Capital Mobility in Developing Countries: Evidence from Panel Data

The purpose of this paper is to show that the use of panel data can shed some light on the FeldsteinHorioka puzzle. The use of panel data would bring in two advantages. First, it would avoid the bias towards low capital mobility brought by the use of time-averaged data. Second, it would make possible to take into account specific effects (heterogeneity) like a country’s size. Pooling annual data for the period 1960-1996 for 29 developing countries, the estimated impact of saving on investment is considerably smaller and it is possible to conclude that there is some degree of capital mobility in developing countries. Therefore, the high estimated saving-investment correlation seems to be due more to the existence of specific individual country effects than to capital immobility. The coefficient stability through time remains a puzzle though. Key woRds capital mobility, panel data, heterogeneity, developing countries Jel classification C23, F31 548 capital Mobility in developing countries Estud. econ., são Paulo, 37(3): 547-561, jul-set 2007 IntRoductIon The empirical evidence on the effective degree of capital mobility in developing countries is extremely scarce when compared to that for developed economies. This happens despite widespread acknowledgment of the importance of the extent of capital mobility in determining, for example, optimal monetary and fiscal policies (Fleming, 1962; Mundell, 1968), the exchange rate (Levich, 1985), the tax rate on income from capital (Feldstein and Horioka, 1980), and to analyze the inflationary tax (Easterly et al., 1995). Feldstein and Horioka (1980) suggested the correlation between saving and investment rates as a measure of the degree of capital mobility. Based on a sample of 16 OECD countries, they obtained evidence that saving and investment rates were highly correlated. They in turn concluded that the degree of capital mobility in industrialized countries was low, going against accepted wisdom that these countries had few restrictions on capital flows. The Feldstein-Horioka puzzle refers to the empirical finding that estimates of the saving-investment correlation have remained high, despite the use of more recent data and different econometric techniques. Murphy (1984), Obstfeld (1986), Dooley et al. (1987) and Wong (1990) also found a close association between saving and investment for less industrialized and developing countries, although the estimated correlation was lower on average. The purpose of this paper is to present some additional empirical evidence on the degree of capital mobility in developing countries, using the procedure based on the correlation between saving and investment formulated by Feldstein and Horioka (1980). Its main contribution is to verify how important are the effects of country heterogeneity in saving and investment rates association. More precisely, it seeks to verify what happens to the FH puzzle when heterogeneity is explicitly accounted for. In order to do so we use three different panel data estimators. The difference among them results from the importance that each gives to the individual and temporal dimensions. The pooled estimator assumes both individual homogeneity and temporal stability. The between estimator, in working with averages, eliminates the temporal variability and emphasizes the inter-country dimension. Just like the pooled indicator, it implies homogeneity among countries. The within indicator, in turn, uses the variability among countries, since it is calculated from the difference between the saving and investment ratio and the averages over time for each country and takes into account the heterogeneity of the data. We also believe that panel data methods can also help us to deal with the dilemma of working only with individual (country by country) data versus temporal (year to year) data that involves the literature. The sample consists of annual data (1960-1996) for 29 developing countries.