A note on the determinants of long-run aggregate state productivity growth

Abstract

We use data envelopment analysis to estimate state-level productivity and then use the pooled mean group approach to investigate its long-run determinants. Results indicate that human capital, infrastructure, energy shocks, a diverse mix of industries and states with large urban populations are associated with higher long-run aggregate state productivity growth, while unionisation has no long-run effects.

Keywords:

• data envelopment analysis (DEA)
• productivity
• pooled mean group (PMG)

JEL Classification:

• R11
• O4

Notes

¹ See e.g. Carlino and Voith (1992), and Partridge and Rickman (1999).

² For example, Carlino and Voith (1992) use OLS, FE and RE models, and find that state industrial composition, infrastructure, level of human capital and metropolitan structure all have significant effects on state productivity. Partridge and Rickman (1999) use similar methods.