No matter how much we pat ourselves on the back about the recent declines
in U.S. crime rates, the fact remains that crime is about three times
as prevalent in America today as it was as recently as 1960. Even now,
our homicide rate remains at about five times the level of that in the
rest of the developed world.
The consequences of high crime rates go far beyond the numbers. Much
of America's social geography is dictated by the desire to avoid victimization.
What else could explain the coexistence of high housing prices and residential
abandonment in the same metropolitan areas? Everyone knows that poverty
leads to crime; less attention is paid to the ways that crime leads to
poverty by driving jobs and services away from the places poor people
live.
Given how much nonsense office-seekers are forced to talk when they discuss
crime, its declining salience as a political issue is a welcome development.
But controlling crime remains a central challenge for American society;
we continue to face the threat that an upsurge in crime (such as the one
the United States suffered in the 1960s or the one Western Europe has
been suffering recently) will give anti-progressive political forces another
great political opportunity.
So what is to be done? Arguments about crime control tend to focus on
how to change the behavior of criminals. In some ways, that's the easy
part; criminals respond when their behavior has immediate and predictable
consequences. The hard part is changing the behavior of officials to make
those things happen. Reducing the crime rate is more a public management
problem than a criminology problem. And one key to management success,
as we're learning from other areas of public policy, is performance accountability.
Nowhere has the accountability principle been more decisively demonstrated
than in New York City's declining crime rates. The crime drop had many
causes, both inside and outside the police department. But the one that
stands out was the bold announcement by Police Commissioner Bill Bratton
that he was willing to be held quantitatively accountable for crime
reduction. Putting his own reputation on the line -- and betting it
on measurable results rather than on a laundry list of specific policies
or organizational changes -- enabled Bratton to hold his subordinates
similarly accountable. Faced with the pressure to perform and to perform
in the foreseeable future rather than in the sweet by-and-by, the New
York Police Department found itself capable of astonishing accomplishments.
The rate of murders went down by three-quarters and is staying down; property
crime rates kept pace.
For some reason, no one seems to have thought of extending this accountability
revolution to corrections agencies. Yet holding corrections officials
accountable for recidivism among their probationers, parolees, and prison
alumni is more obviously justified than holding the police responsible
for the overall crime rate. Many things influence the crime rate in a
city beyond the effectiveness of its police department. But if arrests
for serious new crimes are much less frequent among probationers in one
probation office than among similar probationers in another office, or
if the graduates of one medium-security prison are much more likely to
be back inside within six months than the graduates of another, then the
worse-performing institution might have something to learn from the better-performing
institution. Since most crime, especially serious violent crime, is committed
by repeat offenders, focusing on the offenders whose names we know needs
to be central to any serious crime control strategy.
Current performance wouldn't be hard to improve on. In most states, an
offender leaving prison has a worse-than-even chance of staying out for
three years, and there's no evidence that routine probation supervision
has any crime-control benefit whatsoever. Some specific programs are known
to work. Frequent drug tests for probationers and parolees, with immediate
and predictable, but not severe, sanctions for every incident of drug
use, can reduce hard-drug consumption, a strong predictor of recidivism.
In-prison literacy programs can improve employability, and getting a job
is a strong correlate of going straight.
Part of the reason such commonsense approaches aren't widely adopted
is that the whole project of turning offenders' lives around -- inducing
them to live as citizens rather than predators -- was tainted by its
association with a kindhearted but muddle-headed version of penology.
When crime was low, some influential prison officials and academics tried
to reinvent prisons and probation and parole agencies as primarily service
institutions. The announced goal -- rehabilitation -- became entangled
with what turned out to be a fundamentally wrong idea of how to achieve
it, which might be summed up as "be nice to criminals and criminals
will be nice to you."
Over the past 35 years, a political sea change, driven by rising crime,
made a hard line toward offenders the only publicly acceptable attitude.
Naturally, but unfortunately, this led to the virtual abandonment of the
idea that rehabilitation ought to be a central focus of prison administration.
Somehow, giving up the idea of being nice to criminals came to imply giving
up any intention of making them other than criminal. To talk of rehabilitation
now marks a politician or corrections official as an unrealistic, soft-on-crime
bleeding heart.
The adversarial nature of the courtroom process fosters the illusion
that criminal justice is a zero-sum game. In fact, since both crime and
punishment are expensive, offenders and the rest of us have a mutual interest
in offenders' turning from crime to honest work. In particular, those
who portray themselves politically as advocates for the victims of crime
ought to be more concerned than they seem to be with the fact that unrehabilitated
criminals are future threats. Yet in the current political climate, rehabilitation,
in the hardheaded sense of reducing future criminality, plays virtually
no role in the design, execution, or evaluation of corrections programs.
At the same time, imprisonment has exploded. We now have about 2 million
people behind bars, about four times the incarceration rate of a generation
ago, a level unprecedented in American history and unmatched in the developed
world. Poor urban communities are being hit with a flood of prison returnees,
and that flood is beginning to show up in rising crime rates, but no one
is turning to the wardens and demanding that they become part of the solution.
Perhaps surprisingly, the private prison industry has utterly failed
to demonstrate either cost savings or reduced recidivism rates, other
than by selecting the least-cost, least-likely-to-repeat offenders. Recidivism
ought to be an easy case for incentive contracting, but that isn't the
way private-prison contracts are drawn. And for some reason nonprofit
groups have almost no participation in adult institutional corrections,
though they play very important roles in dealing with juvenile offenders
and in providing some post-release programming. A little not-for-profit
competition based on recidivism rates might help jump-start the interest
of the public systems in installing one-way exits instead of revolving
doors.
Of course, any contracting-out approach confronts its own challenges:
building accountability into the contracts and (even harder) getting that
accountability enforced in the face of the political muscle that contractors,
whether for-profit or not-for-profit, virtually always wield.
But whatever we do about prison privatization, the first step in reforming
corrections is measuring the crime control performance of the existing
corrections agencies.
The case for accountability doesn't rest on any specific theory of what
would bring about rehabilitation. It rests on two basic propositions:
that recidivism isn't completely insensitive to the way our corrections
agencies conduct themselves, and that the people who work in those agencies
and run them have the skill and ingenuity to make better things happen,
given the appropriate authority and the appropriate incentives. Being
accountable for next month's numbers concentrates the organizational mind
wonderfully; it can, at best, create a sort of every-cat-is-good-as-long-as-it-catches-mice
pragmatism. Real accountability can make even ideas "not invented
here" seem worth trying once.
Formally, the problem is similar to the problem of bringing accountability
to public education. Consider the parallels:
Both the tasks are important and both are done unevenly and less than
optimally.
In each case, it's not hard to figure out some key things we ought to
be measuring: reading ability in the schools and crimes committed by offenders
in the corrections case.
In neither case are those easy-to-measure things the only aspects of
performance worth worrying about (we should consider curiosity among students,
for example, or prisoner-on-prisoner assault, or the employment rate among
probationers), and therefore both systems require multiple measures of
process and outcome to avoid creating perverse incentives for managers
(for example, to drop art and music classes from schools or to ignore
infectious disease and other health issues in correctional settings).
Since both students and offenders differ enormously among themselves
and across institutions, any reasonable accountability system in either
domain needs to adjust for the client mix or we just wind up punishing
the institutions that get stuck with the hard cases.
Random variation makes it both essential and hard to develop clever
statistical measures to make sure we're not rewarding and punishing "noise"
in the test scores or the crime rates.
There's a tension between the need for adequate sample sizes to deal
with the "noise" problem and the desirability of moving accountability
down to the lowest practicable level of the organization. The more we
can hold individual teachers or probation officers accountable for results,
the better; but the measurement problems get harder as the numbers of
cases get smaller.
Dukenfield's Law ("Anything worth winning is worth cheating for")
makes it necessary to guard against a variety of ways to game the scoring
system.
Accountability is no panacea. Probation departments, in particular, are
so badly starved for both resources and authority that dramatic improvements
can't really be expected without changes in budgets and laws. But right
now the link between probation budgets and crime rates hasn't been established.
In the absence of outcome measures that elected leaders care about, making
the political case for the necessary changes will remain virtually impossible.
Mayors must learn to ask, and probation commissioners must be able to
answer the question, "How much will it cost me to get the re-arrest
rate down by 10 percent, and what's that going to do for the overall crime
rate?" When they can do that, the battle won't be won, but at least
we'll be ready to start fighting it.
