbers, and the natural defense against a street robber is carrying a gun. Once those guys were carrying guns, then we saw a diffusion of guns out into the larger community. And that was what presented the problem, and guns in an interpersonal conflict represent a radical change in the dynamic of what went on.

Let's look at weapons arrest rates by age. Again, these are the same ages that you saw on that homicide graph. Not much going on with the older folks. With the younger folks—these are 18-year-olds—a major rise starting in 1986 to a very sharp peak in 1993, that peak year, and a very sharp decline since then. Gun arrests are like drug arrests, they're a function of the activity, the level of illegal activity, and the aggressiveness of the police in pursuing it. There's no indication that the police diminished their aggressiveness in pursuing guns in that period. So it's probably the fact that the kids were not carrying them anymore, partly because of the ATF [Department of Alcohol, Tobacco, and Firearms] efforts—federal efforts to restrict the trafficking in guns—partly because of local policing efforts to take the guns, partly because of community efforts to stop them from carrying it. And once we start on the downward slope, it's self-reinforcing because as the kids are discouraged and deterred from carrying it, other kids no longer have to have it because they don't need it for defense. So that is a continuation of that decline. But even as of '99, that was still a bit above where it was when it started in 1985.

So part of the decline, therefore, has been undoing the rise. The characteristics of the rise were kids and handguns, and what we saw was the police, the ATF, and community groups working that issue. The other piece that was critical was that the nucleating factor in this diffusion process [was] the drug markets. The drug markets changed, in the sense that older users may have continued, but new users weren't using [them], so the process of sale could change from the street market, which was very vulnerable, to much more in the way of personal delivery.

And the evidence from the street—and I'm sure you'll hear more about this later—suggests that it's a recognition of the harm of the drugs, not necessarily as a result of any particular enforcement efforts on it.

Fortunately, all those kids who might otherwise have been recruited into the drug markets who weren't needed had jobs in the economy, and when you're in the legitimate economy there's an incentive for conformity, to go straight, and the point [is that] all of these effects, which we'll be talking about, not only operate on their own, but are mutually supportive and reinforcing, and argue vigorously for efforts at essentially prevention getting opportunity for these young folks to have legitimate roles in the society and the economy so that we can continue the really good and supportive trend we're on.

The second theme I want to talk about is the whole issue of incarceration, because it's an important part of the story. For 50 years, the United States had an astonishingly stable incarceration rate of about 110 per 100,000. Is that stuff visible? Barely. Oh, nuts, okay. The United States had this astonishing rate, and we've increased exponentially to four times that in the last 25 years or so. And so the incarceration rate has been very strong. Some people naively look at this growth in the '90s and look at that last portion of the crime rate trend, and they say incarceration rate going up, crime rate coming down; therefore, incarceration is causing it to drop. Well, in the late '80s, it was going up. And that simplistic analysis misses so many of the key issues of when it's going up, when it's going down, and we've really got to sort some of that out. And, again, you'll hear more about that. I think it's instructive to get a sense of where the U.S. stands in this incarceration [rate]. Russia's is the largest in the world, we're next, and then there's a bunch of other countries scattered around the world that are down in the range of 100 to 200 or less with the incarceration rate we have.

An important feature of it is this growth by crime type—and all of it has grown—but the one that's been most dramatic is this growth in drug offenses, for which we've been incarcerating people, and that has grown dramatically to the point where it's now over 20 percent of state prisoners, far and away the largest offense in state prisons, and over 60 percent in federal prisons, with the real question of how much drug transaction is averted thereby, because the market is robust, and if it's losing these guys, it will recruit young people, and they may be far more treacherous and dangerous out in the community. And that's an issue that's going to have to be addressed some how.

We're at four times the 50-year stable rate. We're the highest in the world except for Russia. We have a high rate. And, inevitably, as we dig deeper into the offender pool, the efficiency and effectiveness at the margin is bound to diminish. And we'll hear some more about that. There has been a steady decline in homicide by offenders over 30. As I indicated earlier, that started back in the '70s, so there has been some incapacitation effect of homicide, and we'll hear some comments on that. The median age of prisoners is in the low 30s, 32 to 33, thereabouts. So those are the natural candidates for incarceration, particularly those with a history of violence. And by the time they're 30, if they're still active offenders, even though they are a relatively small portion of the offending population, they represent an identifiable continuing risk, but I think it's important to recognize that the simplicity—X-curve in the crime going down, incarceration going up in the '90s—is just too simplistic an analysis to deal with the issue of incarceration, and we've really got to assess the gross social cost, what are the consequences of the recruitment by markets as replacements for drug offenders, what are the consequences for community, for family, as well as the opportunity cost for all the things we might be doing with the resources required for that level of incarceration.

All of these concerns and interests led to a conference organized by the Harry Frank Guggenheim Foundation, which eventuated in this book that has just come out published by the Cambridge Press, edited by myself and Joe Wallman, with a variety, and resulted from a debate over the factors that are affecting the drop. Ideologues are coming at it on both sides. We've improved social cause in the economy; kids are getting jobs, from the left. From the right, the X-curve argument, as well as administrators who are saying, "Ah, the innovation I used in my police department is really what's driving the crime rate down," with a different answer, obviously, in every police department.

It's clear that there are many policy implications. We'd like to find what can be done to continue this drop as much as it can, and to recognize when conditions change as we've gotten rid of the stuff that's easy to get rid of, what's left may be a lot harder, and it wants to be focused on.

And the press and the public are particularly curious, and when one reads the press one sees so many misstatements of what the primary factors are. The sponsors of the book have been the Harry Frank Guggenheim Foundation and COVER. The chapters in the book, there's a synthesis that Joe and I did, and it was really a delight working with him. He's a very able individual, and we worked on that together. Basically, you heard a 10-minute summary of the first chapter. That looks to disaggregate along a variety of dimensions. In addition to the weaponry and age I talked about, we did it by city size.

There's a chapter on guns and gun violence that Garen Wintemute did that really looks at the variety of efforts that have been impressively successful over the last decade or so. And then the incarceration issue is really crucial. Bill Spelman did a chapter on basically looking at elasticity rates, the estimates of the aggregate effects of incarceration during the up period as well as the down period, because during the up period, if we were locking more people away we might be keeping the growth below where it might otherwise be.

Rick Rosenfeld did a chapter on the older offenders, and there were two major thrusts there, one related to changes in intimate partner homicide rates, which have been notable. The other is a recognition that they represent prime candidates for incapacitation, and he's got some very interesting estimates of the effect of incapacitation taking people off the street so that they can't commit the crimes, and what those effects might be and what it costs.

We mentioned the drug markets as a very important issue, and Bruce Johnson did a chapter with two colleagues, Andy Gallabin and Eloise Dunlap, on the changes in drug markets. Policing has been an important theme in many of the claims, and John Eck and Ed Maguire did a chapter on that. We know the economy has been an important contributor, and Jeff Grogger did a chapter on the economy. And one of the stories that the press keeps talking about as contributing to the decline is changing demography; Jamie Fox will talk about that and you'll see that the demographic contribution, if anything, these days is a negative one. I hope Jamie makes it here; if not, we'll talk briefly about it.

And with that, let me introduce the first speaker from the panel of authors, and that is Garen Wintemute. And his theme is going to be guns and gun violence.

Garen.

Mr. Wintemute: Good morning. I don't have transparencies, so I'll just speak from over here. I'm in the running for one of those grants. So I have 9:05, Al.

I'm a physician. I practice medicine when I'm not doing research on violence in an emergency department. So I see this from a somewhat different point of view maybe than some of the rest of the panelists. One of the keys to controlling an epidemic, and it's very clear that we are talking about an epidemic here, is to move beyond the assessment and treatment of individual instances of the event that's occurring in epidemic proportions, and to determine the patterns of risk and distribution that characterize that epidemic, and then to address, design, and implement interventions that address those patterns.

Al has very nicely laid the patterns out. I'm very briefly going to mention some of the interventions that were undertaken specifically for their potential effect on firearms, and the relationship between firearms and the events that Al has just described. But before I mention some specifics, I want to just highlight a trend that occurred during this time that was not the result of any particular policy or that we know of constellation of policies, at least in a deliberate way, and that is that over time the rates of firearm sales, which we measure by proxy looking at production, correlate very, very closely with crime rates. They go up when crime rates go up, they go down when crime rates go down—there's a chicken and egg problem here. But I think that much of what we've seen may, and I emphasize may, result from an interruption of that cycle. People buy firearms more for protection than for all other reasons put together. If crime rates go up, the stimulus to buy firearms goes up. If crime rates are going down, the perceived need for protection is less, the number of sales diminish, and there may, in fact, be a virtuous cycle.

And just to put numbers on that, the decline in crime rates that Al just highlighted from 1993 to 1998 or so was accompanied by a 50 percent drop in the production of semiautomatic pistols in the United States and a greater than 75 percent drop in the production of inexpensive so-called "junk gun" pistols.

Now, let me just very briefly highlight some things that probably did work, and also mention two things that demonstrably did not. First, changing police practices, you're going to hear a lot about that. I will just mention the Kansas City Gun Experiment. A controlled experiment, a rare event in criminology, in which police patrols focused on firearm carrying and were associated with a substantial decrease in firearm-related crime in a target area without any change in a control area.

Increasing sanctions. We'll also hear a lot about that. I'm just going to mention Project Exile, very well known to everyone. It's important to keep in mind that Project Exile itself, the Richmond program, which is the prototype for the many replications now in existence, began in 1997, too late to have much of an impact on the picture we're describing. I should also mention that no Exile program has been the subject of a controlled evaluation to my knowledge. And not too long ago I was at a House hearing at which a member of Congress from Virginia, where Richmond is located, pointed out that a substantial number of cities in Virginia have had drops in crime rates comparable to or greater than Richmond's, without implementing Exile or anything like it. So I think that there needs to be an asterisk there.

Let me highlight a couple of initiatives undertaken by ATF. One has to do with more widespread and more complete documentation of the pathways by which firearms move from manufacturer to criminal use. I'm talking about gun tracing here. ATF has, beginning in 1994, begun to aggregate tracing results including all known stops on a gun's path from manufacture to criminal use, from the manufacturer down to the possessor, and as many points as could be filled in in between. And just to give you one example, one of those points, obviously, is the person who first purchases the gun at retail from a federally licensed firearms dealer. One can, with these aggregate data, ask of the computer, "Yo, computer, give me the names of the people who were the first retail purchasers of firearms used in crime and submitted by local law enforcement agencies in New England." And there are a half-dozen people or so whose names show up about 50 to 100 times a year as people who are the first retail purchasers. You've heard the rest of the sentence already. These people are not aggregating their personal collections. And enforcement has become much easier as a result.

The impact, let me say, of a second initiative was to substantially reduce the number of federal firearms licensees in the United States. In the early 1990s, it was trivially simple to get a license to sell guns at the federal level: $30 and an application form on which you could lie because nobody checked your answers, and you had a license for three years. At that time, there were literally more federal firearm licensees in the United States than there were gas stations. That is not hyperbole. There were 244,000 retail federal firearm licensees. There are now somewhere on the order of 60,000 to 70,000, and the number is still falling. From a prevention standpoint, it is to me good news that it remains the case that just over 1 percent of those outstanding dealers are still associated with over half of all the crime guns that are traced by ATF. That sort of concentration of risk makes it easier to focus resources.

The industry argues that these are simply dealers who sell a lot of guns and, therefore, have a lot of traces simply as a normal course of doing business. We have data from California to suggest the opposite. We have sales data in California and have been able to examine the number of traces associated with each dealer as a function of their sales volume. And I can tell you that dealers in California who have more traces than you would predict based on their sales volume account for 30 percent of all the sales, but nearly 85 percent of all the traces, which makes it even easier to focus resources, which makes us optimistic for the future.

My eligibility is slipping. A couple of comments briefly: Screening buyers, Brady and its state-level analogs, waiting periods and background checks, work. They reduce rates of criminal activity involving guns and violence among people who are screened out, who are denied by about 25 to 30 percent. And it's a specific association. They do not affect rates of criminal activity among those people for crimes that do not involve guns or violence. They appear not to have had a substantial, if any, effect on crime rates. And the resolution of that apparent paradox is that under current criteria so few people are denied the purchase of a firearm under Brady and its state-level analogs relative to the number of people who purchase guns every year that an impact on that select group is too small at the population level to be noticed.

Let me just briefly mention comprehensive initiatives. David Kennedy, one of the authors of the Boston Initiative, is here; [it's an initiative] familiar to all of you, undertaken at the local level as a response to locally identified patterns characterizing the epidemic in which efforts to reduce demand for and criminal use of firearms were coupled with efforts to reduce the supply of firearms, particularly the illegal supply of firearms, and were coupled, again, with efforts to provide alternative things for people to do. Boston and, to my knowledge, the other local programs, have also not been subject to controlled evaluations. But the circumstantial data, I think, are pretty compelling.

And finally, and very briefly, two things that don't work. Number one, buybacks. Very popular, perhaps a loss leader as a way of focusing a community's attention on its problem. Buybacks don't reduce crime rates, and I think we know why. Young people never participate in buybacks. The firearms that are returned in buybacks tend to be old revolvers. The firearms that are used in crime tend to be new pistols. They are two different populations of guns. Here in Washington, nearly 3,000 guns were turned in last August at $100 a pop, no questions asked; most of the guns were worth much less than that, and the thinking is that people were making a rational economic decision. The average age of the guns that could be traced in the D.C. buyback was 15 years. New guns get used in crime, and many of the guns that couldn't be traced could not be because they were even older.

And, finally, it doesn't work to solve the violent crime problem by throwing guns at it. Making access to permits to carry concealed firearms on the street doesn't work. And, again, we know why. To very briefly summarize research, there have been probably half a dozen variably well-done studies addressing the question of whether converting a state-level CCW [carrying concealed weapons] policy from "may issue" to "shall issue"—I think most people here know that difference—whether that's effective. A couple of studies have found modest benefits. A number of other studies have found modest to, in some cases, substantial, worsenings in crime rates.

Here's the situation on the street: Some 7 percent of adults say that they carry firearms on an occasional or regular basis, 7 percent, on average 900,000 people a day perhaps. In states that have gone "shall issue," no more than 1 to 2 percent of the adult population has gotten a permit. And it's a reasonable assumption, not proven, that some percentage of those people were carrying previously and simply acted to make legal something that they were doing prior to that.

If changing CCW policy doesn't change carrying rates, arguably it won't change crime rates. And if crime rates are unchanged, well-conducted studies will, in some cases, show no effect, in some cases show a benefit; and in some cases show a detriment, which is, in fact, what we have.

Thanks very much.

(Applause.)

Mr. Blumstein: The next presentation is by Bill Spelman of the University of Texas. I'm not going to give any biographical information, because it's in your packets.

And, Bill, you're here.

Mr. Spelman: I had the benefit of working in a field where there has been a substantial research background over the course of the last 25 or 30 years, measuring the marginal effect of an increase in the prison population on the crime rate. It's only in the last few years that we've gotten a good number for that, but at least we've been ambitious in trying to get a good number for that, and that's actually been helpful in answering the research question.

What I've actually tried to do is come up with a real answer to this question: To what extent—meaning what percentage of the total crime drop—can be attributed to prison expansion? Let me cut to the chase: It's about a fourth, somewhere around 25 percent.

Let me tell you a little bit about where that came from. Here's the true crime rate, right yonder, it's gone up, it's gone down, it peaked around 1991 [or] '92, and then has dropped. This is the violent crime rate. What I wanted to do was to figure out what would [have] happened had we not dramatically expanded prisons over the course of the '70s, '80s, and the '90s. Now, it turns out there are two ways of answering this question; both of them require knowledge of a thing called the elasticity. Elasticity measures the percentage change in the crime rate associated with a 1 percent increase in the prison population. There're two ways to estimate it. One of them is what is referred to as a bottom-up estimate, and it's based on a simulation. You basically simulate 1,000 offenders running around in the computer. They commit crimes, sometimes they're arrested, sometimes they're convicted, sometimes they go to prison, you estimate what happens to them in prison, and the effect of increasing their prison population on those 1,000 little guys running around in the computer. That's a good estimate for the incapacitation effect of prisons, but it doesn't measure the deterrent effect of prisons.

If you want to look at both incapacitation and deterrence, you've actually got to do what's called a top-down estimate, looking on a state-by-state basis over a 25- or 30-year period—well, what happened when we increased prisons in Louisiana by 6 percent, what happened when we reduced prisons in Ohio by 2 percent, and so on—and seeing what the effect is on the crime rate.

Looking at the best of the top-down estimates, you can estimate that the elasticity is anywhere between negative 0.15 and negative 0.45. That means that a 1 percent increase in the prison population will reduce crime by anywhere between 0.15 percent and 0.45 percent. And you can use that to estimate what would have happened had we not expanded prisons dramatically over the '70s, '80s, and '90s. And this is basically what would happen.

It depends on the elasticity; obviously we would have had a lot more violent crime had we not done all that prison expansion, but even so we would have seen a dramatic reduction in the crime rates even with this tremendous prison expansion over the 1990s. How much, then, of that crime drop can be attributed to prison expansion? You can determine that by looking at the difference between what the prison expansion would have been had we not expanded prisons, and the true reduction.

Let me simplify the graph some. Briefly, it looks like this: We had a true crime drop of about 145 crimes per 100,000 people, and had we not expanded prisons dramatically over the '70s, '80s, and '90s, we would have had a crime drop anyway almost as large, about 80 percent as large as the one we actually got. Therefore, I think we can, at least on this first cut, suggest that about 20 percent of the crime drop can be attributed to the prison expansion.

I was a little concerned about this number for three reasons. There are some reasons for believing that prisons are becoming more effective, that that elasticity is, in fact, increasing. The first reason for believing that is scale. We're all talking about a dramatic increase in the scale of imprisonment, we've quadrupled the total amount of people in prison over the last 30 years, and that has actually two kinds of effects. One effect is that we are beginning to scrape the bottom of the barrel, when you have only a few thousand people in prisons, you can fill them up mostly with violent offenders. As they get bigger and bigger, we're putting in more property offenders, more drug offenders, the proportion who are violent has been steadily decreasing over that period, which means that the effect of adding one more bed on violent crime is going to be less than it was 30 years ago.

On the other hand, the scale is going to have an opposite effect in the following way. The elasticity is measuring the effect of a 1 percent increase in the prison population; 1 percent back in 1973 was about 2,500 beds, 1 percent today is about 11,000 beds, more than four times as many. So, if you're talking about a percent of a percent, you can expect the elasticity is going to go up, even though on a bed-by-bed basis probably the violent crime reduction associated with each of those beds is going to be going down. On balance, percent-to-percent stuff is probably more important.

Second, the age distribution of offenders, as Al has been talking about, has been changing. By and large, we've got more violent criminals [who] are adults now than [was] true in the 1970s. What that means is because prisons are for adults only, with a very small number of exceptions, the incapacitation effects and, to a lesser extent, the deterrent effects should be greater today than they were back in the '70s when we had a lot of kids doing violent crimes.

Finally, we may have actually learned something through police repeat offender programs, prosecutorial career criminal programs, through judicial sentencing and parole decisionmaking, we may have learned something about how to focus our scarce prison resources on the most frequent and dangerous offenders. There's a reason to believe that it could be that the elasticity has increased for that reason.

Well, this came to a point where I couldn't work off of previous research. I had to actually get a data set. So I talked to Steve Levitt, who actually produced the best of the top-down studies in the past; he sent me the data, I might add, within 72 hours of my asking for them—go Steve—and I actually replicated his estimates for that elasticity, adding in the effects of scale, age distribution, and selected criminal justice operations.

Briefly, scale is extremely important. There is a statistically significant effect. We can be sure that elasticity really does depend on the scale of imprisonment, and it does drive it up slightly. We can also be fairly sure that the age distribution has an effect on that elasticity. There's no reason to believe once you take scale and age into account that, despite all the sturm und drang and sound and fury associated with career criminal programs and repeat offender programs, that this really is making much of an impact on the effectiveness of prisons.

When you cut through all that, there is reason to believe that the elasticity has, in fact, risen gradually over time. I've got two estimates here, this dotted line I've labeled Model 2, that's the model associated with the elasticity depending only on increases in the scale of imprisonment. This other solid line measures the effect of elasticity on both scale and the age distribution. Either way, you show that the elasticity has been rising gradually over time.

What that means is that since the effect of prison expansion on the crime drop is focused on the prison expansions of the 1990s, and since those prison expansions are a little bit more effective than the prison expansions of the '70s and '80s on a percent-by-percent basis, that that 20 percent figure I gave you a minute ago may be wrong. So I had to go back and do the whole thing again.

That's what happens when you do it all again. It looks very similar to the graph I showed you a minute ago. Here's the true crime rate, dropping in the '90s. Here's what the crime rate would have been had we not done that prison expansion. Again, dropping in the '90s by almost as much at almost about the same time. And simplifying the graph again, in the absence of prison expansion, that crime drop would have been a little bit smaller. Before, based on constant elasticity, assuming each percent is going to be exactly the same as the last percent, I was talking about a 20 percent reduction. Now we're talking about something like a 27, maybe even as high as a 28 percent reduction. And by that what I mean [is that] about 27 or 28 percent of the total crime drop can be attributed to prison expansion.

Now, this is as far as I went in the paper. Let me add one more thing. You can't do a whole lot about the economy. And so you might reasonably think, well, we know something about prison expansion. It looks like it's improving, therefore we ought to continue to build prisons. Don't do it.

Briefly, remember that rising elasticity is on a percent-by-percent basis. We don't pay for prisons on a percent-by-percent basis, we pay for them on a bed-by-bed basis. If you want to know the effectiveness of that next bed, you've got to look at what reduction in violent crime you can expect to be associated with the next bed you build. That's roughly what it looks like.

In the early 1970s, each additional bed we built reduced a violent crime number by about somewhere between three and three-and-a-half. That guy we put in jail, either through incapacitation or by deterrence of other offenders, reduced the number of violent crimes on the street by anywhere between three and three-and-a-half. That number stayed fairly constant throughout the 1970s, since 1980 or so it has been dropping, and especially since 1990 it has been dropping like a rock. That means right now each additional prison bed is going to reduce a violent crime number by about 1.1, about a third as much as it did in the early 1970s. To cut to the chase, prisons—each prison bed is about a third as effective in reducing violent crime as it was in the early 1970s.

One more point I want to add: If you decide today to build more prisons, you will not have that prison capacity tomorrow. You will have that prison capacity in somewhere between five and seven years from now, because you've got to do the architecture and engineering, you've got to construct the thing, you've got to staff it up, and so on. That means that if you're making a decision today about whether to expand prisons or not, you're not dealing with that number, you're actually dealing with whatever that number is going to look like seven years from now.

Given what this entire long term trend has been throughout the '80s and '90s, and particularly what it's been since the '90s, I think we can reasonably expect it's going to go down, and that prisons will not be only one-third as effective as they used to be, they will be something like a fourth or even less effective than they used to be. This suggests that prison expansion is about—we've gone about as far as we can go with prison expansion. There's really no good reason for us to go any further.

(Applause.)

Mr. Blumstein: The next presentation is by Rick Rosenfeld, who will be talking about adult homicide, and with a particular attention adults being 25 and beyond.

Rick.

Mr. Rosenfeld: I'm going to stick to my text, so that I can try to stick to my allotted time. And I'll forego slides.

My story is the story of adults, us. Adult homicide rates have fallen continuously for over 20 years. The decline in adult homicide has not figured prominently in recent—all the attention devoted to violent crime trends in the U.S., which has been dominated by the issue of youth violence. That wouldn't be a big problem for crime research, or homicide research, or policy if adults contributed little to the overall homicide rate. But persons 25 years old or older make up over 60 percent of all homicide victims and nearly one-half of the offenders.

Moreover, with the aging of the baby boom, adults are not only the largest but also the fastest growing segment of the U.S. population. Without the adult contribution, reporters and policymakers would not be pressing crime researchers to provide explanations for the crime drop, they would be asking why the homicide rate was continuing at record levels in spite of declines in youth homicide.

The time trends in adult homicide rates tell a fairly consistent story across the differing victim offender categories for both blacks and whites, males and females. With few exceptions within race, trends are negative, and in all cases declines are greater in the spousal category than for killings involving other acquaintances or strangers.

Two factors, it seems to me, are rather strongly implicated in the general decrease in adult homicide and the specific drop in intimate killings. The explosive growth in incarceration rates, which you've already heard about, and the changing living arrangements among young adults associated with the declining rates in marriage. I won't belabor the prison story.

The rate of imprisonment in the U.S. has nearly quadrupled over the past two-and-a-half decades. It seems difficult to believe that an increase of that magnitude in a prison population would have no impact in the crime rate, including the rate of homicide. Confidence in the results of the so-called top down studies, which you just heard summarized, of the impacts of incarceration growth on homicide trends would be heightened were they confirmed by the results of a distinct estimation procedure.

I employ such a procedure, though I don't have the time to explain the details of it here; perhaps during questions [I will]. And I estimated the impact of recent growth of incarceration on yearly change in homicides from 1980 through the mid-'90s. The procedure I used is a bottom-up procedure, and it produced estimated incarceration effects on homicide that are strikingly close to the recent national level investigations of the sort you just heard about.

Prior bottom-up so-called studies of the incapacitation effects of incarceration on other types of criminal offending have relied on arrest records or offender self reports to estimate the number of crimes averted by imprisonment. Whatever the merits of approaches with respect to other offense types, they are of doubtful reliability for estimating the effects of incarceration on homicide, which even imprisoned offenders may be loath to report in a confidential survey and which are very rare events. In the absence of reliable data for individuals then, an alternative is to use data from the populations from which imprisoned offenders are drawn as the basis for estimating their homicide rate, and therefore the number of homicides averted through some given quantity of imprisonment.

Prison inmates are drawn disproportionately from high-risk, urban communities. A reasonable basis for estimating their homicide rate is the rate for those communities. And that's the approach that I took. And based on my estimates of the so-called offender homicide rate, which are based on the homicide rates of the most dangerous neighborhoods in selected U.S. cities, I concluded that by the early 1990s, roughly one-quarter of the yearly decline in homicides was associated with growth in the number of imprisoned offenders, a bottom-up estimate that's very close to the estimate you just heard. Increases in incarceration do appear to reduce the number of homicides in the U.S. To my mind, that must be counted as good news, but it comes with a huge price tag, and, perhaps, compounding effects that once factored in would dampen our enthusiasm for incarceration as the nation's main crime control policy.

The decrease in adult homicide rates has been extensive; it cuts across all victim-offender relationships. But, as I noted, it has been especially sharp in the family category, and it's not obvious why the incapacitation effects of incarceration should be stronger for victims and offenders who are related to one another than for those who are less intimately acquainted. Although the incarceration growth may have played a part, other changes are also involved. And my focus in the chapter in the book is on the decline in marriage. Falling marriage rates reduce intimate partner homicides through a simple mechanism, opportunity.

The decline in the spousal homicide rate dominates the trends in the intimate partner homicide more generally, and therefore it's reasonable to ask how much the plummeting rate of marriage in the U.S., especially for young adults, might have to do with the decline in intimate partner homicide, spousal homicide in particular. I evaluated the extent to which the change in the number of married homicide victims is attributable to the change in the number of married persons between 1980 and the mid-'90s. And I found that the drop in marriage rates had sizable effects on the victim trends, accounting for 40 percent to as much as two-thirds of the decline, but only for the younger adult age groups.

The number of spousal homicides, however, has also decreased for the older adult groups, those between the ages of 30 and 44 years old, and none of these declines is explained by changes in the number of married persons. Some of the drop in intimate partner homicide among the baby boomers might be due to the influence of domestic violence services and improvements in women's status. The expansion in incarceration also might have played a role, both by removing violent men from families and by reducing marriage and family formation in the most violent communities. Additional research is needed to test such speculations about the demographic social and policy changes associated with the decline in intimate partner homicide. Researchers, however, should also consider the possibility that cultural transformations have, in effect, reduced the tolerance for interpersonal violence in the U.S.

The decline in adult homicide has been substantial, widespread, and enduring. Although such predictions are always hazardous, it does not appear to me that this decline is part of a cyclical process that is likely to turn up soon. If anything, it is a countercyclical movement. The adult rates dropped during periods of high and low unemployment, both falling and rising income levels, and both the growth and decline phase of the youth firearm violence epidemic. It follows that the conditions responsible for the downward trend in adult homicide themselves must be relatively enduring features of the social and political order.

Alterations in the living arrangements of young adults, brought about by declining rates of marriage and the incarceration experiment of the last quarter of the 20th century, are, at most, contributing causes of the decline in adult homicide. Neither alone nor in combination do they fully explain the breadth, depth, and length of the decline. Nor are alternative accounts based on standard economic or criminological theories readily apparent. It has proven less difficult in retrospect to explain the abrupt and unexpected youth homicide epidemic of the late 1980s and early 1990s than the long and steady decline in adult homicide.

If church is the last refuge of scoundrels, culture is the final recourse of social scientists in search of explanations when existing economic, social, and political theories have been exhausted. Cultural theories are not unknown in criminology, but they're usually invoked to explain high or escalating levels of violence. An important exception is Norbert Elias's theory of the civilizing process. Elias's thesis has been invoked to explain the centuries-long decline in interpersonal aggression in European societies. Some debate exists regarding its application to the United States with our traditions of self-help and resistance to centralized state power.

However, a modified version of his argument does seem consistent with what two observers suggest is a growing intolerance for interpersonal violence, lethal violence in particular, in the [United States]. Americans have enjoyed steady declines in traumatic injury and death in recent decades on the job, on the roads, and in their homes. Those declines, in turn, have created heightened social expectations of progress in public safety.

In such a climate, it would not be surprising to observe elevated levels of fear, even if overall rates of crime decline, or the application of so-called zero-tolerance policies in schools where levels of violence are already at or near zero. Nor is the expanded use of incarceration as a means of social control inconsistent with a growing intolerance for violence in everyday life. The modern prison substitutes for more visible and disorderly forms of violent punishment and, in fact, is one of the major achievements of the civilizing process.

Finally, we should expect a society with strong democratic traditions that is undergoing the civilizing process to elevate freedom from interpersonal violence to the status of a right, including a woman's right to be free from male violence. With rights, however, come responsibilities. If something like the civilizing process is finally at work in the U.S., we should expect it to influence the behavior of adults before it affects youth. The baby boom generation has enjoyed the benefits of declining levels of lethal violence since attaining adulthood in the 1980s. It now must extend those benefits to the next generation.

The challenge for adults is to find ways of insulating themselves from violence that do not compound the problem or shift its burden to their children. Massive incarceration may avert homicides in the short run, but at the expense of diverting resources from longer term prevention efforts. Avoiding marriage may reduce the risk of violence for adults, only to increase it for children who grow up in communities devoid of stable, two-parent families. More prisons and fewer families are, at best, temporary, limited, and negative remedies for violence. The baby boomers must leave a larger legacy of nonviolence that befits a genuinely civilized society.

Thank you.

(Applause.)

Mr. Blumstein: The next factor that we want to consider is the changes in the drug markets, which have been an important stimulus, and that's a paper being presented by Bruce Johnson, based on a chapter he wrote with Andy Golub and Eloise Dunlap.

Mr. Johnson: Thank you, it's good to be here.

I want to talk to you about the contributions of changes in drug patterns, and in particular changes in patterns of drug markets, and how that's impacted upon violence in America. And in particular, too, both its rise in the late '80s and its decline in the 1990s.

In particular, I want to start about early 1980s with the fundamental problem that American society had, and still has, which is the institutionalization growth, and ghettoization of the inner-city, where you have large concentrations of poor people living in relatively poor conditions, and in particular, a particular subgroup of that population that grew up in what I call seriously distressed households, which is defined in more detail in the paper. But for the most part, people growing up in such households in the inner-city had remarkably few—acquired few skills for functioning in conventional society. They had low numeracy, low literacy, low interpersonal skills that would allow them to function in the general society.

At the same time, they lived in households that were quite good at teaching them how to survive in what Elijah Anderson called the street culture, the street codes. They knew how to hustle, they knew how to get drugs, they knew how to manipulate people, all skills that are particularly what the criminal justice targets. And that's provided a setting for the explosion during the crack era.

Now, I would say much of our chapter is organized around information that's been developed from an ongoing set of studies by looking at changes in primary drug preferences among persons arrested across the 1987 to 1997 period, in terms of what the primary drugs people report using, or are detected in their urine. And the key findings skip over heroin and marijuana a lot, and focus mainly on crack for a variety of reasons.

The important thing is that the birth cohort, or what I'll call the generation that you grew up in, largely determines, at least in the inner cities, and among arrestee populations in New York City, these patterns have been replicated fairly similarly in many other cities, with different trends and troughs, I might add, that people born during and after World War II primarily got heavily involved in heroin, but people born around 1950 who were then reaching—1955 who were then reaching their adulthood in the late '60s and early '70s, that was what I call the heroin generation, and after that persons born primarily in the '60s show this massive drop in heroin injection. You can see the line coming down there among those born in '57 through the '60s.

So now in New York City, at least, less than 10 percent of the persons who are arrested in the 1990s self-report that they've ever injected heroin. Heroin injection has become a very rare behavior among very high risk persons. It's not because there's no heroin, there's plenty of heroin on the streets, it's not because there's no needles, there's plenty of needles on the streets in New York. It's a function of the choices that people are making of which drugs they use.

Now, the second and the main focus of the paper is on crack. Now, crack is a substance that came along, just a brief history of crack, it grew out of freebasing that came out of the early 1980s, and hit New York very heavily starting in 1984-85, and became a major issue and explosive growth, which I've documented in lots of other papers and products, during the last half of the 1990s, and continues well into the current time. And it's particularly people born roughly 1955 through about 1969 that have become the primary consumers of crack cocaine. They are the main people, and that's true down to the current time. Now, they've just gotten older. But, in 1980, in the late 1980s, those persons were usually in their late teens and early 20s, and they got into the crack market, which I'm going to describe briefly, shortly.

The persons born—but, for reasons that are still unclear to me, but I'll come back to it—persons born in the 1970s turned off to both crack, which you can see this dramatic decline in crack use among people born starting roughly in the 1970s down to the current time, and that continues to the current time, as far as I know, though we have yet to update these. At the same time marijuana has been strikingly on the increase, and, in New York, the main focus of that is around the use of blunts, and I'll explain that a little later shortly.

Now, that's sort of the empirical picture. And the fundamental question is if this is true, why is it true. And that's what the rest of the paper is basically devoted to trying to describe and understand. And it's probably too far beyond me to do any more than trace the very core of what my argument about its link to violence is. But basically there's things that I call subcultures which are focused on very narrow behavioral areas. These subcultures set the rules that people follow and how they do it. And I've got 15 to 20 different sort of subcultures that are discussed in the paper. I won't go through them all now.

The key element, however, is that one of the key things fueling the link to violence and to homicide was the rapid growth in markets for illegal drugs. And the development of crack cocaine, and especially crack markets, led to an explosive involvement of a very sizable segment of the youth in the inner-city. The crack market became like a whole new illegal drug market, that came to dominate almost every other drug that was out there. Almost everybody who sold drugs illegally got involved in selling crack, at least at some point in their career.

And that led to an explosive growth over territory. They had a lot of impact. They got themselves armed with those handguns that were being imported and purchased. They had a lot of disposable cash. And so violence rates and homicide rates went up considerably in New York City. And I might add, that led to a major and explosive growth in the demand for increased public order on the part of everybody in American society. And during the late 1980s and early '90s, the New York City police department and the entire society mobilized against drugs. The mantra was to take back the streets from drug dealers. And I would say during the last half of the 1990s that has largely been accomplished.

It doesn't mean drug dealers aren't out there. It doesn't mean that crack selling is not occurring. But it does mean that there has been considerable effort made to restore public order. Unfortunately, the evidence I've got indicates that it's not really changing people's patterns of consumption, so that the people who continue to be chronic crack users, or continue to use heroin, or who use marijuana, are still quite capable of getting their drugs, it's just not happening so much in public places. And police are making it very uncomfortable, and they're not very accommodating to the drug sellers out there in the streets. But the drug sellers still move around and marginalize themselves.

So the key element, however, is that the generation born in the 1970s, who's taken up marijuana with a vengeance, is one who has made a distinct decision not to shoot heroin and not to use heroin and not to smoke crack, and they're, for the most part, not using cocaine, which is very good news. Indeed, the fundamental element—and I want to go back to the theme that Rick just mentioned of civilizing effects—we could do far better if we could civilize the population now that is now most marginalized.

Our public policies are mainly designed to push people not into employment, that's their intent, but in fact people who are coming out of these seriously distressed households aren't doing that very well. They're not making that public market.

So with that, I'll end and we can talk more later.

(Applause.)

Mr. Blumstein: Thank you very much, Bruce.

My sense of this transformation of the drug markets is that it has a metaphor like the transformation of street prostitution from street walkers to call girls. Street walkers are an affront to the community, they represent targets for policing. Whereas, the door-to-door delivery that we're seeing in many drug markets now represents far less of a threat and far less of a problem to the broader community because of the market. There is still the concern about the users, and so on. But, I found that graph that Bruce presented a really striking illustration of the generational change.

The next chapter is on policing. It's been done by John Eck and Ed McGuire. John is from the University of Cincinnati and Ed is at the George Mason University. And policing is one that has had just an enormous array of boasts and assessments that this is what's doing it, and this chapter really tries to lay those out.

John?

Mr. Eck: The fact that I'm up here instead of Ed has to do with the fact that Ed lost, or won, the coin flip that determined the precedence of authorship on this. So we're equal partners in this. Interestingly enough, we didn't know each other before we started this. By the time we finished it, neither one of us had met each other. If I'd run across him in the hall I would not have known who I was looking at. We did the whole thing by e-mail. It's I think a testament both to the power of current technology and to having a great partner like Ed McGuire.

We faced a major challenge in looking at the police, several major challenges. One is the police have multiple strategies that they use. They could all plausibly be linked to the declining crime rates. And second, there's no single data set or group of data sets that one could go to and produce marvelous little graphs showing police actions at every year and mark that against crime. It's a hodgepodge of a whole bunch of different things, and we had to put these together in some meaningful way.

What I'm going to try and do is summarize very quickly what we know about seven different known strategies. There might be some other strategies we haven't looked at, but I suspect these are the biggest ones. Let me, following my friend Bill Spelman, cutting to the chase, our bottom line is we don't know.

For the most part what we're looking at here are things which are local initiatives. They're things which individual cities, police agencies have undertaken, and sort of become aggregated up to a national trend. Not that police chiefs and mayors don't talk to each other, and not that the federal government does not have some role in promulgating ideas, but with a couple of exceptions, they are not part of a national policy in any meaningful sense.

Now, in order for us to make sense of this hodgepodge of information, we decided to set a set of relatively conservative criteria for determining whether some kind of police intervention could plausibly be linked to the reduction in crimes since roughly 1991. First, we needed to have something that was a relatively coherent strategy, that you could say, yes, that department is doing it. You could know when they're doing it and know when they're not doing it; it wasn't too broad and fuzzy.

Second, the next hurdle was that it had to be a strategy which there was some research or evaluation evidence that suggested that it worked. That there was some evaluation, at least one, out there that said, when you do this, crime went down; or, if you do this, crime doesn't do anything. But there had to be empirical research out there linking these things up.

The third criterion was that if police were to be in some sense a cause, or a partial cause, of this drop in crime, whatever the police were doing had to really start before the drop started, right. It doesn't make any sense for the cause to come after the effect. So we said the things that are most plausible are likely to be things that were started roughly before 1991, or at least had a large police involvement before 1991.

And the final hurdle that a truly plausible police strategy would have to confront was that it had to be adopted nationally, it could not be idiosyncratic to one department, even if it was one very large department. And unfortunately, there's not a lot of information, certainly systematic information, about the take-up rates of police departments on these various strategies. So a number of these strategies sort of failed that test.

Now, I need to make it clear here, failure to clear any one or all of these hurdles does not mean that a particular strategy doesn't work if it's used. It might be very effective at controlling crime. What we're saying is—we're saying one of two things. Either it made no contribution to the national decline; it may have made a major contribution to a particular city's decline; or what it means is that we have no information. It did, but no one bothered to collect the right data, so we just don't know. And it's hard for us to distinguish which of these are out there. But what we can say is there are some things which are probably implausible, not good candidates for attributing—for making claims that the police made a contribution, other things which make a little bit more sense.

In an absence of great graphs, we have generated an uninterpretable chart. This chart has roughly the same relationship to our findings that a political cartoon has to a political debate. It gets a few of the major features here, but it doesn't really do justice to some of these things. You do have a copy of this in your packet if you cannot read the thing up here. By the way, I tell my graduate students in the methods class about preparing charts, don't do things like this. Fortunately, I don't think any of them are here.

If we go down to the list of criteria, the tactics or strategies which are not plausibly [thought to] have contributed to the decline in crime due to weak evidence. In other words, there were few or no empirical studies suggesting that these things work. Falling into those categories [are] three things.

One, adding more cops, in this area there is a huge amount of research. And it seems to be an arms race between statisticians on both sides of this issue. One group of statisticians will come up with results that say more cops reduces crime; that will be followed within a couple of years by another group who will say just the opposite. And then somebody will increase the rigor and sophistication and reduce the number of people who can understand the technique, and this will repeat itself. And there doesn't seem to be any reduction in that. So given that, the evidence there is equivocal at best.

Also zero-tolerance policing, there's a fair amount of—well, it's some evidence about some of the background theory about whether the police can actually have an effect on these things through zero-tolerance policing is unclear. And community policing we throw into there, largely because community policing doesn't meet that criterion of being a coherent strategy. It may be a coherent strategy in a particular police department. You can go to Chicago or any number of other cities and talk about their version of community policing. But, when you put all these things into the same group the number of common elements diminishes dramatically. And it's hard to say what's going on. It's safe to say that anything reasonably innovative, or any change that a police agency has done since 1990, has been labeled a community policing strategy, so it's very difficult to say that community policing could be the—could be parts of it, perhaps.

Candidates that we could claim are not plausible due to late start. They began sometime after 1991, and there's some overlap in these candidates. The national effort through the COPS office to put 100,000 officers on the street. It clearly did not start before 1991. That can be ruled out very easily. Efforts like the zero-tolerance policing concept, particularly in New York City, started in 1994, well after the decline in crime. Now, there's an equivocal aspect to that. Police had been experimenting with things like directed patrol and crime analysis for quite some time.

So it is possible—though we don't have great evidence to this—it's quite possible that police agencies have been adopting something that would fit this model for many years. So it might be that police have been doing some kind of directed patrol thing long before this, and that as the police have become more efficient this has helped. Again, we don't have any evidence to that. Also, I would throw in firearms patrol. Not, again, [that] this is not effective, but that they seem to have a late start in this. We don't have any evidence of national implementation of this prior to 1991. And that may be just due to lack of information.

Limited implementation, in other words. We don't have any evidence that it was implemented throughout the United States on a wide scale. Zero-tolerance policing. A lot of talk about it; not clear what's been going on. COMSAT also. A lot of agencies put labels on this. Whether they're doing it, not clear. I mentioned firearms. Problem-oriented policing falls into this category, although it has been developed and implemented before 1991 and gotten into a lot of police agencies before 1991, but the quality of the implementation is pretty weak. So unless this is one of those few things which you can do poorly and get great results, that doesn't seem to be a plausible candidate.

What we have left here is drug enforcement. That clearly was implemented before 1991, there's some empirical evidence that that kind of thing worked. It was done nationally. So we're left with that, by process of elimination.

Let me close here by saying that the police contribution to changing national trends is going to be very difficult to discern, partly because of lack of information, but also because how the police become involved in attacking crime, they don't do this by themselves. At the same time city councils, city managers, mayors, states, the federal government is pushing the police to do something about a particular issue. A lot of other institutions are being pushed. They reinforce each other. So police work with a lot of groups. If you look at what the police do in the drug area you see a whole variety of things, partnering with a variety of groups long before partnering and coalition building became a positive buzzword.

So in many respects, it would be silly to say the police had no contribution. But it's also difficult to make a strong statement [that] the police had a major, independent contribution apart from all these other institutions out there. But the bottom line is, one can't really put a lot of credence to the strong statement about the police having a huge independent role in reducing crime, particularly homicide.

Thank you.

(Applause.)

Mr. Blumstein: I think the issues that John raises are really endemic throughout the attempts to get an assessment of effects within the system, and whether it's policing, and particularly as you try to assess any particular technology that's associated with intervention by the criminal justice system. It's very tough to isolate those, compared to the others.

The next presentation is by Jeff Grogger of UCLA, and he's going to be talking about the changes in the economy, and economic incentives that have been an important part of the change.

Mr. Grogger: My chapter in this volume presents an economic analysis, the rise and fall of what I'm going to term instrumental violence. It focuses on the role of illegal markets, particularly the cocaine market, and it incorporates the observation that in illegal markets such as the cocaine market, such as the crack market, violence is instrumental. This is an idea that has been out there for a long time. It's expressed in different ways, depending on the analyst's persuasion. But the basic idea goes like this. Traders in all markets need some means of resolving disputes. And where participants in conventional markets have access to conventional means of dispute resolution—the courts and the civil justice system more generally—participants in illegal markets for obvious reasons don't have recourse through these types of dispute resolution mechanisms. What they need is an alternative, and violence turns out to be an alternative that serves that purpose well.

So we can think to some extent at least that violence plays the role in illegal markets that's played by lawyers in conventional markets. Now, this observation implies—if we link violence to illegal markets, the observation implies that when illegal markets expand, violence expands along with it. Now, why would illegal markets expand in the 1980s? The obvious place to look is at the emergence of crack. This has been conjected for some time. I can say there has been recent quantitative evidence by me, as well as by others, that suggests that the effect of crack not only gives credence to the idea that crack had something to do with the rise in violence, but actually suggests the emergence of crack was very important in terms of its magnitude in explaining the rise in violent crime in American central cities during the 1980s. So in other words, the model explains the increase in violence that we observe in this time by the emergence of crack cocaine.

Okay. So much for the increase, what about the decline, which is after all what we're here to discuss. The explanation for the decline in violence turned on the following observation, that is that even though violence may provide dealers, drug trade workers, however you want to think about them, with an advantage over their rivals, it also raises risk, and it raises the cost of doing business. And it raises risks not only for the people involved, but basically for everyone else in the market. It's possible that when these risks are high enough drug trade workers may leave the market, choosing other alternatives, and potential entrants may forego the drug market, as well.

So the implication here is that as risks rise the supply of workers to the crack trade falls. Buyers may beg off as well if the risks associated with particularly localized markets become too high. The result being that the market, and in turn the violence associated with it, declines. So in other words, the rise in violence itself may give rise to negative feedback that eventually raises risk and causes violence to fall.

Well, that's very well and good. How does the economy play into this? The economy plays into this through the labor market. And the reason is, we think of the labor market, particularly the labor market for low-skilled workers, as being the alternative to the crack trade. Another way to think about this is, who is it that may be lured into the crack trade? Well, typically it's going to be young, disadvantaged men with little in the way of tangible labor market skills. Now, economic logic tells us that the poorer the labor market the more effective the drug trade, and conversely, the better the labor market the less effective the drug trade.

This leads to a question: How do we characterize the attractiveness of the labor market, which may on the one hand lure people away from the crack trade or conceivably push them into it? Now, this sounds like a side issue, [but] it actually turns out to be important. The reason is that in the past analysts have typically thought of the link between the labor market and economically motivated crime, of which the drug trade would be one example, as working through the unemployment rate. There's been a lot of research on the link between unemployment and crime; much of it has concluded there isn't much of a link. And, in fact, many reviews of the research literature in this area have concluded there isn't much of a link, which has led people to discount the importance of the economy.

There's been a lot of recent research lately, however, that thinks about the link between crime and the labor market as working differently. Rather than working through the unemployment rate, they think of it as working through wages. And the distinction turns out to be important for a reason that I think is well illustrated by the first table in the chapter.

Without going into a lot of details, let me just tell you that this is data from a particularly at-risk group of young men, men between the ages of 16 and 23, who were sampled in the 1980 wave of the national longitudinal survey of youth, which is an ongoing survey. Now, 1980, because even though this is a longitudinal survey, 1980 is the only year in which these young men were asked whether they had participated in crime, or specifically had any income from crime in the past year. At the same time they were asked whether they had worked in the past year.

Now, what do you see in this table. There are two striking features. The first is, 24 percent of the men indicated they had earned income from crime in the past year, which seems like a pretty high prevalence rate; 76 percent said they hadn't. The more important observation here stems from the next line on the table, however. Roughly 95 percent of these young men indicated they had worked in the previous year, and that number is independent of whether or not they had committed crimes. So if we think about this as—if we ask ourselves how much of an effect do we expect on the basis of this information between unemployment and crime, the answer would be relatively little. Work seems to be ubiquitous. People don't specialize, they work, [and] some commit economically motivated crimes as well.

Now, what the wage theory of crime tells us is that the higher wages are, the less likely young men are going to be to engage in economically motivated crimes. And this data provides a little bit of evidence for that. What we see is that the noncriminals, the self-reported noncriminals, earned about $5 an hour in their previous jobs, whereas the self-reported criminals earned only about $4.34. So there's about a 15 percent differential in the wages of the criminals and the noncriminals.

Now, there are a couple of other issues that remain, because after all we're interested in explaining whether the labor markets, through wages, can explain at least part of the decline in crime in recent years, in particular violent crime. Well, there are two things we need to know. We need to know what the strength of the relationship is between wages and property crime, and we also need to know what has happened to wages, particularly wages available to relatively unskilled young men in recent years.

As for the strength of the relationship, when I subject these data to a fairly substantial bit of econometric machinations, I end up with a number that says the elasticity of criminal participation with respect to wages is about minus one. So what that means is a 10 percent increase in real wages would lead to a 10 percent decrease in participation in economically motivated crime among these at-risk young men. Now, that sounds like a large number; however, as I said, there is a lot of recent research in this area. Most of the research—this puts the elasticity at a number of minus one, or even larger, as high as minus three. So there seems to be a really important link between the wages that are available in the low-skill labor market, and economically motivated crime among young men.

Now, the next question is, what has happened to wages recently? And it's interesting to contrast what has happened recently to what happened for many years before. What this shows is that for the period 1979—what I have graphed here is median hourly wages, and these are in constant 1997 dollars, paid to 16- to 24-year-old males. These tend to be a relatively low skilled group, relatively low education, and relatively low labor market experience. The wages that [were] available to these young men between the period 1979 and 1993 fell, and they fell a lot, from about $9 an hour to about $6.75. There's some deviation from that downward trend, but, on the other hand, if we were to push the axis back further, we would see that this decline really began in about 1973. So for a 20-year period the wages that were available on the labor market to young men fell and fell quite precipitously.

What has happened since 1993, well, since 1993 we have seen a rise. What this suggests is that we—I think this goes to the longest peacetime expansion in the United States's history. As of 1993 the expansion reached into the bottom of the earnings distribution, began to raise the wages that were available to low-skilled young men. So what does this say about wages as an explanation for the decline in crime. There are two things that suggest an important role. One is the strength of the relationship between wages and economically motivated crime; the second is the timing. We see the increase of wages beginning around 1993, which is precisely when we see the decline in youth homicide. All the evidence together suggests that the expanding economy, particularly the improved wage opportunities for young men, may have played an important role in the recent decline in violence.

That's it.

(Applause.)

Mr. Blumstein: The next paper, which is the last chapter in the volume, is by Jamie Fox, who as I gather at the moment is up in the air in the fog, waiting to be able to land at National Airport.

As a weak substitute, let me at least try to convey some of the issues associated with demographic change, and it's the one that I think is most frequently misstated in the press, because as the press reports on factors contributing to decline, demography is always listed as one of them, as a contributing factor, and this chapter will at least try to cast some light on that.

So unless Jamie shows up within the next few minutes, I want to revert back to some of the material I showed earlier. You'll remember that there was a peak in violence, and one of the interesting factors in this graph is how close the homicide rate and the robbery rate track each other. Both having—my concern is that this machine doesn't get any better, it's at the limit. And I apologize for the machine.

You'll notice, if you squint carefully, that in 1980 there was a peak, and then it came down. One of the more commonly accepted explanations for that shift was a composition shift. The baby boomers were moving into the high-crime ages in the '70s, and moving out of them in the '80s, and I'll say more about that with some demographic data. And this decline was attributed in large part to that demographic shift, that decline in the early '80s. The crime in the late '80s was very much attributable to what was going on in crack markets. That cut across some of the discussion that we just heard. And the decline of the '90s, gee, demographics was the explanation in the '80s, why isn't it equally an explanation in the '90s, and I just want to say a bit more about the weakness of that explanation.

Let me revert to a graph I showed you earlier. For homicide, the peak had traditionally been 18 through 24, lower rates below 18, lower rates after 24, and what you see here is the flatness of the peak. This graph is 18-year-olds in red, and 24-year-olds in yellow, and the other in-between ages in between. The rates were fairly flat, and the flatness of the peak is reflected by the fact that those colors intertwine. The fact that they're all roughly the same level suggests that the peak was rather flat; again, consistent with the demographic explanation, within each individual age there wasn't much change going on, but the number of 18-year-olds and number of 24-year-olds increases. So there were more people at those peak ages. And that's the basic thrust of the demographic argument. Are we seeing more changes.

This is reflected in a graph I didn't show you, known as the age-crime curve; that is, what's the arrest rate of people of any particular age. What you have on this horizontal axis is age, and the traditional graph—if you can make out that light green graph down here, with a fairly flat peak from 18 to 24, that's the graph that prevailed from 1970 to 1985. Fairly flat, relatively low at the younger ages, under 18, and falling off fairly sharply, coming down to about half the peak at about age 35.

If you look at other crimes, like robbery, the peak is somewhat earlier, at about 18 or 19, falls off to half the peak by age 25. And if you look at burglary, the peak is a little earlier, about 16 or 17, and it falls off very sharply, reaching half the peak by age 21. So that the age composition in those high-crime ages, the number of people in those high-crime ages is really a key factor in the demographic shift.

Just as an elaboration, this is what it went to in '93, as if someone grabbed the 18-year-olds, this peak here is at age 18. Someone grabbed that graph and pulled it up by a factor of 2.5, dragging all the other adjacent ages with it. The changes since '93 have been close to a restoration of that earlier story. As that peak has been brought down—and the '99 results are what you see in this vaguely yellow graph—that is relatively close to the picture that prevailed. And through all of this, these over-30s have been coming down.

With that background on the interaction between particularly age composition and criminality, I want to show you a graph of the age structure of the U.S. population. The peak of the baby boom are the people who are now about 40. Those are the people who 20 years ago in 1980 were 20, right at or near that peak crime age. These are the 40-year-olds. If you want to look at demographic shifts, just think of this rough picture of the population—of the age structure moving through getting older a year at a time, and moving through the population. So if you want to look at the picture 20 years ago in 1980, then you just subtract 20 from each of these ages, and that moves folks through.

There were the baby boomers, they were the folks contributing to that major growth around 1980. The smallest age group in the post-baby-boomer generation is now—this is the graph for this current year. [The] smallest age group is about 24 or 25. Those folks therefore have moved—the goodness of this decline has already been experienced. We're now on the upswing of people in the high-crime ages. So that we're seeing a growth that will persist from about now, and have been experiencing that for a number of years, when one thinks of age 20 as the high-crime ages.

So for at least the last four years we have been suffering the changing age composition in terms of crime. There are more people in the high-crime ages. Why hasn't that driven crime up? Because that's changing at roughly 1 percent a year, whereas those other changes that you saw in this graph, for example, these changes, these declines are occurring at the order of 10 percent a year. So the criminality of 18-year-olds is getting better at about 10 percent a year, and the availability of 18-year-olds is increasing by about 1 percent a year. So that the netting out benefits the crime rate, because this dominates the changes in age composition.

As we settle out and see the factors contributing to the decline within ages, then demography starts to become an important issue. So that we can see this growth, which is about 1 percent a year, that will run for about 20 years, contributing to an increase in crime, but that increase won't show up until we saturate on all the benefits coming from the reduction that have been addressed in the previous chapters. So that demography is an important part of the story, and at this point it's working in the wrong direction. After another 15 years or so we might see it working in the right direction.

Now, if you were to partition this by race, you would see a larger growth in the numbers, because of differential fertility, but that the growth rate of black young folks, African-American young folks, will be increasing at a slightly higher rate, about 2 percent a year. So that those demographic changes represent considerations in the future.

It's not something we can do anything about. But it does represent targets where the prevention theme becomes an important one, to be focused on the individuals showing up in the growing numbers of people in the younger age groups, prevention in the sense of giving them the opportunities to function in the economy that at one point downstream no one can predict terribly well, may be the stock market is, in an economy that's going to become tighter, tougher. And when that happens, that represents the kind of forces that might contribute to a turning around of the good news that we've been seeing.

This really covers the themes that we wanted to address in the book. Let me turn the conference back to Jeremy.

(Applause.)

Mr. Travis: I don't know if you're as fascinated as I am by drug courts. But there's this wonderful moment in a drug court graduation ceremony where the judge asks the audience to applaud some conforming behavior of folks who have been in the drug court program for a while. And I'm not saying anything about the tendencies or the addictions of researchers to talk longer than their allotted time, but would you join me in applauding them for their conforming behavior.

(Applause.)

Mr. Travis: So what we want to do now is to take a few minutes to ask not the presenters, but the others who have been sitting and listening on the panel, to do one of two things before we take a break. And that is to nominate a major social trend or force observable within our society, or an area of policy that has not been addressed in the presentations that we just heard. This is, after all, a book of seven chapters. And as expansive as they are, they are inherently limited. The second thing is to say that within one of the presentations there's something that wasn't sort of adequately attended to in the presentation, something that we need to sort of think of in slightly different terms.

So this is an opportunity to sort of get some things on the table, before we take a break, that have not been put on the table. It's not a chance to talk about elasticity much more, nor methodology. It really is just to say, look guys, you just missed something that's really important. So I'm looking for nominations and quick introductions of a new topic.

Yes, Wes Skogan.

Mr. Skogan: One is the role of alcohol and crime, probably the oldest and best-understood factor. Studies on the homicide victims in the '50s and '60s, for example, would find that either the victim or the offender was drinking in 80 percent of homicides. And [drinking is] a well-known fueler of assaultive violence and domestic violence. And there have been, in the last 20 years, precipitous declines in the volume of alcohol sales in this country. And all of the factors that we talked about—about intolerance, possible intolerance of violent behavior—it seems to me the evidence for intolerance of alcohol-related behavior is actually stronger, and that probably a whole chapter on the role of alcohol and crime and changing patterns of sales and norms about alcohol abuse, which would also spill over into traffic and other areas as well—it seems to me would have been a really important contribution.

Mr. Travis: I'll get to all of you before we break.

Bev Davis.

Ms. Davis: Well, one of the things, and I'm glad you said that, I think not only in terms of—I think there's got to have been an incredible investment by the Robert Johnson Foundation, Casey Foundation, et cetera, into addressing underage drinking and the whole alcohol thing. But I think one very key thing that we missed or that could be added to is the fact that we did not look at the environmental prevention efforts, and the really active, very, very active efforts of crime prevention coalitions, anti-drug coalitions, drug courts, et cetera.

I mean, when you look at it, there's 3,000 crime prevention coalitions; there's 5,000 anti-drug coalitions. These are people who are working on the ground level to really do environmental prevention, to change norms, attitudes, behaviors, and to really drive policy at the local level. And I think that's had an incredible effect about changing what people will tolerate in their neighborhoods.

Mr. Travis: Alcohol issues, energizing of community. To the other end, Peter Reuter.

Mr. Reuter: Immigration is my nomination for something that over the next 20 years might be important and is missing in the volume, and there are lots of ways in which immigration can have an impact. Social capital is the flavor of the year, and so one can certainly talk about increasing heterogeneity and increasing conflict in inner-city neighborhoods that can arise from influx of poor immigrants and the changing mix of those immigrants. So that's my nomination.

Mr. Travis: I see we're developing the next book.

Tracy Litthcut, did you have one?

Mr. Litthcut: One of the focuses that we had a clear focus in Boston was around the local and national influx of gangs and how that associated to weapons, the increase of crime, [and] drugs, on the local level as well as on the national level.

Mr. Travis: We'll come down to this end. Elijah Anderson?

Mr. Anderson: Yes. I think it's very important to consider the nature of the social organization of the inner-city black community, and how the economy is impacting on that community. But also the fact that faith-based organizations, faith-based movements, have had a tremendous impact on the nature of that community, particularly the Million Man March. We haven't really grappled with the impact, the effect, of that kind of movement, not just that but also the faith-based organizations that are having an impact. But I think the nature of the social organization of that community has to be appreciated.

Also, the fact that people that I've been able to understand and relate to are beginning to see what a game is being run on the community. And that is, if you commit crime, if you commit violence and go to jail, you're subsidizing in some sense people who might be your enemies. When you go to these communities in southern Pennsylvania or upstate New York, the people who are working there who have no sympathy for your plight are getting benefits. They have jobs, they have benefits, and the people in the community who have begun to see this and there's a movement to not be this kind of pawn, if you know what I mean. And so these are just some considerations that I think should be researched.

Mr. Travis: On the first one, you would argue that there's been an increase in the involvement of the faith-based community in the African-American community over time that might have a relationship to the crime decline. It's not just a constant factor, but it's an increased factor.

Mr. Anderson: Yes.

Mr. Travis: Jack Calhoun.

Mr. Calhoun: Thank you, Jeremy, both for the book and for this discussion. I think, to me, a rather stunning lack, and picking up on Beverly's theme, is the comprehensive planning efforts, where mayors and chiefs are saying nothing will change unless all key segments of the community have rolled up their sleeves making specific commitments—the faith community, business, schools, cops, and everybody changing the way they do business. There's been a lot of talk about these sort of large objective structural issues making a modest difference—economy, demographics, crack, markets. But nobody has wrestled with the absolutely breathtaking drops in such cities as Fort Worth, 56 percent; San Diego, 46 percent; Boston, 30 percent, where Tracy is working—where it has been a citywide effort to co-produce safety and co-produce prevention. And I think that frontier, writ both small, in weed and seed neighborhoods, Garfield and Phoenix, Milwaukee, some of your research at NIJ has shown, and also writ large in jurisdiction-wide efforts. It is an enormous gap that I think has got to be addressed.

Mr. Travis: So a confirming observation here about coalescing of community capacity, and you add to that sort of an in essence top-down organization of that as well as bottom-up.

Demetra Nightingale?

Ms. Nightingale: One additional issue that I would be interested in hearing more about as someone who is outside of what most of you here already know is disaggregating the trends by gender particularly, and also looking at the family status in terms of parents. Because even in some of the cities that were just mentioned that have had large declines in overall crime rates, what we're hearing in regards to welfare reform is that there's an increasing number of young parents, young mothers, in some of those same cities that are having overall declines in crime.

Mr. Travis: And Lawrence Sherman.

Mr. Sherman: There's a perspective that is tying together a lot of the comments that have been made that is missing from the book, and that is this notion of geometric as opposed to arithmetic change, as expressed by Malcolm Gladwell in his book The Tipping Point, which, of course, came out after this book had gone to press, although some of the ideas have been expressed in some earlier New Yorker articles.

The essential idea is taken from water, which at 90 degrees is still water, and if you drop that substantially to 34 degrees, it's still water. But when you hit 31, all of a sudden it's ice, and we don't really understand where we reach these tipping points in terms of contextual change.

Jonathan Crane shows on the issue Eli Anderson raised of social composition of African-American neighborhoods that you can drop the percentage of professionals in African-American neighborhoods from 40 percent to 6 percent, and not a whole lot happens. When you hit 5 percent all hell breaks loose in terms of teenage pregnancies, drug use, crime, and everything else. And to the extent that we haven't been focusing enough longitudinal research on changes in African-American inner-city neighborhoods, where high proportions of all of the data that we've been trying to explain here are being produced, I think we don't have vital information for understanding how small contextual changes of the kind Jack Calhoun was talking about through comprehensive planning efforts, and even what John Eck was wrestling with in terms of what exactly was going on in policing. They could have been these very small contextual changes that just pushed over a tipping point that created a geometric effect that then spiraled into an epidemic of noncrime which won't show up when we look for linear effects, when we look for sort of standard dosage response curves, because they may not be what really is going on.

Mr. Travis: Yes, Martha Burt.

Ms. Burt: A couple of comments. One has to do with the type of crime that was examined, I believe it was the homicide, violent crime, what could be considered street crime. We do have spousal violence, which you should continue to call spousal and not intimate-partner, if that's what it is, because there's a big difference. And if part of the explanation of the decline in spousal homicide is that there are fewer spouses, I don't know that there's necessarily a decline in intimate-partner violence to go along with that.

My major connection on the criminal side is with violence against women, with domestic violence and sexual assault. I would be very curious to see whether the same analyses can apply to that. I believe that the declines have been lower. My understanding of the latest publications from Bureau of Justice statistics is that the sexual assault rates have not, in fact, changed, at least recently. I'm not swearing to that. But, in other words, different kinds of crimes which might have different levels of motivation or different resistances to reduction by things that happen on the streets would be extremely important to look at.

Mr. Travis: We'll do three quick ones—Hubert Williams, Mark Kleimann, Mercer Sullivan—and then we'll take a break.

Mr. Williams: I would just like to make a couple of brief observations, and one is the statistical categories like the incarceration. I was wondering whether or not we had considered the private incarceration rates of these institutions that the data doesn't seem to come from those institutions, just the public, and that's a significantly growing phenomenon in our society.

The other issue that I wanted to—

Mr. Travis: That's a clarification point. Bill, that was overall incarceration.

Mr. Spelman: Overall.

Mr. Travis: Overall, public and private.

Mr. Williams: Did that include the private?

Mr. Spelman: It does, yes.

Mr. Williams: It does, okay. The other question was that most of the focus of what we've looked at seems to be predicated on external sources influencing behavior. Do the police, for example, have a significant effect to control behavior of people whether or not they're going to violate the law? What we've done very little with is to look at the impact and effect of a culture.

There's been some discussion of that, and I was particularly focused on these discussions associated with the inner-city families tending to teach their children, or the environment in the inner-city family is such that young children grow up with great street skills, but they lack the skills of cohesiveness that would enable them to embrace societal norms and standards.

And as you'll notice around the table, there's somewhat of a difference when you look at the people that come up in the inner cities with respect to the effects of what may work there. Some people, for example, have said, we ought to take a look at faith-based organizations and their impact, in effect. I didn't notice any criminologist particularly focused on that Million Man March issue and the pledge that the people took there not to commit crimes, not to do violence against women, and any of these other things. Obviously, that was somewhat controversial, and I think people tend to shy away from controversy.

But we've made a great stride here, a great step. I compliment the Urban Institute and everybody who has participated in the book and the presentations. But if there's anything I think to look to the future, it's to learn more about the impact of culture. And the culture in the inner cities in the African-American community is a different culture. And I think that that culture we don't understand very much of it or its impact on behavior. And that's something we could look to in this future.

And also, another thing that we're very, very ignorant about, because we haven't done much research in that area, [is] the culture of police organizations and institutions. When we see race of the officer, it's really a smaller factor in terms of racial profiling, abuse of authority, and things of that nature, culture is the issue.

Mr. Travis: Thank you, Hubert.

Mark Kleimann.

Mr. Kleimann: One thing the book doesn't mention—and I think this is not a critique of the book, but if you're looking for explanations of the decline in crime—the book quite reasonably ignores the role of community corrections, probation and parole, because nothing that's been done in that area could plausibly have contributed to the decline in anything.

If we could have a meeting 10 years from now discussing why crime has continued to go down, we'd better have a chapter on community corrections, because if we don't learn how to punish people and control their behavior while we're not paying their room-and-board bill, if we don't learn to deal with the 5 million people who are now under nominal supervision not in prison or jail, then crime is not going to continue to go down.

The second thing I want to pick up on [is] a point that was made earlier about alcohol—there's nothing in the book about alcohol policy; again, that's right, there hasn't been any. But there've been a number of efforts directed at attitudes about alcohol. There's been nothing directed at availability of alcohol, at the price of alcohol, particularly its availability to people who are known to commit crimes while they're drunk. It still confuses me why somebody drinks and drives, we take away his driving license, but regard his drinking license as an irrevocable right. And it seems to me we might want to think about changing that.

The third thing is—

Mr. Travis: Only a third short one, Mark.

Mr. Kleimann: Third short one, there's discussion of drugs in Bruce Johnson's chapter, but it seems to me there's not a discussion of the intelligent use of police and prosecutorial authority to reduce the violence in drug markets. And, again, that's because there hasn't been much of that. But if we want to have another such conference 10 years from now, we'd better start focusing our drug enforcement on reducing crime in a way that hasn't been done.

Mr. Travis: And the final observation and addition from Mercer Sullivan.

Mr. Sullivan:: Well, Hubert Williams got there first. I think there really is need for more attention to the role of cultural change. I believe I'm the only trained cultural anthropologist at the table, and I particularly enjoyed getting to hear Rick Rosenfeld say that cultural explanations are the last refuge of an intellectual scoundrel, and then proceed to offer just such an explanation. So on behalf of myself and my colleague from the Columbia Graduate School in Anthropology, Joe Wallman, who co-edited the book, and Eli Anderson, whose license is in sociology, but who does this kind of work also, I want to say welcome to the fold, Rick. You can hang out with us scoundrels anytime.

I think it's not just among adults. And I think it's not just the commendable efforts such as the faith community and the Million Man March. I think there has been a cultural change among young people themselves. I would particularly point to the deaths of the famous rappers Tupac and Biggie and the big impact that it had on communities around this nation. I think there's been a stop-the-violence kind of attitude that has spread among young people without a whole lot of help from adults. And I think that needs more attention.

Mr. Travis: Thank you.

Please thank our late entrants here into the discussion, and we will take a 15-minute break, talk amongst yourselves; we'll come back and talk about the future.

Thanks.

(Applause and break.)

Mr. Travis:—In the future, not just the way that Al mentioned, that it can't continue to go down at the current rate, but there's some indications actually from 2000 data that some of the favorable trends we've seen have in fact been leveling off. Researchers at the Justice Policy Center at the Institute took a quick look at homicide trends in some of the nation's largest cities and found that for comparable periods, homicide rates this year, compared to last year, are on the increase in Los Angeles, Washington, Dallas, and New Orleans. The homicide rates this year compared to comparable periods last year are flat in New York City, Baltimore, and Detroit. And homicide rates this year compared to comparable periods last year have declined in Chicago, Houston, Atlanta, and San Diego.

So we need to ask what policies make sense in this new environment. So we will structure this part of our discussion for the next hour or so around three broad domains that cut across the discussions we heard in the first portion from the chapter authors. And those three domains are as follows. First, we want to talk—appropriately, I think, given the last discussion we had—about communities and young people. We want to talk about police and other responses to crime, some of the comprehensive and coordinated responses that we heard about as well as the police response. And, third, we want to talk at a more macro level about some of the policies that are relevant here regarding, in particular, on the one hand, sentencing and offender supervision, and a separate domain of policies on substance abuse and guns.

And for each of these, I've asked a mini-panel within this panel to be, in essence, our discussion leaders, and to focus on some of the reasons to be hopeful and the reasons to be concerned within each of these domains. After we have a discussion within that mini-panel within the panel, we'll turn it to the larger panel, and have free-flowing discussion where contributions can come from any sector. And then we will wrap up at the very end with some observations from our panelists about what we've learned, and where we should be headed.

So that's the plan. And we want to start—and let me be Oprah here for a moment. We want to start by focusing on—in fact, this theme that came up in Elijah Anderson's, Hubert Williams', Mercer Sullivan's, a bit in Lawrence Sherman's observations—about what's going on within communities and young people and to tie in some of Beverly's observations and Jack Calhoun's observations about the nature of sort of coordinated responses. We're seeing something happen. Is the community energized and organized in a different way, and what's happening with kids today?

So the question that Beverly is going to start us here, and then I will bring in, in this order, the three researchers who I think are doing work here, Mercer Sullivan, Elijah Anderson, and Demetra Nightingale, then we'll ask Tracy Litthcut and Jack Calhoun, who are particularly interested in youth issues, to be our wrap-up on this mini-panel.

The question to you, Beverly, is you, we know, travel a lot around the country. You work in San Antonio with San Antonio Fighting Back, so you have your ear to the ground in that city. But you're also involved in community coalitions against drug abuse, and the consortium on violence prevention. So as you look around the country with your ear to the ground and [pay] attention to communities and young people, what do you see that provides grounds for hope, and what do you see that provides grounds for concern as we think about whether we can produce greater levels of safety in communities?

Ms. Davis: Two very good questions. What I do see, I'll start first with concerns. I think one of the concerns that we will have to address now is the fact that we will have a large-scale number of people who are going to be returning from prison back into communities. And basically, I think, communities across this country are not ready for that. There aren't programs in place, there aren't systems in place, there's not even a service delivery in place to be able to provide the type of support services that are going to be needed really to address that population.

And I say this always as just sort of a warning shot, that someone who comes back or any person who is without hope is incredibly dangerous. And if we do not address this issue, it will be—I don't even want to think about what's going to happen. The main reason is because communities have changed. And you really do have incredibly broad-based planning efforts at the local level. And being a weed-and-seed site as well, I can't tell you. Again, it's like the tipping point that Larry talked about. People are there. It took them to get to the tipping point to change their communities. But once they have changed [the communities], they are going to fight to keep them that way. And when you introduce an element, a very large-scale element that may be anti to that, simply because there are not systems and services in place that cause [those people] to change their behavior, you're going to have an incredible clash with people who have been quite empowered and are not going to let their communities go back to where they were before.

So I think that is one that is cause for concern. I think it is one to take even more planning efforts, and for us to really take a real great look at what we can do on the local level to address that issue. What I have hope for is the fact that there is so much going on across the country with, again, thousands, at least 8,000 or so communities that I know of that are truly outdoing comprehensive efforts. And these are rural, suburban, inner-city, and I recognize these are hard things to measure. But when you look at some of the local ordinances and the local practices that are changing, we are honestly seeing that what these planning efforts are doing, and this is the impact on crime that I think that it's absolutely having.

Let me give you a good for-instance. This was done actually in San Antonio, but it's also being done across the country. You have funding agencies that normally were just—United Way is a good example—[people] would normally just give out dollars because so-and-so came to their door and said, we serve at-risk communities, and we want to be able to provide services to high-risk youth. If you literally geo-code it, and I learned this from Jeremy how to do geo-mapping, and I learned, went back and took this to my community. But if you literally took the incidences of crime, health indicators, emergency room data, et cetera, and you geo-map that and you put it on a map, and then you literally took—ask your United Way agency, or anybody who funds things in your communities, even your cities—and then you asked them who they funded, often what you will find is a great disconnect between where people are, and where the service providers are located.

Now, granted, people don't necessarily have to be located where the problem is. But if you do not take the funding that you do and direct it to where the need is—because when we literally went back and peeled the onion at United Way, and asked our service providers to give us zip code data of the clients they served, guess what we also found, here's where the areas of need were, guess how many people had to travel 30, 40, 50 minutes on a bus, transfer three times, et cetera, to get to services. Guess what we found? A lot of the people who we thought we were funding, less than 10 percent of those people were actually getting services.

So because of that comprehensive planning effort, we changed the way we did business. We worked with the city to begin to target efforts, and then, in all honesty, I think that had a great deal to do with why we began to see great drops in crime, because we began to finally put the resources where they needed to be instead of just shooting it out there and hoping that it got to the populations. That was an absolutely comprehensive community planning effort. All types of efforts to deal with vending machines, under-age drinking, code enforcement, what can go in neighborhoods, and what they can sell and do in neighborhoods has been a great impact in communities. If you do not allow lots of liquor stores—because I can tell you, we had one on every block—if you change the code enforcement so you cannot have liquor stores next to churches where youth congregate, et cetera, guess what, you begin to reduce the availability and the accessibility to those issues.

Lastly, I have great hope, because I do see the faith-based movement engaged. I see the culture and norms changing that have been set by movements such as the Million Man March, the Million Mom March, the Mad Dads March. I mean, I can tell you all kinds of coalitions across the country that are doing things. Little by little, street by street, block by block, they are changing the norms, they are changing the attitudes, and ultimately they are changing the policies in their cities that are governing how the behaviors in their communities are going to be dictated. And so I think that is the great hope, because I do think local people solve local problems best.

Mr. Travis: We have three individuals here who have thought about these interactions at the community level. Mercer Sullivan, scoundrel, anthropologist, has studied community organizations and young people and attitudes towards violence and work. Elijah Anderson has thought a lot about, turn the phrase, street culture particularly in the African-American community. Demetra Nightingale, who raised, I think, the very important observation about another major shift under way in our society, which is the impact of welfare reform, and has thought about work availability at the local level.

Is what Beverly just described, which is a reawakening of community capacity, community awareness, and effectiveness and efficacy on the production of safety, is that something that you observe in your own work, and is there a reason to think that that has a relationship to crime reduction, and what are your hopes and concerns as you think about the well-being of communities around the country?

Mercer first.

Mr. Sullivan:: I think that does make a difference in some places. I don't think it is happening everywhere, and I think some of the cultural change that goes on is simply a reaction to what's obviously destructive behavior that young people begin to avoid. In the early '80s, when I began studying criminally active young men in Brooklyn, they had grown up seeing their older brothers and older people in their neighborhoods struggling with the heroin epidemic, and they made fun of heroin addicts. They didn't go there. And then, a few years later, they got stuck with crack, another epidemic hit that community, and I think Garen Wintemute's calling our attention to the fact that these things can accurately be described as epidemics is important.

I think we also have to realize that they happen in certain communities, and an extraordinarily small number of communities are extraordinarily vulnerable to these epidemics. So that's what I worry about. What's the next one going to be? Will the learning curve maybe be a little faster as a result of some of these community organizing efforts? I hope so.

I also think that cultural change doesn't happen in a vacuum. I think this one has very importantly been sustained by changes in the labor market. We know on the basis of good research that young black men have been entering the labor market in considerably larger proportions than was true in the recent past. And I worry very much that those who are the last hired will be the first fired when the next economic downturn comes.

So I think that in order to—that what we need to do is reduce the vulnerability of these communities, both by encouraging development from within and by removing barriers in the labor market and in the housing markets for people to live where they want to live, so that we provide a variety of mechanisms for allowing this extraordinary concentration of disadvantage and vulnerability to epidemics—that we begin to remove that, root and branch, from our society.

Mr. Travis: Eli Anderson, has community resiliency been enhanced recently, sustained in the next epidemic?

Mr. Anderson: Yes, well, I think I like what Mercer had to say, and I like what Ms. Watts had to say. See, I spent a lot of years on the streets in the community looking at life at ground zero, so to speak. And a big issue is poverty itself, poverty, persistent urban poverty. And this poverty increasingly is structural, you see. And, yes, things have improved. The economy is booming, and that's important, and there's a trickle-down effect.

I see the inner-city ghetto economy at ground zero based on three prongs. On the one hand, you have low-income jobs, low-wage jobs. You have welfare payments, what's left of them. And you have the underground economy and drugs and street crime. Those are the prongs. And if you take one away, you put pressure on the other two, you see. And within the community, at ground zero, that is so poor to where open-air drug dealing is allowed to exist, where the police do racially profile people, where there's profound alienation, and people in that community at ground zero feel increasingly isolated and that they're on their own. And this is not missed [by] the young people, you see.

And given the abdication of the economy, the abdication of the city fathers and the police, and all these institutions, especially law enforcement—I mean, people in the community feel that there's a law for white people, and there's a law for black people when it comes to violence. That if you hurt a black man, if you kill a black man, it may not be investigated the same way it would be if you kill a white man. People have that understanding. And so this gives rise to this code of the street, so to speak. And most people in the community, and these are their words, are decent, and trying to be decent even at ground zero. But you've got people at ground zero who are into the street, people who will violate you, who will take advantage of you, people for whom feasibility is the main issue. And this is a minority.

But given the abdication of the police, and the city, and the isolation in the community, the decent people feel they have to mimic the street people in order to survive. So the young people have to look that way and act that way, and act tough, and basically code switch, you know, to look street when it's the right time, and to look decent when it's the right time. And this is extremely important. Those conditions at ground zero are horrible, and it's a story that is off the radar screen for the politicians right now. This did not get talked about in the, if you want to call them [that], debates. It did not get talked about. It is off the radar screen. But there is something terrible happening at ground zero. We have the absence of civil law in so many of these communities, and when we have the absence of civil law, street justice and people's law come into play, you see, and that's what I'm talking about with the code of the street. And this is not [lost] on young people.

This is extremely important because the way people dress, the way they act, the importance of rap music and hip hop, all of this is about alienation. And it's oftentimes, and maybe increasingly, confused with being black, so that middle-class kids who begin to negotiate the system and go to colleges and places and what-have-you have to at least remind people that they have not lost their blackness by sometimes acting out or presenting themselves as people who are connecting with the street, and they code switch, if you follow what I'm saying.

And then the wider system, the wider society that tends to have an untextured view of the wider community, does not know how to distinguish between the kids who are decent and the kids who are of the street. And this is especially so with those institutions that increasingly racially profile black people, associating middle-class black people with ground zero. And this is very frustrating to middle-class black people, and it contributes to racial dissension, and lack of comity between middle-class black people and the wider white community, you see. It creates the racial tension that is so much a part of this discussion that Clinton tried to have over the past few years, and these are the tensions. And those are some of the things I see.

I think, if we want to solve this, we've really got to take this into account. And I don't see us doing that really at this point.

Mr. Travis: Thank you for that very powerful observation.

Demetra Nightingale thinks about kids in a different way with workforce development issues, and welfare reform, and what's your take on the changes at ground zero at the community level with young people in particular?

Ms. Nightingale: Well, I think, as I was thinking last night about this conference, and I know much more about the welfare and labor market side than I do about the crime side, but some of the same patterns that we have seen and that you understand well in terms of the decline of the crime rate mirrors what we're seeing in terms of the decline of the welfare caseloads over the past seven years.

What I am concerned about listening today is the same thing I'm concerned about as regarding welfare reform—and I endorse what the previous three commenters mentioned—[and that] is that it may encourage a complacency, that if the crime rate is declining just as welfare reform is declining, and the economy is strong, then everything is well in the world. And we know that that is not true, as Dr. Anderson just reminded us.

And particularly on issues such as the economic opportunity and the labor market, while it is true that in the past couple of years we've seen increasing wages for low-skilled workers, the trend over the past 25 years has been down, and the minor increases that we're seeing today in terms of the wages that low-skilled men can command in the labor market, I think, are minimal and small, and do not represent a long-term trend.

We're very fortunate that we're in a period of strong economic recovery, and I think it does provide us an opportunity to think about what we might do when the bubble bursts, which it will. And what I see at the community level is that many at the community level are not—and at the state level and at the national level—are not thinking very deeply about preparing for an economic downturn. Nor are they thinking about reaping benefits of investments that could be targeted on the communities that are being left out of the economic boom, and the individuals, which [are] primarily still in the African-American communities and inner cities.

In terms of the community effects also, I just wanted to mention one other point that struck me this morning while I enjoyed hearing about the effectiveness of prisons in terms of crime reduction, and that the further investments in prison construction, or number of beds, whatever the proper term is, would not necessarily lead to continued reductions in crime and, therefore, is not a cost-effective investment. I think that sort of begs the point that we're seeing in a number of communities, and in rural communities where they're not as concerned about the effectiveness of the prisons on reducing crime as they are on the effectiveness of the prisons as a source of employment and economic development in those communities, so that the counter-acting effect, I think, needs to be brought into play a little bit more so that the effect on communities from a human dimension is borne more heavily by the low-income inner-city concentrated poverty areas, and the positive benefits in terms of economic development of the incarceration [are] occurring to not those communities but [to] other communities that are seeing the economic windfalls.

And, then, [there is one] final point that I just wanted to mention, also in terms of the labor market. Just as we have seen, over the past 10 years or so, research confirming what many parents have always known: that if teenagers work too much it hurts their education. And by moving individual young people as their first experience in the labor market into the low end of the labor market, we reduce their potential for increasing economic viability in their lifetime. So that I think that what we need to be thinking of is both prevention for young disadvantaged, economically disadvantaged individuals to provide them with some long-term economic opportunity that will be a viable alternative to illegal activity when the economy declines in the future. And, again, there, I think that the effects are disproportionately borne by inner-city communities that are outside of the economic mainstream in terms of the access to anything other than low-wage jobs.

Mr. Travis: Let's get Jack Calhoun—I'm going to save you for last—but I want Tracy Litthcut to come into this conversation. Tracy has experience directly on point here in a number of these organizations, working as a street worker, youth organizer in Boston, associated with David Kennedy and the folks who thought about what some called the Boston Miracle. Let's talk about ground zero. You're close to ground zero, listening in.

Mr. Litthcut: I had the opportunity to come in today to address researchers, my mouth just watered, I couldn't believe it. I said, yes, I definitely have a lot that I want to say to researchers. Anything that I want to say here today, I'm hoping it's going to relate to the success we've had in Boston. At one point we went two and a half years without a juvenile homicide. But I kind of want to shift that a little and go somewhat toward the problems with research.

As I was listening to the—

Mr. Travis: You have to keep it short, that could go on forever. That's a lot—

Mr. Litthcut: As I was looking at the authors of the book talk about specific things that they had written in the book about the homicide rate in Boston had reached this all-time high in 1990. The crack epidemic definitely played a major role in crime and violence. We definitely know about the teenagers with guns. We know African-American and Latino males of a certain age are the ones committing a lot of these crimes. We know that a lot of these individuals went to jail in the '80s and are now coming home, where some of that might be a part of the violence that's occurring. And at that time actually I attended over 100 funerals for kids involved in gang violence, but that was the problem.

That's the problem with research: People like myself, practitioners who are working in the community, we think we know it all. We think that everything that we've seen or done on the street level is exactly what we know and what is occurring in our community. The only problem was, we couldn't figure out how to stop the homicides. As us knowing everything about our community, about what's going on, we wouldn't open the doors to researchers to come in to support us in what we wanted to do. And I must admit, initially I had a love-hate relationship with research.

My hate—I just finished my graduate school work, and no matter what I tried to do, I couldn't get these formulas or nothing down with research to get through this course. I finally did it, but I hated taking the course itself. But in regards to loving the research, David Kennedy, a gentleman from Harvard who came in—and when I was asked to come in and sit down with David Kennedy with Harvard at a sit-down, and to do gang mapping, and gun trafficking, my initial reaction was, what is he going to tell me about my community that I don't already know? Really a well-respected police officer, the superintendent of police, asked me to come in, and I went and I went back a second time because Harvard serves excellent lunches, so I kept going back.

The one thing about it, though, when I sat down and I started going over this gang mapping and strategy and gun trafficking, it taught me a lot about things that I didn't know. One thing that it did, it gave us a clear focus on where we should be heading, and what we should be doing. It told us, too, we can't solve all the problems. We tried to solve every single problem that related to the community, and it's not going to happen that way. We had to figure out what we wanted to do and what was the biggest issue and priority in our community: That was homicides. In 1990, like they said, that crime rate rose and we experienced our all-time high, just like the researchers said. One thing that we realized, too, [was that] we couldn't save everybody out there. There were just certain individuals who didn't want the resources that we had. And sitting down with law enforcement and probation and some of the other partners—that was another thing, having to drop our egos.

The researcher, by bringing us in—for me it made me believe that the researcher really had no agenda but to get this data. He wasn't worried about who was going to get funded, and who was going to be the lead name on the flyer, who was going to get to speak to the press. Dave Kennedy loved research, and he wanted to get this data out of us.

The other thing, too, is collaborations in the community. You go to these meetings and the collaborations and meetings, you have 900 people and everybody has a different agenda. We kept it very small, we kept it quiet. We worked with the people who had—you have a lot of people that talk the talk, but we had a group of people who had walked the walk. They were out there day and night, at homicide scenes, 2:00 in the morning, 3:00 in the morning. And one thing that I believed at that point, which I didn't believe then, years ago, [was that] you didn't have to be black to work in a black community, didn't have to be white to work in a white community. All you had to do was be dedicated, committed to the cause and we could make change. And the most important thing, like I said before, was dropping the egos and coming together.

The other thing, too, just like the researcher said, [was that] it wasn't just happening in Boston. When you get isolated in a city and you're fighting and you're working these long hours, you think that this is the only place where it's happening, and you're the only one that's being affected in this way. We started to realize through David Kennedy [that] it was happening all over the country. So we wanted to hear about what other people were doing, as well, so we could improve on our strategies. The toughest thing that we had to deal with was what we were doing wrong and the changes we had to make. It's very difficult to tell somebody who is working 12 to 14 hours a day that they're not doing the right thing.

And it also helped save careers. It helped save my career. I was at the point I was attending so many funerals, I had got so cold and callous that when I attended funerals I wouldn't cry anymore. I oversee 100 youth workers, I had to make sure I was strong to keep my youth workers straight, and make sure that they could focus on what we had to do at these funerals, and keep families and people together. And you know what, families didn't want to hear about policies and what wasn't happening on the national agenda. When they experience a death in their community, they want things to change immediately.

The thing that was so difficult was to figure out [was] how we need to make these changes. And for me, I was ready to walk away from this field. I couldn't take it anymore, going to these funerals and seeing these kids die, reporting to homicide scenes and seeing kids dead at the age of 16, 17 years old, kids that I had been working with for two and three years. The research said to us, "You know what, Tracey, did you realize that you had gone at this point a year and a half without a juvenile homicide?" I didn't even realize it, because I was dealing with all of these things on a consistent basis.

The one thing, like I said, the gun trafficking, when we sat down with the researchers, it told us that two rival gangs were being sold guns by the same gun trafficker. Something that we knew nothing about, which told me that my outreach work as a street worker, which is one of the number-one violence prevention programs in the country, it made me focus our attention on those two gangs, because if that's where the guns are, we need to give close intervention to those particular groups.

And it also told me—and when you work for a government, political agendas are the most important thing when you work for governments. And when you go and you're in front of the press, you have to make sure you know everything that's going on in the community. But, you know what, research and Dave Kennedy told me I didn't know about all the rivalries that existed between these gangs. And I had to learn more about what I didn't know.

What I'm trying to say to people, more than anything, [is that] we have to bridge this gap between researchers and outreach workers, and people, clergy, whoever it might be in the community. For one thing, it took us to another level nationally. I've been all over the country talking about the strategies that have been very effective in Boston, and just recently we went almost another two years without a juvenile homicide. Funding. Funding started coming into Boston like you wouldn't believe. First grant came in. It was $2.2 million over a couple of years. And then we just received a $24 million grant over a five-year period where we can hire 40 case managers. It decreased homicides. And I love Dave Kennedy for coming in and helping. What he did to help decrease homicides, you might not hear everybody mention that at times, but that research and the research community was a part of it.

And I know it's difficult in dealing with communities. A lot of times white males and individuals coming into the community, you hear that racial thing, "We don't want you here, what can you tell us." We need you, and I know it's difficult dealing with some of us at times, no doubt about it. I have an ego, and I'm tough to deal with at times. I can be a real pain in the ass. But we need you in our community, we need you to continue to work with groups to bridge that gap. In Boston, research has made a major, major difference, and I can't emphasize it [enough]. And if you want to get people to the table, start like Dave Kennedy did: Serve good lunches. It keeps me coming. I know that.

Thank you.

Mr. Travis: Good lunches, there's a lesson.

(Applause.)

Mr. Travis: And Jack, before we bring in other observations from the panelists, if you could give hopes and concerns?

Mr. Calhoun: Thank you for that, especially the link of food to progress and research. Just underscoring your point, Tracey, very passionately about getting these wonderful Olympian researchers into some of the messier stuff in the community, where it may be hard to measure, but it's so very hopeful. And I don't want to put words in Beverly's mouth, which I believe very strongly about a galvanized community. But those galvanized communities work in not just a romantic way, but they are connected positively to those entities that make a community work, like cops, schools, the faith community.

Second is touting those programs that work, but with the caveat that a program alone won't save us. It's a galvanized community that does not produce crime. Another hinted at, but I don't think really said, Jeremy, was I think we've got to look at youth somewhat differently. Most of our policies would either control youth, or fix youth, but very rarely do we say to youth, we need you, you're part of us. And there are some stunning statistics out there. Our own Harris poll showed that 90 percent of kids would be willing to do something about crime and violence if only they knew what to do.

UCLA just released a study: 75.3 percent of all incoming college freshman will have volunteered three or more hours per week. They're a huge resource. They don't want to go to school in messy areas. I know there are a lot of tough kids, I work with them. But I think we've got to have policies that encourage the better angels. Which leads me to my next point: I would like, as a piece of the research agenda, to help states realize how much money they are spending on the correctional system. What is happening is, to the extent that policy is a promise, we have a very clear promise to the next generation. If you screw up, we're here with hugely expensive solutions. California, I think, is spending 16, 18 percent of its budget on corrections, and about 2 percent on higher education. Connecticut, $500 million corrections, $400 million higher education. We need corrections, but that proportion is just one hell of the wrong message to the next [generation]—but a lot of states don't know how to tease these numbers out. That we could help on.

Two quick final points, Jeremy. One is, we've looked hard at cultures of collapse. I think we've got to be smarter about cultures of resilience. What about these anomalies in places that shout high crime, where the objective indices shout massive trouble and yet there's low crime. Why? There's been some very good research on individual resiliency skills, but let's look at community resiliency. And finally, Elijah, just to underscore what you would say about the ground-zero work, I'm not being complacent because it scares me, all this diminution in crime which we all are excited about. It's off the screen politically. And yet it's stubbornly persistent, not just in those communities we know about, the core city, new Americans, Native American communities, with the attendant structural challenges around guns and the economy, and the disproportion of wealth and poverty.

So thank you.

Mr. Travis: So here we have this mix of optimism, Tracey and others; real deep concern, Elijah and others, about the state of communities; optimism about the current situation; and Demetra and others said, this may not last long, and particularly that it's tied to the economy.

So who else wants to make an intervention here to help us understand from their perspective the role of communities and young people as engines of safety, ways that we can produce safety over the next few years? Have we missed the point here about communities and their role, particularly young people and their role in safety production?

Ms. Burt: I think the problem with the economic boom is that it feeds into the all-American fundamental belief that it's all up to us individually. And it lets a lot of people believe that anybody could—well, anybody pretty much can get a job if you can walk and talk. But whether that job is going to get you anywhere, whether to keep that job is going to leave your kids unattended: All these other things are not—it encourages us to not consider what it actually takes to help somebody be part of—to be able to sustain themselves in a family. And that includes not just their own efforts, but the efforts of the larger set of people and environment in which they live.

And what I really worry about is a lot of the kind of blaming behavior that Americans are totally fond of when somebody doesn't make it, can't make it. And that we have an incredible amount of effort going on from individual communities to pull it together at the same time that we have sort of a national ideology that doesn't put a lot of resources into that effort and doesn't really believe that it is necessary. We still persist in this [belief that] anybody can do it, and if you don't do it, it's your fault. And then we'll pay for prisons. But we won't pay to actually invest in youth development. We won't pay to invest in making sure that every kid can read before they get out of first grade, and not to hold them back. There's an incredible amount of stuff we won't pay for, because everybody is supposed to do it themselves. And the only time we'll pay attention to you is when you blow it.

And that really does concern me, and the last eight, nine years of economic expansion, and everybody thinking the stock market will never crash just feeds that set of attitudes that says, we don't need government, we don't need collective effort, we don't need investment in these communities. I mean, that's where the money comes from; whether the government ends up being the people that actually have to run anything isn't necessary. But the fact is the investment as a whole collective has to come from that direction. And I get very, very worried at present attitudes that manage to deny and ignore that altogether.

Mr. Travis: How do we link this discussion with Lawrence Sherman's observation about tipping points and the idea of communities rebounding? Tipping points can work both ways. But it's easy in a way to talk about macro policies and the economy and the like. But there seems to be, from some of these observations, something else going on within communities. Is that an observable, tangible, get-your-hands-around-it phenomenon? And is it related to policy, or is it more sort of organic and bottoms up, to use a phrase?

Larry?

Mr. Sherman: I think it could be heavily related to Peter Reuter's issue, and that is immigration policy. We will have the second edition of this book drawing on the year 2000 census, which can tell us a lot about the extent to which immigration changes have been correlated with crime drops. New York City would have lost population at the same rate of Philadelphia, which is 10 percent of the population over the last 10 years, if it hadn't been for immigration. And I think that we can hypothesize based on a lot of evidence about diverse communities doing a better job of organizing, because there's more conflicts that surface that have to be worked out that energizes people, it starts epidemics of controlling youth, and things that Rob Sampson and Tony Earls are finding in Chicago, that can affect social capital, the capacity to prevent crime, and to be a resilient community.

But I think that Doug Matthews's data showing that American cities as of 1990 were more segregated by race than ever before; even during the heyday of immigration before the 1925 cutoff, [when] you had African Americans, Italian Americans, Jewish immigrants from Eastern Europe, all these folks were mixed up geographically. And we've seen this incredible geographic sorting, and homogeneity grow since the advent of highways and other things, which really got accelerated in the '80s and the '90s. If we had a national policy that would encourage immigration back into cities that have lost population, I think we would create a context for resiliency against crime in which you would have less African-American poverty, isolation from other kinds of people, other kinds of cultural visions. And it's by bringing the visions together that I think we can create the resiliency we're looking for.

And I'll just end with one proposal that links to the problem of getting high-tech workers in from other countries right now. Congress is about to raise [the number of] high-tech visas to 200,000. There's no discussion about limiting those visas to cities that have lost more than 10 percent of their population in the past decade. But if there was, you could transform the neighborhoods that are currently disinvested—property values are down, vacant houses, and the whole context of despair,—if you had lots of people pouring into those cities, which is what they grew up with in the turn of the century. But especially with this emphasis on people who are going to be able to draw jobs into the cities, because if they were required to live there—and somebody could be an urban homesteader—if they live in a city like that for 10 years, work in a city like that for 10 years, we could make them citizens, I don't think you'll have any shortage of people who want to come to these cities from around the world, even if Americans tend to want to flee them with the first automobile they buy.

Mr. Travis: Let's bring the police issues into our discussion. Unfortunately Chief Ramsey had to leave at the time we were about to do that. But he deputized Ed Flynn—did you know this?—to say everything. Every pearl of wisdom that he was about to utter, you now have to utter. He told me what they were.

Mr. Flynn: I'll speak twice as fast.

Mr. Travis: So the question we want to ask our next mini-panel here is, same question: [What are some] reasons to be hopeful, or reasons to be concerned about our capacity to organize responses to crime? Police is obviously a piece of that, but we also want to talk about some of the things that have been put on the table here about comprehensive, interdisciplinary responses. Marti Burt mentioned response to domestic violence, which I think is a classic example of that.

So I want Ed Flynn to start the conversation here with discussions about sort of the policing environment, and do we have reasons to be hopeful about the capacity to respond to crime, reasons to be concerned? Just to note, speaking of immigration, that not your current department—I remember a highly immigrant populated city, and Chief Flynn, as you know, is the chief in Arlington.

Then I'm going to ask our sort of more academically oriented panel—Wes Skogan, Lawrence Sherman and Martha Burt—to come in and help us understand what we've learned about policing, the relationship between the police and the community, the relationship between the community and crime conditions. And then if I could ask Hubert Williams to be our sort of closing commentator on this panel.

So, Ed, as you look at the police profession, what's going on that gives us reasons to be hopeful about crime in the future, or concerned?

Mr. Flynn: I think it's important at a time like this to really take a long view. And it's with mixed emotions that I relate that I am now capable of taking a long view. My police career will be 30 years long next year. And I still feel—I'm still in touch with my inner patrolman. But I make that point because I remember when I was still in college, and no respectable person majored in criminal justice then, and neither did I. But I remember trying to read what was available. And one of the few things that was available was the president's commission on law enforcement and the administration of justice. And believe it or not, I read it in a paperback edition on the beach in 1969. I found it fascinating. It kind of informed a vocation, that I thought maybe this was worthy work for young people who thought maybe we had a responsibility to our country to make things better where they were so obviously bad.

I started out as a foot patrolman in a city at about the same time that Hubert Williams was becoming the director of his. A program that I was a part of would later be evaluated by Wilson and Knelling, and formed part of their broken windows thesis, as they looked at the impact of foot patrol officers on conditions in their neighborhoods. So what I did last night is I pulled out my old copy of the task force report on police, and I looked at the recommendations of what I would call the liberal, professorial canon circa 1965, for what the police should do in order to relate better to communities, be more effective dealing with crime, and presumably reduce some of the community tension that was resulting in the conspicuous urban disturbances of those years.

What was remarkable to me as I went down the checklist is, except for the creation of metropolitan police agencies absorbing small towns, industrywide we've embraced virtually everything that was recommended back then, in terms of higher educational standards, increased police training, more professionalism—

[TAPE CHANGE.]

Mr. Flynn: That same constituency is saying, you did all that stuff, but we're not really sure it mattered. I feel a little bit like Roseanne Rosannadanna in Saturday Night Live, you know, "Well never mind." I think those of us in the profession who have been in it that long can certainly anecdotally say that the station houses that we entered in 1970 or '71 internally are profoundly different places, having profoundly different discussions, with profoundly more integrated agencies than existed then.

Our response to crime of the early '70s, which was random patrol, rapid response to emergencies, and reactive criminal investigations was called into question over the course of a generation when it came—over the course of a generation of policing when it became clear that those interventions weren't having an effect on the direction of overall crime. That despite the fact that every year we could go to community meetings and demonstrate clearly our efficiency, measured by the number of arrests and tickets we were writing, those same communities would say to us, yes, but I'm afraid to go outside my door.

We would go to community meetings and produce impressive numbers of drug arrests, and impressive numbers of robbery arrests, and our communities would say in the early '70s, as we shook our heads, saying, "What's wrong with these people, they just don't get it," they'd say, "I don't care about all that data, all I know is those kids with the boom boxes on the corner, smoking grass in the hallways, double parking outside my house, honking their horns in the middle of the night, are making my neighborhood unlivable, and I don't want to be here anymore."

When we think in terms of historical approaches to what we're doing, I find picking 1991 as the year before after, you know, B.C., before crime, after crime, or however we want to—before crime, after drop, okay, 1991 A.D., after drop begins. It tends to create a false distinction, as though nothing had been happening since 1965. I think unfortunately much of the rhetoric in New York City got overheated because there were dramatic drops over a short period of time, and these became, I think, political issues as much as anything else. I think waters got muddied.

I think people at that time weren't perhaps willing to give enough credit to the work that Lee Brown had done as commissioner, that David Dinkins had done as mayor, that a number of people had done in beginning to change the community's attitudes towards disorder and crime and their linkage, and beginning to beef up the size of the police department, and experimenting with CPOP and various forms of community policing, and tactical narcotics teams, as has been alluded to in a lot of these articles. A lot of experimentation and ferment was going on in New York City before 1991, that kind of crystallized in a lot of the cultural changes that department embraced after 1991.

I think from a macro point of view, one of the things I've noticed in policing is a variation of the Hawthorne effect. And I don't mean this cynically. Those of you who had to read books, I remember promotional exams about management; for four years I had to read about the Hawthorne effect. But that notion that there was an experiment done in an industrial plant once, trying to tie the lighting conditions to the productivity of the workers. And after several times of brightening the lights and productivity increasing, they started to dim the lights, and productivity kept increasing. They finally realized the workers were responding to change, to the fact that they mattered, to the fact that somebody was paying attention to their plight.

Well, I apply that to a lot of this scattershot policing changes and innovations under the enormous umbrellas of problemsolving, and community policing, or zero tolerance, or what have you. The last 15 years, American policing has been trying an enormous number of interventions almost invariably involving collaborations with communities and other institutions of local government.

Now, 20 years ago the schools wouldn't let the cops in unless a teacher had gotten shot. [Schools] were a form of sanctuaries, like a medieval sanctuary. If a crime happened in the high school it didn't count, because people didn't trust how the police were going to respond to it. The relationships between school boards and police departments, and schools and police are profoundly different now, and much more seen as a positive intervention, rather than negative sanctions. I cite that as an example to show that in terms of the Hawthorne effect a lot of people changed in response to the changes, not necessarily the specific dynamics of the intervention chosen.

The second observation I'd make, again in macro terms, and in some ways this is like the economy, when you think about it, did government produce the eight-year economic expansion we've gone through? Well, if you're Republicans, no; if you're Democrats, yes. And if you're a researcher, well, more research is needed. What I would assert is that there are certain things government did at the margins of the economy that were profoundly important for investor confidence. And when the government tinkers with the tax code, when it tinkers with the interest rates, when it makes gestures in how it decides to spend, it has a very large effect on people's conduct that I think can speed up or retard positive or negative social forces affecting the economy. I think that same dynamic has affected crime, that the police, which happens to be what I'm being asked about, can certainly, on a neighborhood-by-neighborhood level, accelerate those social processes that are moving communities in positive directions and I think with partnerships retard the negative ones, which will probably be our next challenge soon, as we write part two of this book. But I think that linkage has been essential and I think the experimentation police have engaged in, based on 30 years of evolution, has been a very important process in strengthening neighborhoods.

Mr. Travis: Wes Skogan, you've been thinking about policing, and policing in communities, and particularly in Chicago; is something going on within policing that's a cause for optimism? Community engagement: a real thing or just a passing fad?

Mr. Skogan: I'd actually like to start with something different, because during the break I heard some grumbling. I heard grumble, grumble, grumble, yet another researcher reports that nothing works, grumble, grumble, grumble. The chief grumbler just had to leave too early. There we go again. Police don't have anything to do with crime. And so I think for at least two prominent members of the audience, I wanted to give my take on what the book says, because I don't think the book says that.

I think my reading of the book, and the stuff that they summarized, says that there are, in fact, a number of promising policing strategies that demonstrably work. And that what the book also says is they can't find evidence that these policing strategies had been widely adopted by enough departments to have had the significantly new impact that we could observe and could account for that shift, let's say, between 1989 and 1992, when the precursors to that crime rate drop had to start.

So that's not saying that nothing works, that's not saying that police don't have anything to do with crime; it has to do with the positioning of policing precisely in the historical nexus that took place at the turn of the decade. And their conclusion about that is the old English conclusion, the English jury conclusion of not proven, which is not to say it's either guilty or not guilty. So I think they're saying something quite different than the "grumble, grumble, there go the researchers, nothing works again" kind of conclusion. It was a difficult one to communicate, because as John pointed out, there were no beautiful charts and graphs that could be displayed, because of the enormously decentralized nature of the beast.

Now, one of the things that's in that report is, I think, that, for example, sustained, highly focused, intelligence-driven policing missions can indeed, when targeted intelligently, drive down without [causing] too much displacement, many of the serious kinds of crime problems about which we are concerned. And the real problem in American police departments is that too many of them are not capable of sustaining that kind of focused, intelligence-driven missions for very long or across very many parts of the city.

I guess in my view, my optimism about police is when I look at lots of these departments, I see lots of room for improvement. I see lots of departments that are poorly managed at multiple levels, with enormous amounts of slack, and a considerable number of slackers to be found at all levels of the organization. They make very poor use of the information that they have, and many of them have[not] been able to actually take advantage of the expensive new machinery that they bought with money from the cop shop—it's still sitting in boxes. They don't have much capacity to learn from each other. And they actually don't have much capacity to learn from themselves. They sort of don't know how it is that they're solving the little problems, and having the little successes that they have, because they have no way of accumulating and redistributing those successes even within their own departments.

So I guess I see positive things on the horizon, only because things—there is so far to go, but there is so much—there is so much room for tightening up the slack, and so much room for better use of information and technology, and so much room for better leadership, that things can only get better.

Mr. Travis: Thank you, Wes.

Lawrence Sherman, we're going to call on you next, because we know that you have another engagement that's going to pull you away in a moment. And police learn something about their effectiveness in crime reduction; is that cause for hope, or cause for concern?

Mr. Sherman: There's lots of cause for hope, but I actually disagree only with Wes's final statement; things can get worse. They have in the past, and what's likely to be the case, in fact, is a continuation of this great unevenness around the country, that may have to do with mayoral elections, choice of police chiefs, rise and fall of police union leaders. There will continue to be lots of diversity, but I would absolutely agree with Wes's basic point that there's a growing gap between the capacity to do knowledge-driven policing and the culture of especially street police personnel in terms of their support for this kind of computer-driven, management-driven policing.

And it's evidenced in medicine. So I don't want to say that this is anything that's particular to policing. There is this growing conflict between information systems analysis that says, okay, everybody ought to do things that way, and the culture of professionalism that says, my individual creativity is what matters the most. We see it in teaching with directed instruction to raise literacy rates. And teachers may say, I don't care if it works better for the studies. When I do it, it's better if I teach "Jack see the dog run" creatively with my particular take on it.

And, you know, from a workplace environment standpoint, this is actually a very important quality-of-life issue. People don't like to be treated like robots, even if they're told that that's going to lower disease rates, increase reading rates, lower crime rates. It's still an issue of what Professor Csikszentmihaly (University of Chicago) likes to call flow, and people getting satisfaction in their work because they have a creative experience.

I think we can solve that in policing. I think if we can enlist the culture of street policing in the whole idea of problemsolving, which really does create lots of flow if you get into it. And the police officers who do get into it absolutely love it. And there's this huge turnout, thousands of police going to San Diego every year to talk about their problemsolving. Although where they do tend to wind up sort of losing Herman Goldstein's hope for what they can do is in the area of evaluation, and looking systematically at what effects their efforts are having on the problems they're attempting to solve.

So I think we've got a big problem of cultural change to think about in policing, which is partly a matter of leadership of the current generation of police, but it's also about thinking about how we recruit and even how we change laws regarding recruitment.

The one issue I do want to take with Ed Flynn's comment about the president's crime commission is that we have not implemented the full dose, and perhaps gotten to the tipping point on particularly the college education degree. That 1967 report says every police officer hired should have a bachelors degree at the time they're hired. We haven't done that in American policing. We have done maybe an average of two years of college education among all police officers, some of which is acquired after they enter the department.

But I think if you took the premise of the '67 report, which is that you would create a radical change in the culture of policing in terms of its capacity to use information, to be much more focused on the tools we now have that they couldn't even imagine in '67, like computer mapping analysis of crime, it would take that kind of a radical shift that seems inconceivable at the moment, because the labor market is so good we can't even fill police positions. New York City has just lowered education requirements in order to keep the number of uniforms on the street.

And maybe that's our plan. The next time we have an economic downturn and crime goes up, the police job will be so attractive that we could raise—radically raise educational credentials, try to change the culture of policing so that everybody is driven by a focus on results, and then maybe individualizing it in ways that we've never done before, so that we could do rankings of police officers by the repeat offending rates of the people they arrest, which we have evidence that it's related to how they treat those people, and how much respect they show them, or the repeat victimization rates. So that when they take an offense report their goal isn't to write up a report, their goal is to prevent that victim from being victimized again, which is very strong in England, as John Eck points out in his chapter, but really hasn't taken hold in the United States.

So I think we've got lots of opportunities, lots of reason to hope that if we can play our cards right, these things will turn out better. But I think we've still got significant risks that a lot of what we've gained could go south in a kind of backlash to the racial profiling attack. And I have to tell you, there's an awful lot of angry police officers out there who are furious at being labeled as racial profilers, and deny that that's what's going on, even though they might admit, well, "There's a few officers that do it, but I don't, and I don't like being labeled." That's an issue we've got to pay a lot of attention to.

Mr. Travis: I'm going to pick up on your repeat victimization issue to bring in Martha Burt, who has been doing evaluations of, in fact, one of those problemsolving efforts, nationally supported, to bring cops, and service providers, and advocates, and folks together around violence against women. And you've also done work on child abuse. As you look at that—

Ms. Burt:—that I hope is relevant, and that is that new things tend to work or feel like they work, and some of what people said about, it doesn't really matter what you do, as long as it's new and people are enthusiastic about it. And so you see waves of new program ideas coming in, in a sort of maybe 10- or 15-year cycle, and they start to look very similar after a while: "Yes, we tried that 15 years ago by another name." But there's always a great deal of enthusiasm about [how] this one is, in fact, going to work. And I think there's actually even some evidence that whatever it is does work for a while, and I think that the reason that it does work is because the staff feel better about it, they're enthusiastic about it, assuming they've been convinced that this is the way to go, and that there's leadership around it. I think more than anything else, that is what I have seen to make a huge difference in anything.

I don't care. I mean, my career is all over the map. I've looked at stuff on teenage pregnancy, prison reform—not together—violence against women, homelessness, child welfare—I mean, you name it. And ultimately what I end up telling people, which they hate to hear, is that what you need to do is hire the right person. And if you hire the right person that person will know how to bring people together, will get different organizations working together, will organize the community, will inspire volunteers, and they will. And I always get told, I can't do that, I either have civil service hiring whatevers, or I can't find that person, or the mayor wants me to hire so and so, or whatever. And, in fact, you can watch programs just go down the drain. You might as well not spend the money if you don't hire the right person.

But I would like to see how we can generate more right people, how we can mobilize that enthusiasm and experience in the policing area, in the community development area, no matter what it is. And I think part of the way you do that is the commitment to using everybody's ideas. You can talk about Japanese management, you can talk about—I just heard a great thing on how Fresh Fields manages, and why they've expanded 100-fold, because everybody's department is its own thing. Everybody is responsible. Each department is responsible for organizing itself, bringing in new ideas, and they get a piece of the pay action for their increases in sales. I mean, that's what we really need to do, except we need to figure out an equivalent to sales for community development, for payoffs for youth, for payoffs in terms of safety for the overall community, so people have that immediate feedback of their involvement and of the use of their ideas, and of their commitment, so that they continue to stay committed.

Now, on the violence against women side, I've been doing an evaluation of the STOP program, which is an acronym, but it really doesn't matter what it's an acronym for, because STOP was made up, and then the words were applied, figured out to make it fit. The whole idea is to stop violence against women. And this is a huge amount of money, billions of dollars now that have actually gone out to states for distribution to local communities over the past five years. Now in its sixth year, [the program] just got renewed in the last week after an incredible struggle.

And that legislation and, to some extent, local efforts have had a major emphasis on collaboration, trying not just to have this victim services program, that police department, that prosecutor over there, work on their own thing, but to try to get them to work together. And I am totally convinced that the more they do work together the more they establish what in that particular trade is called coordinated community responses, whether it's to domestic violence or sexual assault. The better the experience is for women who are victims, the more quickly they recover and live safe lives, and also the more the community is able to reduce repeat victimization. And there's some beginnings of starting to work culturally on the attitudes of men, especially in smaller communities where everybody knows each other, and where you really can start leaning on people who are the individual violent culprits.

However, I should say that it is incredibly difficult to do. It does take that leadership that I was talking about. It takes somebody paying for coordination, which most funders hate paying for. They think about it as administration, or the fax machine or whatever. I can't tell you how important fax machines are. If you can fax 30 people all at once a notice of a meeting and organize that, you save somebody an incredible amount of time. If you can pay for a coordinator who makes sure that the commitments made at meetings are, in fact, carried out on the ground, in individual organizations; that the protocol that is developed jointly by every police department in the county is then turned into training for every police officer in the county; if you can make sure that the victims services people know where a case is in the police prosecution process, they can help women know what to do in relation to that case, [then] you can, in fact, transform how that community relates.

I mean, I can point you to communities that have become known in their environments as—some of the cops in some of these places talk about drive-by DB, because anybody would rather report a DB in their community than in the neighboring community, because they get treated well, and they get treated safer, and the results are safer. Those can be repeated, but not without the investment and the commitment to that collaboration, and to creating a community where there hasn't been a community before.

Mr. Travis: Hubert Williams looked at policing and related issues from the national perspective, as well as the New York perspective, for many years.

Mr. Williams: Thank you, Jeremy for giving the opportunity to close. I don't get that very often. I was reflecting on some of the comments that have been made, and also looking back at a career that spans three decades. And one of the things that has impressed me most is the democratization of police. We may not look at that—we may look at that as just like an overused term, but police were always ancillary. We didn't want outside involvement. We always called it interference. It took time before that evolutionary change began to set in policing. And I do think that Ed Flynn hit on a pivotal point. It was the challenge of crime in a free society, that study, that pointed out a lot of the big issues that we're confronting today. This whole movement in community policing—the Police Foundation did a study in 1998 that looked at the use of force in an era of community policing. And one of the things that came out of that was that 96 percent of the people that were surveyed—and this was a stratified random survey of the ranks throughout the country—96 percent thought that the most effective way to deal with crime was when police and community are working close together. That was a philosophy we didn't have years ago. We also have the leadership in policing all attesting to community policing as the most viable and effective strategy to deal with crime control.

Sometimes I get calls from the media that get tired of this community policing. They want an objective, balanced approach, so they say, "Can you tell me a police chief that's not in favor of community policing?" Well, yes, I can name a few, but they won't say it. They won't say it. And so the movement and the thrust now is towards outreach, towards the police becoming more involved. We have become more diverse as a force. The leadership in police have become more educated, and we do focus and try and target these issues.

The second point I wanted to raise was the contributions made by the research community, which are large. When you look at the—I think it was the funding with NIJ, NIJ and Chip Stewart, the Police Foundation actually did the study on fear reduction. That was a big, big move that took a lot of chutzpah to fund something called fear reduction in those days because our concern was crime reduction. And the media took them over the coals for fear reduction. We know it causes crime. We know what causes fear, crime. What is this fear reduction business? But that did open up the door for us, because one thing it told us, the fear reduction study said that any positive interactions by the police with people in the community of a nonthreatening nature will enhance the reputation of the police force and make people feel safer. Now, tell me that's not the foundation predicate for community policing.

And if you go back, as some people often like to do, to 1869 with Sir Robert Peale, the first class graduating from the police academy, when their role was defined, it was, you must be of the people, understand the people, be in tune with the people—big, big critical factors.

Now, the third point is that there is a large movement, and you hear it coming from so many different people, that the police are really not the only force here. This coming from community groups and people, organizations that have diverse interests, that are working towards some common cause without having a common plan, but everybody is working to try and galvanize support for making the community safer, and improving the quality of life in the community. I think those efforts have grown. I remember about seven years ago, we profiled some of those efforts of community-based organizations that were having an impact on crime doing positive, constructive things. It was anecdotal, it was not like a scientific-based study, but it was very interesting to see what community organizations are doing around the country.

Now, if I look at where the shortfalls are, what the concern is, I tell you that it's our insufficient temerity, chutzpah. We have not been—we have not had enough guts to face up to the big issues. Well, maybe the politics won't allow us to deal with the big issues. Whatever the reasons, we've come a long way, and there's a great change and improvement in policing. But we have not yet dealt with the most profound issue. I mentioned it earlier. There's some problem with the police house, and it's culture, and we've got to find out more about that culture.

Now, like Larry Sherman said, the police are pissed off at this constantly being accused of racial profiling. Well, the minority communities, communities of color are pissed off by constantly being targeted by the police. What does that mean? You know, what are we talking about here? Well, I think if we really looked at this time closely, one of the reasons people get pissed off is because they get accused unjustly. And if you look at the cops in the country, if you know them like I do, I was in there for so long, most cops aren't doing that. They don't believe in that, they're professionals, they do their job the right way.

But there's a great failure in policing to deal with accountability. We allow the guy on the bottom, that cop out on the street, to take all the heat, and when he goes over the line, it's his problem. He shouldn't have beat that person, he shouldn't have shot that person, all these things, right? What about the supervision? What about the management? What about the resources? What about the department? What about the direction and control provided by the institution itself at every hierarchical rank in the chain? And what about those things I used to learn about when I was a cop? They told us watch out for the three M's: misfeasance and malfeasance—the feasances—and there's a nonfeasance. When are we going to apply the law in ways that it's going to affect the big guys as well as the little guys? If we do that, we deal with this issue of culture, we deal with the issue of race, we deal with the issue of being accused unjustly, and we need the chutzpah, you know, to face up to these issues as researchers and take it to the next level so that we can truly leverage the gains that had been made and improve the profession.

Mr. Travis: We're going to keep moving right along. And David Kennedy, I want you to join the conversation here. Thank you for providing lunch, whenever that was, to Tracey. That clearly was better than the NIJ grant to get things going. But we do want to broaden the scope a bit from policing to community on the one hand, police and other responses, to just talk about the relationship between the behaviors we're concerned about: crime and social behaviors, a larger sort of regulatory environment that involves prisons—we'll talk to Reggie Wilkinson in a second—alcohol and substance abuse policy, and Peter Reuter and Mark Kleimann I'm sure will want to weigh in on the supervision of offenders in the community. Mark has already put that on the table. And we certainly want to talk about incarceration as part of the equation here.

But you've personally been involved both in the Boston story and in some of these other planning efforts around the country. Are you hopeful? Do you have concerns about whether we can reduce crime in the near future?

Mr. Kennedy: I have to begin by saying that I honestly believed that I had heard every utterable explanation for why the killing stopped in Boston. I missed lunch.

So I actually want to step back a step or two, and begin with an idea that's been raised a couple of times already. The serious crime story we're talking about is primarily one of youth violence, as has been said; that's where the action was in homicide. And I think a lot about what the truest thing that one could say about youth violence was, and what I keep coming back to, is that it was an accident. It's a product of the crack cocaine epidemic; that crack cocaine epidemic didn't have to happen, and it's certainly rooted, and much of it has been talked about. But it was a contingency. Nobody saw it coming, it is a technological development that didn't have to occur. And one of the interesting things about today's conversation is that a lot of what's being talked about really represents or is evidence of very interesting, relatively new set of ideas in the business of most of the people on this side of the table, which is that these kinds of epidemic or contingent processes really might be able to explain a lot about how crime happens. And that's to distinguish them from the way researchers have usually talked about crime, which is as rooted in more structural and persistent characteristics like structural aspects of communities, or structural aspects of continuing aspects of individual character, opportunity, and sanction structures.

But this says that crime could be a whole lot more like the weather than we've been willing to believe, and that small things could start an avalanche that then can play out in very profound ways. And I'm personally convinced that that's true of the youth violence epidemic, that if we hadn't had crack, then we wouldn't be here talking about either the rise or the fall, and that that's an important thing to realize. Now, matched into that comes the idea that there are various kinds of strategic interventions, usually these days bundled in policing under the problemsolving rubric that Jonathan talked about, but also visible in community activities and elsewhere. There might be ways to reach into these epidemics or contingent processes in ways that stop the avalanche early or are sufficiently powerful to kind of draw a wall across it. And that for relatively small investments you can get these big, maybe persistent, maybe not, but certainly worthwhile gains. And that is, if they're to be believed—which is another question—that's the story in New York, it's the story in Boston, it's the story in Richmond, it's the story in a lot of places where we're now looking for evidence of success over the last 10years, that by doing street enforcement, and doing something about gun carrying, you can essentially haul New York City back from the precipice of complete loss, that you can work hard over a year in Boston and drop the homicide rate quite dramatically, that you can get the word to chronic offenders in Richmond that there's going to be federal prosecution for gun offenses, and have big impacts on the homicide rate.

That's where I think the action lies. It's not to take away from the traditional roots, causes, or criminal justice agenda, but it's opening up this big new area where there's a lot of interesting thinking and work to be done. So that's the first point.

Second point is that we need to remember that no matter what the judgment is about how any of these things work, whether they did work, whether they might work in the future, that doesn't in any way solve the debate [or] answer the question about whether we ought to do them, and we need to keep those things separate.

Norms do and ought to drive decisions about criminal justice policy, enforcement policy, [and] crime control policy, and it fundamentally doesn't matter whether prisons, in fact, are responsible for the drop in crime, or whether zero-tolerance policing saved New York, or whether gun enforcement could do useful work at the national level. There are going to be good and sufficient reasons outside any possible impact to do or not do those things. And that's where the political debate runs. That's where the community debate often lies. It was absolutely explicit in Boston where we said, oh my God, we've got 1,300 chronic offenders here, ought we to give them all to the U.S. Attorney? And the answer was, hell no, for various reasons. Nobody involved wanted to do that, even though it might well have done the job.

So it's just to draw some bounds around an implicit assumption in the way we sort of get drawn into these impact questions. It's not either the beginning or the end of where these things come out.

All right, now, to Jeremy's text.

Mr. Travis: Quickly.

Mr. Kennedy: Corrections, comedogenic commodities like gun control and alcohol, firearms, I put in that basket. These are all areas where I think on both those above dimensions there is tremendous potential [that is] oddly unrealized for reasons I still don't entirely understand. It's been noted, as Mark always does, and properly, that there's been no conversation here about community corrections, probation, parole. And just, again, to say why this matters so much, there are three plus people out in the community under judicial supervision for every one inside of a locked facility. We talk about probation and parole supervision. When I teach about this stuff, I forbid the use of that term—no disrespect to any of the community corrections people here. But when you're talking about caseloads of 150 on up for each correctional officer, and situations in which it is often literally the case that we don't even know where people under supervision actually live and have no capacity for finding out, using the term supervision is Orwellian in the extreme. And I think there are two places here where there is—and I felt confident in saying this before Larry pointed out another connection that things could, in fact, get worse, and I guess that's true. But it seems to me like this can only get better.

It can, I think, in two ways improve. One is fundamental systemic changes. And if the money were available, and the will, then very obvious things like lowered caseloads, more drug testing, more follow-up on drug testing, all sorts of things that any of us could think of would make enormous differences in what we could actually do under even traditional probation and parole strategies.

Second possibility is kind of special projects, problemsolving, strategic intervention stuff. And those are harder to characterize because almost by nature they're all different. But there are lots of possibilities there, too. And fortunately there was something off the shelf in the work we did in Boston, which involved home visits for selected probationers, which turned out to be enormously valuable, I think both in and of itself and because when we started working closely with offending communities in meetings with gang members and such, we could say to them, look, we promise if your group offends, we will make sure that you guys get home visited. And that meant a lot. And it was clearly the case that having a bed check was much more of a concern to a lot of these folks than serious state or federal prosecution was. Clearly the case. And there is, again, research evidence. But also we could make sure it happened on Tuesday, which is not something you can guarantee for these other things. So, lots of possibilities there.

I would say the same thing about firearms trafficking. Again, this is—and just a final thought on that—this is something that ought, I think, to be uncontroversial. Nobody is against this. Say almost the same thing about the firearms stuff. Here you go to almost any community—gun-owning, not gun-owning—and say, who amongst us believes that gang members and domestic offenders ought to be able to buy guns on the street, and very few hands go up.

This is also something we can do—Garen has alluded to this—there are all kinds of hot spots which clearly ought to be attended to. There are dirty gun dealers, there are dirty gun buyers. My favorite idea here would be to eliminate the secondary market entirely. We don't let people sell heroin on the street if they say [they're] not selling enough to rise to the level of a pharmacist's license; we make them get a pharmacist's license. We make them get a pharmacist's license. And I really can't figure out why guns should be any different.

Mr. Travis: Thank you, David.

Mr. Kennedy: And I'll leave alcohol for others.

Mr. Travis: When you run something like this, you sense that people—no one is leaving here. I'm willing to go another 15 minutes, if people are willing, and we haven't heard from some very important people who were promised some time. So, we're going to go another 15 minutes.

I want to quickly get Reggie Wilkinson and Mark Kleimann both into this conversation. Reggie told us over dinner last night that his Department of Corrections in Ohio, which runs both prisons and parole, has just made an organizational decision not to build any more prisons, right? How does that—is that a crime policy decision? Is it a fiscal decision? How does supervision of Beverly's issue, reentering offenders, play into this next five years or so in crime reduction?

Mr. Wilkinson: I think more so than all the research rationales, or political rationales. I just think it's the right damn thing to do. You know, we have been building prisons like it's been going out of style. You can't build your way out of prison crowding, you can't build your way out of crime. You can't do a lot of different things. And so, I think I even have our governor and our legislature convinced, especially now that they understand that education is something that needs to be focused on more so than corrections, that those billions of dollars that we spend could be better appropriated to other kinds of functions.

We've reduced our prison population in Ohio by 6 percent in the last two-and-a-half years, so there is a method to our madness when we're talking about not building any more prisons. And so we are on record as saying that we think it's important.

Now, I do think prisons play an important role. And I do think prisons have played a role in terms of the declining crime rates over the course of time. But not just because the bad guys are locked up, and some bad guys are locked up, and that's important. You know, we've got to also talk about the adultification of the juvenile and who is going to prison, and who is not committing crimes on the streets of Boston and Chicago and other places around the country.

But for another reason I think, you know, prisons are playing a role—and my middle name is hope and optimism—I think we have done a good job with helping to rehabilitate offenders, and that word has not been used today. And I'm not afraid to say it because I think it's a very important focus that we have to have in this business, whether it's in prison, or parole, or probation, and it's happening. And I can name you a number of jurisdictions that are making a big difference. We are disproving the Martinson theory that nothing works from the middle-1970s, and saying now we understand what best practices are all about. We understand what works. We understand also what's promising, and we are focusing on those particular dynamics. We are integrating concepts like restorative justice and community justice into the corrections business.

We understand that we cannot have the kind of dichotomies that existed in the past, such as new prisons against the community corrections, and treatment against custody and the adult system against the juvenile system. We know that those systems have to be tied together, there has to be partnerships established that's going to make all of this work. We understand that we have to pay close attention to those persons with special needs, those persons who have retardation, those offenders who have a mental illness, those older offenders, those younger offenders, and a lot of people in between if we're going to have an impact. And whether the doomsayers and the naysayers like it or not, prison systems are looking very seriously at these issues. We know we're a target for a lot of people and a lot of research, but we are doing a great job in terms of making or in having a major impact on public safety. But public safety, even in our minds, isn't enough any more. We're trying to make people understand that we have to have offenders in a mindset where they're actually making a contribution to society, not just us having a mission of having them to be crime free or drug free. We need to raise the bar in terms of what our expectations are for both the offenders and the community.

We also need to look at technology and look at a number of different things that we think are going to have a major impact on the future of corrections. There's a lot more to say, but in deference to time, I'll end it there.

Mr. Travis: Mark Kleimann, you think about offender supervision and drug policy and gun policy, incarceration; what's the prognosis here for the future on the value of any of these for crime reduction?

Mr. Kleimann: We'll have to see. Any time my use is being taken away from Reginald Wilkinson it's not very cost effective. I'd much rather hear him talk about what he's doing than my thoughts about what might be done.

But let me suggest what the principles ought to be for reinventing community corrections, with some deference to the fact that it seems to have already been done in Ohio, but not on either coast, that doesn't get any—that doesn't count, it doesn't get any attention. We, the community corrections systems, have the capacity to set—let me say, most community corrections systems outside Ohio currently set very large numbers of rules for probationers and parolees. [The systems] have almost no capacity to find out whether those rules are being complied with. And when they detect violations, they have almost no capacity to do anything about it except to take the guy back to the judge and say, please put him in prison, we don't want to deal with him anymore.

So we need to reverse that. We need to set a small number of rules that we know are linked to crime and other behaviors we care about. We need the capacity for individual community corrections officers to know whether those rules are being complied with. So, if there's a rule that says you can't use hard drugs, then there needs to be drug testing. If there's a rule that says you have to do community service, then there has to be a data system that gets that information to the community corrections officer. [If there is] a rule that says you have to stay away from the drug market you used to sell in and the apartment of the girlfriend you used to beat up, then we need position monitoring to know whether that's being done. No points to any rules you can't enforce.

And then when a violation is detected, there needs to be capacity for the community corrections officer to impose a sanction, and I'd say probably a formulaic sanction, that may involve confinement and might not, but that's still within the community corrections context. You don't get to give up on the guy, you don't get to send him off to prison. You have to figure out a way to discipline him, get him back on the street to try again. So that seems to be like the agenda for community corrections.

I want to say one more thing about that. Steve Tallus just wrote an article which he signed my name to—I'm grateful for it—where he argued that getting community corrections better is the solution to the over-imprisonment problem. That if we don't have a nonprison alternative that's real, we're never going to convince anybody to start shrinking their prison population. I've heard the argument today, I've heard it before, that we ought to reduce our prison population because it's so expensive. That may be a winning argument politically, although I'd have to say I doubt it, it strikes me as not really the right argument, not really the argument anybody in this room believes. We ought to reduce the number of people in prison because having people in prison imposes unnecessary suffering on prisoners in the regiments. And we ought to be prepared to say that, that sending people to prison is a cost that we're trying to minimize, but I do believe that could be done within the constraint of continuing to reduce crime if we get the community corrections system to work better.

Mr. Travis: Thanks.

And our final word from a very patient man, and I thank you, Peter, is from Peter Reuter, who is going to reflect on what he's heard.

Mr. Reuter: I wasn't going to. And I think I will stick to what I do better, which is what I was thinking about the last 15 minutes. Which is, we've heard a fair amount about what police can do that's different, and what Mark has talked about, community corrections. I think that we also have to pay attention to factors that are perhaps like the weather, not particularly susceptible to policy interventions, and my particular piece of this action is understanding drug markets. And the sort of interesting piece of ambiguous news from the last 10 years, which is, the good news is that there's been very little initiation into use of hard drugs, certainly cocaine and heroin, even with methamphetamine, it turns out to be a fairly small number nationally, although large in some places locally, and that's unambiguously good news. We don't have much idea why initiation has gone down beyond the sort of standard David Musto theory, which is a one-line theory, called generational forgetting, that is in some sense it is the model of bad consequences of drug use that tends to reduce initiation, and when the memory of that fades, then initiation will start up again, because these are, in fact, attractive drugs.

But, anyway, we've seen very little initiation into drug use. Nonetheless, our drug problem has actually stayed rather stable. That is, if you measure it in terms of the number of people who are frequent users of expensive drugs and [who] cause a great deal of crime as a consequence—less violent crime, but a great deal of crime—as a consequence, that's actually been declining fairly slowly.

And what that reminds you of is that, in fact, these are, as the treatment community says, people with a chronic, relapsing, lifetime disorder. And one can overstate that, but these are very long addiction careers, and the addiction in our society is associated with very high levels of crime. Indeed, that's probably true in almost every Western society.

So the good news is that we seem not to have faced another epidemic. The bad news is that, in fact, we still live with the consequences of an epidemic that happened more than a generation ago. The heroin epidemic is basically something from the late '60s and early '70s, and yet still in the year 2000 the people who were infected then are still—"infected" is taking that analogy too seriously—those who initiated into frequent heroin use are still causing problems or themselves and the rest of society.

And the issue then is, well, what happens if we actually have another epidemic, because that will be layered on this long-standing population about whom we seem to be able to do very little, for whom se seem to be able to do very little. And I think in terms of the future, thinking about whether it is, indeed, possible to prevent something like a crack epidemic again seems to be fairly important. I'm less optimistic by nature than some of the others here about whether, in fact, that's possible.

Mr. Travis: I've asked Al Blumstein to give us some closing observations. We've talked about the president's commission here a couple of times, the 1968 volume that was very important to all of us; Al was the director of the science and technology unit of that commission, and here he is as a co-editor of this book.

So we expect you here in the conference 50 years from now to talk about this, since you have a lifelong interest in it. So give us a sense of what you think the next 10 to 20 years might hold in crime rates, and what might influence their increase or decline.

Mr. Blumstein: Twenty-year projections in this area are pretty tough. Let me make a confession. When Joel and I sent the manuscript out to the publishers, we had somewhat of a bidding war on their part, and there were two items that were important to us. One was that they get the book out fast because we wanted it out before the UCR in '99 came out, to make sure that it wasn't going to be contradicted very quickly, for good or bad—mostly for good, the crime decline continued, and so we weren't embarrassed there. The other was that we negotiated for a price that I suspect is a loss leader for Cambridge Press. It's under $20 for a fairly dense book with a lot of graphs and so on.

And a number of you have said you'd like to see a copy of it, and unfortunately the Cambridge folks didn't have the foresight to send a bunch of copies down. But I'm sure they will be available.

First, I thought some of the discussion in the second half of this session was really very insightful. There are lots of ideas out there of what we can do, what we can do better. I think the real difficulty is translating ideas into practice. It is clear that many of the ideas are associated with practice, but practice in the communities that the individuals who are speaking here were able to implement, and implementation in the face of lots of political restraint is very difficult.

I think it's important to note that our crime rates, however precipitously they've declined to the rates that prevailed in the late '60s, are still significantly above the rates that prevailed in the 1950s. So there's still a long ways to go.

What are the clouds on the horizon? We've all heard various comments about the economy turning sour, and as someone pointed out, there are really three means by which people get income. One is through working, one is through welfare, and there is some real concern about what happens as we finish the skimming of the people who do have viability in the labor market. And the promise of the welfare reform [is] that they're going to dump people off the welfare rolls; where they're going to get the income becomes another problem. And the third means of earning revenue is through theft or crime, or illicit market. And so, if the economy turns sour, if we find lots of people dumped off the welfare rolls who don't have economic skills, those represent serious concerns.

A number of people have made mention of the fact that we're going to have—we now have roughly almost 600,000 people coming out of prison every year. Reggie Wilkinson talked about rehabilitation, and Ohio and Reggie are impressive examples of where rehabilitation is important, but I think that's much more the exception than the rule, and there's a real concern about the degree to which prisonization and criminalization through the incarceration process will result in their contributing to a crime problem.

Right now, we're putting more people in than we're taking out, but it's not going to be very long, if current trends continue, before that number reverses and then more people start coming out.

Indeed, one of the reasons so many are going in is not that we're getting a lot more criminals sent in, it's that we have become really efficient at urinalysis of parolees, and so that given the propensities of addiction to keep people on drugs, whether or not they're committing crimes, they end up contributing to what in most states now is a larger number of people that are in prison than are entering under new crimes, new court commitments. So that we're seeing an increased investment in incarceration in an environment that views incarceration as the dominant solution to whatever problem we're facing, regardless of whether it works, as it certainly does for incapacitating violent offenders, or whether it doesn't, as is so markedly the case in locking up drug offenders. So that the hope for the future is that we get some of the wisdom and rationality that we've seen by the panelists this afternoon, or this morning, get some of that wisdom and rationality into the public debate.

I'm delighted personally that we haven't seen crime become a major issue in the political campaign, because it's most likely that if it were, we would see a whole rash of irresponsible policies that one of the candidates would feel impelled to do something about, because the nature of campaign drives one to simplistic solutions that the public will resonate to. And what we heard this morning, what was really sophisticated assessment of the complexities, of the tradeoffs, of the difficulties of using the simplistic solutions, and in the theme of the book, there's a whole variety of things that contributed to this crime drop. Any one of them alone would not have done it. It's the combination of them, and in any political setting people find a simple one that people will respond to, and the press in particular: "Don't bother me with all that complexity, tell me what's really behind the crime drop." And as long as that's the attitude that prevails in the country, it's going to be very difficult to get the political environment to deal with the real complexities associated with continuing the crime drop, and we do have lots of room for it, and we've just got to find much more enlightenment in the implementation, and hopefully that might happen anew in January.

Mr. Travis: We are at the close of our time together. I'm reminded—early in his term, President Kennedy invited to the White House all of the living Nobel laureates, and said when they were sitting around dinner, that this is the greatest assemblage of brain power in one room, with the possible exception of when Thomas Jefferson dined alone. I feel that we've assembled not just here, but in the audience as well, the greatest assemblage of brain power on this topic. So, I want to thank everybody for participating, and thanks for coming.

Return to The Decline in Crime: Why and What Next?



Topics/Tags: | Crime/Justice | Families and Parenting | Race/Ethnicity/Gender


Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact publicaffairs@urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.

Email this Page