Search About Newsletters Donate
Support independent, nonprofit journalism.

Become a member of The Marshall Project during our summer membership drive. Our journalism has tremendous power to drive change, but we can’t do it without your support.

Gov. John Bel Edwards signed a package of bills to overhaul Louisiana’s criminal justice system in June.

Nine Lessons About Criminal Justice Reform

What Washington can learn from the states.

Adapted from remarks to the Ninth Circuit Judicial Conference, San Francisco, July 17, 2017.

Since November, a kind of fatalistic cloud has settled over the campaign to reform the federal criminal justice system. With a law-and-order president, a tough-on-crime attorney general, and a Congress that has become even more polarized than it was in former President Barack Obama’s time, most reform advocates say any serious fixes to the federal system are unlikely.

Reformers have been consoling themselves by looking to the states. After all, most of law enforcement, most of criminal jurisprudence, and most incarceration takes place at the state or local level.

“Reform” is one of those ambiguous words that mean different things to different people. I think of reform as something that aims to reduce the numbers of Americans who are removed from society and deprived of their freedom, and to do it without making us less safe. In 1972, when I was starting my newspaper life at The Oregonian, 93 out of 100,000 Americans were in state or federal prisons. By 2008 the incarceration rate had grown nearly six-fold, to 536 per 100,000, and it has hovered in that vicinity ever since. That’s not counting the hundreds of thousands held in county jails on any given day or those confined in the juvenile justice system or immigrant detention.

Every year about 650,000 of those prisoners are released back into the world. We know that most of them will be unemployed a year later, and that two-thirds of them will be rearrested within three years. We have a corrections system that fails to correct.

Here are a few lessons Washington can learn from the states.

Lesson 1: It is possible to reduce incarceration and crime at the same time.

Between 2010 and 2015, 31 states reduced both crime and imprisonment. In the 10 states with the largest declines in imprisonment, the crime rate fell an average of more than 14 percent.

This obviously does not mean that reducing incarceration necessarily leads to a drop in crime. Correlation is not causation. The question of why the crime rate declined is a subject of heated debate among social scientists. One of my colleagues at The Marshall Project wrote a piece we called “10 Not Entirely Crazy Theories Explaining the Great Crime Decline.” One thesis our writer examined is that after Roe v. Wade the legalization of abortion meant fewer unwanted children who were more likely to become delinquents. Other researchers have surmised that removing lead from paint and fuel has made for a less criminogenic environment. Another theory credits technology: anti-theft devices in cars and the spread of online banking made it harder for criminals to profit. Most experts give some credit to the increased deployment and improved equipping of police. And, of course, some of the decline is a result of the fact that more bad guys were locked up — though that is an expensive way to keep communities safe.

Whatever the factors responsible for the relatively low crime rate, the evidence from the states is that reducing incarceration is compatible with reducing crime. Obviously, a lot depends on how you reduce prison populations, which is where the states have much to teach us.

Lesson 2: The embrace of criminal justice reforms is bipartisan.

This is one of those rare issues in our polarized country where activists on the left and right have found a patch of common ground. On the right, we have fiscal conservatives who see our prisons as wasteful, libertarians who see our handling of crime as another manifestation of oppressive big government, evangelical conservatives who see aspects of the system as inhumane.

There are of course issues where left and right still part company. Controlling the proliferation of guns remains a political third rail. The left wants to talk about race, and the right mostly does not. But on issues such as pre-trial diversion, indigent defense, sentencing, parole, rehabilitation, bail and asset forfeiture, you will find the Koch brothers arm-in-arm with the ACLU.

In fact, conservatives rightly boast that red states have often led the way, starting with Texas during the governorship of Rick Perry. In the past decade, that state has closed four prisons, reduced its incarceration rate by 20 percent, and invested $240 million in alternatives such as drug treatment. The Texas experience is often cited as evidence that politicians can support so-called smart-on-crime reforms and live to tell about it.

Two caveats regarding the Texas Story: First, Texas started out with one of the highest incarceration rates in the United States, so it had a long way to go; it is still the seventh most incarcerated state. Second, Texas accomplished its reductions by redirecting money, not by changing the legal infrastructure. The state invested in alternatives, which meant judges had greater confidence that when they diverted someone to drug treatment, there would actually be drug treatment.

Other conservative states — Georgia, South Carolina, Utah to name a few — have tackled the structure of criminal justice, reducing some felonies to misdemeanors, revising mandatory minimum sentences and three-strikes laws, expanding eligibility for parole and removing barriers to reentry.

The most recent convert is Louisiana, a state long known among criminal justice reformers as a contender in every race to the bottom. Last month Louisiana, passed a remarkably comprehensive legislative overhaul. That feat was a product of strong political leadership, intense lobbying by reform groups across the political spectrum, and a corrections system that was bursting at the seams.

Lesson 3: Probably the most effective way to reduce incarceration is not to lock people up in the first place — at least not so many, and not for so long.

I’ve already mentioned sentencing reform; in the last decade, 23 states have relaxed their sentencing laws — something Congress has so far been unable to do for the federal system. But there are other front-end measures that have been employed by states to keep people out of prison.

One is less reliance on money bail. The people most likely to spend time in jail awaiting trial are not the worst offenders but the poorest offenders. Even a short stint awaiting trial in jail increases the odds that an offender will ultimately end up doing prison time. A number of jurisdictions have curtailed the use of cash bail — most notably New Jersey, which now requires judges to hold hearings shortly after arrest to determine whether a defendant can be safely released before trial. Since the new procedure began in January, the average daily jail population has dropped 19 percent.

A second measure aimed at reducing prison intake is raising the age at which juveniles, whose brains have not yet developed mature impulse control, are thrown into the adult system, which too often subjects them to predators and leads many to careers in crime. In the last two years Louisiana, South Carolina, New York and North Carolina have raised the age to what is now the national norm: 18. Connecticut may be the first state to raise the cap to 21. In March, Gov. Dannel Malloy, announced the opening of a special corrections unit for young adults as old as 25.

A third way to slow the traffic into prisons is to provide better — and earlier — indigent defense.

And a fourth is to elect prosecutors who don’t regard maximum prison sentences as the main measure of job performance. Within the past year several jurisdictions — including Chicago, St. Louis, Houston, Denver, Tampa and Orlando — have elected prosecutors who campaigned on reform platforms.

Lesson 4: While the front end is important, don’t neglect the back end.

There is abundant evidence that supports college and vocational programs behind bars, regular contacts with family, reentry and parole and probation programs that have the resources and the mandate to land their clients safely back in society. A RAND Corporation study in 2014 concluded that “inmates who participated in correctional education programs had 43 percent lower odds of recidivating” — a verb that makes my inner English major cringe but suggests a pretty good return on investment.

Lesson 5: Be wary of reformers who suggest you can cut incarceration drastically by releasing low-level, nonviolent offenders.

On the homepage of #Cut50, a movement that advocates cutting prison populations in half, you will not find the word “violent.” But more than half of those incarcerated in state prisons have been convicted of crimes classified as violent. Only 16 percent are in for drug crimes, not all of them nonviolent. (The #Cut50 activists know this very well and at the state level have advocated measures to reduce draconian penalties for violent crime. On Thursday, they updated their homepage, saying they support reforms for "those convicted of both violent and nonviolent crimes.")

Decriminalizing marijuana will reduce incarceration, but any hope of restoring the incarceration rates of the 1990s means reducing sentences and stepping up rehabilitation for people convicted of violent crimes. Given the political difficulty of that, reduction of incarceration is likely to happen incrementally. The state that has been downsizing its prison population longest and most aggressively — the state where we are gathered today — has cut a bit more than 25 percent, and no other state has come close.

Lesson 6: Prison reform doesn’t necessarily mean a huge windfall for taxpayers.

Whether reform means massive savings remains to be seen, for two reasons: first, the alternatives to prison aren’t free. To keep crime in check, money not spent on actually confining offenders has to be spent on mental health and addiction treatment, more hands-on probation, and, ideally, education, job training and housing support. Moreover, some states have found that the beneficiaries of prison — the corrections staff, the contractors, the politicians, the unions — are ferocious defenders of corrections budgets.

But there is another way of looking at the cost of mass incarceration.

The most commonly cited estimate of how much it costs to maintain the country's prisons and jails is $80 billion a year. If you throw in things like health and pension benefits for prison staff, the cost to governments is more like $90 billion.

But a report last year by researchers at Washington University in St. Louis, attempted to add up the "social costs," a toll that includes: lost wages, the cost of visitation, the higher mortality rates of both former inmates and their infant children, child welfare payments, evictions and relocations, divorces, diminished property values and the increased criminality of children with incarcerated parents. The bottom line they came up with was one trillion dollars a year, nearly 6 percent of GDP.

Lesson 7: Metrics matter.

It’s impossible to know what works and what doesn’t without reliable data, and reliable data is often in short supply. Our data guru at The Marshall Project, Tom Meagher, wrote a piece last year entitled “13 Important Questions About Criminal Justice We Can’t Answer” — can’t answer because the data is unreliable or unavailable. They include such basic questions as how many juvenile offenders graduate to become adult offenders, how many people have served time in prison or jail, how many people in America own guns and what percentage of inmates eligible for parole are actually granted release from prison.

The Washington Post and the Guardian set out separately to count the number of civilians killed each year by police in the line of duty. The number they came up with was about 1,000. That is roughly double the official estimates from the Department of Justice — an astonishing margin of error.

As important as having good data is knowing what to do with it. That brings us to the debate underway in many states over the use of risk-assessment tools, basically tests aimed at helping make wise judgements at critical moments in the handling of the accused or convicted. Risk assessment tools are algorithms that examine a subject’s history in order to mitigate the chances of re-offending. There are various tools for various applications: to help determine whether a defendant is a flight risk, how severe a sentence should be, whether an inmate is a fit candidate for parole and what kind of supervision he requires upon release.

The left generally hates risk assessment, because the inputs may include factors like employment stability and past encounters with the law that weigh more heavily against communities of color. Advocates of risk assessment tests, respond that a) they are getting better, both more accurate and less biased; b) they are meant to assist judges and parole boards, not preempt professional judgement; and c) properly used, risk assessment tools can assure people in the system get support they need to stay out of prison, such as placement in anger management courses and addiction treatment. (Full disclosure, the national leader in this field is the Arnold Foundation, which is a funder of The Marshall Project.) I’m not a worshipper at the shrine of data, but if I think if I had to make these profound decisions I’d prefer to have a sense of the odds.

Lesson 8: Many states are finding that incentives work better than mandates.

A good example is an approach being used in about a dozen states. Take a defendant who is probably not a threat, who would do fine returned to the community under proper supervision. But the judge knows “proper supervision” is unlikely, because the local probation system is threadbare. Suppose the state agrees that for every dollar it doesn’t have to spend locking people up, it will send 40 cents to the county to pay for more robust supervision? The state saves money, the county improves its oversight of former inmates, and the judge has greater assurance that the subject will be supervised.

Lesson 9: The states have wide latitude to experiment and borrow ideas from one another, and they are seizing it even now. But they are not immune from the new regime in Washington.

The federal government sets a tone, and you will hear complaints from several states that the new administration has had a chilling effect on state legislatures. When the attorney general instructs federal prosecutors to charge the maximum, as Jeff Sessions has done, when his response to a national opioid epidemic is to yearn for a revival of a discredited 1980s anti-drug program, that sends a message to state legislators contemplating new approaches — and a warning to states that have legalized marijuana. So far, since the new sheriff arrived in Washington, reform measures have lost momentum in Kentucky and Oklahoma. An advocate involved in the reform battles in both states said of the rhetorical barbs from the White House and DOJ: “To say that those things haven’t had an impact on the environment is just incredibly naive.”

Moreover federal programs, especially Medicaid, which was expanded to include former inmates under the Affordable Care Act, can be essential to getting released offenders up on their feet.

In other words: What happens in Washington doesn’t always stay in Washington.