When production is an O‑ring (multiplicative) technology, tasks are quality complements: automating one task alters the marginal value of others, can force discrete bundled adoption choices, and may increase earnings for workers who retain control of remaining bottleneck tasks. Simple linear task‑exposure indices therefore mismeasure displacement risk and policy should focus on bottleneck structure and time allocation.
— This reframes automation policy and labour forecasting: regulators, firms and retraining programs should target where automation changes the structure of bottlenecks, not average task vulnerability, because the social and distributional outcomes can be qualitatively different.
Tyler Cowen
2026.01.05
100% relevant
Gans & Goldfarb’s recent working‑paper model (reported by Tyler Cowen) formalizes an O‑ring multiplicative task technology, derives discrete/bundled adoption and the possibility of rising labour income for remaining tasks, and explicitly warns that linear exposure indices overstate displacement.
← Back to All Ideas