Reply to “A Replication Study of ‘Openness, Country Size, and Government Size’ (Journal of Public Economics 2009)”

Andrew Musau is able to replicate almost exactly Ram’s estimates from Penn World Table 6.1. Like Ram, Musau is also unable to show a significant negative association between country size and openness and between country size and government size of the kind postulated by Alesina and Wacziarg. To shed further light on the issue, fixed-effects estimates of the three core regressions are obtained from the widely used World Bank data for 174 countries covering the period 1990–2015. The estimates show a highly significant positive association between openness and government size, further reinforcing Ram’s estimates and those from Musau’s 189-country Penn World Table 7.1 sample. While the new country size estimates in openness and government size regressions are negative, the latter lacks significance at any meaningful level. Therefore, the substantive conclusion remains the same as indicated by Ram; while the predominance of evidence supports Rodrik’s thesis of a positive association between openness and government size, statistically significant evidence is lacking to support the Alesina–Wacziarg proposition that the aforesaid positive association is due to a combination of the negative association between country size and openness and between country size and government size