Re-examining the Feldstein-Horioka and Sachs' views of capital mobility: A heterogeneous panel setup

•We test for OECD capital mobility using both the F-H and Sachs' approaches.•Time aggregation does not appear to bias findings towards capital immobility.•Cross sectional dependency and endogeneity warrant a heterogeneous panel approach.•We apply Pesaran's CCEMG estimator•Capital mobility greater than hitherto found, though not perfect

We re-examine two complementary views of international capital mobility using data for 25 OECD countries over the period 1970-2011. Estimation of the original Feldstein-Horioka and Sachs' equations provides mixed evidence of capital mobility, though we do not detect a significant bias towards finding in favour of capital immobility in using time-averaged data. However, potential bias in cross-sectional estimation motivates us to examine the data as a heterogeneous panel which allows us to control for the effects of cross-sectional dependence and endogeneity. In addressing the Feldstein-Horioka puzzle, application of the CCEMG estimator of Pesaran (2006) to the Feldstein-Horioka and Sachs' equations points towards greater (though not perfect) capital mobility than hitherto found.