Speed of Policy Reform and Outcomes
研究1961-1993年间西半球发展中国家实际人均GDP增长与五项经济政策指标的关系，检验政策偏离华盛顿共识程度越大、人均GDP停滞或衰退期越长越严重的假设。
Economists working in developing countries often advise the governments that policy reform is a necessary condition for realization of potential growth. The governments generally agree but often argue that must be introduced slowly, and one or a few at a time rather than comprehensively within a short period as in a stand-by arrangement with the IMF. Yet, as Robert J. Barro and Xavier Sala-i-Martin (1995 p. 8) have noted, there is surprisingly little empirical evidence on the relationships between specific policy reforms and growth. The research underlying this paper seeks to join this issue by identifying empirical regularities in the movements of real per capita GDP growth and five measures of economic policy. The proposition examined is that periods of stagnation or decay of per capita GDP will be longer and more severe, the greater the departure of policy from norms such as those outlined in John Williamson's now famous Washington consensus. We focus specifically on Western Hemisphere Developing Countries (IMF classification) over the period 1961-1993, and we rely on IMF International Financial Statistics and World Bank Indicators CD-ROMs for the data. We have been influenced by the nonparametric approach taken in Michael Bruno and William Easterly's (1995) study of relationships worldwide, between price inflation and growth before, during, and after episodes of exceptional price inflation. Their results include a generally robust pattern in which growth of real per capita GDP falls substantially during the episodes of high inflation and then rises to rates above even the pre-episode rates. These findings conflict with those of cross-section studies that have found no robust relationship between growth rates and rates of price inflation. Bruno and Easterly's interpretation of the difference is that parametric cross-section and linear time-series regressions are unsuitable for detecting the relationship between high inflation and growth. Similar reasoning seems to argue for nonparametric tests of the relationship between good policy and good growth performance. Arguably, monetary expansion rates (or fiscal deficits, foreign-exchange flows, or investment/GDP rates) can be very different among countries with similar growth rates. Historical patterns and regularity of these variables may be as important as their magnitudes.