Facets of Balassa‐Samuelson Thirty Years Later*

Abstract Prior to the important Penn studies of Kravis‐Heston‐Summers, statisticians relied on exchange‐rate conversion estimates that would be correct only if naive Gustav Cassel versions of purchasing‐power parity were true. By contrast, correct real‐income estimates, using actual local prices and incomes, exhibit the systematic Penn effect: real per capita income ratios between poor and rich are systematically exaggerated by conventional exchange‐rate conversions. Bela Balassa and Paul Samuelson, independently in 1964, explained why. And, as the Penn authors cited, David Ricardo and Roy Harrod had already similarly argued. It is shown here how subtle must be the theoretical analysis addressed by Jagdish Bhagwati and mainstream economists in tackling this problem.