RESEARCH

How do FDI characteristics and labor market conditions affect productivity spillovers? Evidence from Moldova, with Shawn W. Tan. (revise and resubmit at The World Economy)

We use firm-level administrative panel data and annual region-sector survey indicators of labour market conditions in 2005-2014 to estimate productivity spillovers from foreign to domestic firms through the labour market in Moldova. Baseline regressions show that domestic firms benefit from backward FDI spillovers, while we find no evidence of horizontal or forward spillovers. Spillover effects are heterogeneous and depend on foreign firm characteristics, such as ownership structure and age. Domestic firms in upstream sectors benefit from both fully foreign firms and joint ventures. However, joint ventures need less time in the market for positive spillovers to materialise, while fully foreign firms only lead to signicant spillovers when they are older. In regions and sectors where firms experience fewer labour market restrictions, backward FDI spillovers are larger. Interacting foreign firm characteristics with labour market restrictions, we find that older fully foreign firms generate positive backward and forward FDI spillovers, and the latter are larger for more flexible labour markets. Our findings suggest that in developing countries less labour mobility reduces the size of inter-, rather than intra-industry FDI spillovers. (Latest draft version here)

We examine the effects of corruption on firm productivity in Moldova in 2005-2015 using administrative firm-level panel data merged with two sector-region measures of corruption. The paper finds that bribery affects labour productivity directly, while inspections prevent rms from investing in more productive capital. A unit increase in the difficulties with bribery score decreases labor productivity by 5.2 percent, while a unit increase in the difficulties with inspection score decreases tangible assets investment by 7.2 percent and intangible assets investment by 9.6 percent. The results also show that state-owned enterprises are insulated from the negative effects of corruption, while there is no heterogeneous effect for large or foreign firms.