The Supreme Court Provides a Dissertation Topic for a Budding Economist

Photo: dbking

Last week, the Supreme Court ordered California to release at least 30,000 prisoners due to poor prison conditions caused by overcrowding.

This is what economists call a “natural experiment,” or what I prefer to call an “accidental experiment.” The Supreme Court order will be a “shock” to the California prison system, leading to roughly a 10 percent reduction in the prison population there. I used this sort of accidental experiment in a paper I published back in 1996, finding a large impact of mandated prison releases on state crime rates.  If my estimates remain relevant to the current time period, I predict that California violent crime rates should rise about 4 percent relative to the rest of the U.S. over the next few years. That adds up to about 80 extra homicides a year.

Five years from now, no doubt, an economics graduate student will analyze the data and tell us what the actual numbers look like. Unless, of course, I beat them to the punch!


Does your model take in to account the "types" of convicteds released? I'm going to assume anyone in for Murder 1 or 2 won't be getting a supreme court-ordered release; but rather those in jail for marijuana-related offenses, prostitution, etc.

joey k

How many people do you really think are in state prison for pot and prostitution? Both probably misdemeanors I'm sure.This is cali we are talking about.


About 20% of the current California prison population are there on drug convictions:'s_attempt_at_prison_reform_looking_like_an_attempt_to_pass_the_buck

That's more than enough to meet the court-mandated reduction right there. And help with California's budget problems, too.


The court order was to "reduce the state prison population" not necessarily to release them. There are a couple of options: They could move prisoners to another state, or move some into county jails. Politicians will do everything they possibly can to avoid being responsible for "releasing prisoners onto the streets," that's why the problem is so bad in the first place.
If they do have to release some, they will be very careful about whom they release (i.e. not Willie Horton).


Interesting natural experiment. However the prison release stems from California's relative economic troubles. Any increase in the crime rate could be due to that instead of the release of prisoners. Cointegrated variables imo.

Steve Bennett

Ok, let's assume 80 extra homicides per year. Question is: what can you spend the money saved by not incarcerating all those prisoners on that would more than compensate for those 80 violent deaths?


I just can imagine all the crime economists stirring in their armchairs, adjusting their thick glasses and saying "Those grads and freaks are not going to be fast enough, I will beat them all" and the flag on the finish line is going to have black and white stripes.


First off the order was to reduce, so releasing is one of a few options available. I believe that there are a large number of non-violent drug offenders, along with parole violators whose only infraction was testing dirty, that could and should be let go. Does this affect the statistics at all. Clearly all criminals are not the same. I tried to read the 1996 study but not being an economist nor an academic it became a prison sentence unto itself.
Help on a few other things please..1. Do your statistics take into account the potential reduction in violent crime, both reported and unreported, inside the prisons themselves that could result from a 10% prison population reduction? After all crime is crime.
2. How relevant are crime stats anyway? They are so easily manipulated and falsified, and even when accurate tell an incomplete story at best. Of those 80 extra homocides, and in fact the 4% overall spike, how much of that will be criminal on criminal violence. Of those 80 bodies how many will be violent criminals themselves. Gangbangers killing gangbangers drug-dealers killing drug-dealers. Obviously even in this scenario there is collateral damage, but if it lowers the heat in these thug factories we call prisons, if due to jacked conditions we are not turning out so many new supercriminals, maybe the net result is more positive than you think. Of course yelling out 80 EXTRA HOMOCIDES is alot more scary.