Monday, April 11, 2016

How to make mass incarceration disappear: John Pfaff, magical statistics and sentencing reform

Grits enjoys reading Fordham law professor John Pfaff, including his informative Twitter feed, because his best observations are true but utterly counter-intuitive, sometimes in ways that are as difficult to explain as if one had watched a magician perform a trick. I enjoy the process required to figure out if I agree; sometimes yes, sometimes, no.

Take, for example, Pfaff's observation that sentence lengths have not increased greatly since the '90s and the real source of mass incarceration is prosecutors using their discretion to pursue cases to conviction more frequently. This surprised me. Many of us have taken as an article of faith that longer sentences drive incarceration numbers, in Texas at least since the Criminal Justice Policy Council's 2000 report, "Goal met: Violent offenders in Texas are serving a higher percentage of their prison sentences." After all, there are more than 8,400 Texas prisoners serving life or capital life sentences (nearly 700 more serving LWOP), with another 32,500 Texas prisoners facing sentences longer than 20 years (as of Aug. 31, 2014). TDCJ houses thousands of sex offenders, most of whom serve their full sentences, or close to them, because the parole board won't let them out. A lot of the DNA exonerees had been in prison two decades or more. We've had an inmate spend more than thirty years in solitary, for heaven's sake! Those folks all exist. How could their long-term presence not be driving mass incarceration?

And yet, Pfaff's observation about time served apparently holds true for Texas - average age of prisoners at release has risen slightly, but the actual length of sentences served by those getting out has gone down since the late '90s. Instead, increased prison admissions are the main driver of mass incarceration. Texas prisons accept and release more than 70,000 people per year now, with the number of new "receives" remaining roughly constant over the last decade despite a marked decline in crime. Grits cannot explain why the length of sentences served has not increased given the large number of lifers and others with long sentences - that's the magician's trick, it's like he made them disappear! - but Pfaff's mathematical observations about the amount of time inmates serve hold true for Texas.*

Instead of long sentences, Pfaff blames prosecutors for mass incarceration. Parsing the data, Pfaff has found that "longer sentences explain [prison population] growth during the 1990s, and increases in [prison] admissions per arrest [explain] the growth in the 2000s." That's why Texas prisons remain full even though we release more than 70,000 prisoners per year and crime is down dramatically from two decades ago: Prosecutors have been more aggressive, securing convictions at ever-higher rates from a smaller pool of offenders. Pfaff found that, between 1994 and 2008, "the probability that a district attorney files a felony charge against an arrestee goes from about 1 in 3, to 2 in 3."

Unfortunately, Prof. Pfaff so far has been short on solutions aimed at the core problems he's identified beyond the feds spending $4 billion per year to subsidize local public defender offices. (He's writing a book so maybe we'll get more magical suggestions there.) Grits suspects bail reform might help, giving defendants more leverage in negotiations at the plea-bargain stage. Expanding funding for community supervision programming might make prosecutors and judges less likely to seek prison time, one supposes. But in general, Pfaff identifies few public-policy means to check prosecutors' discretion besides elections. And in my experience the R or D in front of an elected DA's name typically doesn't mean much - the job is the job.

Pfaff's criticisms of sentencing reformers routinely suffer from a significant omission, particularly when one contemplates state-level reform, as he agrees we must. While it's true that focusing on long sentences alone won't eliminate mass incarceration, there's one aspect of sentencing reform that very much will help: Whenever an offense shifts from felony to misdemeanor status, that takes people out of the state prison system and puts them under the county's jurisdiction, making them much more likely to serve a short incarceration stint or receive probation. (Most county-jail incarceration happens pretrial; the max misdemeanor sentence in Texas is one year.)

For example, last year the Texas Lege boosted felony property theft thresholds so that now theft becomes a felony at $2,500 instead of $1,500. That means stealing $2,000 was a felony in August and a misdemeanor in September. So, those offenders never go to prison or have a felony on their record (for that crime, anyway). To that extent, sentencing reform does matter, but it matters most along that margin between felony and misdemeanor offenses. It's not lessening long sentences where decarceration advocates get the most bang for the buck; it's reducing the number of people sent to prison in the first place. Part of that must involve sentencing reform.

In a similarly dismissive fashion, Pfaff poo poos the impact of the drug war on mass incarceration, and indeed only 16 percent of Texas state prisoners are drug offenders (24,005 people as of 8/31/14). You could release them all and we'd still incarcerate more people than any other state, so that part's true enough. Pfaff further argues that the majority of prisoners are incarcerated for violent offenses (fifty-five percent in Texas, so, check), so reformers should focus on restructuring sentences for those crimes.

I've given this argument a great deal of thought and, for Texas anyway, I don't particularly agree. For starters, the parole board here is responsible for release rates, which have been rising slightly in recent years, so the actual length of violent offenders' sentences can only be indirectly affected by the political process (the Governor appoints them, the Legislature approves their budget). OTOH, sentencing reform for lower-level drug and property offenders contributes much more directly and quickly to achieving decarceration goals because it results in fewer people sent to prison. By contrast, most serious violent offenders would still go to prison even if max sentences were reduced, so admissions (which in Pfaff's view is the only metric that matters) would not go down.

Pfaff does grant that the drug war boosts incarceration by generating more defendants with prior criminal records, which may increase the likelihood prosecutors file future charges or make the defendants subject to enhancements. "[B]y producing cohorts of offenders with longer records, [the drug war] generated a pool of offenders that may face tougher sanctioning outcomes even when all else is constant, thus helping prison populations trend upwards even as crime rates fall." To that extent, making fewer people felons because of drug offenses over time would mitigate prison population growth.

Bottom line, if it's true that mass incarceration has been "driven almost entirely by increases in felony filings per arrest," then reducing common charges like petty drug possession and low-level property theft from a felony to a misdemeanor is a legitimate avenue for diminishing that trend. Is it a panacea? Definitely not. Nothing ever is. But it's a starting point. And journeys of a thousand miles must begin with first steps.

* One thought: One can only judge time served based on offenders who are released, both because that's the datapoint available in the TDCJ Annual Statistical Report and because, by definition, that's when "time served" is defined. However, lifers and people with very long sentences who were convicted in the last quarter century are unlikely to get out in any given year. So maybe they're not included in that dataset because they're not getting released yet. ¿Quien sabe? I can't tell where he hid them. Is there a trap door? A secret compartment? It's a head scratcher.


mike connelly said...

Two questions on "sentence length": do the data account for increased use and/or requirement of consecutive sentences, particularly important since count stacking became a popular pastime in the same period of sentence length increases, and have "times served" changed proportionately or gotten longer (including any revocations after initial release)? Make that three: do the data take into account states like WI where there is a prison sentence and a "supervision" sentence afterward wherein "supervision" increased in length after the practice was adopted and revocation restarted the original prison sentence from the point of "supervision" release?

As for policy proposals, you hit on the key point: take away the ability of prosecutors to impact the policy changes as much as possible. Downgrading felonies to misdemeanors requires more creative accounting to overcome on prosecutors' parts and can be audited if the will is there. The popular initiatives that legalize and decriminalize offenses should be used more and extended past pot and felony threshold amounts on property crimes (which should be indexed to inflation each year, btw, too, to be lasting reform). The initiatives can also provide a way to get away from the too often self-defeating "commissions," "workgroups," and "committees" that insist on lowest common denominator "stakeholder" policymaking derived from its participants which always include prosecutors and other representatives of constituent groups heavily invested in the status quo and who are given effective vetoes when "consensus" is placed above good policy.

In addition, doing away with the "bargaining power" provided for and explicitly defended by prosecutors that comes with mandatory minimums would hit at both prosecutor political power and the sentence lengths, especially when mandatories often include multiple offenses be sentenced consecutively. Basic and frequent recodification efforts by states to review and revamp the state criminal codes has been used in WI and MO to bring more light and consideration what the system is accomplishing as well, although that candle can obviously be burned at both ends.

Most important is simply to include in your "reform thinking" that outcomes must consciously be designed to limit, not increase, prosecutor control over the system. That's implicit in the few really successful reforms that have been achieved, but full-scale changes in the system to get true decarceration going will take acknowledgement and political leadership to pull off. TX did it with the refusal of its Speaker in 2007 to build more prisons, GA is doing it with clear and involved participation of its governor, CA and the pot law change states have done it by going directly to voters. It can be done, but the famous box, paradigm, narrative, whatever that has driven the last 2+ decades of reform efforts has to be consciously not just stepped outside of but denied. As crime rates and public fear of crime numbers seem to be rising a bit again, the time to do so may be getting shorter before the policy window that has let in the reform that has occurred, such as it is, closes again as it historically has. Pfaff better get his book and better ideas for change out before it's shut.

Anonymous said...

Another question - does the data apply only to prisons, or does it show how many people are in jail for felony non-payment of child support thanks to the Bradley Amendment (42 US Code 666) because they are 'dead-broke' and can't afford to pay child support?

I don't have access to the data that the author used, but I think that it's a question worth asking, especially after the problems identified in Ferguson, Missouri associated with the unintended consequences of flawed child support enforcement.