Endogenous Task Bundling, Skills and Automation / Joshua S. Gans.

Format:

Book

Author/Creator:

Gans, Joshua S.

Contributor:

National Bureau of Economic Research.

Series:

Working Paper Series (National Bureau of Economic Research) no. w35211.

NBER working paper series no. w35211

Language:

English

Physical Description:

1 online resource: illustrations (black and white);

Place of Publication:

Cambridge, Mass. National Bureau of Economic Research 2026.

Summary:

An occupation's task list records what its workers do, but not why those tasks form one job. We model a competitive labour market in which tasks ordered by required expertise are bundled into jobs, and ask how technology moves the boundaries between them. A job's wage prices its most expert task. Bundling less expert tasks into a job therefore uses expensive worker time, whereas splitting tasks across workers forces each receiving worker to reconstruct prior work; equilibrium job boundaries balance these costs. Automation inside a job splits it when the machine has standardised interfaces and, locally holding the job count fixed, widens an interior non-top job when operating the machine requires its holder's context. Artificial intelligence that merely reduces context loss can rebundle jobs without automating anything. Away from job-count transitions, smooth technological change moves occupational task content at first order while the unit-cost saving from adjusting boundaries is second order; standardised interfaces and job-count changes instead produce discrete reorganisation. Holding the expertise-price schedule fixed, wage changes track movements in a job's most expert task. And bundle protection can preserve a top job while eliminating the less expert role that previously performed the automated task.

Notes:

May 2026.

Print version record