Learning objectives

Methods¶

Students sketch the history of returns-to-education estimation since Mincer and articulate why “the return to schooling” is not a single, model-free number
Students explain how the Liu–Mogstad–Salvanes (LMS) reduced-form strategy uses the Norwegian compulsory schooling reform as an instrument and recovers the joint distribution of potential earnings, employment, and their dispersion for compliers
Students explain why a Mincer regression and an Imbens–Rubin-style LATE answer different questions, even with the same instrument
Students locate the precise points at which LMS’s reduced-form IV results enter the parametric life-cycle earnings model and identify which structural objects each IV moment helps to pin down
Students explain why the structural step is needed at all — i.e., which quantities of economic interest (precautionary saving, willingness-to-pay for education) are not identified from IV moments alone

Students relate the LMS findings on heterogeneity in earnings risk by ability and education to the standard Mincerian story
Students recognise how the progressive Norwegian tax-transfer system enters as a partial insurance mechanism, attenuating the gross return
Students explain, at the level of intuition, why a welfare-relevant return can fall well below the certainty-equivalent return when markets are incomplete — a higher-variance or steeper earnings stream is worth less to a prudent agent who cannot fully insure or borrow against it

Students judge for themselves where they trust reduced-form-only answers and where a structural step adds value at the cost of stronger assumptions

Students debate IV-vs-structural trade-offs in small groups, using LMS as a concrete example rather than an abstract methodological debate