شروع بزهکاری زودرس و خط سیر اختلال در رانندگی الکلی در میان مردان جوان

Abstract Building upon the literature in developmental and life-course criminology, the present study assesses the possible association of age onset of delinquency with the trajectory of alcohol-impaired driving using data collected from the three waves of the Buffalo Longitudinal Survey of Young Men (BLSYM). It is argued that as a unique form of delinquency, alcohol-impaired driving among adolescents may be better understood in a broad context of adolescent delinquency involvement. The study adopts the general approach for the analysis of early onset of delinquency and criminal careers in developmental and life-course criminology and hypothesizes that early onset of delinquency is associated with a higher growth of alcohol-impaired driving over time among adolescents when age onsets of alcohol-impaired driving, drinking, and drug use are controlled. Our analysis with the HLM growth modeling method provides support for the hypothesis. Respondents who had an early start in delinquency were likely to have a faster growth of alcohol-impaired driving over the three waves of BLSYM, which implies that these respondents were likely to have a longer path of alcohol-impaired driving in their transition to adulthood. The implication of this finding is discussed.

. Introduction Alcohol-impaired driving among young people is still a serious public concern in the United States. The National Highway Traffic Safety Administration (2008) reported that in 2007 more than two-thirds of drivers in fatal crashes who had a BAC (Blood Alcohol Content) level of .08 or higher were young drivers at ages 21 to 35 (35% at ages 21 to 24 and 29% at 25 to 34, respectively). Because alcohol-impaired driving involves drinking behavior, a number of studies have examined association of adolescent daily drinking patterns with the risk of alcohol-impaired driving (e.g., Barnes and Welte, 1988, Berger and Snortum, 1986, Bingham and Shope, 2004, Calvert and Bucholz, 2008, Copeland et al., 1996, Donovan, 1993, Lapham et al., 2006, Usdan et al., 2005 and Valencia-Martin et al., 2008). Some studies have also assessed the possible effects of other drugs such as marijuana and smoking on alcohol-impaired driving (e.g., Bingham and Shope, 2003, O'Malley and Johnson, 2007 and Sabel et al., 2004). The common wisdom is that drinking, drug use, and smoking may be different types of substance use that could be interrelated as significant factors in alcohol-impaired driving. However, alcohol-impaired driving per se is criminal behavior and both drinking and driving have legal limits. Underage drinking or driving is illegal. Therefore, alcohol-impaired driving as a delinquent or criminal behavior may be associated with other delinquent or criminal acts. It may be part of a large syndrome of antisocial behavior. Studies have indicated that offenders tend to be versatile in delinquent or criminal acts ( Farrington, 1991 and Farrington, 2003). Given these considerations, it may be strategic to study alcohol-impaired driving among young people in a larger context of their delinquency involvement when they are making a transition to adulthood. The present study is aimed to assess the possible association of early onset of delinquency and the trajectory of alcohol-impaired driving among young males using data collected from the Buffalo Longitudinal Survey of Young Men (BLSYM). Because both drinking and driving have legal limits and their implementation largely depends on available resources and related skills, alcohol-impaired driving commonly start later than other delinquent acts among adolescents. It would be important to examine whether early onset of delinquency affects the pathway of alcohol-impaired driving in understanding the developmental trajectory of alcohol-impaired driving among young people.

. Results The results are reported in Table 3. The table has two models, one for the growth rate of alcohol-impaired driving and the other for the growth curve of general delinquency. Each model has unconditional and conditional submodels. The results for the unconditional submodel of alcohol-impaired driving show that there is a significant increase in the average rate of alcohol-impaired driving across the three waves (b = 0.15). The results for the conditional submodel indicate that the age onset of delinquency significantly affects the growth rate of alcohol-impaired driving. If respondents had an early onset of delinquency, they were likely to have a faster increase of alcohol-impaired driving (b = − 0.01) across the three waves. The age onset of alcohol-impaired driving also has a significant effect on the average growth rate of alcohol-impaired driving across the waves. However, the direction of the effect is opposite. Respondents who had an early onset of alcohol-impaired driving were likely to have a slower increase in the average rate of alcohol-impaired driving across the three waves. Interestingly, age onsets of drinking and drug use have no effects on the growth curve of alcohol-impaired driving. Overall, the significant effects of age onsets of delinquency and alcohol-impaired driving explain about 47% of the variance in the growth rate of alcohol-impaired driving across the three waves.3 Table 3. Slope models of early onset of delinquency and the trajectories of alcohol- impaired driving and delinquency. Model Alcohol-impaired driving Delinquency Variable Unconditional model Conditional model Unconditional model Conditional model Slope across three waves 0.15** (0.02) − 0.79** (0.27) − 1.83** (0.15) − 13.35** Age onset of delinquency ------ − 0.01* (0.004) ------ 0.17** (0.03) Age onset of DWI ------ 0.06** (0.02) ------ 0.24* (0.11) Age onset of drinking ------ − 0.01 (0.01) ------ 0.13* (0.06) Age onset of drug use ------ − 0.004 (0.01) ------ 0.14* (0.06) Race (White) ------ 0.06 (0.04) ------ − 0.26 (0.29) Family SES ------ 0.01 (0.01) ------ 0.15 (0.09) Random effect Variance component Variance component Variance component Variance component Intercept1 U0 0.71** 0.38** 69.19** 44.72** Level-1 R 0.89 0.83 54.14 50.47 Notes: Standard errors are presented in parentheses. N = 563. *p < .05 **p < .01. Table options The results for the unconditional submodel in delinquency trajectory show that there is a significant decline in the average rate of delinquency across the three waves (b = − 1.83). The results for the conditional submodel indicate an opposite finding for the effect of age onset of delinquency on the delinquency trajectory (b = 0.17). Respondents who started delinquent acts at an early age were likely to have a faster decline in delinquency, which is difficult to interpret and inconsistent with previous studies. To explore this unusual finding, we conducted additional analyses. We broke down the variable of age onset of delinquency into three groups 1 = age onset group below 10 years old, 2 = age onset group of 11–14 years old, and 3 = age onset group of 15 years old or above. Using this new variable, we assessed and compared the mean scores of delinquency across the three waves for these three groups. For a comparative purpose, we also conducted the same analyses of the mean scores of alcohol-impaired driving across the three waves for the three onset groups. The results are presented in Table 4. Table 4. Mean scores of delinquency and alcohol-impaired driving at each wave by age onset groups of delinquency. Mean scores Delinquency Alcohol-impaired driving Onset group Wave 1 Wave 2 Wave 3 Wave 1 Wave 2 Wave 3 Below 10 years old n = 188 18.84 12.10 10.08 0.62 0.93 1.19 11–14 years old n = 211 14.08 8.95 8.26 0.66 0.71 0.96 15 years old or above n = 164 5.96 5.02 4.31 0.26 0.48 0.57 Note: N = 563. Table options The results in Table 4 indicate that for the comparison of delinquency mean scores across the three waves, the age onset group below 10 years old has the highest mean score (mean = 18.84) at the first wave of the survey, followed by the age onset group of 11–14 years old (mean = 14.09). This group indeed has larger drop of delinquency across the three waves (mean = 12.10 at Wave 2 and 10.08 at Wave 3) than other two groups because of the highest starting point at Wave 1. Although it has a larger drop, the group still has a highest average level of delinquency at Wave 3 (mean = 10.08 compared to mean = 8.26 for the age onset group of 11–14 and 4.31 for the age onset group of 15 years old or above). The latest age onset group of 15 years old or above does not have a significant drop of delinquency across the three waves because it has the lowest starting level of delinquency at Wave 1 (mean = 5.96) and remains a relatively flat, low level of delinquency across the three waves (mean = 5.02 at Wave 2 and 4.31 at Wave 3).4 It is interesting to compare these results with those for alcohol-impaired driving across the three waves. As Table 4 shows, the earliest age onset group of delinquency has much more significant increase of alcohol-impaired driving across the three waves (mean = 0.62 at Wave 1, 0.93 at Wave 2, and 1.19 at Wave 3) than the latest age onset group of 15 years old or above (mean = 0.26 at Wave 1, 0.48 at Wave 2, and 0.57 at Wave 3). In comparison, this group also has a much higher starting level of alcohol-impaired driving than the latest age onset group (mean = 0.62 compared to 0.26 at Wave 1). Finally, the results in Table 3 also indicate that the age onsets of alcohol-impaired driving, drinking, and drug use have similar effects on the delinquency trajectory as the age onset of delinquency does. The earlier age onset of alcohol-impaired driving, drinking, or drug use is likely to have a faster decline of delinquency across the three waves. We conducted similar analyses to explore these opposite findings as we did for the effect of age onset of delinquency on the delinquency trajectory across the three waves. The results are similar to what we have presented and discussed above for the effect of age onset of delinquency. Overall, about 35% of the variance in the declining rate of delinquency is explained by the age onsets of delinquency, alcohol-impaired driving, drinking, and drug use.