« Bankruptcy court as government intervention | Main | Assorted links »
*When Brute Force Fails*
That is the new and excellent book by Mark Kleiman and the subtitle is How to Have Less Crime and Less Punishment.
The reader learns many interesting facts. In 1974 the average burglary resulted in about four days behind bars (that's the average sentence adjusted by the chance of being caught). The average 1974 burglary yielded about $320 (in today's dollars), which amounts to about $80 a day payment for time behind bars. Even forgetting about the fixed costs of getting a criminal record, or a longer criminal record, that screams out to me: "It isn't worth it!"
Today, for burglary, the average return per day in jail is $22. That is one reason why crime has gone down but the real question is why crime has not fallen even more.
Many burglars are not risk-averse or they underestimate their chances of being caught.
There are many excellent bits and segments in this book.
Posted by Tyler Cowen on September 15, 2009 at 07:28 AM in Law | Permalink
Comments
What about opportunity cost? Many burglars are desparate people with nothing to lose. Four days behind bars might be a positive compared to their normal accomodations. Also, don't forget the signaling benefits. Not all social circles consider a criminal record a bad thing.
Posted by: Steve R at Sep 15, 2009 8:06:11 AM
It's a lottery isn't it? "This time I'll hit the jackpot - the unexpected Picasso, the box of gold - and my fortune will be made." It never happens. Just as you "don't" win the lottery. Yet people still play fervently. Either that or burglars are cognitively impaired.
Posted by: Sebastian Franck at Sep 15, 2009 8:24:48 AM
I think that restitution should be the default punishment for property crimes.
Posted by: Floccina at Sep 15, 2009 8:46:22 AM
At least some burglars are "professionals." They spend their lives being burglars and that is the primary source of their income. And they aren't looking for a box of gold.
I have read news stories about burglars who spent over a decade doing what they did for a living. Now maybe these folks are outliers, but burglars clearly aren't a monolithic class.
Posted by: Seward at Sep 15, 2009 8:46:50 AM
Dope heads who need money for a fix aren't known for doing a lot of critical thinking. An addict will steal from their own family to buy more drugs. They are thinking purely in the short term "How am I going to buy more drugs", not some elaborate economic calculation.
Also, it is important to remember that average jail times for the United States are just about meaningless. In Detroit, for example, home burglary are not investigated nor are they punished (assuming that it was a simply burglary and not armed robbery or something!), the police won't even bother come to your house to take a report (you have to go to the police station). For all practical purposes, burglary has been decriminalized - the only real deterrent to burglary is self defense (almost universal gun ownership, and pitbull ownership). In one of the Oakland County suburbs of Detroit, in contrast, a burglary conviction can get you years in prison! Averages are useless in reviewing U.S. crime statistics. There is just too much variation in crime and punishment depending on your location.
Or perhaps burglars are just showing "irrational exuberance" in the market.
Posted by: Vehical Driver at Sep 15, 2009 8:59:48 AM
Expected value is a poor model for decision making within outcome distributions having a strong mode of zero. Put another way, no one expects to be caught.
Posted by: Noah Yetter at Sep 15, 2009 9:56:36 AM
"The average 1974 burglary yielded about $320 (in today's dollars), which amounts to about $80 a day payment for time behind bars. Even forgetting about the fixed costs of getting a criminal record, or a longer criminal record, that screams out to me: "It isn't worth it!""
This screams to me that Tyler has never tried to pay bills on a couple of part-time no-benefit minimum-wage jobs.
$7.25/hr x 10hrs/day < $80 a day
Posted by: Alger at Sep 15, 2009 10:39:34 AM
Dope heads who need money for a fix aren't known for doing a lot of critical thinking.
Criminals in general don't do much critical thinking. They can't - the average I.Q. score of prison inmates is 85.
Posted by: Peter at Sep 15, 2009 11:02:37 AM
maybe urban crime in the 70s and 80s was just a bubble and when it popped crime rates collapsed.
Posted by: babar at Sep 15, 2009 11:16:55 AM
I'm with Vehical Driver here; using national aggregates is horribly misleading.
It's all about location, location, location. You can combine the mean average rate crime in the combined group of Vermont and Washington DC and nowhere else if you want to, but that mean average is absolutely meaningless (so to speak).
I recall a news article from the 80's or early 90's, breathlessly stating that our SAT scores were dropping, and that this meant that our schools were underfunded. The same article then went on to point out that there was some good news though, that scores for blacks were rising significantly, and the same with Hispanics. It also mentioned in passing that scores were up a bit for whites and also for Asians. The story never connected the dots that our demographics were changing and that by no means could one deduce that we needed more education spending, at least not from the given data.
The bottom line is that aggregate statistics lie. You need to dig down and break down each subcategory and view the stats and the trend in each category.
The same thing is happening in national income averages too, with the aggregate data showing that only a very few at the top have improved, while data that tracks individuals showing that the biggest gains were for those in the bottom quintile for any decade measured.
Posted by: happyjuggler0 at Sep 15, 2009 11:30:34 AM
1. $7.25/hr x 10hrs/day < $80 a day
2. There is a wide distribution of probabilities of getting caught. A skilled burglar can work for years without capture. A unskilled burglar is too dumb to accurately estimate his probability of getting caught.
If your options are to make $70 a day at Mcdonalds or steal stuff then you have to figure in that you will be paying payroll taxes right off the bat plus other withholding. If you steal some stuff you can evade payroll taxes. If you go to jail 4 days then you are also getting free meals/workout equipment/libraries and networking opportunities. They can discuss various techniques with people in the same line of work, learn about the most frequent errors that criminals make and then try to improve their methods by eliminating those simple mistakes. They can also learn about the legal system and figure out better ways to play the lawyer/evidence/court game.
Posted by: Gabe at Sep 15, 2009 11:46:06 AM
Happyjugglers statistical points are well taken, particularly when the composition of the aggregate changes over time. That doesn't undercut the expected value of the haul, unadjusted for the risk of detection. What you might want to do is look at the percentage of unsolved criminal complaints to get an adjustment for the probability of getting caught.
Posted by: Bill at Sep 15, 2009 12:06:33 PM
that scores for blacks were rising significantly, and the same with Hispanics. It also mentioned in passing that scores were up a bit for whites and also for Asians
But SAT scores were failling??
Posted by: athEIst at Sep 15, 2009 12:32:20 PM
athEIst,
Yes, the aggregate was falling, despite each subgroup going up. The reason is that the lower scoring groups were:
1) much lower to start with
and
2) increasing in population size
It is like having Steve Jobs alone in a room, and adding extra middle class people to the room, each of whom have rising net worths. What happens to the average net worth per capita of the people in the room as each new person enters? It goes down, even though everyone's net worth is rising.
Posted by: happyjuggler0 at Sep 15, 2009 1:10:51 PM
Invisible hand(cuffs) ?
Posted by: Rama at Sep 15, 2009 1:21:59 PM
Does the book talk at all about rationalizing prison sentences for various crimes? I feel like I'm constantly reading both about people who serve six months in prison for home invasion and people who serve life for shoplifting penny candy.
Posted by: Scoop at Sep 15, 2009 1:36:47 PM
If the rewards of burglary and the rate of it have declined, perhaps its because most people do not have much in their homes that is worth stealing these days. Electronics are dirt cheap (except for the big flat screen TVs, and those are getting cheaper too). Furniture is hard to move and much of that, even the good stuff, is a lot cheaper today than it was in the 70's. Most people do not keep large amounts of cash in their home, with debit and credit cards being so common. Gold and Silver, if not tableware, these are usually kept in a bank vault.
In short, residential burglary does not pay any more.
Posted by: kurt9 at Sep 15, 2009 2:16:09 PM
Any thoughts on when the backlash against subtitles and/or cute titles will finally happen for real?
Posted by: Greg at Sep 15, 2009 2:44:32 PM
Looks like an example of "missing markets in almost everything", a useful parallel series to Prof. Cowan's fascinating "markets in almost everything" series!
Posted by: PQuincy at Sep 15, 2009 4:44:48 PM
I don't think it's a matter of risk preference or misunderstanding the odds of getting caught. A declining return still suggests a risk preferring individual would engage in less crime.
I think the more obvious explanation is to apply Becker's human capital model to his crime and punishment model. The returns to crime are declining, but I suspect today's low wage and low skill occupations require more human capital to succeed at than the low skill occupations available in 1974. And perhaps, there is more human capital to being a criminal in 2009, when criminality - especially imprisonment - more often leads to gang affiliation, drug use, and a wholesale plunge into the "thug life". I could be wrong about this... it's obviously an empirical question, but suppose you're released from jail after committing a burglary in 1974 or 2009. I think I might prefer to be released in 1974.
Posted by: MikeDC at Sep 15, 2009 5:39:25 PM
How about unusually large discount rate for the future? 320$ cash now may be worth it if you pay for it at a time you discount to 1% of its value now...
Posted by: Mikko at Sep 16, 2009 2:13:24 AM
I assume outcomes vary significantly among burglars. Possibly there are some "super burglars" whose chances of getting caught are close to 0. Hoping to emulate, less competent burglars are the ones who get caught. But since they see it is possible to succeed they keep trying. I guess it's sort of like professional sports in that respect (the median outcome for professional sports is pretty terrible).
Posted by: Andy at Sep 16, 2009 8:22:19 AM
I suspect most burglars *hugely* discount the future. Jail time is usually paid farther in the future than the procedes would be gained.
Along the same lines I suspect that many burglars are drug addicts who need money ASAP to avoid withdrawal. People who need to pay off debts with high consequences for failure also come to mind. Without a good credit score these people probably don't have other means to get the money immediately.
So if I highly discount the future and I get huge utility gains from having cash within the day the lack of competing ways to make a quick buck makes it perfectly rational to be relatively non-responsive to decreases in the cost per prison day.
Posted by: TruePath at Sep 16, 2009 1:42:30 PM
This is a very good security feature, however if a person is bruteforcing a site.
Can their true ip be logged if they are using proxy severs to bruteforce ( I dont think so)
also if someone wants to really piss someone off they can brute force every 15 minutes ( write a program to do this) and they will be for ever out of an account
Posted by: firewire at Sep 19, 2009 3:36:49 AM
