The US has experienced a joint decline in corporate tax rates and startup rates in recent decades. We study how changes in corporate taxes affect startup rates in a firm dynamics model with occupational choice. A change in profit taxes has ambiguous effects on firm selection, and therefore on the startup rate. Quantitatively, the observed fall in US corporate tax rates generates, at most, one-fifth of the startup rate decline. Sensitivity analysis suggests that this number is best interpreted as an upper bound, indicating that corporate taxes did not play a major role in the startup rate decline.

© 2017 Elsevier B.V. For a sample of OECD countries, I document a systematic positive relationship between (i) aggregate productivity, (ii) the employment share by large firms and (iii) the proportion of large firms in the economy. I propose that differences in bankruptcy procedures can explain this relationship. In a model of financial intermediation and informational frictions, I show that as bankruptcy procedures worsen—measured by the amount a lender can recover from bankrupt borrowers—lenders respond by (i) shifting their portfolio of loans to smaller (less productive) firms and (ii) lending less. This finding is supported by empirical evidence: across countries, efficient bankruptcy procedures are associated with a higher proportion of new bank loans allocated to large firms. In the model, moving the level of recovery rate from the U.S. level to that of the lowest recovery rate country in the OECD sample reduces TFP by around 30 percent.