"Smart people (like smart lawyers) can come up with very good explanations for mistaken points of view."

- Richard P. Feynman, Physicist

"There is a danger in clarity, the danger of over looking the subtleties of truth."

-Alfred North Whitehead

September 27, 2010

"Nothing Works" in Reducing Recidivism is False

In the mid 1970’s, a summary of 231 studies of correctional programs were reviewed and evaluated based on their effectiveness in reducing criminal recidivism. The ultimate conclusion of the researchers was that “nothing works” in reducing recidivism of offenders.[i] Other researchers have criticized these reviews and conclusions. The conclusion that “nothing works” to reduce recidivism probably wasn’t true back in the 1975, and is certainly not true today.

One group of researchers argued that the Martinson review cited above, demonstrated knowledge destruction as the conclusions were simply not empirically based. However, at the time, the Martinson conclusions were politically acceptable from both ends of the political spectrum. Researchers now argue that the empirical research had identified programs that work to reduce recidivism then, and the research has identified programs that work now.[ii]

Recently, there have been several comprehensive meta-analyses and analytical reviews of the literature to attempt to identify correctional programs that are effective in reducing recidivism and those that are cost effective. Of course, the goal is to use the effective programs and eliminate the ineffective programs.[iii]

I will discuss some of the findings of what works in corrections to reduce recidivism. I will warn anyone that is looking for a silver bullet program that is going to reduce recidivism by close to 100%, that these programs don’t exist. Recidivism reduction in the 10% to 15% range compared to control groups is on the high end. The question is do the benefits of the programs outweigh the costs.

Throughout the treatment literature, one principal stands out. To increase the efficacy of any treatment modality, one first needs to identify the group of offenders at highest risk to re-offend, identify the criminologic factors of those high-risk offenders that contribute to their offending, and then design treatment to address those factors. Treatment resources should not be wasted on offenders with a low risk to re-offend or those lacking the problems that the treatment modality is designed to meet.[iv]

Vocational education of offenders with job placement was shown by all three groups of researchers to reduce recidivism. According to Drake et al (2009), recidivism is reduced about 10% with vocational education in prison, and it provides the greatest return on investment of any program. Job placement increases the effectiveness of vocational training by about three-fold.[v]

Adult basic education again has been shown to reduce recidivism. Andrews and Bonta (2010) and Drake et al (2009) both concluded that adult basic education reduced recidivism. MacKenzie (2006) found adult basic education to be promising. Drake (2009) estimated a reduction in recidivism by 8.6% from adult basic education.

Working in correctional industries in prison has also been shown to have a positive effect on reducing recidivism. Andrews and Bonta (2010) did not address correctional industries. The other two research groups found that work in correctional industries was effective in reducing recidivism. Drake (2009) estimated that correctional industries reduced recidivism by 6.4% over nonparticipants.

The above examples show that educational programs and work programs have been shown to be effective in reducing recidivism. I will discuss more structured correctional treatment in my next entry.



[i] Martinson, R. (1974). “What Works—Questions and Answers About Prison Reform.” The Public Interest, 35, 22-54; Lipton, D., R. Martinson & J. Wilks (1975) The Effectiveness of Correctional Treatment: A Survey of Treatment Evaluation Studies.New York: Praeger.

[ii] Andrews, D.A. and James Bonta, 2010, The Psychology of Criminal Conduct 5th Ed., New Jersey, Matthew Bender, p. 351-356.

[iii] Ibid; MacKenzie, Doris Layton (2006), What works in Corrections-Reducing the Criminal Activities of Offenders and Delinquents, New York, Cambridge University Press; Drake, Elizabeth K. Steve Aos and Marna G. Miller, 2009, “Evidence-Based Public Policy Options to Reduce Crime and Criminal Justice Costs: Implications in Washington State” Victims and Offenders, 4:170-196; Aos, Steve, Marna Miller, and Elizabeth Drake, (2006). “Evidence-Based Adult Corrections Programs: What Works and What Does Not.” Olympia:Washington State Institute for Public Policy.

[iv] Ibid ii above, p. 111 and chapter 12.

[v] Ibid, p. 265.

The views expressed in this blog are solely the views of the author(s) and do not represent the views of any other public official or organization.

September 20, 2010

Prison and the Tragedy of the Commons

On Sunday, the New York Times reported that, in Missouri, judges are now given estimates of the cost of potential sentences before they make their sentencing decisions. Missouri Tells Judges Cost of Sentences . As someone who was indoctrinated with economics for six of my formative years, looking at the cost of a decision seems not merely reasonable, but the only responsible way to make a thoughtful decision.

As trial judges in Wisconsin have tremendous discretion in setting the length of a sentence, and that discretion has broadened over the last approximately 20 years, the consideration of the cost of a sentence is especially important. First, sentences are reviewed under the most deferential appellate standard—the erroneous exercise of discretion—resulting in little appellate oversight of sentences.

Secondly, over the last twenty years the legislature has increased the maximum penalties for most crimes giving judges a wide range of penalties for most crimes. For example, at the extreme, for a class B felony, a judge can sentence someone to a period of initial confinement, without the possibility of parole review, somewhere between zero days and forty years. Based on my experience, in the past, the legislative maximum had more often provided the cap on sentences. I can remember when prosecutors were always concerned about getting convictions for multiple crimes so that the maximum possible penalty could be increased with consecutive sentences.

Now, obtaining multiple convictions is rarely a concern as a conviction for one crime, with the higher penalties, is often sufficient to provide an appropriate sentence range for a defendant. Judges are rarely constrained by legislative maximums, as they are sufficiently high. The effect of legislatively increasing the maximum penalties has been shifting the responsibility of setting the effective (as opposed to the legal) maximum penalty from the legislature to the judiciary. On top of the longer sentence potential, we also have truth-in-sentencing, where, with some exceptions, judges decide when a defendant can be released from prison at the time of sentence. All of which makes the judge’s initial sentencing decision all that more important and which argues for the use of consideration of the cost of a sentence.

Incentive and cost mismatches occur with our sentencing. A prison sentence, for example, has the indirect cost to the defendant in loss of freedom, lost wages, and lost contribution to family and society, along with the direct cost of incarceration (the cost being considered by Missouri judges). Those costs are considered against the benefits of locking someone up. When a judge sentences someone to prison, he or she has decided to spend somewhere in the range of $25,000 per year of state taxpayer’s money for each year of incarceration. A problem arises from mismatched economic incentives.

In Wisconsin, judges are elected by citizens of each county; whereas the costs are born by citizens of the State of Wisconsin as a whole—most who will never be able to vote for or against the judge. Only a very small portion of the costs of incarceration will be paid by the taxpayers of the judge’s county whereas most of the benefits of incarceration, in terms of community safety and retribution, will accrue to the residents of the judge’s county as it is likely that the defendant and the victims of the crime came from the judge’s county. Judge Richard Posner has written that elected judges, because they will be subject to a performance review in the form of a re-election requirement, have to be somewhat sensitive to the electorate when making decisions.[i] (The elected judges may also more closely match the values of the electorate than appointed judges and make decisions that are more in line with this common value system.)

While judges are electorally held responsible for considering the benefits and the indirect costs to the defendant and his/her family, because most of the direct cost of incarceration is not born by citizens of the judge’s county, judges are not electorally held responsible for considering or not considering the direct cost of incarceration.

This mismatch of the signals of the benefits and cost of incarceration create a negative economic externality—a cost born by someone other than the decision maker. Prisons become the commons—the place in the village where everyone’s cows can graze without cost to the individual cow herder resulting in the pasture being overgrazed. Everyone’s cows starve. We have the tragedy of the commons.[ii] By not even considering costs of incarceration at sentencing, and not being held electorally accountable for these costs, the prison system becomes the commons. Are our overcrowded prisons, partially the result of the tragedy of the commons?

Like many people, I am concerned about our society being able to provide needed government services at a cost the citizenry are willing and able to pay. Would the system be better if criminal sentencing would be handled like juvenile delinquencies, or mental commitments, where the county receives a block of money to spend on corrections, and when people are sent to prison, that money comes from the county coffers? Would sentences be different if judges would have to justify spending county dollars for incarceration? I think an explicit consideration of the costs of a sentence is appropriate.

Although a fair and just sentence is much more than a straight economic decision, economics will always provide a tether as to what a society can or cannot do. That maxim has held for all societies in all ages. We in the criminal justice system ignore it at our peril.




[i] Posner, Richard A., (2008) “How Judges Think”, Harvard University Press, p. 135.

[ii] See Hardin, Garrett (1968) "The Tragedy of the Commons", Science, Vol. 162, No. 3859 pp. 1243-1248



The views expressed in this blog are solely the views of the author(s) and do not represent the views of any other public official or organization.

September 13, 2010

Hedonics and Punishment

Defendants are sentenced to prison for either incapacitation purposes or for punishment. Theoretically, punishment serves the joint goals of retribution (just deserts), general deterrence of other potential offenders, and specific deterrence of the sentenced offender. Longer prison sentences are assumed to equate with increased punishment of the offender. Psychological research suggests that this assumption may not be correct.

Robinson and Darley discuss research on different aspects of the “hedonic treadmill.”[i] The hedonic treadmill theory posits that while an individual’s level of happiness is initially affected by a positive or negative change in life circumstances, with time, the individual’s level of happiness returns to its original state. A defendant sentenced to prison will initially experience his change of circumstances as a negative event. However, as time passes, he adapts to his situation, prison becomes his life, and his level of happiness returns to the level it was before he went to prison. Because of the hedonic treadmill, the prison sentence has lost its bite as punishment and, hence, as a negative reinforcement. [ii]

Furthermore, humans also have the ability to become desensitized to changes within their environment. When conditions in prison change for the better or the worse from one day to the next, prisoners adapt to these changed conditions and experience them as minor changes in their levels of contentment. The prisoners become hardened to prison life. [iii]

Recent psychological research also undermines the assumption that longer sentences imply more punishment. Robinson and Darley discuss research that shows that duration of a punishment has little effect on the amount of remembered pain. Research showed that after individuals experienced a short period of intense pain and the same short period of the same intense pain followed by a longer period of less pain, they remember the short period of the intense pain followed by a longer period of less pain as less unpleasant than just the short period of intense pain. This researcher opined that individuals remember a negative experience as an average of the most extreme pain during the experience and the pain at the end of the experience.[iv]

Robinson and Darley point out that a shorter sentence has a greater likelihood of being felt as aversive at its end as a longer sentence. They conclude as follows: “The startling realization is that this short sentence will be experienced as more aversive than a much longer sentence that is equally aversive at the beginning but less so at the end! There are two reasons for this. The first is that, under the duration neglect account, the much longer duration of the long sentence contributes little or nothing to the reconstructed negativity of the remembered sentence. The second reason is that the ‘end-point intensity’ of the short sentence comes before it has had an opportunity to decay, while the end point intensity of the longer sentence is reduced at the end. The point here is that lengthening sentences may actually reduce their recalled negative character if the end experiences are relatively less aversive!”[v]

Robinson and Darley point out that society is getting a diminished “punishment” bang for their buck with longer sentences. (Of course, incapacitation is still effective.) The cost of incarceration remains constant as the value of the incarceration in the currency of punishment declines. [vi] The marginal punishment declines with sentence length.

This body of research implies that a shorter period of incarceration may be remembered as more unpleasant than a longer sentence. If a negative reinforcement for specific deterrence purposes is what was intended by the sentence of incarceration, then a shorter sentence may be more effective. Further research is required to outline the contours of the relationship between the lengths of incarceration and the levels of punishment. The study of hedonics has alerted us to other reasons why deterrence through longer sentences has not been shown to be a very robust factor in reducing recidivism.



[i] Robinson, Paul H., and John H. Darley, 2004, “Does Criminal Law Deter? A Behavioral Science Investigation”, Oxford Journal of Legal Studies, Vol. 24, No 2, p.187.

[ii] Ibid. p. 188.

[iii] Ibid p. 188

[iv] Ibid p. 190; D. Redelmeier and Daniel Kahneman, 1996, “Patients Memories of Painful Medical Treatments: Real Time and Retrospective Evaluations of Two Minimally Invasive Procedures”, 116 Pain 3.

[v] Robinson and Darley, Ibid p. 190.

[vi] Ibid, p. 189

The views expressed in this blog are solely the views of the author(s) and do not represent the views of any other public official or organization.

September 4, 2010

Why Deterrence Has Minimal Impact

In my last entry, I discussed the empirical research regarding the efficacy of deterring crime through increased sentences, as opposed to preventing crime through incapacitation from increased sentences. I want to discuss the psychological research as to why deterrence is not very effective.

Paul H. Robinson of the University of Pennsylvania Law School and John M Darley, a psychologist at Princeton, discuss what the research says about the ability to use increased punishment to deter crime.[i] First, to be clear, no one is arguing that eliminating punishment for crime would not result in all sorts of problems. These researchers, as well as myself, believe that a functioning criminal justice system that applies appropriate punishment for criminal acts is required for a functioning society. The only question that is being examined is the ability to deter crimes by increasing punishment for that crime, given the crime is already being punished.

Deterrence theory posits that individuals undertake a cost-benefit analysis before a crime is committed. If the perceived benefits exceed the costs, the crime is committed, if not, the crime is deterred and not committed.

Robinson and Darley discuss their research showing that while individuals have an intuitive feeling of what is wrong, they do not know the nuances of what is and is not legal within the individual’s jurisdictions.[ii] They also discuss another study of incarcerated felons. Twenty-two percent of these felons knew what the punishment would be for the crime, 18 percent either had no idea what the penalty was or thought they knew but were wrong, and 35 percent didn’t even think about the penalty.[iii]

Robinson and Darley also argue that the research shows that individuals that involve themselves with crime are less inclined to think about the consequences of their actions, be risk takers, and are more impulsive than the average person. Further, alcohol and drug use alter these individuals ability to make rational decisions. Superimposed on these qualities of an offender are often temporary states of mind such as the desire for revenge, rage, paranoia, and other mental illnesses or disorders that do not rise to the level of a legal defense. [iv]

The authors of this article also explain that crimes are often committed with other individuals creating an arousal effect, and that group decisions and actions increase the risk-taking of the individuals in the group over what they would do individually.[v]

Robinson and Darley further discuss the research on the perceived cost of crime by criminals. First, conditioning research would say that changing behavior through punishment is difficult if not impossible at low rates of punishment. For example, research shows that a ten percent punishment rate results in no reduction in the behavior. If crime is not being detected and punished at a high rate, the behavior will be difficult to deter.[vi] Further, research has shown that people regularly over-estimate their ability. Criminals will probably under-estimate the likelihood that they will be caught and punished.[vii]

These researchers also discuss that the amount of punishment is not likely to deter. They discuss a study in which graduated punishment results in the necessity to inflict large amounts of punishment to obtain deterrence. Robinson and Darley suggest that this research implies that by punishing first offenders lightly, we may be reducing our ability to ever deter these individuals, as they have learned to tolerate punishment. [viii]

Robinson and Darley also discuss research regarding the “hedonic-treadmill”. That discussion deserves a separate blog entry itself.



[i] Robinson, Paul H., and John H. Darley, 2004, “Does Criminal Law Deter? A Behavioral Science Investigation”, Oxford Journal of Legal Studies, Vol. 24, No 2, pp. 173-205.

[ii] Ibid., p 175-76

[iii] Anderson, David (2002) “The Deterrence Hypothesis and Picking Pockets at the Pickpocket’s Hanging”, 4 Amer. Law and Econ. Rev. 295

[iv] Ibid, p 179.

[v] Ibid. p. 180.

[vi] Ibid. pp 183-84.

[vii] Ibid p. 185.

[viii] Ibid. p. 186.

The views expressed in this blog are solely the views of the author(s) and do not represent the views of any other public official or organization.