Globalization and the Feldstein–Horioka puzzle

Abstract

Capital account liberalization and the integration of world financial markets should increase capital mobility across countries. This article uses the Feldstein–Horioka savings–investment methodology to examine the impact of economic globalization on the degree of capital mobility in 99 countries over the period 1970 to 2005. Our findings suggest that economic openness and financial market integration have led to increased capital mobility in developed as well as developing countries. However, their effect appears to be larger for the latter. This also implies that countries with more financial openness can run higher current account deficits due to better access to external capital markets. Our results also support the previous findings that foreign aid supplements domestic savings for investment in developing countries.

Notes

¹ Later research finds high savings–investment correlation in large as well as small economies, although this correlation for small economies is weaker (Dooley et al., 1987; Wong, 1990; Mamingi, 1997; Vamvakidis and Waczairg, 1998; Coakley et al., 1999; Kasuga, 2004). Sinha and Sinha (2004) examine capital mobility for developed as well as developing countries and find that capital is more mobile for countries with a low per capita income. The findings of relatively higher capital mobility in developing countries have mainly been attributed to the integration of world financial markets and capital account liberalization.

² The coefficient on savings rate is also called savings retention coefficient as it shows the extent to which an increase in domestic savings is used to finance domestic investment or, in other words, the fraction of a dollar savings that is invested domestically (Georgopoulos and Hejazi, 2005).

³ The time trend would be a weak indicator of economic openness and financial integration as it may also capture other factors such as business cycle effects, changes in policy regimes, etc.

⁴ Despite the advantage of this technique in terms of correcting potential endogeneity in the model, dynamic panel GMM has hardly been used to study capital mobility in the FH literature.

⁵ In the presence of fixed effects, the lagged endogenous determinants will correlate with the error term, resulting in biased and inconsistent estimates for a panel with large cross-sections and short time periods.

⁶ One may plausibly argue that savings is endogenous as it not only affects but may also be affected by investment. Therefore, in order to get consistent estimate we need to employ Two-Stage Least Square (2SLS) method. The problem with this approach is nonavailability of valid instruments and their data, especially for the developing countries. Isaksson (2001) used government consumption expenditures and dependency ratio (sum of the population ages between 0–14 and 65 and above divided by labour force of a country) as instruments for the savings rate. However, they are not valid instruments in our study due to their very weak correlation with the savings rate. The dynamic panel estimation technique in our study overcomes this concern.

⁷ For every regression, we also checked the conditions of both validity of instruments and the absence of serial correlation in residuals.

⁸ Isaksson (2001) points out two reasons for using gross savings over net savings: (i) it is gross savings that moves between countries and (ii) the accounting definitions of depreciation differs across countries.

⁹ Gross fixed capital formation, as defined in WDI (2007), consists of outlays on additions to the fixed assets of the economy, net changes in the level of inventories, and net acquisitions of valuables. Past studies consistently use gross fixed capital formation as a measure of investment because it does not include highly procyclical components of inventories (Bayoumi, 1990; Sinha and Sinha, 2004; Payne and Kumazawa, 2005; Younas, 2007).

¹⁰ Data for all variables is broken into separate 5-year data averages. Therefore, there are a total of seven time periods.