The Minimum Legal Drinking Age: a Time-Series Impact Evaluation.

After Michigan reduced its legal drinking age from 21 to 18 in 1972, several studies reported significant increases in alcohol-related traffic crashes among drivers affected by the lower drinking age. Michigan raised the drinking age from 18 to 21 in December, 1978. This investigation measures the effect of the 1978 increase in legal drinking age on the frequency of alcohol-related motor vehicle crash involvement among young drivers. A 20 percent r and om sample of all reported motor vehicle crashes in the State of Michigan from 1972 through 1979 is analyzed, stratified by the condition of the driver as reported by police officers ("had been drinking" vs. "had not been drinking"). An alternative indicator of alcohol-related crash involvement is also examined, consisting of late-night, single-vehicle crashes with a male driver. Monthly time-series of these measures are analyzed for drivers aged 16-17, 18-20, 21-24, and 25-45 using Box-Jenkins time-series analysis. The analyses controlled for long term trends, seasonality, and other patterns in crash frequency over the baseline 1972 through 1978 period, expected to influence the 1979 data. The frequency of police-reported "had been drinking" crash involvement among 18-20 year old drivers decreased 31 percent when the drinking age was raised to 21; the frequency of late-night, single-vehicle crashes with male drivers was down 18 percent after the increase in legal drinking age. Alcohol-related crash involvement for drivers aged 16-17 also decreased after the drinking age was raised, but the results are less conclusive than the findings for drivers aged 18-20. Reductions in general non-alcohol-related crash involvement, for all age groups, were much smaller than the substantial reduction in alcohol-related crash involvement among drivers aged 18-20. As a result, the alcohol-related crash reductions among 18-20 year old drivers cannot be attributed to factors causing a decrease in general non-alcohol-related crashes, such as an economic recession, fuel shortages, and fuel price increases. It is concluded that the legal minimum drinking age has a significant effect on alcohol-related crash involvement among young drivers. Recommendations for further research, and implications of the findings for beverage alcohol availability theory and public policy concerning the prevention of alcohol-related problems are included.