jobdownupSOURCES
Every stat, sourced.
Credibility is the brand. Every number displayed on jobdownup.app lives here with its citation, the period it describes and the date we checked it. The market figures below are rendered from the same records the rest of the site renders, so a statistic cannot appear on a page here without appearing on this one.
Where a number could not be traced to the organisation that published it, it was removed rather than reworded. Three were.
Labour market
National statistics offices and the organisations whose own data it is. Primary sources only.
UNITED STATES·July 2026·checked 2026-08-26
The US unemployment rate was 4.1% in July 2026, and non-farm payrolls fell by 23,000 over the month.
Hiring is not merely slower. It went slightly backwards.
UNITED STATES·July 2026·checked 2026-08-26
1.8 million Americans had been out of work for 27 weeks or more in July 2026 — 25.5% of everyone unemployed.
A quarter of the people looking have been looking for over six months. That is the norm now, not the tail.
UNITED STATES·July 2026·checked 2026-08-26
The average spell of unemployment in the United States ran 24.9 weeks in July 2026 — nearly six months.
Budget for half a year, not for a fortnight.
GERMANY·July 2026·checked 2026-08-26
German unemployment passed three million in July 2026 — 3,007,000 people, a rate of 6.4%, up 71,000 in a month.
A threshold Germany had stayed under for a decade.
GERMANY·2026, year to June·checked 2026-08-26
Around 1.07 million people in Germany have been unemployed for a year or longer — roughly 35% of everyone registered unemployed.
Germany counts long-term at twelve months, not six. On that stricter test, a third of the queue is in it.
UNITED KINGDOM·April to June 2026·checked 2026-08-26
There were 2.5 unemployed people for every vacancy in the UK in April to June 2026, with vacancies down to 707,000 — below pre-pandemic levels.
The queue is not imaginary. It is two and a half deep, per opening.
UNITED KINGDOM·April to June 2026·checked 2026-08-26
474,000 people in the UK have been unemployed for more than twelve months — the highest figure since January 2016.
The searches that stall are stalling for longer than at any point in a decade.
UNITED KINGDOM·April to June 2026·checked 2026-08-26
The UK unemployment rate was 4.9% in April to June 2026, up 0.2 percentage points on the year.
AUSTRALIA·July 2026·checked 2026-08-26
Australian youth unemployment stood at 10.4% in July 2026, against 4.5% for the workforce as a whole, with underutilisation at 10.8%.
Young Australians are looking more than twice as hard as everyone else.
NEW ZEALAND·June 2026 quarter·checked 2026-08-26
440,000 New Zealanders wanted more work in the June 2026 quarter — an underutilisation rate of 13.8%, with unemployment at 5.6%.
171,000 unemployed, 154,000 working fewer hours than they want, 115,000 who want a job and had stopped actively looking.
GLOBAL·Q4 2024, reported January 2025·checked 2026-08-26
Between 18% and 22% of jobs posted on the Greenhouse platform in any given quarter are ghost jobs — roles the employer is not actually trying to fill. In 2022 the figure was 12–15%.
Roughly one listing in five was never going to hire anyone.
GLOBAL·Published 12 August 2026·checked 2026-08-26
Employment among workers aged 22–25 in the most AI-exposed occupations now sits about 19% below where it would be had it kept pace with similarly aged workers in less-exposed occupations.
The researchers are explicit that this is not economy-wide displacement, and that these are descriptive patterns rather than causal estimates. The gap is concentrated at the entry level, and it has widened from 13% in November 2025.
GLOBAL·December 2024·checked 2026-08-26
61% of job seekers have been ghosted after a job interview — up nine percentage points in eight months.
Not silence on an application. Silence after you took time off, prepared, and sat in the room.
GLOBAL·2024 research year·checked 2026-08-26
29% of North American candidates had not heard back from an employer one to two months after applying.
Worth stating precisely, because the inflated version of this figure is everywhere. It is roughly three in ten, not three in four — and it improved by 19% on the year.
GLOBAL·Spring 2026·checked 2026-08-26
About 42% of recent US college graduates were underemployed in spring 2026 — working a job that did not require their degree — the highest share since 2020.
Measured here
Read from our own crawl database on the date shown. There is no external citation to give, so the method is written out instead.
VERIFIED·retrieved 2026-08-23
70,000+ live roles, open right now, read from the employer.
jobdownup crawl database — count of board_roles with no close date recorded. Measured 76,396; stated as a floor because the corpus grows daily.
Used on: / (scale stats)
VERIFIED·retrieved 2026-08-23
1,200+ employers, career pages read directly.
jobdownup crawl database — count of resolved employer boards. Measured 1,242.
Used on: / (scale stats)
VERIFIED·retrieved 2026-08-23
16 hiring systems read natively.
jobdownup crawl database — distinct ATS platforms with at least one live employer board (Greenhouse, Workday, Ashby, SmartRecruiters, Lever, SuccessFactors, OracleHCM, iCIMS, Phenom, Recruitee, Personio, Radancy, SnapHire, LiveHire, Workable, Teamtailor).
Used on: / (scale stats)
VERIFIED·retrieved 2026-08-23
Re-read daily — a filled role disappears within a day.
jobdownup crawl scheduler — 982 of 1,242 boards re-read in the 24h before measurement; every board holding an open posting had been read within 7 days. A posting is closed when it vanishes from the employer's own page.
Used on: / (scale stats)
VERIFIED·retrieved 2026-08-23
Roles are scored a second time for how gettable they are, from how long they have stayed open against the median for that profession, market and employer.
jobdownup winnability model — computed from posting history (first seen, employer's own posted date, close date) against per-market and per-employer medians. Deterministic; no model call. Repost history is part of the model but is not yet carrying data, so current scores rest on the age signals.
Used on: / (second axis section)