Definition of Heckman Two-Step Estimation: Heckman two-step estimation is a way of estimating treatment effects when the treated sample is self-selected and so the effects of the treatment are confounded with the population that chose it because they expected it would help -- the classic example is that college educations may be selected by those most likely to benefit.

Taking that example, we wish to advance past the following regression:

w_i = a + bX_i + dC_i + e_i

where i indexes people, w_i is the wage (or other outcome variable) for agent i, X_i are variables predicting i's wage, and C_i is 1 if i went to college and 0 if not. e_i is the remaining error after least squares estimation of a, b, and d. (Econterms)

Terms related to Heckman Two-Step Estimation:

• Treatment effects