Rates of recidivism were examined for DWI offenders arrested in each of the years 1997-1999. In accordance with the definition of DWI offense used in this chapter, recidivism was defined as a subsequent arrest that resulted in an administrative penalty, a conviction in court or both, as recorded on the driver license files. For persons with more than one offense in a given year, the focal offense was the first offense that occurred in that year. For persons arrested in 1997 and 1998, first subsequent offenses were tallied for each of the 24 months following the month of arrest. For persons arrested in 1999, first subsequent offenses were tallied for each of the 12 months following the arrest. (The discussion of results focuses on the recidivism rates after one year and after two years, which are shown in Italics font in Tables 9-11). As specific BAC information was unavailable before 1998, comparisons of recidivism before and after the law focused on total first-time offenses, total repeat offenses, offenses involving test refusals, and offenses involving test failures.
For each of the years 1998 and 1999, the rates of recidivism were examined for repeat and first-time offenses for offenders who refused the alcohol test, had BACs at or above .20, and had BACs less than .20. In addition, the rate of recidivism among high-BAC offenders (BAC > .20) was contrasted to the rate among a “comparison” group defined as persons with BACs .17-.19. These offenders represent persons who also recorded a “high” BAC but should have been unaffected by the law.
Differences in rates of recidivism were examined with the chi-square statistic (p < .05). Survival analysis was used to examine the association between the rate of recidivism and the results of the alcohol test, in the presence of other predictors.
As shown in Tables 9-10, the overall rates of recidivism after one year were significantly lower in 1998 than in 1997 for total first-time offenders (6.7 percent vs. 7.3 percent, X2 = 6.51, df = 1, p = .01) and total repeat offenders (7.9 percent vs. 9.0 percent, X2 = 6.51, df = 1, p = .01). The rates of recidivism among first-time offenders after two years were similar for the years 1997 (12.9 percent) and 1998 (12.6 percent), and did not reach statistical significance. The rates of recidivism among repeat offenders after two years were significantly lower in 1998 than in 1997 (15.7 percent vs. 17.5 percent, X2 = 9.74, df = 1, p = .002). The overall patterns of the one-year recidivism rates for first-time offenders and repeat offenders arrested in 1999 were similar and not statistically different from the rates in 1998.