Busy day today, but I can't let this topic go without commenting. Kevin today tag-teams off of one of Barry's posts earlier this week about the hesitancy of the Durham Police Department in going after speeders. I meant to get to Barry's post earlier, but I can't let yet another one on this topic go by without commenting on it.
My master's thesis was on crime in Durham and various location-based strategies for addressing crime, and what their results are. In the process, I did a pretty extensive review on the criminology literature surrounding order maintenance policing, better known as the "broken windows" theory, based on an article of the same name that James Q. Wilson and William Kelling wrote in the Atlantic in 1980. Back in early '06, when I was in the thick of finishing the thing, we got into quite the conversation on the PAC2 mailing list about the panhandling ban, and whether it represented an effective "broken windows"-style approach to crime prevention. I thought it was not, and put forward, oddly enough, residential speeding as something that the police should crack down on instead.
This message came in the middle of a conversation, so it refers to e-mails written by Chris Sevick and Ken Gasch previously. I think it's still pretty apparent what's going on, and this post laid out my case for how broken windows applies to Durham better than I think I could write now. And since I don't have a whole lot of time, I'll just quote the thing in its entirety below the fold, edited slightly for formatting:
To respond to a couple of posts quoted down at the bottom:
Ken is essentially coming out and advocating what I call the strong broken windows position. Chris thinks the whole broken windows idea is "silly." In order to respond substantively without making everyone read another long one, I'm going to put the main points here at the top, and the longer discussion underneath. (I'm trying to be concise, I promise.)
The main things I want to say here:
1. I disagree with Chris when he implies that petty issues like jaywalking aren't worthy of some degree of attention, but I also disagree with Ken's implication that cutting down on panhandling could have a similar impact as the fare jumping crackdown in NYC. The two are very different.
2. There's been a lot of work done on why the order maintenance policy worked in New York, and there's still plenty to be done. That said, there's pretty good evidence that whatever else happened, it's highly unlikely that the drop in crime happened because of decreased disorder. That's not to say the effort did nothing -- the reason the fare-jumper crackdown worked is because it lead to the arrest of criminals with outstanding warrants for more serious crimes, or arrested them for fare jumping before they went on to do something more serious. The reason here is because there was a *large overlap* between fare jumpers and serious criminals.
3. Because of this, I find it incredibly unlikely that cracking down on median panhandlers will cut down on the number of murders, assaults, robberies, rapes, larcenies, or other serious crimes. Why? Because I really find it hard to believe that in arresting median panhandlers, we're going to turn up felons. Call me crazy, but I just don't see the guy in the wheelchair at Duke and Gregson as the gangsta type.
4. "Broken windows" is all about enforcing norms of conduct and of physical maintenance, so that people feel comfortable and at home in public spaces. With the existing panhandling laws, we've already required median panhandlers to follow codes of orderly behavior. They must obtain a permit, wear a reflective vest, stay out of travel lanes, and stop at night. As such, *we've already incorporated them into our notion of orderly existence.* The way they are behaving is, in fact, quite orderly.
5. I'm all for decreasing the disorder in Durham, but let's pick some targets that will have the most impact. For instance:
- On the subject of physical disorder, why is so much city property a mess? Why are city sidewalks so badly cracked and warped? Why are city roads in such bad shape? Fixing these would increase residents' pride in their neighborhoods, make them feel ownership of the space, and perhaps encourage them to fix up their own properties, and we don't have to worry about stepping on anyone's rights.
- How about people speeding through residential neighborhoods? I'll put money that if you set up a kind of police sting operation for residential speeders, particularly in higher crime areas, you'd pick up a fair number of serious criminals.
- Or, for that matter, Ken's anti-parking-in-yards kick. I think he may really be on to something.
Now, the longer discussion:
The fundamental question under debate here, as I see it, is to what degree cracking down on these minor crimes will prevent major ones. We're not alone here -- there's a lot of people out there spending a lot of time trying to figure out how we can generalize on what happened in New York. Did subway crimes fall because the increased order inspired everyone to look a little harder at criminals, or was it something else?
The strong "Broken Windows" hypothesis, as laid out by Wilson and Kelling in their 1982 Atlantic Monthly article, is that allowing disorder to creep leads to more serious crime. In sociological language, this is a strong causal theory; in other words, disorder *causes* serious crime. Sociologists address problems like this with a statistical technique called multiple regression. Roughly, this lets you see how multiple variables mix, and in effect, using levels of correlation and removal to determine the likelihood that one thing causes another. The problem with using this to address the strong "Broken Windows" hypothesis (which I'm going to call the SBWH from now on), is that disorder is very hard to quantify in any sort of objective fashion. Thus, it was 17 years after Wilson and Kelling's 1982 article that someone finally figured out how to do it.
The breakthrough came from Robert Sampson and Stephen Raudenbush in a landmark 1999 article in the American Journal of Sociology. The method they used was, in my opinion, brilliant. They rigged up several cars with cameras inside, and tinted windows to hide the cameras, and drove around Chicago videotaping neighborhoods, for a total of over 15,000 street blocks. They then took all of these tapes back and sat undergraduates in front of them, and told them to score each block for the presence or
absence of a number of types of disorder, as defined by Wilson and Kelling, such as litter, gang graffiti, condoms and needles on the sidewalk, empty beer bottles, people drinking alcohol in public, prostitutes, or adults or youths congregating on the sidewalk. On top of that, they went through the
neighborhoods and did surveys of adult residents, then took this all back and compared it with crime data from the Police Department, and fed it through a number of multiple regressions. They found many fascinating results, but most importantly, they found strong evidence that disorder is not a strong causal factor causing serious crime (in this case, murders, assaults, and robberies). Instead, it appeared that a third causal factor was the driving force behind both crime and disorder, and that the two simply tended to co-occur. In other words, ONLY DECREASING DISORDER WOULD NOT CAUSE A DECREASE IN CRIME.
So, what happened in New York City, with the turnstile jumpers? Well, my guess is that Ken is getting his story from Chapter 4 of Kelling and Coles' book, "Fixing Broken Windows." (I had to wait a day to write this e-mail so I could go home and get my copy.) Fortunately, the answer to how this happens is right here in their book:
During the early days of the farebeating effort police discovered that a high percentage of those arrested for farebeating either were carrying illegal weapons or had warrants outstanding for their
arrest on felony charges, many for crimes committed in the subway. In certain neighborhoods, as many as one arrestee in ten was either wanted on a felony charge or carrying an illegal weapon. . . Consequently, when action was taken against farebeaters, serious crime dropped. (p 134)
Which is why, in my bullet points above, I point out why it's so unlikely that taking panhandlers off of medians is going to do anything. The criminals in our community are simply not going to pay $25/year, buy an orange vest, and pay attention to the size of their road sign just to get a few bucks.
So, since I really don't want to be seen as someone who's only here to criticize and never come up with any constructive ideas, let's look for ways of cleaning up social disorder in our neighborhoods that would actually have a positive impact on serious crime. Why did the subway farebeating effort work so well? Simple: the subway is how people get around in New York City. How do they get around in Durham? In their cars. So, to apply similar logic to the NYC example, if you're already wanted on felony charges, are dealing large amounts of drugs, and are involved in some sort of gang or other turf struggle against other thugs in the city, are you really going to be careful to drive 35 through the 'hood? I seriously doubt it. Which is why I say have unmarked police cars set up for stings on neighborhood streets. Is anyone seriously going to object to cracking down on people burning rubber on the streets that our kids play in, our cats have to cross, and we have to walk down? Wouldn't it be a major quality of life improvement? And again, don't you think we're a lot more likely to pick up serious offenders that way than by cracking down on the guy in the wheelchair?
I'll close with the immortal words of Tom Waits:
Seen them fellows with the cardboard signs
Scraping up a little money to buy a bottle of wine
Pregnant women and Vietnam vets
I said, Beggin' on the freeway's 'bout as hard as it gets.