Skip to content
Slow Variables

Ask the data

Answers come from the same store as the site; every number is checked against the record it cites.

Labour & consumers

Consumer surplus from generative AI (willingness to accept)

dispersing65/95 confidence, mixed or hard to operationalisegrade Clagging

1What this measures

Aggregate annual US consumer surplus from generative-AI chatbots, estimated by Stanford DEL from willingness-to-accept surveys (users × mean WTA to give the tools up for a month, annualised).

Why it matters. The leak from the stack. If consumers keep more surplus than every layer earns in US revenue, the capture story is about who does not capture it.

Proxy types
welfare
Unit
USD
Cadence
annual

2How we track this

Direction over 1 periods, dead-band $10.0B; higher = dispersing.

Two annual waves so far, so the trend is one step; moves under $10B are within survey error. Rising surplus disperses value away from producers to consumers.

Applied to stanford_del.us.genai_consumer_surplus_usd.pt.

3Tracker interpretation

Consumer surplus is growing faster than any layer's revenue; the Nordhaus pattern (innovators keep a sliver) is so far holding.

4Evidence

Direction over the last 1 periods; the band is ±$10.0B around the window's first point. Higher reads dispersing.

Latest point
$172Bas of 2026-03-31
us
Value the bands apply to
$172Bas of 2026-03-31

4 observations. Hollow points are disputed (see counterevidence). Every point links to its observation.

5Status and reasoning

dispersingsince 2026-09-10 · claude-initial-seed

Stanford DEL WTA estimates: $116B (Jul 2025) to $172B (Mar 2026), one step above the $10B dead band; mean WTA $98 to $124.50 per month. Surplus is moving to consumers faster than any layer's revenue grows. Initial seed.

6Timeline notes

  • 2026-03-31 us · $172Bas of 2026-03-31
  • 2025-07-31 us · $116Bas of 2025-07-31

7Counterevidence

What cuts against this reading

A single survey series with a stated-preference method; WTA estimates run higher than willingness to pay. No high-frequency version exists.

8Update history

  1. 2026-09-10unmeasured to dispersingconf — → 65 · claude-initial-seed

    Stanford DEL WTA estimates: $116B (Jul 2025) to $172B (Mar 2026), one step above the $10B dead band; mean WTA $98 to $124.50 per month. Surplus is moving to consumers faster than any layer's revenue grows. Initial seed.

    obs:2d481469

9Confidence

65 / 95 — mixed or hard to operationalise

Confidence is independent of status: 90–95 multiple strong independent sources; 70–89 good evidence, some ambiguity; 50–69 mixed or hard to operationalise; below 50 limited or vague.

10Related