1. Including between-country effects, by using OLS or random effects estimation, reinforces the impressions from
the preceding robustness tests.
2. The one- through four-year lags of GDP and FDI in the secondary sector are all insignificant when replacing their current values in the specifications of Table 3, while one- and two-year lags of democracy, and one- through three-year lags of institutional quality are significant.
3. Summing up, we find evidence of a relationship between institutional quality and FDI in services which is robust to
changes in specification.