Re-Imagining Criminal Justice

I met recently with a nonprofit organization that seeks to reduce prison usage in the United States.  After our meeting, I tried to come up with a list of ideas and policies that I think should have wider circulation amongst the body politic that would help to promote a deep re-thinking of mass incarceration.  I thought I would share these ideas here, knowing that I’ve probably left a bunch of things out and knowing also that I probably need to make some other things clearer.  I welcome everyone’s feedback!

Click for the list….

1.  Fund all responses to crime, not particular responses to crime.

States currently subsidize some responses to crime and not others.  Prison is free to local officials, but drug treatment and other responses typically use money out of local budgets.  States should subsidize everything equally, or make every locality bear the cost.  The current system distorts decisionmaking in favor of the most expensive and most destructive response to crime: prison.  (Note: it’s not true that crime leads to prison use.  Different localities use prison at different rates, even controlling for crime.)  We should also consider not only efficacy (whether something works) but efficiency (whether it works using the least resources).  Prison is definitely inefficient, and possibly ineffective.

  1.  Collect data

Collect data on everything.  You measure what you manage and you manage what you measure.  Right now we have scattered databases that are antiquated.  We don’t get a sense of the system as a whole, just particular parts of it, but we need to understand how the different parts of the system interact: how policing affects courts, how prison affects crime, and how people perform longitudinally as they make their way through the system.  Too often we end up duplicating intake interviews with offenders, leading to needless waste and lag time in treating their underlying needs.  We also need to track police and prosecutorial behavior to see how that affects the populations of the justice-involved.  We then need to open that up (subject to privacy protection) so that people can analyze the information.  The bottom line: we need to know what we’re doing, especially given how much we’re doing.

  1.  Understand that various parts of the system impose costs on and provide benefits to other parts of the system

We can’t think about individual agencies and their policies without thinking about the “downstream” effects those agencies and policies have. Zero tolerance policing imposes costs on jails, courts, and prisons.  If prisons rehabilitate, that reduces costs to police and society.  Right now we don’t think about these interactions, nor do we reward agencies for their positive effects or punish them for their negative effects.  A prison should be incentivized to rehabilitate prisoners because it’s good for society.  Currently, prisons have no incentive to rehabilitate: it’s more work for them and the benefits don’t accrue to their bottom line.  Similarly, arresting a minor offender imposes some costs on police, but the total costs aren’t internalized.  We need to think of the system as a whole.

4,  Make sure that we’re funding things that work, including prison

We need to start requiring that our various responses to crime and criminal behavior work and do so using the least resources possible, and that we create financial incentives to ensure that this is the case.  Just as medicine has moved from paying for procedures to paying for outcomes, so, too, should the criminal justice system.  Don’t pay as much for ineffective surgery, and don’t pay as much for ineffective prison rehabilitation.  Don’t give someone every test under the sun for a hangnail; don’t pay for maximum criminal responses for minor behaviors.  We need to collect longitudinal data on outcomes and then start to channel money only towards those policies, agencies, and institutions that get both effective and efficient results.

  1.  Invest on the front end

We need to accept the relationship between social deficits and prison usage.  Prisons are full of people who can’t read, who were raised in violent neighborhoods, who have poor mental and physical health.  We can spend money on quality preschool and reduce crime in 20 years, or we can fail to invest in preschool and pay ten times the amount per year to imprison people for the offending behavior that results–a cost that does not include the accompanying crime and victimization.  Social welfare programs are public safety programs.

  1.  Understand that “recidivism” can mean different things

People returning from prison fail because they have immense social deficits–including the stigma of imprisonment reinforced by counter-productive policies–but not all failures are the same.  If someone is arrested for loitering and someone else is arrested for assault, those are different–and yet, by some measures, they would both count as recidivists.  We need to disaggregate the notion that recidivism means “any failure”.  Instead, we should start to measure the failures that we care about from the failures we don’t, and adjust our responses to them accordingly.  Success or failure is not binary; there will be failures.  People die of cancer at Sloan Kettering, but that doesn’t mean the hospital doesn’t do a great job with difficult cases.

  1.  Re-entry is too late; think diversion and decriminalization

Prison can be criminogenic; fixing people coming back from prison is, in some ways, too little too late.  We should think about conditions that shouldn’t be part of the of the criminal justice system at all.  Rather than treat people with mental illness in prison, let’s treat people with mental illness outside of prison and understand that dollars you spend inside a prison are dollars you’re not spending elsewhere.  Prison makes the mentally ill sicker, it’s expensive to treat them there, and they don’t do well in terms of reoffending.  We should seek to treat them outside of criminal justice as much as possible. The same is true for offenders whose criminal behavior is a result of drug usage.

  1.  Remind prosecutors that their duty is to promote public safety, not just to get convictions

Prosecutors should not only be concerned with how many sentences they get; they should be thinking about whether what they are doing is the minimum effective dose needed to protect public safety.  Data transparency and changes in budgetary incentives would encourage prosecutors to think about the big picture, not just about their individual caseload.  Sometimes a “win” for an individual prosecutor is a loss for society.

  1.  Think very carefully about whether we want judges and prosecutors to be political

Most criminal justice activity takes place at the local level.  Prosecutorial elections and judicial elections tend to promote policies that are seen to be tough on crime, but which are not necessarily smart on crime or resource-efficient.  We should explore alternatives to elections–there are ways of keeping these positions accountable to the public without encouraging candidates to take positions that are penny-wise and pound-foolish.

  1.  Deal with violent offenders

Too much of criminal justice reform centers around nonviolent offenders, but violent offenders are the ones who cause the crimes we care about the most.  We should spend our time and energy focusing on how to make them better.  Moreover, there are a lot of violent offenders: even if we let out all drug offenders in the United States we would still incarcerate more people than almost any other country on earth.

  1.  Understand that people age out of crime

Most crime is committed by men ages 18-35.  Older offenders are generally less likely to continue offending.  We should consider whether it makes financial sense to keep older offenders in prison, particularly since they are much more likely to need expensive medical treatment.

  1.  Understand that less is sometimes more

Throwing the book at people can sometimes make them more likely to commit crimes.  We are not “playing it safe” by treating everyone as though they were all high-risk individuals–for low-risk offenders, we are generating a self-fulfilling prophecy.  Arrested youth are at greater risk of future criminality; we would do better for low-level offenders not to label them as delinquents.  Using high-risk treatments on low-risk individuals makes them more likely to fail, in what is known as the risk-responsiveness principle.  Just as you can make someone sick by giving them medicine that isn’t necessary to treat them, so, too, can you “overprescribe” criminal treatments.

  1.  Be precise about “drug offenders” and “violent offenders” and “sex offenders”

These terms encompass a wide variety of behaviors; being precise can sometimes resolve differences.  Drug offenses, for example, encompass both those arrested for possessing a joint for personal use and someone who tries to sell methamphetamine to children.  When you use the phrase “drug offender” some people are going to think of the former and some people going to think of the latter.  We would treat those cases differently; we should make sure that when we set policies, we are going to the underlying behavior, not the general label.

  1.  Get the lead out

There is plenty of evidence that lead poisoning contributes to violent crime.  Lead poisoning is also bad for child development, so even if it later turns out that lead’s contribution to crime is overstated (or non-existent), we should still get rid of it.  It would cost us about $20 billion, once, to get rid of lead poisoning.  This is a fraction of the annual cost of prisons and jails.

  1.  Eliminate the commercial bail industry and encourage release

Money bail allows dangerous arrestees with financial resources to gain release and keeps poor but safe arrestees in jail.  If we are to sort out who is safe for release and who is too dangerous to be let out, most people would agree that poverty is not the ideal sorting mechanism.  We are the only nation besides the Philippines to have a commercial bail industry; the states that have done away with commercial bail have not seen crime rates or failure to appear rates increase.

  1.  Understand that crime cannot be eliminated

We cannot eliminate disease.  We cannot eliminate accidental deaths.  We cannot eliminate crime.  All of these are, of course, problems in society, but, at least when it comes to the first two, we realize that there are limits to the amount we can do as a society and limits about how much we should pay to address these problems.  There has been crime since Cain and Abel and there always will be.  The question is what we can do with the resources we have.  The existence of crime, in other words, is not a sign that society has failed.  We need to figure out what crimes we care most about and what crimes we can live with.  If we seek zero crime, we will always fail.

Author: W. David Ball

W. David Ball is an Associate Professor at Santa Clara School of Law. He writes and teaches primarily in the fields of criminal law and criminal procedure, with a special focus on sentencing and corrections. He also serves as the Co-Chair of the Corrections Committee of the American Bar Association.

13 thoughts on “Re-Imagining Criminal Justice”

  1. As someone who, over 30 years ago, wrote a book on recidivism (entitled, strangely enough, "Recidivism"), I feel that we should be concerned about the language we use to portray this type of failure. The word “recidivism” implies full agency on the part of the individual. That is, saying that a person recidivates puts the entire burden on the ex-offender, ignoring the part that the external environment plays in his/her failure. Instead, we should look at the other side of the coin and talk about the ex-offenders probability of survival in the face of the difficulties s/he faces. That is, ethical issues arise even in the terms used.

    I'll look at your other recommendations as time permits.

    1. Thanks for the feedback, Mike! I agree. That was part of my recommendation to move away from the binary measure of recidivism to something more scalar. In fact, in a recent article that I've just posted on ssrn ( I talk about survival times being something we should start to measure. I'll look forward to checking out your book!

    2. Behavior Analysts view all behavior as a product of environmental contingencies in reaction with genes (ontogenic + phylogenic). We take a deterministic view that seeks to analyze the functional relations in a person's life. This is a scientific philosophy that is sadly at odds with much of the criminal justice system, from laws to sentences to prison environments.

      The use of Applied Behavior Analysis (ABA) would be a game-changer in the prison environment, considering the level of control we have over the population's environment. By emphasizing differential reinforcement of socially appropriate and meaningful behaviors in prison, I have little doubt we could effect enormous positive changes in the lives of the incarcerated.

      In larger society, it's more difficult, because of the diffuse contingencies involved, and many inappropriate behaviors are reinforced by the basic structure of our society. Behavior change is easy, with the right design. Unfortunately, I think there's a lot of distrust of behaviorism on both the right and left, for different reasons (romanticism vs. authoritarianism).

    1. Agreed! I really think that drug abuse, for example, is much better dealt with via the public health approach than via the criminal approach. The same is true for property crimes (e.g. trespassing) that essentially criminalize homelessness.

      1. "The same is true for property crimes (e.g. trespassing) that essentially criminalize homelessness."

        I think you have to be very careful, though. Law enforcement isn't just about enforcing the state's idea of what people should be allowed to do. It's also about displacing private defense and vengeance.

        It's one thing to repeal laws which just criminalize consensual behavior that directly involved parties want. (The classic victimless crimes.) It's quite another to repeal laws that criminalize behavior that directly effects unconsenting parties. Trespass can certainly be in this latter category, and a cession of enforcement of trespass laws could lead to private enforcement, which the former criminals might find even more unpleasant than the legal system.

  2. 8 is probably the most important and the least doable. If we could change the incentives for prosecutors so they stopped seeing their job as a game where putting people behind bars is a "win" that would do more for the cause of criminal justice than anything else I can think of.

    1. Thanks for the feedback–you might be interested in some of the budgeting/forced choice/ordinal approaches that I talk about in some of my writing, as well as the many other approaches that I cite in it (see, e.g., this article… as well as the work that it cites). In addition to changing prosecutorial incentives–and performance measurements for the job–I also think that it is important to change the conversation about what prosecutors do. So reframing the conversation might start with using the term DA more often. "Prosecutors" implies that their job is to prosecute, when their job is actually defined more broadly than that. We need to talk about public safety in a way where it isn't just a byword for punishment. DA's should be praised, promoted, etc., for important cases they didn't bring. I agree that it will be a tough slog, but we can start the way some police departments have started–by measuring and rewarding alternative dispositions, referrals to social services, etc., and not just plea bargains obtained.

  3. A lot of these suggestions interact with each other — for example, if you're thinking about releasing older prisoners because they're less of a risk for violent crimes, then you have to beef up all the training and re-entry services, because otherwise life on the outside will be nasty, brutish and short.

    Also, how does the child-welfare system play into this? Are convicted criminals under sentence automatically considered unfit parents, even if they're not in a prison?

    1. Hi Paul–

      Yes, I agree that these overlap with one another, and that we would certainly have to start beefing up programming and re-entry before release. This is part of the problem with moving away from prison as the default–in the short term, we're likely to have to spend more money as we shift from pure incapacitation to another model. In a future article, I'm planning to talk about how we might use social impact bonds to bridge the gap.

      I also think we might find that if we start to collect data, we find out that things are much worse than we think they are. I wrote about this here ( but there have been jurisdictions that have discovered that some of their recidivist "wins" are due to people dying or moving.

      As for the unfit parents question, that's a matter of state law, but one that points to another really important issue: reforming the foster care system. Here's what I do know about the parental situation. Most prisoners have minor children, most women in prison were the primary caregivers for their minor children before they were arrested, some police departments have no plans for what to do with minor children who are present when the parent is arrested (!), and prisons are hard to get to for in-person visitation. There have been some important steps in the right direction in terms of capping phone call rates (both inter- and intra-state), but in-person visitation is being phased out in some jurisdictions and replaced with video visitation, which is not subject to rate caps. We really need to see prisoners not as a captive audience for wealth extraction and to embrace the social science that suggests (quite convincingly) that, except in cases of parental abuse, keeping families in contact helps prevent recidivism on the part of the parent and makes justice involvement less likely for the children.

  4. A more general thought: Saying that prison usage is due to prosecutorial zeal puts the starting point at trial. Saying that it is due to too much policing puts it on the street, and perhaps to peer pressure [see the articles on co-offending]. Saying that it is due to inadequate schools puts it earlier in the offender’s life course. Or due to family interactions (or lack thereof) puts it yet earlier. Or due to housing patterns and redlining puts it before the offender was born.

    In other words, all of these are true to some extent, but a strategy aimed at each of these has a different “injection point” in a (potential) offender’s life course and involves a different organization, be it a municipal, county, state, or federal agency or other organization. So it’s obvious that a mixed strategy is needed, one that focuses on current prisoners or offenders, on those at risk of offending in the future, and on circumstances that put individuals at possible risk in the future.

    1. Again, completely agree. This is part of what I was trying to get at with 7 (and with the other points working in concert). There are numerous injection points (policy points?) at which new thought should be aimed. It's clear that some policies will have to target the behavior of offenders; it's equally clear that some policies will have to target the behavior of criminal justice actors (be they prosecutors, police, corrections, etc.). In my own view we need something like unified corrections–expanded to include police and DA's–at the local level, as I've written about here ( Thanks again for the thought–and for the link (above) to the book!

Comments are closed.