Fidelity-scored labor index

Pharometric

An index of entry-level hiring contraction — measured at county × occupation-family, monthly, against each cell's own pre-2020 baseline and hiring cycle. Where contraction coincides with high AI task-exposure, we flag the cell an AI-attributable candidate — never a causal claim.

Pharos, gr. lighthouse — the instrument the ancient world built to make a hazard visible before it was felt.

What the index measures

Pharometric fuses Census QWI administrative hires data, JOLTS, and O*NET task data through generalized least-squares estimation into a single per-cell reading of entry-level hiring contraction — at county × occupation-family resolution, monthly. Every cell carries a fidelity score: an explicit statement of how much that reading should be trusted.

The index measures contraction. It does not, by itself, measure cause — attribution to AI is a separate, deliberately weaker layer, published only as a labeled candidate signal.

Read the full methodology →

Designed for the calculation-agent role

An index used to settle a contract has to be objective, independent of both parties, and reproducible from a disclosed method. Pharometric is built — immutable timestamped releases, a pre-registered falsification test, a disclosed revision policy — to carry that governance from day one, not add it after the fact.

Resolution

County × occupation

Cadence

Monthly

Method

GLS fusion

Anchor source

Census QWI

What actuaries ask of an index

01
History
Enough back-data to estimate how often a trigger fires, and how severely.
02
Revision policy
A disclosed rule for which timestamped release settles a contract — later restatements don't reopen it.
03
Manipulation resistance
Independent, government-grade sourcing that neither party to a contract controls.
04
Independence & transparency
Computed by a party with no stake in the payout, from a method others can inspect.
05
Objectivity / reproducibility
Same inputs, same output — every time.

In development

Currently in discovery — not yet publishing.

Pharometric is pre-launch: the backtest and falsification test are underway ahead of a first public release. If you underwrite emerging perils, structure parametric products, or hold capacity that prices on diversification, this is a good time to talk.

jdeford@gmail.com · Methodology →