Foreign direct investment and urban concentrations: unbundling spatial lags

Foreign direct investment (FDI) is seen as a way to import technology and catch up with economic leaders. It is therefore important to understand why some countries attract more investments by multinationals than others. We expand the set of common determinants of FDI with urban agglomerations and ask the question whether the accessibility of market potential and the number of potential investment locations, in the shape of urban concentrations, matters, since the importance of urban agglomeration economies for FDI has not been investigated before. We show that countries with several medium-sized cities attract more foreign investment, especially if they are close to main agglomerations, but too much concentration (primacy) reduces the inflow of FDI. Moreover, we unbundle spatially lagged FDI by including spatial lags of the determinants of FDI. It is important to disentangle such third-country effects in order to understand how FDI flows depend on the complex ways in which multinationals fragment sales and production across countries. Using a panel of U.S. affiliates' sales in foreign countries between 1984 and 1998, we find evidence that cities are important drivers of FDI. Furthermore, third-country effects suggest that horizontal and complex forms of FDI coexist.