Did Oregon’s tough mandatory sentencing law “measure 11” improve public safety? New evidence about an old debate from a multiple-design, experimental strategy

Abstract

This research contributes to policy debates about whether mandatory sentences improve public safety and are responsible for maintaining lower rates of crime. The current study used two quasi-experimental approaches, regression point displacement (RPD) and interrupted time series (ITS), to test the effect of Oregon’s Measure 11 on violent and property crimes. The RPD study found that, between 1995 and 1998, Measure 11 was associated with significant increases in most crime rates but decreases in the rates of rape and murder. These findings were not replicated in the more rigorous ITS, which found that Measure 11 had no statistically discernible effect on the incidence of the index crimes in the state of Oregon. Overall, the findings indicate Measure 11 had little to no marginal benefit relative to policies in place before the law was implemented.

Keywords:

• Mandatory sentencing
• deterrence
• sentencing reform
• interrupted time series
• design
• regression point displacement design

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 A full description of the original Measure and the modified versions can be found in Oregon Revised Statute at ORS 137.700, ORS 137.707, and ORS 137.712.

2 We derived this list of “control” states from Harmon (2011, 2013). We also cross-checked for reforms enacted during the 1990s in these states through official legislative reports and the Bureau of Justice Statistics National Corrections Reporting Program.

3 Including other covariates that are related to crime in the model would be redundant with the pretest and reduce the precision of the test. If part of an omitted variable is not controlled by the pretest (e.g., an economic or other shock that occurred between the pre and post-test), this variation becomes part of the residual and will only confound if it is also correlated with the treatment (see Shadish et al., p. 243–45).

4 The following outliers were identified: murder = Dec 1995; rape = Dec 1998, Dec 2000, Oct 1991, March 1993, and Aug 1991; robbery = Oct 1990 and Jan 1989; assault = Dec 1998; burglary = Jan 1989, March 1989, Aug 1989; motor vehicle theft = Jan 1995; and larceny = Dec 1998, July 1990, Jan 1997, and Jan 1996. Replacing outliers in the time series provides a better estimate of the average effect of interventions. Once the outliers were replaced, the distributions of the time series were approximately normal.

Additional information

Notes on contributors

Jody Sundt

Jody Sundt is Professor and Chair of the Department of Criminal Justice at the University of North Texas. Her research focuses on the effectiveness of correctional policy, the professional development of correctional employees, and public attitudes about crime and punishment. Her work appears in Criminology, Criminology and Public Policy, Justice Quarterly, Crime and Delinquency, and Criminal Justice and Behavior.

Breanna Boppre

Breanna Boppre is an Assistant Professor in the School of Criminal Justice at Wichita State University. Her research examines correctional policies and women's system-involvement. Her work appears in the International Journal of Offender Therapy and Comparative Criminology, Corrections Policy, Practice, and Research, Victims & Offenders, and The Journal of Ethnicity in Criminal Justice.