Barro and Sala-i-Martin (1991, 2004) use growth regressions to look at the patterns of convergence across U.S. regions and states They find that there is a slow pattern of convergence across regions and states, and they interpret this through the lenses of the neoclassical growth model. Explain why Corollary 19.1 implies that this interpretation is not appropriate. Suggest instead an alternative
explanation for why convergence across regions and states might be slow. [Hint: should we expect technology or capital to flow more rapidly across regions?]