Previous estimates of the profitability of investment in medical training contain an upward bias. These studies treat as the return to such investment the full difference in trained and untrained earnings, and thus they fail to account for expected labor/leisure substitutions associated with...