The Marginal about Average Returns to Schooling.

The main contribution of this paper are that : the possibility that different instruments may affect different margins in explored, and, because of its potential importance, the linearity in schooling assumption is tested. The paper finds a large and significant downward bias in the least squares estimate of the returns to schooling; the returns to schooling are significantly different from being linear, and that the estimates are stable with respect to the choice of instrument