Acemoglu, D., Johnson, S., Robinson, J.A. (2001). The colonial origins of comparative development: an empirical investigation. American Economic Review 91(5): 1369-1401.
Results reported in Tables 4 and 7 of AJR (2001) are replicated. However, the test of weak instruments reported by AJR (First-stage R2) is inappropriate in the presence of more than one regressor. Appropriate tests of weak instruments (First-stage partial R2, Cragg-Donald F-statistic) reveal that the null hypothesis of weak instruments cannot be rejected in the presence of more than one regressor.
The Stata code below generates detailed estimation results and a summary table.
Bond, S.R., Hoeffler, A., Temple, J. (BHT) (2001). GMM estimation of empirical growth models. CEPR Discussion Paper, 2048, Centre for Economic Policy Research.
Work in progress
Rodrik, D., Subramanian, A., Trebbi, F. (2004). Institutions rule: the primacy of institutions over geography and integration in economic development. Journal of Economic Growth 9(2): 131-165.
Results reported in Table 3 of RST (2004) are replicated. My IV estimates ignore the correction of standard errors in Panel A considered by RST. The difference between corrected and uncorrected standard errors (t-statistics) appears to be small.
RST estimate separate first-stage regressions for each endogenous regressor (Panel B), which ignores the possible collinearity among the instruments when used in the second stage. Two of the first-stage statistics reported by RST (F-statistic, R2) are potentially misleading; the Partial R2 s reported by RST differ from my estimates. The Kleibergen-Paap F-statistic reveals that the null hypothesis of weak instruments cannot be rejected in three of four cases (with a borderline estimate in the fourth case).
The Stata code below generates detailed estimation results and a summary table.