Using wage data developed by Smith (1984), this paper compares human capital and institutional explanations of trends in the racial wage gap since 1890. While a regression including daily schooling variables is consistent with the human capital theory, so is an institutional model which omits schooling variables but adds the influences of isolation in the rural South, discriminatory hiring and promotion practices, unemployment, and government policy. Including schooling and institutional variables together casts doubt on the relative importance of schooling compared to labor market demands and political variables.