Predictors of conviction and sentencing of DWI offenders in a New Mexico County

OBJECTIVE: To analyze predictors of conviction and dismissal of individuals charged with DWI, and predictors of the sentences of those who are convicted. METHODS: Data come from the Citation Tracking System of the State of New Mexico's Motor Vehicle Division and includes information on all individuals who were arrested for DWI in San Juan County between August 1994 and December 2000. Independent variables were: age, gender, race/ethnicity, waiver of right to an attorney, court of arraignment, year of arrest, BAC, and number of prior DWI arrests. Dependent variables were: (1) conviction or dismissal, (2) jail or no jail, (3) incarceration/treatment or not, (4) fine or no fine, (5) length of jail sentence, (6) waived right of attorney, and (7) magnitude of fine. Multiple linear and logistic regression was used in the analyses. RESULTS: Use of an attorney is associated with reduced likelihood of conviction and, if convicted, in reduced likelihood of jail sentence and reduced jail time, but greater likelihood and magnitude of a fine. Native Americans were more likely than Hispanics and non-Hispanic whites to waive their right to an attorney. Native Americans were most likely to be sentenced to the detention/treatment program. BAC and number of prior arrests were each positively associated with increased likelihood of conviction and more severe sentences. There is also substantial variability in severity of sentencing among courts. CONCLUSION: Likelihood of conviction and severity of sentences are both determined by extra-legal factors, resulting in inconsistent application of the law. This may in turn contribute to a lack of compliance with laws related to DWI.     

Variations in jail sentences and the probability of re-arrest for driving while intoxicated

OBJECTIVE: The probability of re-arrest for driving while intoxicated (DWI) is compared for four different groups of individuals classified by whether the individual was convicted and, if so, the type of sentence received. METHOD: Subsequent re-arrests for DWI were examined for all individuals whose index arrest for DWI had occurred between 1994 and 2001 in a county in New Mexico. The groups included (1) those convicted as a result of the index arrest and sentenced to a 28-day jail/treatment program (N (#)=(#) 2,703); (2) all those not convicted as a result of the index arrest (N = 709); (3) those who were convicted but not sentenced to jail (N = 1,047); and (4) those convicted and sentenced to jail (N = 1,290). RESULTS: Adjusting for covariates of BAC, number of prior arrests, ethnicity, age, and sex, the probability of not being re-arrested was greatest among those sentenced to the jail/treatment program (Group 1), next highest in the two groups convicted but not sentenced to jail/treatment (Groups 3 and 4), and lowest in the group that was not convicted (Group 2). Length of jail sentence among people convicted but not sentenced to the jail/treatment program was unrelated to the probability of re-arrest. CONCLUSION: Conviction for DWI, regardless of the sentence, appears to reduce the probability of re-arrest, and being sentenced to a multi-modal treatment/incarceration program further reduces the probability of re-arrest. However, the other types of sentences do not appear to differ in their impact on probability of re-arrest.