Job Search Pressure by State: Methodology

The measure

Unemployed people per job opening: a state’s unemployment level divided by its job openings, both seasonally adjusted, both for the same month. Openings are published in thousands and unemployment in whole people, so the openings figure is multiplied by a thousand first.

A reading of 1.00 means there is one unemployed person for every advertised vacancy. Above 1.00, unemployed people outnumber openings. In December 2025 that was true in 27 of 51 states.

Why this and not a state Friction Index

The national Job Search Friction Index combines three standardised dimensions, one of which is how long people stay unemployed. No comparable monthly state series exists for that dimension. Producing a state “Friction Index” would mean substituting different inputs while keeping a name readers would reasonably assume meant the same thing.

A simple ratio has a further advantage: anyone can check it. Both inputs are published, the arithmetic is one division, and there is no standardisation, weighting or rescaling to take on trust.

What is ours and what is not

The ratio itself is not an original measure. The Bureau of Labor Statistics publishes unemployed persons per job opening by state, and this atlas is built from the same two published series anyone else would use. Presenting the ratio as a Jobsearch.ing measurement would be a straightforward misrepresentation.

What is added is the treatment: all 51 places ranked in one table, 119 months of history for each, a percentile against each state’s own past rather than only against other states, the year-on-year change and rank movement, a comparison with bordering states, a single downloadable file carrying all of it, and the rank-correlation checks showing what the ranking does and does not restate.

Whether that treatment is worth citing over the source is a judgement for the reader. It is stated here so the judgement can be made with the facts.

Why the period is December 2025

State JOLTS is released well behind the national series. At the time of retrieval, state openings ran to December 2025 while state unemployment ran several months later and the national index reached June 2026.

The atlas uses the latest month both state sources cover. Pairing a state’s openings with a later month’s unemployment would produce a fresher-looking number that compared two different periods, which is not a measurement. Where a figure is eight months old, the page says so rather than implying currency it does not have.

Historical percentiles

Each state’s current ratio is ranked against its own history — 107 months drawn from the 119 months since January 2016, excluding March 2020 to February 2021. That window is excluded for the same reason the national index excludes it: a mass layoff and rehire makes the ratio describe something other than job-search conditions.

A percentile is against that state’s own past, never against other states. A state can rank near the top of the country and near the bottom of its own range at the same time, and both statements are true.

Does it just restate the unemployment rate?

No, though it is related to it. Across the 51 places, the ranking by unemployed-per-opening and the ranking by unemployment rate correlate at 0.679 (Spearman). That leaves substantial disagreement: a state can have few unemployed people and even fewer vacancies.

More striking, the ranking is almost unrelated to the state hires rate (0.006). Competition for openings and the speed at which employers actually hire are close to independent across states — the same pattern the national Job Market Disconnect describes over time.

The national comparison

The national figure shown alongside the states (1.15 for December 2025) is computed from the national JOLTS and household survey series, not by summing the states. BLS models state and national estimates separately and they do not add up exactly.

What it cannot support

• Any claim about a city, county or metropolitan area. State averages hide large internal differences.
• Any claim that the openings in a state suit the people looking for work there — by skill, seniority, industry, location or pay.
• Any claim about how quickly people are hired. The hires rate is published alongside and ranks differently.
• Any claim about people not counted as unemployed, including those who have stopped looking and those searching while employed.
• Any forecast. The measure describes a month that has already happened.

Revisions

Both sources revise. State JOLTS estimates are re-benchmarked and LAUS figures are updated routinely, so a figure published here may differ from the same month retrieved later. Every page carries the retrieval date, and the published data files carry it per row.

Source series

• JTS0000000<ST>0000JOL — Job openings level by state, Job Openings and Labor Turnover Survey, U.S. Bureau of Labor Statistics
• JTS0000000<ST>0000JOR — Job openings rate by state, Job Openings and Labor Turnover Survey, U.S. Bureau of Labor Statistics
• JTS0000000<ST>0000HIR — Hires rate by state, Job Openings and Labor Turnover Survey, U.S. Bureau of Labor Statistics
• LASST<ST>0000000000004 — Unemployment level by state, Local Area Unemployment Statistics, U.S. Bureau of Labor Statistics
• LASST<ST>0000000000003 — Unemployment rate by state, Local Area Unemployment Statistics, U.S. Bureau of Labor Statistics

<ST> is the two-digit state FIPS code. All series are seasonally adjusted and were retrieved from the BLS public API on 2026-08-22, covering 51 states and the District of Columbia.