The Net Labor Capacity index is demand for hours relative to the hours that will exist, at each horizon, indexed to 100 at a base date. Above 100, forward labor scarcity is tightening relative to base; below 100, loosening. It is the family’s flagship analytical product and its eventual settlement objective — and it is staged deliberately behind the two legs it is built from.
Status: The composites launch after $BIRTH, the S-series and the demand leg, and their settlement is offered only after twelve consecutive official prints each carrying a fully seasoned settlement A-R. This page describes the construction as specified in the methodology paper.
The economic reading is direct: NLC is a forward wage-pressure gauge. A rising NLC-10 says the economy a decade out demands more hours relative to the hours that will exist than it did at base. The ratio form is chosen over a difference because it is unit-free, scale-invariant, and robust to proportional errors that strike both legs equally.
The horizon grid is chosen because the composition of uncertainty inverts across it. Supply carries two terms with different properties. The cohort term is how many people there will be. The schedule term is how much labor each supplies. The cohort term is closed at every horizon on the grid: everyone the core series counts at any horizon through twenty years is already born and counted, the youngest cohort entering through realized births. The schedule term is frozen, so it adds nothing to index variance between rebalancings, and it is never certain about future behavior, at five years or at twenty. At five years the schedule has had little time to drift, so nearly all index variance comes from the demand leg’s automation term. At twenty years the schedule and automation terms are both materially uncertain while the cohort term stays closed and is fed by a live signal: a person aged twenty at that horizon is being born now. Realized births, published through $BIRTH, are the only input in the system that updates the twenty-year supply leg in real time. Each print carries a variance-attribution table that quantifies this inversion input by input.
Settlement composites use D-R only. Research composites publish in both D-R and D-X versions, with credible intervals on the D-X versions; the monthly print shows NLC-5, NLC-10 and NLC-20 in the settlement construction, their research D-X counterparts, and DRR. The base date for the composites is set separately from the supply product’s, because the composites depend on an A-R history the supply product does not.
Composite settlement inherits the full defensibility burden of A-R — attribution rules, corroboration, seasoning and the public event log — which is why it is staged: no representation is made that the composites are settlement-ready before the A-R event system has produced twelve consecutive official prints each carrying a fully seasoned settlement value, one complete pass of the twelve-month aggregation window.
The composites do not claim to forecast wages or employment; they measure the balance of hours demanded and hours available under published constructions, and publish the sensitivity of that balance to every frozen schedule. The demand leg’s limitations — attribution bias in D-R, overstatement in D-X — pass through to the composites and are disclosed rather than netted away, which is why the D-R and D-X versions are published side by side.
The methodology paper is available on request; the FAQ answers the common ones on coverage, revisions and what each tier permits.