« Split Eleventh Circuit ruling on reasonableness of long sentence for one co-conspirator | Main | China changing execution method from firing squads to lethal injection »

June 15, 2009

"Explaining the imprisonment epidemic"

The title of this post is the title of this editorial introcution to this issue of the February 2009 issue of Criminology & Public Policy.  Here are the other articles on imprisonment in this issue:

June 15, 2009 at 05:29 PM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference "Explaining the imprisonment epidemic":


Thanks for posting that, and thanks to Wiley for providing access to the full text. This is heavy stuff that will take some time to digest, but it looks quite interesting on a quick skim.

Posted by: Kent Scheidegger | Jun 15, 2009 6:38:56 PM

I had a brief item about Spelman's piece recently here where I pulled what I thought was the money quote from his meta-analysis of prison cost-benefit studies:

"Estimates vary widely, but the marginal prison bed seems to prevent somewhtere between two and seven crimes, which saves potential victims between $4,000 and $19,000 per year.

"But note the details: If each prison bed reduces costs by no more than $19,000, but costs us $20,000 to $40,000, then do we need this many beds? Clearly not, and it's not (too) difficult to use current estimates of the crime-control effectiveness of prison, the costs of crime to victims and nonvictims, and the costs of prison to show that we overshot the mark sometime in the early 1990s. Enormous cutbacks - reductions of 50% or more in the prison popoulation - are not difficult to justify and would probably save the US public billions of dollars earch year. Certainly there is little economic justification for continuing to build."

Spelman's conclusions are especially relevant because, in previous work, he's provided some of the highest estimates among peer-reviewed researchers for the effectiveness vis a vis public safety of expanded prison spending in the 1990s. But even using the highest estimates for costs of crime, each new marginal bed is an economic loser from the perspective of a rational cost-benefit analysis.

FWIW, Spelman was just (re)elected to the Austin City Council. He ran unopposed after earlier serving a single term and retiring without running for re-election. In addition to being a top-flight criminal justice statistician, he's one of the most respected city leaders in Austin.

Posted by: Gritsforbreakfast | Jun 15, 2009 7:45:59 PM

If crime were pursued in a tort litigation against the criminal, what would the verdict be for a carjacking? It is likely $1 million. Say, it is $10,000.

The average criminal likely commits 50 crimes a year, many commit 500. This is based on crime victimization, the reliable measure of crime. Most are not reported. Most are not investigated if reported. Most are not prosecuted if investigated. Every conviction stands in for perhaps 100 other crimes.

Take the most conservative valuations. 50 crimes at $10,000 each. The return on investment of $25,000 a year is 20 times, year in year out, guaranteed, with no risk, unless the convict is innocent. Few investments are that good.

Posted by: Supremacy Claus | Jun 15, 2009 10:38:36 PM

SC, "the most conservative valuations" are among the ones that Spelman cited. You're just making stuff up.

Posted by: Gritsforbreakfast | Jun 16, 2009 7:20:09 AM


It would not surprise me at all if people who commit crimes such as burglary also tend to flout other laws, such as traffic regulations. If you were to count these low level, difficult to detect crimes it would not surprise me at all if SC's 50 to 500 figure were correct.

Posted by: Soronel Haetir | Jun 16, 2009 9:05:19 AM

There are 23 million FBI Index felonies a year, 5 million being violent. Only 1 in 100 of these ends in any prosecution. Most prosecutions do not end in a prison sentence. The busier the criminal, the likelier the chance of ending in prison. So, most of the convicts are quite busy.

Don't take my word. Fill in the numbers below for yourself.

Does crime generate medical costs, police costs, personal and business security costs, social services costs?

Does crime drop real estate value?

Does crime drop the productivity loss from the loss of trust, and the deterrence of commerce? Does crime impair educational aspirations of kids in bad neighborhoods, suppressing educational achievement?

What would you pay to prevent your being carjacked? If someone did carjack you, and they had assets (criminals are quite rich), what settlement or verdict would be a fair to make you whole after the assault (inducement of fear), assuming no physical injury? Would exemplary damages apply to such an intentional tort?

Posted by: Supremacy Claus | Jun 16, 2009 9:18:57 AM

Soronel, now we're including TRAFFIC VIOLATIONS in a discussion of prison populations? Please! SC is ignoring the real data and making up numbers to suit herself.

SC, regarding your comment at 9:18, you need to read Spelman's article and you'll find he's asking exactly those questions and building a statistical model to answer them. The difference is, he's constructing his model from actual, real-world data that he thoroughly documents instead of pulling the numbers out of his behind.

Posted by: Gritsforbreakfast | Jun 16, 2009 10:28:01 AM

I could not find where he interviewed criminals about their activities with certified immunity, and counted the reported rates. When you do that, the rate is about one a week. A junkie from the 1980's would be doing a 100's of felonies a year to support a habit. Put in abuse of kids and women, property damage, disruptive behaviors, stealing from everywhere, bringing down real estate prices to zero. Where is all that?

Posted by: Supremacy Claus | Jun 16, 2009 10:50:50 AM

For state offenses the decision about the use of alternatives to incarceration is made at the county court level and there can be very large variations in incarceration rates from county to county. If you use state and regional data the local factors are reduced by averaging.

The crime venue is also very important because rural crime differs greatly from urban crime and within the urban crime class there are large variations between domestic and street crime. It is very important to take time into account because some types of crime are cyclic and criminals may change their theater of operations in response to increased police presence. My main concern about these models is that they tend to discount the local nature of crime.

Posted by: John Neff | Jun 16, 2009 10:54:01 AM

Then you didn't read his footnotes, SC. The cost estimates he draws on account for both reported an unreported crimes.

Why don't YOU perform the study you describe, publish it in a peer-reviewed journal like Spelman did, then we can continue debating the merits of your made up guesstimates.

Posted by: Gritsforbreakfast | Jun 16, 2009 10:56:38 AM

Peer reviewed journals are all Commie front propaganda organs, devoid of any validity. They are garbage science, expressing the big government preferences of their academic editors.

I invited your guesstimates. I own a crack house. I go there to ask for the rent, they will kill me with a Nine held sideways. As in this documentary quality fictional movie.


I recommend this movie to all the criminal lovers here. The police is nearly irrelevant, as the crime meter whirs at maximum speed. These folks have nearly total immunity granted by the criminal lover lawyer. One may not even verbally criticize them without risking losing one's government job.

Back to the property I want you to buy. They have destroyed everything of value in it, including all copper wiring. This house was in a good neighborhood before the lawyer destroyed it by forcing sales and rentals to criminal elements, on the pretext of anti-discrimination.

Name a price. An equivalent home in good condition in a lawyer neighborhood is worth $250,000. Make me an offer. I know no one would buy it. The criminals you love have caused it to have a negative value. What offer would I have to make to get you to take possession?

Posted by: Supremacy Claus | Jun 16, 2009 3:33:57 PM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB