We Can Do Better at Fighting Crime
©Michael Marcus
September 27, 1997
An edited version of this article appeared in the Fall, 1998, issue of Oregon's Future under the title
"Improving our Judgment: Knowledge-based Sentencing";
another appeared in the July-September 1999 issue of Law Enforcement Review,
a quarterly publication of the Centre for Law Enforcement Education (CLEEN) in Nigeria

     Debates about crime and punishment often capture public attention. Although the criminal justice system consumes enormous resources, most of us agree that its performance is far from satisfactory. Legislatures and electorates are considering and passing "tough on crime" measures like the "three strikes" laws of several states. A recent ballot measure in Oregon imposes minimum sentences of 70 months or more on even first-time juvenile offenders who commit certain offenses. Opinions on "what's wrong" with criminal justice abound. Some suggest that we are still too "soft on crime," while others argue that there's something wrong when we imprison a higher proportion of our population than almost any other nation in the world. Commentators attribute our crime problem to misled juries, bad parents, deteriorating public schools, liberal judges, plea bargaining, the loss of family values, or the Miranda decision and other exclusionary rules.

    At a recent presentation to Oregon lawmakers, the National Institute of Justice suggested that the biggest return from anti-crime dollars would come from investment not in law enforcement but in high school completion and parenting classes. Many people argue that other human service expenditures also reduce crime. I don't doubt that there are many social programs whose success or failure substantially affects crime rates, but current public sentiment and fiscal frugality pose real barriers to any expansion of social services. And looking for the roots of crime will not answer the challenge of offenders who continue their criminal behavior after repeatedly being convicted and completing their sentences.

    After some years of experience as a trial judge, I have become convinced that there is a relatively simple and inexpensive step we can take which would greatly increase our effectiveness in reducing crime. Essentially, we must agree that the criminal justice system has the responsibility to do its best to divert criminals from criminal behavior. To do this, we must keep track of which sanctions seem to work best on which offenders, and rigorously apply what we learn through the sentences we impose. If this sounds obvious, it should be. But in practice we almost never take this approach to sentencing decisions.

    Part of our problem is that we don't seem to agree which of these functions criminal justice can or should actually perform. Prosecutors are fond of demanding that we "send a message" with our sentences. Rhetoric aside, this demand is usually a reference to one or more of the various theoretical functions of criminal punishment. These are best considered individually.

Deterrence assumes that potential offenders are convinced by punishments meted out to others not to violate the law themselves. Deterrence probably works at the level of parking fines and with a few more serious crimes. The likelihood and consequences of getting caught should figure into a business decision whether to dump toxic waste illegally, for example. But the vast majority of crimes that frighten and anger us most are not committed by people who plan ahead, weigh their options and make good choices. When Jeremy Bentham first proposed the creation of penitentiaries near the end of the Eighteenth Century, everyone knew that the best location to watch pickpockets at work was at the gallows -- while the crowd was distracted by the hanging of a pickpocket. Today, we still have plenty to do with the enormous number of offenders who are not deterred by the prospect of being caught and punished for their crimes.

Incapacitation is a reliable but limited function of those punishments which actually put an offender behind bars. With very rare exceptions ("organized crime" and Charlie Manson come to mind), prisoners can't commit crimes on the outside while they're on the inside. At the very least, we do protect the public from further crimes by an offender for as long as that offender is locked up. What we haven't sufficiently explored is whether and to what extent incarceration has a "criminogenic" effect on offenders who serve substantial periods of time in custody so that they end up committing more and worse crimes during their criminal careers than they would have had we done something different to them along the way. My guess is that there are some categories of prisoners for whom this is often so, and that recent mandatory imprisonment laws are increasing their number. I am certain that we can learn a lot more than we now know about this as soon as we separate the public safety component of custody from the retributive. Meanwhile, most of us agree that prisons put people together who are likely to share criminal values and techniques. We also know that crimes are commonly committed by offenders who have previously completed prison sentences.

Retribution is the oldest function of punishment. More recently clothed within such concepts as "accountability" or "consequences," retribution is punishment as a matter of moral just deserts. It is the infliction of a sanction to restore moral equivalency or to assert public values. This is what people usually have in mind when they discuss whether a sentence is "just" or "fits the crime." Retribution is the closest to a religious objective in our criminal justice system, but it has at least a theoretically secular function. For purposes of civil order, we take the right of retribution from victims and their families, and -- in theory -- exact it upon offenders in their stead. In theory, this prevents victims from taking the law into their own hands.

Restitution is the relatively new function of attempting to compel convicted offenders to repay their victims at least for financial losses caused by their crimes. Before victims' advocates convinced legislatures to authorize restitution orders as part of criminal sentences, victims of crime were relegated to the civil justice system for any hope -- often slight -- of recovering their losses. If the offender is not producing a steady income (or is behind in child support), enforcing a civil judgment is often impractical. "Exemption" laws in many states spare debtors and taxpayers the consequences of adding to the ranks of the destitute. Tools of trade, furnishings, certain amounts and sources of income, and, within limits, a car and a residence are typically beyond the reach of civil creditors. In practice, restitution as part of a criminal sentence improves things only a little. Although there are no rights of exemption, a court must find an offender in wilful violation of a restitution order before imposing a sanction for failing to pay restitution. Most people who commit crimes -- at least most of the ones who get caught and sentenced -- tend not to be successful people. And without a steady income, few can be compelled to pay restitution. In any event, courts are often far too busy trying to process the offenders who continue to cycle through the criminal justice system to give enforcement of restitution the attention crime victims deserve.

Reformation is the process of reducing the likelihood of future criminal behavior by an offender. This is the most glaringly unsuccessful mission of criminal justice. All of us in the criminal justice system know that apart from drunk drivers, men caught trying to hire prostitutes, and a very few other criminals, the first offender is rare in our system. When I sit in high volume arraignment or pretrial courts, I occasionally ask the defendants in the courtroom how many have been through the system before. Eighty percent typically raise their hands. We can do what we want to alter statistics by fiddling with the definition of "recidivism" or by modifying guidelines for parole and probation revocations, but we cannot escape this simple reality: whatever the source of criminal behavior, the vast majority of crimes are committed by people who have already been convicted of a crime and subjected to some form of sentence. More often than not, we find that a serious offender's career started with a variety of petty offenses and escalated into more serious crime in spite of whatever sanctions the criminal justice system imposed along the way.

Restorative Justice is an even newer purpose of criminal sentences, and the only one almost never intended by the plea to "send a message." Restorative justice seeks to repair the rift in a community which a crime typically creates, often by requiring the offender to respond to the victim in a way which makes amends to the victim or to the community for the offense. "Victim-Offender Mediation" is one device employed by restorative justice. It involves a structured meeting between victim and offender which is hoped to convey to the offender that there is a human consequence to the crime, while restoring a sense of control and resolution to the victim.    That we know the various theoretical functions of punishment does not mean that we have any consensus as to which we should pursue. On the eve of the McVeigh death penalty verdict, I heard a national radio discussion of the functions of punishment. A trial judge, a theologian, and a former prison warden gave their views. I agreed with the former warden that society seems uncertain about what it wants from prisons, and with the theologian that criminal justice in part performs some of the functions of religion -- reinforcing values and providing a form of morality play. On the other hand, the trial judge articulated a view which I think is quietly held by many, but which is antithetical to any progress in our fight against crime.

    The judge argued that retribution is the only proper function of criminal justice. He suggested that reformation might be inconsistent with the individualism which underlies a democratic society, and instead characteristic of a totalitarian state. I think that is absurd; surely a democratic society has the right to make and enforce laws and to attempt to change behaviors which violate those laws. Combating thievery and assault is hardly anti-democratic. It is not the enforcement of law but its scope and content which distinguishes a totalitarian state from a democratic one. But I had heard such arguments before. The day I was appointed to the bench, I was on a panel which included a law school dean who insisted that any notion of protecting the public from offenders in the future, or even of meeting the needs of crime victims, was the fault of politicians whose meddling had distorted the true role of criminal justice: to determine guilt or innocence, and then simply to impose whatever sanction the legislature prescribed for the crime in question.

    I acknowledge that there is plenty of room for reasonable debate about crime and punishment. I also agree that politicians willing to exploit fear and anger have done much to damage the effectiveness of criminal justice and to reduce the rationality of public discussion. I submit, however, that those of us who hold positions of public trust in the criminal justice system have an obligation to do what we can to reduce crime. Whether or not deterrence is a viable objective or retribution a proper one, surely reformation should be a consideration in most criminal sentencing decisions -- all but those involving offenders too dangerous and crimes too heinous to contemplate anything short of life imprisonment without parole. Surely reformation is one of our goals. Statutes in most states say so, and the public continues to support (or at least to pay for) a criminal justice system which includes in its arsenal of tools alcohol and drug treatment programs, work release centers, domestic violence and other anger counseling programs, and even sex offender treatment programs. Whether or not they do any good, these programs demonstrate by their existence that we are supposed to try to reduce criminal behavior by offenders once they return to their communities.

    So why would a judge or a law school dean argue otherwise? I think this is at least in part a form of "mission creep" -- an erosion of performance standards so that objectives become attainable; we assure ourselves of success by redefining success. We in the criminal justice system know that our attempts to reform offenders have been alarmingly unsuccessful. It is much more of a challenge to reverse that failure than to conclude that it isn't our responsibility to reform offenders in the first place. Ironically, this form of coping with failure by adopting more easily achievable goals is extremely common in prisons. Prisoners want first to survive, and then to reduce their discomfort on a day to day basis, so they adapt to the inmates' culture instead of meditating on their poor choices or values and finding a path to better behavior as Jeremy Bentham had hoped. So judges and corrections personnel sometimes adopt as their goals processing cases quickly and avoiding chaos to escape the more difficult challenge of trying to improve offenders' behavior when they return to their communities. Treatment programs in the criminal justice system often do the same thing. Rather than being "held responsible" for their charges' future behavior -- arguing that they are not to blame for the sources of that behavior -- these programs measure success by such attainments as keeping "clients" in the program until they complete it, or achieving 30 days of sobriety, or at best staying out of trouble during the short period of the program itself. Almost all programs connected with the criminal justice system measure success by anything but reducing offenders' future criminal conduct. This misses the point. As one victims' advocate suggested, in some cases we may be treating the addiction of some thieves only to end up with more competent thieves!

    But reformation is not an unachievable goal, at least for many in the offender population. Academia has generated thousands of pages from hundreds of studies about what does and does not work to reduce crime. Academia's mission is apparently to generate studies and academic debate about those studies, as academia has done virtually nothing to implement any of its findings. And courts have done virtually nothing to learn from or implement what academia has discovered. After over seven years on the bench, I continue to be astounded that almost never does an advocate for either side at a sentencing hearing volunteer any argument about which of the available choices is most likely to reduce an offender's future criminal behavior. We make sentencing choices based on the public relations skills of the available programs, with no more precision than rumor and folklore. Incredibly, though we now have computerized data which could enable us to see how a given sanction's graduates perform after graduation, and to narrow the focus to graduates who are significantly like the offender before the court for sentencing, we have not provided this tool to lawyers or judges or to correctional professionals to help them make more intelligent decisions about their charges.

    Any business in a competitive industry which spent so little effort to monitor its performance would soon parish. Had our military been so inattentive to its ability to achieve its goals, we would have lost the war in Grenada. The cynical could argue that the criminal justice system has tolerated our dismal performance in reducing recidivism because the status quo creates demand for more police, judges, prosecutors, and defense attorneys. Unlike most enterprises, our failures increase the demand for our services. But we can and should hold ourselves to a higher standard of performance.

    I cannot believe that we achieve by accident as much as we could with some careful attention to what works on which kinds of offenders. I am convinced that there are sanctions which do no good at all, and that there are many which could do a lot more good if we stopped exhausting their resources on the wrong offenders.

    I appreciate that there are many in the corrections population who are not susceptible to improvement. I appreciate that most offenders have been shaped by social factors which we cannot hope to correct within the criminal justice system, and that wisely spent crime prevention dollars buy may buy far more public safety than bricks and mortar. But we continue to see the offenders who are committing the bulk of the crimes; few are anything but prolific in their criminal behavior. I am certain we can do a better job of diverting many from criminal careers. And because most offenders commit many more than one crime, diverting even a small proportion would save the public a great many crimes. At the very least, it is irresponsible in the extreme not to try.

    Here is what I propose. First, we should make it known to our policy makers and to our judges that whatever our disagreements about the other functions of criminal justice, we all agree that we must reasonably attempt to reduce the criminal behavior of offenders. We do that with some by imposing life imprisonment without parole (or the death penalty); as to the rest, it is our responsibility to do the best we can to reduce their criminal behavior after they complete their sentences.

    Second, we must use our technology to keep careful track of which sanctions and programs, in and out of prison, seem to work best in reducing criminal behavior, and on which offenders. Yes, some kinds of offenders are harder to improve than others, but keeping careful track means being able to see how certain categories or profiles of offender compare in their ability to avoid criminal behavior after completing each of the various kinds of sanctions or programs we impose on them.

    Third, we must insist that sentencing decisions be made with due consideration of this information. In some cases other concerns may override the objective of avoiding future criminal behavior, and the magnitude of the offense may properly limit what we are prepared to do to a particular offender. But in the great bulk of criminal cases, judges should select from among the legally and practically available choices based on which seems to work best on offenders like the one before the court in reducing criminal behavior.

    Finally, we should learn from the process so we can improve our ability to profile offenders in terms of their susceptibility or not to various forms of sanctions. With this information, legislatures can adjust what range of sanctions they prescribe for various offenses and offenders, and policy makers can make more intelligent choices about which programs or sanctions to support with public safety dollars. Ultimately, we would divert many more offenders from crime before we had to put them in prison to protect ourselves from them, and we would focus prison resources more accurately on those from whom we can only be safe when they are in custody. And within prisons, programs would focus on those inmates whose post-prison behavior the programs would likely improve.

    These goals are attainable, but they won't come without strong leadership and persistence. When the issue is raised, these goals quickly draw broad support. Criminal defense attorneys, judges, prosecutors, corrections officials, and victims' advocates supported changes the Oregon Legislature has just adopted to facilitate this approach. For example, the Oregon Department of Corrections is now directed to "Provide central information and data services sufficient to . . . permit analysis of correlations between sanctions, supervision, services and programs, and future criminal conduct." The Oregon Judicial Department has recently resolved "that in the course of considering the public safety component of criminal sentencing, juvenile delinquency dispositions, and adult and juvenile probation decisions, judges should consider and invite advocates to address the likely impact of the choices available to the judge in reducing future criminal conduct." Further, "judges are encouraged to seek and obtain training, education and information to assist them in evaluating the effectiveness of available sanctions, programs, and sentencing options in reducing future criminal conduct."

    State and local committees are at work and have much to do. Information systems specialists are crucial, because numerous databases must be accessed to extract the information necessary to give us the information we need, and it must be made accessible to all of us involved in sentencing decisions in a way that results in it actually being used. This change will not be a success until and unless advocates and judges routinely use accurate information about what works to help make sentencing decisions. Once the information is available and actually used in courtrooms, lawyers will demand access to it just as they do to a new case or statute which they need to know about to make (or meet) arguments in court.

    At this stage, I see two very real hurdles that we must overcome to achieve real improvement in our performance in reducing criminal behavior. First, "mission creep" is always a risk. All of the systems which make up the criminal justice process have organizational needs in their own right, and none are used to making reduced criminal behavior their dominant performance standard. Not all should. For example, before guilt is determined, the goals of  fairness and efficiency should not be sacrificed to this effort. We should not make it easier to convict people just because we'd like to be able to use corrections resources to improve their behavior. But once there is a criminal conviction, we should not allow the systems' own needs to divert us from our task. Treatment programs will generally be wary of any performance standard that insists that we look at post-treatment criminal behavior, in part because programs fear losing funding, and in part because many generally hope to serve their populations' real needs. But we can fairly insist that corrections dollars be expended for corrections purposes, and that they serve social service functions only to the extent that those functions reduce crime. We also must take care that information systems are not totally dominated by operational needs other than effectiveness measurement. Tracking offenders and cases is important, of course, but information systems must be held responsible for providing convenient access to data about our effectiveness in reducing crime. Otherwise, those information systems will remain useless for evaluation even though they contain the very information we need to monitor what works and what does not.

    The second hurdle is the risk of being overrun by traditional notions of research. Professors, students, and social scientists have been warmly supportive of the notion that we can and should do a better job of monitoring what works in sentencing. They agree in principle that the information must be accessible to be used by those who need it. But they are very wary of "untrained" evaluators. They fear that lawyers and judges will be unable to handle "raw data" to improve their decisions, and we would be best advised simply to help social scientists process the data, draw the appropriate conclusions, and deliver it to advocates and judges for their use.

    This is a very real dilemma. On the one hand, we have many decades of academic research about what works which has had virtually no impact on sentencing decisions and even less on legislative and public debate about crime and punishment. And judges who want to serve public safety have virtually no information to go on. For example, I regularly have many drug and alcohol treatment programs to which I can send an offender as a condition of probation, and at least two "theft talk" programs. I have absolutely no information on whose graduates stay crime free longest, let alone how offenders like the one before me for sentencing have fared in those programs. This is simply unacceptable.

    On the other hand, the experts do have a point. To make the best decision, I need to know more than which programs' graduates have stayed crime free longest. I need to know how their populations differ in criminal history, drug and alcohol problems, and perhaps even psychological makeup. One program may have a more challenging clientele than the other, yet still be more likely than the other to prevent my particular offender from stealing again -- even though the other program, with a less challenging population, has fewer active thieves among its graduates. And the greatest potential of this quest is to learn why some programs work on some offenders, and how more accurately to profile offenders in terms of which of them are susceptible to improvement and how. Ideally, experts can help us refine sanctions and programs, and design new ones, with what we learn from carefully monitoring what works on which offenders. In short, we could all use the help of competent social science in refining our understanding of what works and why.

    I believe the answer to this is that we need both raw data in courtrooms and social science. We must urge lawyers and judges to look at the actual track records of programs with various kinds of offenders. Until and unless they do so routinely, there is little hope that social science will have any more relevance to what happens in courts and legislatures than it now has. Only when what works is routinely debated in courtrooms will there be day-to-day demand for academia's input. And even without that input, we can't possibly be better off with no information than we would be with a good flow of "raw data" about which programs' graduates are doing what. The adversary process, after all, is very good at finding the flaws in arguments. In the civil context, we often have debates between experts through cross-examination. Judges and juries are routinely called upon to decide which opinion or analysis is most convincing. There is virtually no science which cannot occasionally be subjected to scrutiny in court, because courts are called upon to resolve controversies involving virtually every field of human endeavor. Courts deal with the magnificent intricacies of DNA evidence, medical malpractice, product liability, and human relationships in families and businesses. If a court proceeding is capable of resolving disputes about who was responsible for a flaw in a critical part of a nuclear reactor, it should be able to handle debates about which theft programs' graduates steal less and why. It has been my experience that the adversary process is quite capable of testing the validity of academic experimentation and reasoning, and that without something approaching this real world contact, experts are occasionally apt to go quite astray.

    I hope that the experts will assist in designing and building the information delivery systems which we need to make useful data available for better choices by courts, probation officers, and corrections workers. Once the process is in regular use, it will generate massive amounts of useful data for further research, and dramatically increase the demand for the assistance of social scientists in learning even more about what works and why.

    If all of this comes to pass, I predict two major beneficial consequences. First, we will actually do a far better job of being "tough on crime" by reducing criminal behavior and diverting offenders from criminal careers. We will increase public safety, reduce the frequency of victimization, and spend our crime and corrections dollars far more efficiently than we do now. Second, when the media routinely encounters and reports on debates about what works in high profile sentencing hearings, public debate will benefit from a decrease in fear and anger, and an increase in reasoned consideration of how best to respond to crime. Policy makers will then be permitted and encouraged to do the same. We will all benefit from a much wiser approach to the design and execution of criminal justice strategies.

    We have an enormous opportunity to bring our criminal justice system into line with modern notions of performance and public responsibility. There's no reason to wait until the Twentieth Century is over.
Return to Smartsentencing homepage