World Bank economists are mostlypractical people, people who try to answer the question,'what exactly should this particular country do rightnow?' But if they had hoped that the growth regressionlessons summarized in William Brock and StevenDurlauf's article would enhance their practical advicegiving, they might feel some dissatisfaction. How would theychange their advice to, say, Brazil? But that is why thisarticle is important conceptually. It goes to the heart ofthe matter by proposing a change in the empirical growthliterature's fundamental methodology, from modeltesting to decision theoretic. The article's valiantbut flawed attempt reveals the difficulties in making thisshift, however. The reader likes to make three points: thereis a tension between the interests of academics andpractitioners in growth regressions; output responseheterogeneity is a huge practical problem; and policydecisions can be guided only in broad outlines by growth regressions.
PDF
download
878987797
028-85220240
