Do Targeted Business Tax Subsidies Achieve Expected Benefits?

We examine the association between thousands of state and local firm-specific tax subsidies and business activity in the surrounding county, measured as the number of employees, aggregate wages, per capita employment, per capita wages, and number of business establishments. Using three different matched control groups, we find a positive association between subsidies and the employment measures. However, we show that local information--measured based on subsidy-specific​ disclosures, public awareness, and local press coverage--plays an important role in the effectiveness of subsidies. We also demonstrate that (i) receipt of multiple or subsequent subsidies in the same counties is critical for these employment outcomes and (ii) results are concentrated in the largest subsidy packages by dollar value. In addition, we observe mixed evidence for the relation between subsidies and business establishments and find little to no local effects for over 1,000 subsidies that cost approximately $99.8 million in aggregate. By providing a large-scale empirical analysis of the relation between firm-specific tax subsidies and aggregate economic activity at the county level, we extend a literature that generally focuses on the real effects of statutory tax policies that impact all firms in a jurisdiction. We also contribute to the accounting literature by examining the role of the local information environment in subsidy effectiveness.