Roads, exports and employment: Evidence from a developing country

Domestic road programs are often justified on the basis of their presumed positive effects on firms' exports and accordingly on firms' employment. In this paper we evaluate this policy claim for Peru, a developing country whose regions were exposed to an asymmetric infrastructure shock. In so doing, we take advantage of detailed geo-referenced data on firm-level trade for the period 2003–2010 as well as on recent and historical road infrastructure. In particular, to identify the impacts of interest, we first exploit the dimensions of this dataset to account for regional-sectoral and even firm-level confounding factors through extensive sets of fixed effects. In addition, we conduct placebo exercises and carry out instrumental variable estimations whereby we instrument recent changes in the road network with the pre-Columbian Inca road network. Estimates concur in suggesting that improvements in transport infrastructure had a significant positive impact on firms' exports and thereby on firms' job growth.