Prisons are terrible, and there’s finally a way to get rid of them

In any given year, 3.2 percent of those in jail, 4.0 percent of state and federal prisoners, and 9.5 percent of juvenile detainees report having been sexually abused. A 2005 study found that the rate of physical assault was over 18 times higher for male inmates and 27 times higher for females relative to the general population.

The system that sends people to these inhumane pens where they're likely to get beaten and raped is baldly racist. For the same crimes, black men are typically sentenced to 20 percent longer prison stays than white men.

And taxpayers are paying through the nose for all this. About 2.2 million Americans were incarcerated in 2012, accounting for one in every 108 adults, a figure that has more than quadrupled since 1980, leading state spending on incarceration to rise nearly as fast. America imprisons more people than any other country, and has the highest incarceration rate. And this wrecks communities. Mass incarceration weakens the economy and increases teen pregnancy in areas where many residents are sent away, and harms the children of the incarcerated in ways that persist for decades, among many other damaging effects.

So why do prisons exist? In theory, because we need them. They keep bad guys off the street. They give people a reason to not commit crimes. They provide a place where violent or otherwise threatening people can be rehabilitated.

But prisons aren't the only way to accomplish those goals. Technological advancements are, some observers say, making it possible to replace the current system of large-scale imprisonment, in large part, with alternatives that are not as expensive, inhumane, or socially destructive, and which at the same time do a better job of controlling crime. The most promising of these alternatives fits on an ankle.

Why GPS changes things

While the idea of house arrest has been around for millennia, it has always suffered from one key defect as a crime control tool: you can escape. Sure, you could place guards on the homes where prisoners are staying, but it's much easier to secure a prison with a large guard staff than it is a thousand different houses with a guard or two apiece.

Today, we have something better than guards: satellites. The advent of GPS location tracking means it's now possible for authorities to be alerted the second a confinee leaves their home. That not just enables swift response in the event of escape; it deters escape by making clear to detainees that they won't get away with it.

Researchers have tested electronic monitoring as an alternative approach to parole, probation, or other criminal punishments that fall short of imprisonment — and it's been a huge success. An Urban Institute analysis found that electronic monitoring reduces odds of re-arrest by 23.5 percent relative to traditional probation, and a randomized study in Switzerland found major advantages to electronic monitoring compared to mandated community service. Results from a Swedish trial in which prisoners were offered early release under electronic monitoring were similarly promising. The threat of electronically-enforced house arrest appears to provide a strong deterrent effect; when Santa Fe County, New Mexico started threatening drunk drivers with home confinement if they failed to install ignition interlocks (which prevent a car from starting until the driver passes a breathalyzer test), installation rates skyrocketed.

The most intriguing evidence comes from Argentina, where Harvard's Rafael Di Tella and Torcuato Di Tella University's Ernesto Schargrodsky found that electronic monitoring cuts recidivism nearly in half relative to a prison sentence. That raises the possibility that electronic monitoring could be more than merely a supplement to prisons. It could replace many of them. The program evaluated used something a bit less technologically sophisticated than GPS tracking. Offenders wore an ankle bracelet which transmitted a signal to a receptor in their home. If the signal is interrupted, or the device appears to be manipulated, or the vital signs of the individual are not being transmitted from the bracelet, then the receptor calls it in.

Di Tella and Schargrodsky's evidence is particularly compelling because the decision of whether to give Argentinian arrestees house arrest or prison was made randomly. In most countries, electronic monitoring is offered to defendants judged to be less dangerous, so you'd expect those sentenced to it to reoffend less than those sent to prison. "If you show someone released into monitoring has lower recidivism, all you show is that the judge was successful and identified the person who was less dangerous," Di Tella notes.

But in Argentina, judges are randomly assigned to cases, and strict and lenient judges differ wildly in their inclination to use electronic monitoring. The result was that extremely high risk people were sometimes given electronic monitoring and extremely low risk people were sometimes thrown into jail — it was just random. The leniency of some judges meant that there were "people accused for the second time of murder [who] were still given electronic monitoring," Di Tella says. Di Tella and Schargrodsky had stumbled upon a true, randomized experiment, and the result was being monitored instead of imprisoned caused people to reoffend less.

Who do we actually imprison?

So, if electronic monitoring can work just as well as prison — and keeps prisoners from being physically and sexually assaulted by guards and other inmates, and saves money, and perhaps even allows some inmates to earn a living while serving time — why not switch?

Di Tella says he isn't ready to make that dramatic a shift. For one thing, GPS probably isn't prepared to deal with genuinely violent and dangerous offenders just yet. In one case, a GPS assignee named Angel Fernandez escaped and proceeded to murder a family of four, including children aged 8 and 10; the ensuing backlash was enough to end the program in Argentina. Similar incidents have occurred in the UK, which has seen the use of GPS tagging increase considerably in recent years.

But the fact of the matter is that rapists and murderers are a distinct minority of the prison population, at least in the United States. According to the Bureau of Justice Statistics, as of 2011 only 12.6 percent of state prisoners in 2011 were there for murder, 1.5 percent for negligent manslaughter, and 12.4 percent for rape or sexual assault. That's only 26.5 percent of the overall prison population. The numbers are even starker in federal prisons: only 3.8 percent of prisoners committed any kind of violent crime.

If you look at who's being sent to prison in a given year, the share is smaller still. In 2011, only 2 percent were admitted for murder, 0.7 percent for negligent manslaughter, and 5.4 percent for rape or sexual assault. Because they face long sentences, people convicted of homicide and rape make up a considerably larger share of a prison's population than they do of a given year's admissions. But that means that in a typical year, 91.9 percent of the people who are thrown into prison are there for something else. The vast majority of the people getting locked up aren't killers or rapists.

Which isn't to say they're angels, either. 26.9 percent of state prisoners are there for robbery, assault, or another violent offense; 18.7 percent are there on non-robbery property crime charges, 16.6 percent on drug charges, and 10.6 percent on public order charges (which includes everything from weapons charges to drunk driving to contempt of court to prostitution). Even if you support legalizing drugs and sex work, you probably want the government to do something about people who rob or assault each other. The question is whether that "something" has to be prison.

Making punishment predictable

Better technology only strengthens the case. Previous electronic monitoring systems, like the one Di Tella and Schargrodsky studied, involving bracelets transmitting signals to a receptor, fell short in cases of escape. They left police with no sense of where escapees have gone. GPS tagging changes that (at least for confinees who flee with their bracelet still on). "The GPS tagging technology now available enables the development of a tagging regime that works," Rory Geoghegan, a crime policy researcher at the British think tank Policy Exchange, writes. "One that protects and controls offenders, but also aids them to change, because constant supervision of a wearer's location accords with the academic evidence that certainty and swiftness of sanctions is more critical than the severity of any sanction ultimately invoked, and is therefore the best basis for behavior change."

The latter point is crucial. While better technology is making replacing prison with GPS-enforced house arrest possible, it needs to be combined with a system of punishment that guarantees a speedy reprimand if the rules are broken. If escapes aren't answered quickly, either with an immediate return to home confinement with increased supervision or with a return to prison, the system will fail. If people are punished when they have not in fact violated the rules, the system will fail. And if the costs of violating the rules are vague and uncertain — even if they're also quite severe — the system will fail.

This paradigm, pioneered by UCLA criminologist Mark Kleiman, has been gaining traction for years, and has shown great promise when applied to drug probation in Hawaii (and, later, Texas and Michigan), punishment for drunk driving in South Dakota, and parole enforcement in Washington. There are two benefits to "swift and certain" sanctions. The first is that it works. The Hawaii system — which replaced an approach where drug probationers faced lengthy jail sentences for breaking probation, but where those sentences were rarely imposed, with one where regular drug tests were expected and missing or failing one guaranteed a night in jail — reduced recidivism by about 50 percent. The second is that it's way, way more humane. Throwing people in jail for five, ten, even twenty years for a drug violation is obscene. Swift and certain sanction not only reduces crime, it reduces punishment and makes it less horrible.

A solution

So how's about this. The US should:

1. Move those imprisoned for offenses short of homicide or sexual assault to GPS-supervised house arrest as soon as is practicable, with a guaranteed, immediate prison stay for those who violate its terms.

2. Reserve prisons for repeat offenders and those who've committed truly heinous crimes.

There are obviously other details to be worked out. You wouldn't want people convicted of domestic violence to be sentenced to home confinement with their victims, for instance; in those cases, some kind of alternate housing would have to be offered to ensure separation.

But if successful, this plan could reduce admissions by at least half, probably much more. Hopefully, this will just be a temporary measure. In principle, it could get to the point technologically where house arrest becomes as hard to escape as prison is. At that point, abolishing prison outright starts to become imaginable. UK home secretary David Blunkett spoke too soon when he referred to electronic monitoring as "prison without bars," but that dream is attainable. As Kleiman once put it, "My view is that if you know where someone is, you don't have to put them in the cage."

Read more:

• Graeme Wood on privacy concerns with electronic monitoring.
• Molly Carney on using electronic monitoring to replace, rather than supplement, the prison state.
• Mark Kleiman on swift and certain sanction, drug policy, and more.