It’s been a while, and finally… all my work started in grad school is out. Ungated version: raw.githack.com/yangguanyi/W...

@bestofecontwitter.bsky.social #econsky #econtwitter 7/7

Policy: A nationwide tuition-free program raises BA completion from 25 to 78 percent and increases welfare by nearly 40 percent. A targeted tax break for part-time student-workers boosts completion for disadvantaged groups with smaller aggregate effects. 6/7

Why does flexibility matter?
•Investment channel: early college has high returns but fades after the mid-20s.
•Risk-reduction channel: after age 22, college sharply lowers welfare losses from shocks. Risk-reduction channel explains why later access is especially valuable for the less advantaged. 5/7

Main finding: flexibility matters. Preventing people from re-entering college cuts BA completion by 28 percent and reduces welfare by almost 4 percent of lifetime consumption. That is one quarter of the entire welfare value of college. Lower-wealth and lower-prepared individuals lose the most. 4/7

I build a heterogeneous-agent life-cycle model where people can return to college at any age. College raises human capital, helps insure against risk, and interacts with asset choices in general equilibrium. The model matches delayed enrollment, stopouts, and age-30+ completion. 3/7

Many people delay college or step in and out before finishing. Using NLSY79, I show that this pattern is widespread and strongly linked to wealth and human capital at age 18. 2/7