How Much of the Economy Runs Without People? Measuring the Automated Sector by Industry

Is a "fourth sector" of work done by AI without human steering visible in US data? We measure it as a share of task input for 19 industries from three exposure measures, Census firm adoption and one platform's automation shares. Exposure averages 0.30 on a 0–1 index. The performed share is not identified: no survey measures delegation depth, the share of exposed tasks adopting firms hand over. Across scenarios varying this depth (0.05 to 0.30) and four other assumptions, it is 0.1% to 2.4% of task input (central 0.7%); at full adoption and delegation 9.3% to 21.1%. Exposure predicts where firms adopt (rank correlation 0.86; 0.85 excluding usage-built measures); none with adoption speed was detected. One survey question could close this gap. In an assumed firm-level scenario, shared AI workflows could free and reuse time worth USD 430 per employee-year from ad hoc use (USD 269 from regular use), before adoption costs. Working paper; preliminary, comments welcome. The record holds the paper, the Online Appendix and the replication package (code under the MIT licence; raw data are not redistributed). JEL codes: O33, O41, J24, E25, L80, O14.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-13
DOI
https://doi.org/10.5281/zenodo.24208665
Primary Topic
Labor market dynamics and wage inequality
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

How Much of the Economy Runs Without People? Measuring the Automated Sector by Industry

Haluk Gül
Zenodo (CERN European Organization for Nuclear Research)
Labor market dynamics and wage inequality
article

How Much of the Economy Runs Without People? Measuring the Automated Sector by Industry

Haluk Gül
article en

Abstract

Is a "fourth sector" of work done by AI without human steering visible in US data? We measure it as a share of task input for 19 industries from three exposure measures, Census firm adoption and one platform's automation shares. Exposure averages 0.30 on a 0–1 index. The performed share is not identified: no survey measures delegation depth, the share of exposed tasks adopting firms hand over. Across scenarios varying this depth (0.05 to 0.30) and four other assumptions, it is 0.1% to 2.4% of task input (central 0.7%); at full adoption and delegation 9.3% to 21.1%. Exposure predicts where firms adopt (rank correlation 0.86; 0.85 excluding usage-built measures); none with adoption speed was detected. One survey question could close this gap. In an assumed firm-level scenario, shared AI workflows could free and reuse time worth USD 430 per employee-year from ad hoc use (USD 269 from regular use), before adoption costs. Working paper; preliminary, comments welcome. The record holds the paper, the Online Appendix and the replication package (code under the MIT licence; raw data are not redistributed). JEL codes: O33, O41, J24, E25, L80, O14.

Zenodo (CERN European Organization for Nuclear Research)
Villanova University (US)
Decent work and economic growth, Industry, innovation and infrastructure
Labor market dynamics and wage inequality
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon