Abstract:
|
The U.S. Congress regularly considers public policy issues that require insights from the statistical analysis of data. As an agency of the legislative branch, the U.S. Government Accountability Office (GAO) conducts evaluations and audits of federal programs and policies to help provide these insights to policymakers. Our use of statistics is as diverse as the programs we assess, spanning areas such as probability sampling, modeling of observational data, econometrics, and causal inference for program impact. Our talk will illustrate our statistical work with two recent case studies. In the first case, we will discuss models developed to assess whether market participants on Wall Street continue to view the largest financial institutions as "too big to fail," despite policy reforms designed to address the financial crisis of the last decade. In the second case, we will discuss our original estimates of the prevalence of "contingent workers" with transitory attachments to the labor market, which highlighted gaps in federal statistical programs to measure this phenomenon.
|