« Yale LJ Pocket Part piece on Kennedy rehearing denial | Main | New federal sex offender law (with Senator McCain's imprint?) »

October 15, 2008

"On Justifying Punishment: The Discrepancy between Words and Actions"

The title of this post is the title of this interesting article from a social scientist that was recently posted on SSRN.  Here is the abstract:

This article reveals a discrepancy between the actual and stated motives for punishment.  Two studies conducted with nationally representative samples reveal that people support laws designed on the utilitarian principle of deterrence in the abstract, yet reject the consequences of these same laws when they are applied.  Study 1 (N = 133) found that participants assigned punishment to criminals in a manner consistent with a retributive theory of justice rather than deterrence.  The verbal justifications for punishment given by these same respondents, however, failed to correlate with their actual retributive behavior.  Study 2 (N = 125) again found that people have favorable attitudes towards utilitarian laws and rate them as "fair" in the abstract, but frequently reject these same laws when they are instantiated in ways that support utilitarian theories.  These studies reveal people's inability to know their own motivations, and show that one consequence of this ignorance is to generate support for laws that they ultimately find unjust.

I lack the social science background essential to assess carefully the empirical work and analysis found in this article.  But the ultimate conclusions drawn by the author could have profound implications for sentencing structures and policies at both system-wide and case-specific level.

October 15, 2008 at 04:35 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference "On Justifying Punishment: The Discrepancy between Words and Actions":


I do not pretend to be qualified to critique the methodology, either, but here is one red flag. Carlsmith notes the result of one of his prior studies that the participants "generally ignored information about whether the perpetrator had committed similar crimes before and whether he was likely to commit them again in the future."

Um, so all the defense lawyers fighting tooth and nail to keep priors away from the jury can just relax and not worry about it? I doubt you would get many takers for that proposition at an NACDL meeting.

Posted by: Kent Scheidegger | Oct 15, 2008 10:55:50 AM

Maybe I have a flawed definition of of utilitarianism, but I would consider it to extend beyond just seeking to deter crime. A utilitarian should look to deterence, but also to incapacitation and rehabilitation.

The sampling methods seem very reasonable, and it seems like alot of care has gone in to the studies used by the article, but I wonder if it is really possible to create an objective standard as to what is retributist, and what is deterist or utilitarianist. Utlimatly, reasonable people can differ as to what the most utilitarian approach to a problem is. In many casesa a person who is more 'deserving' of punishment, will be percieved as more likely to re-offend, therefor a longer or harsher sentance may be called for by both retributists and utilitarians, even if sentence was unlikely to deter others.

Ultimatly, the article doesn't contain enough information to draw our own conclusions about what the responses mean.

Posted by: Monty | Oct 15, 2008 12:37:24 PM

I agree with Monty above and would add two other points.

(1) I was bothered by the limitation of the research to two seperate theories. There are many theories about why punishment should happen. The author assumes that because the utilitarian and retributive theories are the most common in America, it follows that the participants must be operating under one or the other theory. That's just not true, and I can see nothing in the article that indicates how he dealt with that problem.

(2) In addition, the demographic sample was 40% non-American. Theories of punishment are indisputably culturally bound. It's difficult to know if the observed discrepancy between action and justification is located in the individual psyche or rather the difference has it's locus in different culturally bound legal systems. The author is a psychologist, and so his natural bias is to look at the individual. But his actual sample is so culturally diverse it is at least plausible that all he is measuring is differences in culture. His refusal to deal with this issue in his paper is also troubling.

Posted by: Daniel | Oct 15, 2008 1:20:53 PM

For those without access to SSRN, here's an "uncorrected proof" PDF from Colgate University.


(Prof. Berman, I am an interested lay person.)

Posted by: Branden R. | Oct 15, 2008 3:08:56 PM

“Punishment” is the concept of a whole category of strategies that respond to an offense; e.g., denunciation, condemnation, retribution, restitution and disassociation. It is not surprising that people respond differently at one point in time or another. All of these strategies are legitimate ways of responding to offenses, depending upon what is more salient at the moment.

What this tells us is that a sentencing system should be structured so that it can accommodate all of these strategies and their associated tactics. I believe this can be done by describing the deprivations that carryout these strategies generically, so that the less restrictive can be nested within the more restrictive. In this way, all of the State’s sentencing objectives can be accomplished, because the most restrictive tactics control at any given point in time.

Posted by: Tom McGee | Oct 15, 2008 3:24:04 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