Interlocks for First Offenders: Effective?

Abstract

Objective. Vehicle interlocks have been shown to effectively reduce the recidivism of multiple driving-while-impaired (DWI) offenders; however, the evidence for their effectiveness with first offenders has been mixed. Two Canadian studies found that the installation of an interlock reduced first DWI recidivism, but U.S. studies in West Virginia and California failed to find a significant reduction in recidivism for first DWI offenders in interlock programs. The objective of this study was to determine the extent to which such devices were effective with first offenders in New Mexico.

Methods. This study compared 1,461 first offenders, who installed interlocks in New Mexico between January 1, 2003, and December 1, 2005, with 17,562 first offenders convicted during the same period who did not install the units. Cox multivariate proportional hazards regression (CMVPHR) was used to compare recidivism rates during three periods:

1. while the interlock was on the vehicles of offenders who installed them,

2. after those offenders removed the units until the end of the study period (approximately 2 years), and

3. for the combined period (both while the interlock was installed and after it was removed).

Results. While the device was on the vehicles of the interlock group, their recidivism rate, 2.6% per year of exposure, was significantly less than the 7.1% per year rate of the comparison group (CMVPHR hazard ratio = 0.39, p < 0.0001). After the device was removed, the annualized recidivism rate of the interlock group increased to 4.9% per year of exposure, which was less than the 6.7% rate of the comparison group, but the hazard ratio was not statistically significant (CMVPHR hazard ratio = 0.82, p = 0.16). When the combined periods (interlock on and off) were considered, the interlock group had a recidivism rate of 3.9% per year, which again was significantly lower than the 6.8% rate for the comparison group (CMVPHR hazard ratio = 0.61, p < 0.0001).

Conclusion. The study provides evidence that interlocks are as effective with first offenders (approximately 60% reduction in recidivism when on the vehicle) as they are for multiple offenders. In addition, the benefits of requiring an interlock for first offenders exceed the costs by a factor of three.

Keywords:

• Alcohol
• Interlock
• DUI
• DWI
• First Offender

ACKNOWLEDGMENTS

This work was supported by the National Highway Traffic Safety Administration (contract number DTNH22-02-D-95121, task order 09), the National Institute on Alcohol Abuse and Alcoholism (grant number K05 AA014260), and by the Robert Wood Johnson Foundation Substance Abuse Policy Research Program (grant number 52251). The authors also express thanks to those who provided access to the Citation Tracking System and the Interlock Installation and Removal data: Ken Ortiz, Director of the New Mexico Motor Vehicle Department; Michael Sandoval, Bureau Chief of the New Mexico Traffic Safety Bureau; and James Davis, Director of the University of New Mexico Division of Government Research.

The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the funding agency.

Notes

¹Many first offenders did not install interlocks under the 2003 law because it only required interlocks on all vehicles driven by first aggravated offenders. Aggravated includes those who refuse to blow, those with a BAC ≥ 0.16, and those who are involved in an injury accident. The interlock mandate was optional (up to the judge) for first offenders with a BAC < 0.16. Also the wording allowed offenders to escape the mandate by claiming “no vehicle” or “not driving.” Finally, many aggravated DWI offenders are pled down to nonaggravated . These loopholes were partially closed in June 2005, when interlock licenses became mandatory for all convicted impaired drivers because, to get an interlock license, an offender had to have evidence of having installed an interlock.

²Derivation of formula: Multiplying the numerator and denominator of the formula by the number of persons in the treatment group gives (N_TU − N_TT) ∗ b/(N_T ∗ c), which is the total benefit, B, divided by the total cost, C. The term N_TU is the number of rearrests expected in the untreated group if it had not been treated.

³The Federal Highway Administration's estimate of the economic impact of drunk driving in 2003 in New Mexico was $1,005,000,000 and the number of DWI arrests was 20,313. So the economic impact per DWI arrest prevented is ∼ $50,000.

⁴Interlocks cost offenders about $1,000/year, and this first-offender interlock group had them installed for an average of 0.54 years. So the average cost of treatment, c = $540.