Non-traded goods, factor markets frictions, and international capital flows

The first two papers focus on the relationship between FDI flows, domestic investment, exchange rate risk, and total factor productivity. Every single study of the impact of FDI on domestic investment uses a reduced form regression with domestic investment as a dependent variable and FDI as one of the regressors. The biggest challenge of such approach is, the problem of endogeneity – domestic investment and FDI are jointly determined, and the direction of causality may be reversed. This leads to inconsistent estimates of the empirical model parameters. Typical approach to deal with this problem is to use instrumental variables or Granger causality tests. Our approach is novel. We recognize that at the heart of FDI is the ownership of capital stock that is physically located in a different country. If it does not matter who owns capital, then FDI would completely crowd out domestic investment. If foreign ownership brings something to production that domestic ownership cannot, then crowding out will be limited, eliminated, or even reversed. The key parameter of interest is therefore the elasticity of substitution between the domestic and foreign ownership of capital. Our goal is to estimate that parameter, using an agent-based international business cycle model, where capital used in production is a composite of capital stocks owned by domestic and foreign residents. In the model, both FDI flows and domestic investment expenditures will be endogenous, as they are in the data. Because claims to capital are denominated in local currency units, FDI flows will respond to exchange rate risk, if investors are risk averse. The strength of such response will depend on the aforementioned elasticity of substitution, which we will then estimate using “indirect inference” – making sure that the partial correlation between exchange rate risk and FDI flows in the model and in the data are the same. The model will then be used to evaluated the impact of FDI on measured productivity at the macroeconomic level. The partial correlation in the data will be estimated using bilateral capital flows within the European Union – an environment that is as close free capital mobility as one can hope for – using fairly standard panel data methods.

The third paper will study the welfare effects of global re-balancing, using an open economy version of the state-of-the art macroeconomic model with uninsurable idiosyncratic risk. The model will endogenously generate an equilibrium outcome with surplus and deficit countries, and with a non-trivial distribution of income and wealth within each country. We will then introduce policies whose goal is to eliminate the external surpluses and deficits of all countries (global re-balancing), and simulate their effects on income and wealth distribution (and inequality). Finally, we will evaluate their welfare effects for different groups of people. We will analyze separately policy proposals that force the surplus economies to consume and invest more, and those that force the deficit economies to save more (focusing on both the private and the public sector).