Sector · NAICS public administration

Government

Context first: jobs, pay and output, with adoption measures beside them. Outcomes are defined before looking for favorable examples.

Evidence confidence: insufficientAll sectors

Payroll jobs

23.33M

2026-10 (prelim.), +3.7% vs Jan 2019

Real hourly pay

n/a

Jan 2019 dollars, n/a vs Jan 2019

Real output

+2.5%

Value added, 2019Q4 to 2026Q2

Value added

$3.43T

2025, current dollars

What is being adopted?
No Census estimate; 36.4% of Ramp customers in the sector pay for AI (Oct 2026).
What changed economically?
Context series below. No AI-attributable change has been assembled for this sector.
What happened to workers?
Payroll employment and real hourly pay are shown below. Not attributable to AI.
What did consumers gain?
Not assembled.
What remains unproven?
Any causal effect of AI on this sector's output, jobs, pay or prices. Dossier not yet written: insufficient evidence.

Context, not attribution

Jobs, pay and output.

Payroll employment

Observed

Payroll jobs in Government, +3.7% since January 2019.

0K5.0M10.0M15.0M20.0M2020202120222023202420252026ChatGPT released (timing marker, not a cause)

Caveat. Employment moves with demand, rates, migration and the pandemic recovery. The ChatGPT marker shows timing only.

UnitThousands of jobs, seasonally adjustedGeographyUnited StatesObservedJan 2019 to 2026-10SourceBLS Current Employment StatisticsRetrieved2026-11-18
Inspect the evidence
What this chart does not show
  • Hours, hires or separations.
  • Self-employed and farm workers.
Lineage: source to transformation to chart
  1. BLS series CES9000000001
  2. Monthly values as published; last two months preliminary

Text description. Payroll jobs in Government, +3.7% since January 2019.

Link
Insufficient evidence

Hourly earnings not published by BLS at this sector level in the series retrieved.

Real value added

Observed

Real value added index for Government, +2.5% since 2019Q4.

02040608010020182020202220242026ChatGPT released (timing marker, not a cause)

Caveat. Real output by industry is revised annually. Growth here cannot be attributed to AI without further evidence.

UnitChain-type quantity index, 2017 = 100GeographyUnited StatesObserved2017Q1 to 2026Q2SourceBEA GDP by IndustryRetrieved2026-11-18
Inspect the evidence
Lineage: source to transformation to chart
  1. BEA ValueAdded.xlsx sheet TVA103-Q
  2. Matched by exact BEA line title

Text description. Real value added index for Government, +2.5% since 2019Q4.

Link

Businesses on Ramp paying for AI

Observed

Ramp adoption in Government reached 36.4%.

0%20%40%60%80%100%202420252026

Caveat. Vendor sample. Census (representative) puts this sector at n/a% using AI in late Oct 2026, a different question and universe.

UnitPercent of Ramp businesses in the sectorGeographyUnited States, Ramp customersObserved2023-07 to 2026-10SourceRamp AI IndexRetrieved2026-11-18
Inspect the evidence

Text description. Ramp adoption in Government reached 36.4%.

CSV not offered: Ramp reuse terms not yet verifiedLink
Next · Chapter 05Job accounting: Observed, modeled, claimed and causal, never summed.