Transitional Aid Research Project (TARP), 1976-1977

The Transitional Aid Research Project (TARP) was a randomized field experiment conducted in Texas and Georgia in 1976-1977 that was designed to reduce recidivism among ex-prisoners by lowering incentives for re-engaging in property crime through provision of minimal levels of income support and extension of some unemployment insurance coverage to released prisoners. This study evolved out of an earlier LIFE (Living Insurance for Ex-Prisoners) study conducted in Baltimore, Maryland in the early 1970s. In the LIFE study, 500 prisoners with a high probability of re-arrest were randomly assigned at release from prison to experimental and control groups which varied by the amount of money received (contingent upon employment or unemployment and job placement services provided). The results showed that ex-prisoners receiving payments were less likely to be re-arrested for property theft-related crimes than those who received only job placement or no services or payments of any kind. The United States Department of Labor commissioned the TARP experiment, designed to replicate the LIFE experiment while providing a larger and more representative sample of prisoners, greater variation in treatment conditions, and administration of payments and job placement services through existing agencies rather than by a special purpose project staff. Texas and Georgia were the states chosen for the experiment, and stratified random samples of inmates were assigned, at the time of release from prison, to experimental and control groups. The groups varied in the amount of money and job placement services they received upon their release. Originally, the data were recorded in nine files for each state corresponding to each of the nine different sources of information for each TARP case. The ICPSR data collection combines these into one file for each state: Part 1 for Texas, and Part 2 for Georgia. Each file contains over 1,500 variables, clustered in nine topic areas for each inmate: (1) prison history (e.g., background information, psychological and aptitude test data, and prior criminal and present incarceration activity), (2-5) data from four personal interviews (conducted at the prerelease, three-month, six-month, and 12-month stages and that include living arrangements, employment history, and financial status), (6) state arrest data, (7) records of TARP payments received, (8) social security wages, and (9) parole records.

过渡援助研究项目（Transitional Aid Research Project, TARP）是1976至1977年间在美国得克萨斯州与佐治亚州开展的一项随机实地实验，其设计目标为通过向释囚提供最低限度收入支持、扩展部分失业保险覆盖范围，降低其重新实施财产类犯罪的动机，进而减少释囚的再犯率。本研究脱胎于20世纪70年代初在马里兰州巴尔的摩开展的早期LIFE（释囚人寿保险，Living Insurance for Ex-Prisoners, LIFE）研究。
在LIFE研究中，500名具有高再逮捕风险的囚犯在出狱时被随机分配至实验组与对照组，两组的资金发放额度（与就业、失业状况挂钩）以及所获就业安置服务均存在差异。研究结果显示，相较于仅获得就业安置服务、未获得任何服务或资金的囚犯，领取补贴的释囚因财产盗窃类犯罪再次被捕的概率更低。
美国劳工部委托开展了本次TARP实验，旨在复刻LIFE实验的研究框架，同时采用规模更大、代表性更强的囚犯样本，设置更多元的干预条件，并通过现有公共机构而非专项项目团队来发放补贴与提供就业安置服务。本次实验选取了得克萨斯州与佐治亚州作为实验地，研究人员将囚犯的分层随机样本在出狱时分配至实验组与对照组，两组在出狱后可获得的资金额度与就业安置服务方面存在差异。
最初，实验数据按每个州分为9个文件，对应每个TARP案例的9类不同信息来源。政治与社会研究校际联盟（Inter-university Consortium for Political and Social Research, ICPSR）的数据合集将这些文件整合为每个州一个单独文件：得克萨斯州数据为第1部分，佐治亚州数据为第2部分。每个文件包含超过1500个变量，按每名囚犯划分为9个主题领域：(1) 监狱史（如背景信息、心理与能力测试数据、既往犯罪与当前监禁经历）；(2-5) 四次个人访谈数据（分别在释前、出狱后3个月、6个月及12个月开展，内容涵盖居住安排、就业经历与财务状况）；(6) 州级逮捕记录；(7) TARP补贴领取记录；(8) 社会保障薪资数据；(9) 假释记录。