Friedman (1992) argues that regressing cross-country income changes on their final levels can be informative about σ-convergence (the tendency for the dispersion of income levels to narrow) whereas a similar regression on initial levels of income cannot be. In this note we show that Bliss's (1999) dismissal of this argument is in error.
No takes yet. Share an insight, caveat, or question.
Edmund Cannon (2000) studied this question.