After some years as a trial judge, I still haven't gotten over my impression that criminal justice blunders hopelessly at the disposition phase of criminal cases -- the sentencing hearing. Typically, prosecutors tell me what an ogre the defendant is and how cruel or thoughtless was his behavior; they ask me to "send a message," a plea which always brings to my mind an image of the giant array of telescopes used by NASA in its Search for Extra Terrestrial Intelligence (SETI). On their part, defense counsel tell me of the defendant's remorse, his childhood hardships, and his decision to turn over a new leaf. This debate has always seemed largely misdirected to me. Has anyone failed to notice that our attempts to document deterrence are about as successful as SETI? Or that damaging childhoods tend to impede reformation and predict danger?

Sentencing arguments rarely address the obvious questions - what are we trying to accomplish and how can we best achieve our goals? We continue in a morality play whose roots lie deep in antiquity. Our criminal justice system seems more a secular substitute for religious ceremony than a modern social mechanism for the rational pursuit of social goals. We claim to meet our responsibilities by decreeing a punishment as if it were ringing a temple bell - with no attempt whatever to assess any expected outcome other than the sound itself. Our major progress in the last 100 years or so was to update our vocabulary from "deterrence, retribution, and reformation" to "protection of society, personal responsibility, accountability for one's actions and reformation."⁽¹⁾ It is regrettable that it took a constitutional amendment to embrace public safety as a goal; it is a shame that we discarded the prohibition against retributive justice while we were at it. I expect that notions of just deserts and sentencing directions such as those in Ballot Measure 11 flourish precisely because we have paid so little heed to public safety in sentencing policy. I expect that more responsible consideration of public safety will alter the very culture of criminal justice and avoid many needless cruelties - both to the victims whose assaults and losses we should have prevented and to the offenders we should have diverted from the cycle of crime and punishment.

Thanks to the efforts of many, we will soon see some major changes in criminal sentencing. Essentially, those changes will encourage and enable us to have far better information concerning which sentencing choices are most likely to reduce the criminal behavior of which offenders. "What works?" will become a major issue in most sentencing hearings and plea negotiations, and practitioners will actually be expected to become expert at debating how best to perform their new responsibilities.

The evidence that change is on its way is substantial. In 1997, the Oregon Legislature passed HB 2229 (1997 Or Laws ch 433), which modifies many statutes to make reduced criminal behavior the primary measure of success in juvenile and adult corrections, and which requires that we collect data and exploit technology to facilitate that effort. The Oregon Judicial Conference in 1997 adopted a resolution which recognized the various articulated purposes of criminal justice, and directed 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." Our Governor appointed a working group to recommend how best to implement these principles, and its final report announces our intent to build the technology to enable us to use criminal justice data to make better decisions.⁽²⁾ Oregon's efforts have put us in line for some federal money to get the job done. And Multnomah County's Local Public Safety Coordinating Council has directed well over a million dollars to build a data warehouse and query tools which will allow users to see what works on which offenders to reduce criminal behavior. These tools will reach Multnomah County during 1999. State-wide, the construction of an integrated criminal justice information system, based on a "Public Safety Data Warehouse," will begin in 1999 and continue for years to come. I hope that the Oregon Judicial Department will be able to start delivering this tool to courtrooms throughout the state some time in 2000.⁽³⁾

These new tools will provide opportunities and challenges for defense counsel.

The changes already underway will exploit modern data management tools in ways which have become commonplace in industries in which failure is not rewarded and repeat business is a success. Over recent years, virtually all important governmental and private enterprises have produced enormous amounts of information which is routinely collected and stored in databases in diverse formats and disbursed locations. Our ability to search that data for useful information has been severely inhibited by the crudeness of our tools, the diversity of database formats, and the multiplicity of databases holding information necessary for addressing many important questions.

Modern technology has provided solutions to many of these hurdles. A data warehouse is simply a location into which data from various sources can be deposited to allow analysis which needs data from multiple sources. Extract tools "normalize" the data so that information is rendered comparable regardless of its original format. Modern analytical tools such as those performing "on-line analytical processing" (OLAP) allow users rapidly to perform queries to address typical questions, much as if they were doing traditional research. A major difference is that research that takes months or years without this technology can often be performed in seconds, minutes, or hours. More importantly for our purposes, when the results are challenged, the queries can be re-run in seconds to determine how changes or refinements in the questions we ask affect the answers we get. This is not at all the same as having researchers perform studies for us and send us a report some time in the future - it is looking fairly directly at information we can glean from a variety of sources at once.

As this approach is applied to sentencing, the changes will be quite dramatic. Whether I have a shoplifter facing a Theft II sentence, a domestic violence defendant being sentenced for Assault III or IV, a drug user on his fourth PCS, or an armed robber, under present circumstances I can look all I want at the defendant's criminal record, the circumstances of the crime, and, if I have it, the childhood, mental health, substance abuse, and other social background which accompanies this particular offender. What I don't have at present is any clue how similar offenders have fared after being sent to any of the places I might send this offender. I have no access whatever to any information about how similar offenders who have graduated from prison, jail, theft talk, alternative community service, forest project, batterers' intervention, inpatient treatment, or any other combination of responses have performed after they were sent there. We do know that with few exceptions, most crimes are committed by people who will return with new crimes after we are through with them; that our most common crime is recidivism, and that we do a terrible job of diverting offenders from criminal careers (first time DUII offenders and prostitution "johns" seem to be most likely to stay away once they've met the system, but I suspect that has little to do with sentencing). [click here  for one view of the extent of recidivism, return with the "back" function]

With the coming set of tools, all this will change. Because we will be able to gather information from multiple sources, I will be able to display graphically the performance of graduates of the various sanctions (in the community or in custody) which are available for sentencing this offender, while screening for similar offenders and selecting criteria of success from among various flavors of recidivism. I will expect prosecutors and defense attorneys to have access to the same tools before and during sentencing, and to be competent to argue productively about all of this.

Here's the way this will work: at sentencing, the offender and his crime(s) of conviction will be identified. The software, initially with and eventually without the user's help, will propose a "profile" of characteristics of the offender on the one hand, and a measure of success on the other. From the crime of conviction the software will gather possible sentencing ingredients for comparison. Offender characteristics can include criminal record, substance abuse history, risk level assessments (if the offender has been assessed at some time in his criminal career), and a variety of other possible "fields" such as age, ethnicity, sexual orientation, and mental health flags.⁽⁴⁾ Measures of success will all be related to reducing criminal history, but will allow us to look over varying periods of time for reduction in any convictions, in property crimes, in person crimes, in sex crimes, in drug crimes, in "negative police contacts" (available by reason of linking to state and local police databases), or even in 911 calls. We will be able to compare incarceration with probationary sentences, and even levels and styles of supervision. (The same technology will soon be available for pretrial release purposes). For a look at some sample screens, contact the web site mentioned in footnote 2.

The point here is that we will soon be able to have serious discussions about what sentence is most likely to serve public safety. Since we are now looking at potential criminal careers, it isn't enough to argue that jail or prison will protect people on the outside from people on the inside, even if we had unlimited jail space. For those offenders for whom incarceration is a legally and realistically available option, we can see if similar offenders did more harm over the course of five or ten years with or without spending some months or years in a secure facility. Judging by the number of recidivists I see, it isn't at all clear that prison is always the safest sentence from the community's point of view.

I'll save a lot of this discussion for part two, and hope that you will send any questions or concerns to me in time for me to respond in the next installment. Yes, I am aware that a defense attorney's job is to represent the client, and that a given defendant's interests may be best served by looking somewhere else for an argument to use to mitigate a sentence. I realize that there's quite a range of views concerning how much, if any, an attorney should counsel clients to worry about their future by addressing their problems. And although these new tools will be far more "user-friendly" than the data processing tools some of us have been exposed to, there will be some new skills to master. I will address these and other issues, and any you may bring to my attention, in a subsequent article.

For now, consider the hope that the coming technology offers. We will do a better job of linking offenders with those programs that can actually reduce their criminal behavior - so the offenders can enjoy the benefit of staying out of trouble while avoiding the burden of useless sentences, so citizens are not victimized by the crimes they would otherwise commit, and so the programs and sanctions which can offer some offenders a benefit aren't choked with those they can't help. Yes, we'll still have plenty of need for prisons, but the same approach should help us reduce the violence within prisons and increase the effectiveness of their programs. "What works" will at least begin to compete with "just deserts" in plea negotiations and in public discussions of crime, and the energy which has understandably been expended to "get tough on crime" can start to support policy choices which focus on reducing the crime instead of punishing the offender for the sake of punishment itself.

This series of changes has the potential to drag criminal justice into the Twentieth Century before it's over.

Michael Marcus
Department 35
Multnomah County Circuit Court
Multnomah County Courthouse
1021 SW Fourth Avenue
Portland OR 97204
SMMarcus@Compuserve.com

Return to the Smartsentencing home page

2. The text of all of these measures, some longer articles on the thinking behind all this, and some glimpses at the software we are developing can all be seen at my website, whatwrks.html return

3. Chief Justice Wallace Carson, Jr., and OJD ISD's Carl Ward are strong supporters of this effort. return

4. Yes, looking at ethnicity and sexual preference raises eyebrows and, potentially, concerns for misuse. Our purpose, however, is to be able to see if providers who specialize in gay youth or African American gang-involved are really more successful with their intended clients than other programs and, if they are, to ensure that we can link offenders appropriate for those programs to those programs. return