As a final step in this analysis, I present in Table B.1 the results from alterative estimations of the “baseline” model of preceding sections. To minimize the proliferation of results (and because results are very similar for any dimension of capital mobility), I focus on the liberalization of capital markets. The liberalization of capital markets is significantly related to other dimensions of capital mobility (e.g., 1960–92 correlations of lagged liberalization with total flows, direct investment, and capital market borrowing are statistically significant and substantively moderate or large). Also, in the case of every direct effect and interaction effect from Tables 3.2 to 3.5, two of the three alternative estimators confirm the findings; in the large majority of cases, all three alternative estimators produce the same result.

As discussed in the section on methodology in Chapter 3, I present (along with my “baseline” model) a “full” fixed effects model with a lagged endogenous variable and a generalized error correction model. To make the table manageable, I report complete model results (institutions and general controls) for the equations that estimate the linear effects of capital liberalization on social welfare effort. I present results for the individual interactions that are obtained from individual estimations of the complete model with the focal interaction and display significance levels, but not standard errors for the interactions.

As Table B.1 reveals, the primary findings about the welfare state effects of international capital mobility and its mediation by national political institutions are systematically confirmed.