Vol. 2, Issue 8, Part H (2016)

Abstract
Education is not really a continuous variable but rather an ordered set of different levels. The discrete nature of education is captured in an ordered probit model that is used here as an alternative estimation method. Earnings equations can then be estimated using a selectivity correction. In an ordered probit model, the optimal amount of schooling is not observed. What is observed is the discrete level of education closest to the desired amount. Thus, the actual level of schooling chosen depends on the optimal amount falling between certain threshold values. These thresholds can be estimated with an ordered probit together with the coefficients of the exogenous variables.