Preparation

This week’s session is concerned with returns to education, and more broadly with the interface between IV identification and structural estimation in a life-cycle setting. The running paper is Liu, Mogstad, and Salvanes (2025): “Returns to Education with Earnings Uncertainty and Employment Risk over the Life Cycle”.

As a preparation,

1. Watch the screencasts on:

   • the history of returns to education since Mincer (slides) — what people have meant by “the return to schooling” since Mincer (1958, 1974) and why the answer keeps moving

   • the reduced-form strategy in Liu–Mogstad–Salvanes (2025) (slides) — the Norwegian compulsory schooling reform as an instrument, Imbens–Rubin identification of complier potential outcomes, and what the IV step does and does not deliver

   • what structural estimation is (slides) — a general primer: structural work recovers the parameters of an economic choice model, trades different assumptions than IV, and usually rests on a dynamic programming problem and a stand on agents’ expectations

2. Skim Liu, Mogstad, and Salvanes (2025), focusing on:

   • the abstract, introduction, and Section 3 (everything before “Education and Process of Earnings”) — this is the IV step

   • Section 4.1’s equation (11) and the surrounding paragraphs — the parametric earnings process whose moments will be matched to the IV results

   • Table 1 — note carefully what is identified directly from IV (Panel A) versus what additionally requires the structural employment selection correction (Panel B)

   You may skim Sections 4.2 and 5 lightly; we will come back to those next week.

3. Take the quiz on eCampus.