Working Paper: CEPR ID: DP8079
Authors: Steven Poelhekke; Frederick van der Ploeg
Abstract: A new and extensive panel of outward foreign direct investment (FDI) at the sector level is used to estimate the determinants of non-resource and resource FDI. Since FDI is I(1), we estimate panel error-correction models of FDI with spatial lags for FDI and market potential. Our main result is that subsoil assets boost resource FDI, but crowd out non-resource FDI. The effect on non-resource FDI dominates, so that aggregate FDI is less in resource-rich countries. Spatial lags aggravate this crowding out of non-resource FDI. In addition, we find that (i) resource FDI is mainly vertical whereas other FDI is of the export-fragmentation variety; (ii) trade openness, free trade agreements and institutional quality do not impact non-resource FDI but institutional quality does have a positive effect on resource FDI; and (iii) the short-run dynamics comes mostly from shocks to FDI itself. Our main and ancillary results are robust to different measures of resource reserves and the oil price and to allowing for sample selection bias.
Keywords: Cointegration tests; External margin; Hydrocarbon reserves; Outward sector level FDI; Sample selection bias; Spatial econometrics; Subsoil assets
JEL Codes: C21; C33; F21; Q33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Subsoil assets (L72) | Resource FDI (F21) |
| Subsoil assets (L72) | Non-resource FDI (F23) |
| Resource FDI (F21) | Aggregate FDI (F21) |
| Non-resource FDI (F23) | Aggregate FDI (F21) |
| Institutional quality (I24) | Resource FDI (F21) |
| Trade openness (F43) | Non-resource FDI (F23) |
| Free trade agreements (F13) | Non-resource FDI (F23) |
| Short-run shocks to FDI (F21) | Non-resource FDI (F23) |
| Resource abundance (Q33) | Non-resource FDI (F23) |