Are Sufficient Statistics Necessary? Nonparametric Measurement of Deadweight Loss from Unemployment Insurance
提出一种分解方法，利用州行政数据的实验和准实验估计，比较失业保险两种改革的财政外部性，发现提高周福利比降低隐性收入税更有效率。
Central to the welfare analysis of income transfer programs is the deadweight loss associated with possible reforms. To aid analytical tractability, its measurement typically requires specifying a simplified model of behavior. We employ a complementary "decomposition" approach that compares the behavioral and mechanical components of a policy's total impact on the government budget to study the deadweight loss of two unemployment insurance policies. Experimental and quasi-experimental estimates using state administrative data show that increasing the weekly benefit is more efficient (with a fiscal externality of 53 cents per dollar of mechanical transferred income) than reducing the program's implicit earnings tax.