AFR - Good Schools, Less Crime

Australian Financial Review, 20 July 2010

Notorious gang rapist Bilal Skaf dropped out of school at age 14. Convicted murderer Neddy Smith writes in his autobiography that he last saw the inside of a classroom at age 13. Port Arthur killer Martin Bryant could not read or write when he left school. In Australia's jails, the typical prisoner has under 10 years of education - substantially less than the general population.

Could improving school quality help cut the crime rate? To date, the evidence has been pretty limited. We know from the Perry Preschool program that high-quality early childhood programs for disadvantaged children can reduce crime rates later in life. But some commentators have argued that by the time children start school, the window for life-changing interventions has closed.

Enter David Deming, an assistant professor at Carnegie Mellon University in Pennsylvania, whose thesis work exploits two aspects of American social policy. First, over-subscribed US schools frequently use a lottery system to allocate scarce places (on the basis that this is fairer than other alternatives such as first-in, best-dressed). Second, many US cities have a policy of posting arrest records online (yes, that's where those mugshots of Hugh Grant and Lindsay Lohan came from).

Using data from a county in North Carolina, Deming cleverly matched school lottery results with online arrest records. This allowed him to compare otherwise identical students who had won or lost a school choice lottery. Competition for the best-performing schools was fierce, with around 40 percent of students applying to a school where they were not guaranteed a spot.

Success in the school choice lottery, Deming shows, had a substantial impact on a youth's criminal behaviour. Focusing on the group of students who were most at risk of committing crimes, he finds that winning a school choice lottery reduced the number of felony arrests from 8 arrests per 10 students to 4 arrests for every 10 students. In the seven years after the school choice lottery took place, high-risk students in the losing group spent an average of 56 days in jail, while those in the winning group averaged 17 days in jail. This is particularly striking given that both groups continued to live in the same neighbourhoods.

These results back up the old adage that great schools can save us the trouble of building more jails. But fewer jails is only part of the public benefit of cutting crime. The largest cost of crime to the community is its impact on victims. To put a dollar value on this, Deming uses figures such as jury awards, value of life estimates and victims' medical bills to calculate an approximate social cost for each crime. Crunching the numbers, he estimates that for high-risk youth, each year of enrolment in a first-choice school saves society over US$55,000 in criminal victimisation costs. Simply put, great schools for disadvantaged youth are a bargain for the community.

For the US, Deming argues that the worst schools are not only dangerous on a day-to-day basis - they may also be doing long-term damage to their students. Education administrators, he argues, should provide students in these environments with the chance to move to better schools. As Deming puts it: "For high risk youth on the margins of society, public schools may present the best opportunity to intervene."

Supporting evidence for these findings comes from another study looking at the impact of school desegregation in the 1960s and 1970s. Researchers have long recognised that desegregation had the effect of raising the quality of schools attended by the most disadvantaged students. Now, new work led by David Weiner (University of Pennsylvania), has shown that court-ordered desegregation in the nation's largest school districts led to a significant drop in the homicide rate. Weiner's team estimate that school desegregation reduced youth homicide by about one-tenth. They conclude that some of the best crime-prevention strategies may have nothing to do with the criminal justice system.

In Australia, much of the debate over school quality to date has focused on the impact that a terrific education can have on raising productivity and participation. But Deming's North Carolina results show that for the most disadvantaged students, education also offers a path away from crime. This implies that programs such as the federal government's $1.5 billion program to improve low-SES schools may have an unanticipated payoff. Great schools may be the most effective social policy we know.