NCPA - National Center for Policy Analysis

Does School Choice Reduce Crime?

April 4, 2012

In measuring the effectiveness of current education policies, many studies immediately turn to traditional indicators such as test scores.  However, reliance on this proxy for student outcomes as a result of a given reform is irresponsible: many researchers have shown only a weak correlation between test scores and future success, and test scores are also easily manipulated by teachers, says David J. Deming, assistant professor of education at the Harvard Graduate School of Education.

Recognizing this issue, a new study of the Charlotte-Mecklenburg school district in North Carolina sought to discover the effects of school choice on crime among students.  Following the progress of students from 2003 to 2009, the study compared those students randomly selected (via lottery) to attend a chosen school to those who sought selection but were not chosen.

Broadly, the study found that the chance to attend a chosen school did have sweeping impacts on crime:

  • Among the crimes of property felonies, violent felonies and drug felonies for each of middle school and high school subjects (six categories total), lottery winners had committed fewer crimes than lottery losers in four categories.
  • Among the statistically significant results was that the average number of drug felonies per student fell from 0.33 for high school lottery losers to 0.10 for lottery winners.
  • Additionally, the incidence of violent felonies among middle school lottery losers was 0.25, while it was only 0.08 for lottery winners.
  • The only statistically significant category that suggested that lottery winners are involved in more crime was property felonies among middle school students: 0.24 for lottery losers, but 0.48 for lottery winners.

The differential between lottery winners and lottery losers at both the high school and middle school levels proves more significant when its broad implications are considered.  In terms of aggregate prison sentences and social costs of crimes, lottery winners performed far better.

  • Social cost of crimes for high school lottery winners and losers were $7,084 and $12,500, respectively.
  • Similarly, the cost of crimes for middle school lottery winners was only $4,657, compared with $11,000 for lottery losers.
  • Additionally, the total expected jail sentence falls from 58.6 months for high school lottery losers to 35.5 months for lottery winners.
  • This pattern continues for middle school students, who would receive sentences of 48.3 and 17.3 for lottery losers and winners, respectively.

Source: David J. Deming, "Does School Choice Reduce Crime?" Education Next, Spring 2012.

For text:


Browse more articles on Education Issues