The Whetstone Sign in

From the archive A piece from the site's earlier corpus, kept for reference. New briefings are written by hand. Browse the archive →

Briefing · economics

Does raising the minimum wage cost jobs?

25 June 2026

Thirty years after Card and Krueger, the question still splits economists, and it has narrowed to one point: how competitively does the low-wage market actually clear?

Our read

If forced to bet, the monopsony read is stronger at the wage levels actually proposed.

Full editor's view, and the devil's advocate ↓ Briefing: engine-assisted · view: human

The commentary

Three positions, audited in full

Each makes a different kind of case against the reports: the measure, the load-bearing input, the clock.

Price floors cut jobs
1 Jardim et al. (Seattle) NBER Full audit leans competitive market

The disemployment case now leans on its cleanest natural experiment. Studying Seattle's ordinance, Jardim and colleagues found that

“the second wage increase to $13 reduced hours worked in low-wage jobs by 6-7 percent”

That is the textbook price-floor effect, turning up in the hours data. Read as a general law it says: raise the wage floor and the market buys less low-wage labour.

Audit Hasty Generalisation Logic

It is one city, at a wage ($13) near the top of the range anyone had studied; treating Seattle's result as what every labour market does assumes they all clear the same way. That is the very point in dispute.

Which is where the evidence comes in, and the other camp denies exactly that.

Monopsony absorbs it
2 Cengiz et al. NBER Full audit leans monopsony power

The other camp answers from the accumulated record. Pooling 138 state minimum-wage changes, Cengiz and colleagues report that

“the overall number of low-wage jobs remained essentially unchanged over five years following the increase”

Pay rose, employment held. Empirically it is the stronger hand. Yet it makes a quiet move of its own, sliding from the increases actually measured to increases of any size.

Audit Equivocation Logic

It slides from "increases within the studied range left employment unchanged" to "no increase could cost jobs". Those are two different claims; the measured changes never approached the levels the disemployment case worries about.

The field: who stands where

Each voice in its strongest form. Inclusion is not agreement.

The map · 2 audited positions

competitive marketmonopsony power
12numbers match the cards below · tap a disc to read the source
how to read the map

Tap a numbered marker to jump to that voice’s card; the same number sits on the card. A wider marker means we are less sure exactly where that voice sits on the axis (how placement works).

The shared assumption

Both sides, though, stand on the same floor. Each treats today's labour market as the fixed backdrop; neither asks what a market with sectoral bargaining, or a higher wage held for a decade, would do. That is the question that would actually settle it.


The editor's view

written by hand · no engine text

If forced to bet, the monopsony read is stronger at the wage levels actually proposed. But I hold it loosely, because those proposals keep drifting past the range we've studied.

Devil's advocate

A large enough jump leaves that studied range, and the textbook result could reassert itself.

Other takes · audited

“Our empirical findings challenge the prediction that a rise in the minimum reduces employment.”

Card & Krueger (1994) · NBER

Audit: One industry, one state border, one small step ($4.25 to $5.05). A null result there is read as a general verdict, yet it says nothing about the large increases now on the table. Absence of an effect at the bottom of the range is not its absence across it.

“among the papers we view as providing the most credible evidence, almost all point to negative employment effects”

Neumark & Wascher (2006) · NBER

Audit: The weight of the literature is doing the work, but the weighing is theirs ("the papers we view as providing the most credible evidence"), so the count turns on which studies were let in. A consistent negative sign still fixes neither the size of the effect nor whether it survives at the wages actually proposed.

Does a reading here seem wrong? Tell the editor →


Sources

Positions · 2
Further evidence · 8

New briefings by email

When one publishes. Nothing else.

Your address is stored for briefing updates only. Nothing else, no tracking.

Another briefing

Will generative AI eliminate creative jobs?

Run the engine yourself

The audit behind every card on this page runs free on any text.

Open the Audit →