The Urban Institute, Washington, DC.
Am Econ Rev. 2009 May;99(2):508-11. doi: 10.1257/aer.99.2.508.
In April 2006, Massachusetts enacted a comprehensive health care reform bill that seeks to move the state to near universal insurance coverage. The bill included expanded eligibility for public coverage, subsidized insurance, market reforms, requirements for employers, and, most controversial, an individual mandate. A study of the early impacts of the state's initiative found evidence of a substantial drop in uninsurance--from 13 to 7 percent for nonelderly adults (Long 2008). Because that study relied on a simple pre-post comparison, it is possible that the estimates of the impact of health reform reflect both the changes under health reform and factors beyond health reform that changed over the same period, leading to biased estimates of the impacts of reform (Lawrence B. Mohr 1995). This paper expands on the earlier work to estimate the impacts of health reform in Massachusetts using new data and a stronger research design. Specifically, we rely on data over time for Massachusetts and other states from the Current Population Survey (CPS) to estimate difference-in-differences (DD) models (Jeffrey M. Wooldridge 2002).
2006 年 4 月，马萨诸塞州颁布了一项全面的医疗改革法案，旨在实现该州接近全民保险覆盖的目标。该法案包括扩大公共保险、补贴保险的资格、市场改革、对雇主的要求，以及最具争议性的个人强制规定。对该州计划早期影响的一项研究发现，无证保险的人数大幅下降的证据——非老年人的无保险率从 13%降至 7%（Long 2008）。由于该研究仅依赖于简单的前后比较，因此，对医疗改革影响的估计可能既反映了医疗改革下的变化，也反映了同期发生的、超出医疗改革范围的变化，从而导致对改革影响的估计存在偏差（Lawrence B. Mohr 1995）。本文扩展了早期的研究，使用新的数据和更强有力的研究设计来估计马萨诸塞州医疗改革的影响。具体来说，我们依赖于来自现有人口调查（CPS）的马萨诸塞州和其他州的随时间变化的数据来估计差分模型（Difference-in-Differences，DD）（Jeffrey M. Wooldridge 2002）。
