« More on Brady violations | Main | Religion and the Politics of Eminent Domain »

Thursday, July 09, 2009

Reconsidering the Punishment of Violent Crime

For those of us who believe that America imprisons too high a portion of its population, recent years have seen mixed progress.  Many states have adopted measures that aim to move drug-addicted offenders involved in drug and property crime from jail or prison toward treatment under the threat of jail.  Other categories of prisoners who seem to be enjoying some increase in sympathy from the public and politicians include women  (whose crimes are often tied up with the criminality of abusive partners, and whose loss from the community often leaves a bigger hole than with men), juveniles facing long prison terms, and parolees returned to prison on technical violations (this may be a unique California problem but it includes tens of thousands of prisoners).  This approach has lots of political appeal right now, but it will not work to achieve a substantial and enduring reduction in the US rate of incarceration.  Instead, against all political common sense, I believe we need to directly address the extraordinarily harsh prison sentences we now hand out for violent crime.

The current approach takes aim at what might be called the “soft underbelly” of mass imprisonment; those categories of prisoners who manifestly do not belong in prison (at least for long terms) and could be adequately punished and managed through better use of jail and probation (as well as drug treatment as diversion).  In contrast, violent crime constitutes what might called (at the risk of stretching a metaphor a bit too far) the “hard-back” of mass incarceration (think the Terminator after all his cyborg flesh was burned off in the movie franchise of that name).  These are prisoners who for the most part are good candidates for imprisonment, but for how long? 

Not withstanding progress in attacking the soft underbelly, the overall rate of imprisonment in the US has continued to increase modestly despite crime rates that have remained relatively low over the past decade and a half (Bureau of Justice Statistics summary page on imprisonment).  The reality is that close to half of the national growth in imprisonment since 1980 consists of increased punishment for “violent” crime.  If we are to cut into that growth, and just as importantly, permanently reduce the public appetite to punish drug users and other non-violent prisoners, we need to revisit the policies that send so many to prison for so long.

Punishment for violent crime is now extraordinarily harsh.  Just consider murder.  The average time served back in the ‘70s in California prisons for 1st degree non-capital murder (under Ronald Reagan) was about 10 years (5 for 2nd degree).  Today 1st degree murderers do not get a parole hearing for a minimum of 25 years (less a possible third for good time) and 15 for 2nd degree, but given current parole practices many of these prisoners will never be released, or at best after 30 or 40 years of maintaining an exemplary record in prison. [read my recent Roger Hood lecture at Oxford University on California’s treatment of murder].  In contrast, most of the rest of the world releases murderers after 10 to 20 years.  [For a non homicide example read my blog post on an Oakland man sentenced to 70 years to life for a non-homicide crime spree].

We can debate just how long one should serve for murder of various kinds of seriousness (a debate I believe we need to have), but my main concern is the influence these life trashing sentences have on the who scale of penalty in the US.  Unreasonably long sentences for violent crime help sustain an unreasonably high appetite for imprisonment in the US in three ways.

  • Price Signal

Natural life for non-capital murder sets a reference point that makes it seem reasonable to lock a burglar up for 25 years (under 3-Strikes) or a drug dealer for 50 years.  In contrast, if the sentence for murder were 25 years or less, it would be very hard to justify these sentences for non-homicide crimes.

  • Net-widening Along the Criminological Continuum

American prosecutors, and to a large extent, the public, have long been influenced by criminological thought to believe that minor crime exists on a continuum with the most serious crimes through the causal role of “criminality."  The popular “broken windows” theory is just a rehashing of a criminological common sense that goes back to Lombroso, if not earlier.  The minor offender may well be on their way up the criminality ladder toward violence.  Likewise the prisoner who has already committed an act of violence is deemed capable of committing any crime.  It is this connection between minor criminality and violent crime that has generally led Americans to support the imprisonment of drug users and property criminals (the soft underbelly) when fiscal conditions are good and when crime fear is high.

  • Hardening Prison

Just the shear number of life sentences is beginning to constitute a growing block of indigestible population within our prisons.  In California there are more than 30,000 prisoners facing the prospect of life in prison, with about at 1,000 new cases a year and around 5 to 10 releases.  The existence of a growing core of prisoners with no realistic hope of release will produce a prison culture that is even more anomic and criminogenic than the already exists in states like California.  Not only will such prisons be more dangerous places to work or serve time, they are likely to reinforce the public’s fear of crime.

My conclusion is that we need to undertake the hard work of rethink our legal response to violent crime. Just as a person who wants to lose weight sustainably has to check their appetite for pizza and chocolate (or Vodka in my case) not celery and carrots, a sustainable reduction in imprisonment will require us to check our appetite for punishing the manifestly violent.

Posted by Jonathan Simon on July 9, 2009 at 10:18 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Reconsidering the Punishment of Violent Crime:


One of the important problems with most first degree murder statutes is the felony-murder rule. The notion of distinguishing first degree murder and second degree murder is to distinguish people who are basically irredeemable from those who made a bad choice in a bad situation but who are not fundamentally bad people.

Yet, among groups who we expect to be less prone to irredeemable violent tendencies, like juveniles and women, a large share of first degree murder convictions are for "felony-murder" which requires less of a showing of bad mens rea than conspiracy to commit murder. Simply being part of a group of people participating in a felony (sometimes a mere "technical felony" not within the heartland common law definitions of the felony crimes that trigger felony murder statutes) can create first degree murder culpability if someone else is involved.

What would be a several year sentence for a property crime turns into a life without parole sentence, due to what, from the convict's perspective looks like random chance. One doesn't even have to know that there was a murder or attempted murder until after the fact to have felony murder culpability. While felony-murder grading makes sense for trigger men and culpable participants in a murder (e.g. someone who orders the killing) it does not for a mere gang member.

The numerical impact of this seemingly modest statutory distinction is very great. A decision to charge felony murder rather than the underlying felony can be the equivalent of putting twenty or thirty more people in prison.

Also particularly troubling are cases where consecutive rather than concurrent sentencing for multiple offenses that are part of the same general course of criminal conduct, but have different legal elements, produce very long sentences, that were not really contemplated by the legislative bodies that enacted them. A very large share of all extraordinarily long sentences flow from the imposition of consecutive sentences for multiple offenses in connection with the same course of conduct.

For example, while vehicular homicide while driving drunk is clearly a crime that should receive a serious sentence, it is not at all obvious that a sentence four times as long is appropriate if there are four victims in the car hit, rather than one. The harm may make it more serious, but not necessarily four times more serious. Similarly, while possession of hundreds of child porn images, or repeated instances of sexual assault on a child victim over an extended relationship may be more serious than an isolated instance of either, it isn't at all obvious that a sentence hundreds or dozens of times worse makes any sense.

Posted by: ohwilleke | Jul 13, 2009 2:29:09 PM

Normally state departments of correction provide a snapshot of the prison populations by offense type and in some cases offense severity. If they also provided the number of admissions by offense type/severity it would be much easier to understand what types of offenses were primarily responsible for the cancerous growth in the prison population (which appears to have slowed in general and reversed in a few states).

If the admission date is included in the snapshot one can study the dependence of length of confinement on offense type/severity. I have done this for Iowa prisons and about 75% of the prisoners were confined for less than three years. For about half of this set the admit type was a parole/probation revocation and their most frequent original offense types were drug trafficking, property, sex and assault. The set of inmates with lengths of confinement longer than three years the offense type was primarily violent with serious sex and property offenses minor constituents.

It used to cost $1 million to incarcerate 40 people for a year and now it costs that much to incarcerate 35. I am convinced that over half of the crime is local and that means there are city blocks in high crime neighborhoods where were are spending about $1 million per year to incapacitate some of their residents. If you look at their recidivism rates this has become a fixed cost and I would not be surprised if we have created the most expensive revolving door in history.

Posted by: John Neff | Jul 12, 2009 10:12:49 AM

I've been thinking about this post since I first read it yesterday, but I'm having a bit of a hard time formulating a reply. That is because I can't quite seem to figure out if I agree with Jonathan or not.

Jonathan's position that we need to decrease sentences for violent offenders is counterintuitive. Most people who speak out against the modern increase in incarceration rates do so by identifying how much of that increase is attributable to the incarceration of non-violent offenders (e.g., drug offenders). Indeed, Jonathan seems to rely on this apprach in his price signal and net-widening arguments, which (to some extent) say that we shouldn't increase sentences for violent offenders because it will result in increased sentences for all offenders, including those who didn't engage in violence.

Jonathan's argument also champions those offenders who a liberal democracy feels most comfortable punishing --- those who have intentionally inflicted a physical injury on others. (Of course, I'm ignoring the recent trnd to reclassify various crimes, like drunk driving, as violent offenses under various statutory schemes.) Because hurting someone is the sort of behavior around which there is a broad social consenses that it is wrong, I think it will be difficult to argue that violent offenders do not deserve lengthy punishments (or at least, it would be difficult for me to make such an argument).

But there are also parts of Jonathan's post that I find myself agreeing with. I am sure he is correct that the modern sentencing increases for violent offenders is attributable to public fear of crime. And as I mentioned in my post on future dangerousness earlier this week, there are reasons to distrust such fears. Public fear about crime is not rational --- it has grown as crime rates have dropped. And there are various social sceince studies that demonstrate when individuals hear about crimes (including violent crimes), they assume that the crime was committed in the most aggravated ways, and that those erroneous assumptions may help to explain modern increases in sentencing.

So, I suppose the upshot is that I am skeptical about the need to reduce sentences for violent offenders, but I'm definitely not sure that my skptical instincts are correct.

Posted by: carissa | Jul 10, 2009 2:16:23 PM

My sense is that long sentences really aren't the source of prison growth. My own analysis of the NCRP suggests that such sentences--say sentences of over 5 to 10 years--are rare (at least in those states that provide data, which results in an under-representation of the South but includes California). Perhaps frequently imposed, but rarely actually served, which is what counts. Admissions, not duration, seem to be the primary engine of growth, and admissions are better targeted through the "soft underbelly," since non-incarcerative sentences for many if not most violent offenders are both politically infeasible and perhaps bad policy.

If you're interested, my paper is here: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1338365

Posted by: John Pfaff | Jul 9, 2009 1:42:34 PM

Yes- I'd guess that's right. Perhaps that's the low-hanging fruit that should be gone after first (though I'd rather see more systematic change over time.)

Posted by: Matt | Jul 9, 2009 11:49:45 AM

Our sentences are so long now that I suspect even many who believe incapacitation through imprison did contribute to the crime decline would think it safe to shave a few years or even decades off them. A 70 year sentence ceases to have much incapacitative value after the prisoner is 40? 50? 60?

Posted by: Jonathan Simon | Jul 9, 2009 11:33:52 AM

I'm very sympathetic to the point of this post and think it's clear that we over-punish for many, perhaps most, crimes, including "violent" crimes. Mixed with the work of Paul Robinson and others, I think there's a clear room for improvement here. The one worry I have is related to the the claim, made by several people though I won't look them up now, that a large part of the crime drop we've seen since the 1990s has been due to more and more would-be criminals being incapacitated via long prison sentences. Now, I think that even if this is true (and I think the evidence for it is at best mixed), we probably pay too high a price for this reduction in crime, both in terms of doing injustice (punishing people more severely than they deserve) and in other social costs. I also expect we could get some of the benefits in other ways (we might hope that a more just system might better engender compliance, for example.) But it does seem to me that the possibility that violent crime would go up, perhaps quite sharply, if we had less incarceration is one we must take seriously and be willing to meet head-on if we're going to make progress here.

Posted by: Matt | Jul 9, 2009 11:11:46 AM

The comments to this entry are closed.