Minimum Legal Drinking Age and Crime in the United States, 1980-1987

This collection focuses on how changes in the legal drinking age affect the number of fatal motor vehicle accidents and crime rates. The principal investigators identified three areas of study. First, they looked at blood alcohol content of drivers involved in fatal accidents in relation to changes in the drinking age. Second, they looked at how arrest rates correlated with changes in the drinking age. Finally, they looked at the relationship between blood alcohol content and arrest rates. In this context, the investigators used the percentage of drivers killed in fatal automobile accidents who had positive blood alcohol content as an indicator of drinking in the population. Arrests were used as a measure of crime, and arrest rates per capita were used to create comparability across states and over time. Arrests for certain crimes as a proportion of all arrests were used for other analyses to compensate for trends that affect the probability of arrests in general. This collection contains three parts. Variables in the Federal Bureau of Investigation Crime Data file (Part 1) include the state and year to which the data apply, the type of crime, and the sex and age category of those arrested for crimes. A single arrest is the unit of analysis for this file. Information in the Population Data file (Part 2) includes population counts for the number of individuals within each of seven age categories, as well as the number in the total population. There is also a figure for the number of individuals covered by the reporting police agencies from which data were gathered. The individual is the unit of analysis. The Fatal Accident Data file (Part 3) includes six variables: the FIPS code for the state, year of accident, and the sex, age group, and blood alcohol content of the individual killed. The final variable in each record is a count of the number of drivers killed in fatal motor vehicle accidents for that state and year who fit into the given sex, age, and blood alcohol content grouping. A driver killed in a fatal accident is the unit of analysis.

本数据集聚焦于法定饮酒年龄的变化如何影响致命机动车事故数量及犯罪率。主要研究者确定了三个研究方向：其一，分析涉及致命事故的司机血液酒精含量（Blood Alcohol Content）与饮酒年龄变化的关联；其二，探究逮捕率与饮酒年龄变化的相关性；其三，考察血液酒精含量与逮捕率之间的关系。

在此背景下，研究者将致命车祸中血液酒精含量呈阳性的死亡司机占比作为人群饮酒情况的指标；将逮捕数作为衡量犯罪的指标，并采用人均逮捕率以实现州际及跨时间的可比性；将特定犯罪逮捕数占总逮捕数的比例用于其他分析，以弥补影响总体逮捕概率的趋势性因素。

本数据集包含三个部分：
1. 联邦调查局（Federal Bureau of Investigation）犯罪数据文件：变量包括数据适用的州与年份、犯罪类型、被逮捕者的性别及年龄类别，分析单位为单次逮捕；
2. 人口数据文件：包含七个年龄组各自的人口数量统计及总人口数，同时记录了提供数据的报告警方机构所覆盖的个体数量，分析单位为个体；
3. 致命事故数据文件：包含六个变量——州的联邦信息处理标准代码（FIPS code）、事故年份、死亡个体的性别、年龄组及血液酒精含量，每条记录的最后一个变量为该州和年份中符合给定性别、年龄及血液酒精含量分组的致命机动车事故死亡司机人数，分析单位为致命事故中的死亡司机。