the list of countrymatches underlying, the results can be displayed in small samples (see the next two subsections and the online appendix).
Summing up, applying matching estimators to (small) cross-country samples comes with a disadvantage and an advantage. The disadvantage is that unconfoundedness is unlikely to hold, as it is often implausible to assume that country-specific unobservable characteristics do not play any role in treatment assignment. The advantage is that they allow us to transparently check for the existence of common support. Consequently
Studies of the impact of trade openness on growth are based either on cross-country analysis-which lacks transparency-or case studies-which lack statistical rigor. We apply transparent econometric methods drawn from the treatment evaluation literature to make the comparison between treated (i.e., open) and control (i.e., closed) countries explicit while remaining within a unified statistical framework. First, matching estimators highlight the rather far-fetched country comparisons underlying common cross-country results. When appropriately restricting the sample, we confirm a positive and significant effect of openness on growth. Second, we apply synthetic control methods-which account for endogeneity due to unobservable heterogeneity-to countries that liberalized their trade regime and we show that trade liberalization has often had a positive effect on growth.
B. Openness and Growth in MCD Countries
VI. Conclusions and Extensions
1. Appendix I. Treated and Control Countries by Region
A. Period 1991–2000
B. Period 1981–90
C. Period 1971–80
D. Period 1961–70
1. Openness and Growth, Cross-Country Evidence (I), 1950-98
2. Openness and Growth, Cross-Country Evidence (II), 1961-2000
3. Cross-CountryMatches, Treated Countries, 1991-2000
4. Cross-CountryMatches, Control Countries, 1991-2000
5. Cross-CountryMatches, Treated Countries, 1981-90
where I and I t are the sample size and the number of treated countries, respectively. These nearest-neighbor matching estimators both allow for the identification of the ATE and ATT under unconfoundedness and are very transparent, as the list of countrymatches underlying the results can be displayed in small samples (see the next two subsections).
Summing up, the application of matching estimators to (small
During long periods of history, countries have pegged their currencies to an international standard (such as gold or the U.S. dollar), severely restricting their ability to create money and affect output, prices, or government revenue. Nevertheless, countries generally have maintained their own currencies. The paper presents a model where agents have heterogeneous preferences—that are private information—over goods of different national origin. In this environment, it may be optimal for countries to have different currencies; we also identify conditions where separate national currencies do not expand the set of optimal allocations. Implications for a currency union in Europe are discussed.