Who Keeps the Dividend

A firm gets more productive. Somebody ends up with the gain. Set the conditions and see who.

New output at old prices $4.0M
Capital
—
Workers
—
Customers
—

How much more output the same people produce per hour.

Share of the gain the firm can hold in price. Zero means competitors bid it all away.

Of what the firm keeps, how much reaches workers as compensation.

Hours, not outcomes. Clients need the same work, so they buy fewer hours.

The firm, before and after

LineBeforeAfter

Baseline: 100 people, $20.0M revenue, $12.0M labor cost, $5.0M other fixed cost, $3.0M operating profit. Headcount is held constant throughout — nobody is fired in this model.

Why the aggregate looks calm

The macro series averages adopters and non-adopters together. Set the adoption rate and the adopter's gain, and compare what the economy reports to what individual firms experience.

Census BTOS: 18% of firms, 32% of employment, winter 2025–26.

Non-adopters are held at zero.

+4.8 pts
What the aggregate productivity series reports
15 pts
Gap between an adopter and a non-adopter
68%
Of employment that sees none of it
Employment at adopting firmsEveryone else

One level down: the worker

A salary is hourly billing. A worker who supervises agents and produces more for the same pay has cut their own price. Set what they produce, what they negotiated, and what they own — and see who keeps the difference.

New output the worker created $400k
Employer
—
Worker
—

Same role, same quality, same $200,000 base pay.

What the worker extracted through scarcity or an outside option.

Equity, profit share, or their own business. Applies to whatever is left after pay.

Baseline: the worker's output before agents is valued at their pay. New output is pay × (multiplier − 1). Pay changes come off the top; the ownership stake divides the remainder.