How Cost-Benefit Analysis Might Save America’s Criminal Justice System

The Economist, April 30, 2016

It took decades for politics – that messy, flawed business of gauging the public mood, haggling with interest groups and turning out votes – to build America’s prison system: a behemoth of staggering size and cost, unlike anything else in the developed world. What chance is there that the monster can be tamed by the dry, abacus-clicking discipline of economics?

A non-trivial chance is the surprising answer, judging by a bi-partisan conference of economists and policy wonks at the White House on April 15th, co-hosted by the conservative American Enterprise Institute and the broadly progressive Brennan Centre for Justice. The meeting launched a report into a system that, after decades of relentless growth, holds over 20% of the world’s prisoners, though America is home to less than 5% of the global population. The report’s dry title, “Economic Perspectives on Incarceration and the Criminal Justice System,” should not deceive. Written by the White House Council of Economic Advisers (CEA), an in-house think-tank that has advised presidents since Harry Truman, it is a 79-page catalogue of unintended consequences, failed policies and heartbreaking waste, counted in the tens of billions of dollars and, more important, in human lives.

By tradition, American journalists refer to really startling facts as “Hey, Martha!” moments, imagining a newspaper reader exclaiming to a breakfast companion. In Britain, such facts are “marmalade droppers.” Such moments fill the new report. America spends $80 billion a year locking up 2.2 million people, reflecting an incarceration rate that has climbed remorselessly to more than four times the world average, even as violent crime rates fell sharply. Between 1980 and 2014, the proportion of people behind bars more than tripled, with especially sharp rises among black and Hispanic men. Much growth was driven by “tough on crime” laws passed in the 1980s and 1990s, which sent even non-violent drug offenders to jail. If past trends continue, one in three black men born in 2001 can expect to serve time at some point.

Every year, 600,000 people are released from American prisons. More than half of all prisoners have mental health problems, while about two-thirds did not complete high school. Once out, ex-cons join about 70 million Americans with criminal records, a status which in several states will deny them public housing and the right to vote, and legally bar them from occupations which require a license, such as hair-cutting or plumbing. Recidivism rates are dire: one study following people released from state prisons found that, within five years, more than half were back behind bars.

Small wonder that politicians on both the left and right are increasingly asking whether today’s system is humane or just. Protesters have challenged the Democratic presidential hopeful, Hillary Clinton, to disown a crime law passed in 1994 by her husband, Bill Clinton, which included billions of dollars for new prisons and tougher sentencing rules. She has partially obliged, calling for an end to today’s “era of mass incarceration.” Congress is weighing bipartisan bills to loosen ridged sentencing laws. But for all the talk, politicians seem stuck. In Congress, conservatives vow to resist bills that might release violent offenders. Leftists demand radical change to a system they call inherently racist.

Enter the economists. The CEA report confines itself, explicitly to questions of costs and benefits: whether locking so many people up for so long is an efficient way to reduce crime. Its conclusion is a resounding no. Incarceration does prevent some crimes from being committed, it concedes: a big California study from 2013 suggests that each extra year in jail typically avoided one or two property crimes, with the effects strongest for motor-vehicle theft. But hefty prison spells turn out to be a costly and clumsy way to deter offenders. Explaining this finding, the CEA chairman, Jason Furman, cited the law of diminishing marginal benefits. Once a country’s prison population is large, additional prisoners are more likely to be low-risk offenders. Indeed, several studies suggest that harsher sentences may be counter-productive, for instance as young offenders emerge hardened by juvenile jails. Other research points to better policing and an ageing population as likelier causes of lower crime rates.

Misbehavioural economics
Two former directors of the Congressional Budget Office spoke at the White House meeting, Douglas Holtz-Eakin, a Republican who served on George W. Bush’s CEA, called criminal justice reform a “rare public-policy moment” that offers both parties a chance to save taxpayers money, help more people into the labor force, strengthen families and reduce poverty without sacrificing public safety. Peter Orszag, a Democrat who headed the Office of Management and Budget in Barack Obama’s first term, called the evidence “compelling” that to deter crime, the severity of a possible punishment matters much less that the certainty that it will be inflicted. Alas, politicians have poured resources into incarceration rather than more cost-effective tools, such as hiring more police and directing them to crime hotspots. (America employs two-and-a-half times more corrections officers per person than the global average, but 30% fewer police). The vast majority of burglaries (85%) are never cleared up, while fewer than half of violent crimes lead to arrests.

One strength of this wonkish, follow-the-numbers approach is that it avoids the political challenge faced by reformers: the fact that many voters yearn to feel safe from crime, and do not want to be told that this is a wicked and selfish ambition. The CEA report duly calculates the cost to society, and finds that it is large. But it then shows that today’s policies are a horribly wasteful way of reducing the scourge. Scare-mongering headlines, followed by pandering and guesswork by politicians, have driven criminal justice policy for too long. Enough marmalade-dropping panic: time to give cost-benefit analysis a go.

You can review the CEA Report