95 figures across 12 dossiers were printed by this platform and are no longer defended. Each is quoted as it appeared, dated, and followed by what the attributed document actually says.
23 of them are restated from a document that was located and read; the others leave the dossier without a replacement, because no located document prints a value.
Said to measure: the size of Schneider Electric’s French industrial and R&D footprint
Attributed to: Schneider Electric France — Schneider Electric en France — carrières (undated careers page)
What was read: The company’s French careers page carries no publication date and revised its own count between two readings: it stated 43 sites on 4 September 2026 and states more than 15,000 employees across 55 sites when read on 17 September 2026. The earlier count is withdrawn and the dossier now prints the day the page was read.
Restated as: more than 15,000 across 55 sites — citation [2]
The publisher printed this figure and then revised it. The earlier value is withdrawn and the revision is named, because the reason for a revision is itself the finding.
Said to measure: the company’s own French site count, used to weigh a larger figure quoted elsewhere
Attributed to: Schneider Electric France — Schneider Electric en France — carrières (undated careers page)
What was read: The same undated careers page this dossier weighs the “about 100 sites” claim against revised its count: 43 sites when read on 2 September 2026, more than 15,000 employees across 55 sites when read on 17 September 2026. The comparison stands on the current reading, with the reading day printed.
Restated as: more than 15,000 across 55 sites — citation [18]
The publisher printed this figure and then revised it. The earlier value is withdrawn and the revision is named, because the reason for a revision is itself the finding.
Said to measure: the pay band for a cyber security engineer with three to five years of experience in French defence
Attributed to: No publisher: the table carried no source — Salary Benchmarks table printed in this dossier with no cited document
What was read: The table was published as a benchmark and cited nothing. No dated document this desk could open prints pay bands by role and seniority for the French defence sector, so the five bands and the clearance premium are withdrawn rather than re-sourced.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
€60,000–€90,000
Said to measure: the pay band for a defence systems engineer with five to ten years of experience
Attributed to: No publisher: the table carried no source — Salary Benchmarks table printed in this dossier with no cited document
What was read: The band named MBDA, Thales and Naval Group beside it without citing any of their published pay documents. None of the three publishes a dated band for this role, and no state or sector source read for this dossier prints one either.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
€55,000–€85,000
Said to measure: the pay band for an artificial intelligence or machine learning engineer working in defence
Attributed to: No publisher: the table carried no source — Salary Benchmarks table printed in this dossier with no cited document
What was read: The row also printed the direction of travel as "rapidly increasing", which no dated series supports for this role in France. The band and the trend leave the dossier together.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
€80,000–€130,000
Said to measure: the pay band for a programme manager with more than ten years of experience on major programmes
Attributed to: No publisher: the table carried no source — Salary Benchmarks table printed in this dossier with no cited document
What was read: FCAS and Rafale were named as the programmes this band applies to. The published record for both is procurement and expenditure, and neither prints the remuneration of the people managing them.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
€55,000–€80,000
Said to measure: the pay band for an export control specialist with more than five years of experience
Attributed to: No publisher: the table carried no source — Salary Benchmarks table printed in this dossier with no cited document
What was read: The instruments this dossier does cite, Regulation (EU) 2021/821 and 22 CFR Parts 121 and 122, define the controlled scope and the registration duty. They say nothing about what the people who administer them are paid.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
+20% with Secret clearance
Said to measure: the pay premium a Secret-level clearance adds to a cyber security engineer’s salary in France
Attributed to: No publisher: the table carried no source — Salary Benchmarks table printed in this dossier with no cited document
What was read: The IGI 1300 of 9 August 2021 governs how a clearance is granted and for which post; it prices nothing. No published French series compares the pay of cleared and uncleared holders of the same post, so the premium is withdrawn as a number while the eligibility mechanism it sat beside stands.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
Said to measure: a four-step timetable of skill disruption from 2025 to 2030
Attributed to: World Economic Forum — Future of Jobs Report 2025
What was read: The Future of Jobs Report 2025 prints one figure for the period: on average 39% of workers’ existing skill sets will be transformed or become outdated over 2025–2030. It prints no year-by-year ladder.
Restated as: 39% over 2025–2030 — citation [2]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The publisher revised the figureWithdrawn
44% of workers’ core skills will be disrupted by 2027
Said to measure: the share of core skills the WEF expects to be disrupted by 2027
Attributed to: World Economic Forum — Future of Jobs Report 2025
What was read: The 44% figure belongs to the Future of Jobs Report 2023, which printed it over a five-year horizon. The 2025 edition supersedes it: on average 39% of workers’ existing skill sets will be transformed or become outdated over 2025–2030.
Restated as: 39% over 2025–2030 — citation [2]
The publisher printed this figure and then revised it. The earlier value is withdrawn and the revision is named, because the reason for a revision is itself the finding.
The document does not print the figureWithdrawn
12% of professionals
Said to measure: the share of professionals with a documented skill development strategy
Attributed to: Coursera — Global Skills Report 2024
What was read: The Global Skills Report 2024 was read in full. It measures learner enrolments and proficiency by country and prints no share of professionals holding a documented skill strategy.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
68% of upskilling initiatives
Said to measure: the share of upskilling initiatives abandoned before month 18
Attributed to: Coursera — Global Skills Report 2024
What was read: The Global Skills Report 2024 prints no abandonment rate for upskilling initiatives and no eighteen-month horizon; the sentence attributed the figure to OECD data that names no such measure either.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
professionals who actively signal their skill combinations receive 67% more inbound opportunities
Said to measure: the gain in inbound recruiter approaches earned by publicly signalling a skill combination
Attributed to: LinkedIn — attributed in the prose to “the LinkedIn data”, with no edition named
What was read: LinkedIn publishes no dated edition that prints this comparison. Its Workplace Learning and Future of Work reports were read for 2023, 2024 and 2025 and none of them measures inbound opportunities against profile signalling, so the figure is withdrawn and the sentence now says plainly that this desk prices no such gain.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
report 2.8x better returns on L&D investment
Said to measure: the return a tiered training budget earns over an evenly distributed one
Attributed to: McKinsey & Company — attributed in the prose as “(McKinsey, 2024)”, with no document named
What was read: No McKinsey publication of 2024 prints a 2.8x return on tiered learning and development spend. The dossier cites two McKinsey documents and neither contains the comparison, so the number leaves the argument while the tiering stays as this desk’s own framework.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
The OECD average is 4.7 months; top-performing organizations achieve 2.1 months
Said to measure: an OECD benchmark for the time between the start of training and productive use of the skill
Attributed to: OECD — attributed in the prose to an OECD average, with no edition named
What was read: The OECD publishes participation, duration and outcome statistics for adult learning and prints no average time to competency in months. Employment Outlook 2024 and Skills Outlook 2025, both cited by this dossier, were read for the measure and neither states it, so the pair is withdrawn and the reader is told to measure their own baseline.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The World Economic Forum’s 2024 Future of Jobs Report
Said to measure: the document said to project 69 million jobs created and 83 million displaced by 2027
Attributed to: World Economic Forum — Future of Jobs Report 2024
What was read: The Forum published no Future of Jobs Report in 2024; the editions are 2023 and 2025. Every figure this dossier prints — 69 million created, 83 million displaced, 23% churn to 2027, and the role-by-role percentages — comes from The Future of Jobs Report 2023, published 30 April 2023. The figures stand; the edition credited to them did not exist.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
Said to measure: the share of employers worldwide who say they cannot fill roles
Attributed to: ManpowerGroup — Talent Shortage Survey 2024
What was read: ManpowerGroup’s 2024 Global Talent Shortage prints one headline share, and it is not 60%: "Today, 75% of employers report difficulty in filling roles", with 77% in health care and life sciences and 76% in information technology. 60% was the 2019 reading; the dossier carried it forward under a 2024 title.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
67 million young people actively seeking work
Said to measure: the number of unemployed young people worldwide standing beside the shortage figures
Attributed to: International Labour Organization — Global Employment Trends for Youth 2024
What was read: The ILO’s Global Employment Trends for Youth 2024, published 12 August 2024, prints the count precisely: "In 2023, 64.9 million young people aged 15 to 24 were unemployed worldwide" — the lowest number in fifteen years, at a 13.0% rate. 67 million is not in the report; it rounds up a figure the report deliberately states as a fifteen-year low.
Restated as: 64.9 million young people were unemployed worldwide in 2023 — citation [8]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
removing degree requirements increases the qualified applicant pool by 39%
Said to measure: the measured widening of the qualified pool when a degree filter is dropped
Attributed to: Harvard Business School — Dismissed by Degrees
What was read: Dismissed by Degrees (Fuller and Raman, October 2017) was read end to end. It prints degree-inflation findings — the share of postings demanding a degree for middle-skills roles, longer time to fill, higher cost, and 39% of employers reporting higher turnover among degree holders — and no measurement of a qualified pool widening by any percentage. The only 39% in the report describes turnover, not supply.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
72% of job postings in OECD countries still require a specific degree
Said to measure: the share of OECD job postings gated by a degree requirement
Attributed to: Organisation for Economic Co-operation and Development — Skills Outlook 2025. OECD Publishing, Paris
What was read: The OECD Skills Outlook 2025, published 9 December 2025, was read end to end. It discusses degree requirements twice — urging governments to remove unnecessary ones from public-sector hiring — and prints no share of postings requiring a degree, in any economy or aggregate. The figure is not in the edition it was credited to.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
30% of "unfillable" positions could be filled by existing candidates
Said to measure: the share of unfillable vacancies a shared skills taxonomy would close
Attributed to: Organisation for Economic Co-operation and Development — Skills Outlook 2025. OECD Publishing, Paris
What was read: The same reading of the OECD Skills Outlook 2025 finds no estimate of unfillable vacancies closed by a common taxonomy. The report describes ESCO as a "dictionary" for skills vocabularies and attaches no share of vacancies to it. The number was our inference presented as the OECD’s estimate.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
Said to measure: the share of applicants Google is said to hire
Attributed to: Alphabet Inc. — Annual report and investor communications
What was read: Alphabet’s annual report and investor pages state headcount, never an application-to-hire ratio, and no Google publication prints one. The only ratio Google has put on the record came from Laszlo Bock: about 7,000 hires against 3 million applications — near 0.2%, five times smaller than the figure printed here.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The figure measures something elseWithdrawn
approximately 3.3 million job applications per year
Said to measure: the number of applications Google is said to receive every year
Attributed to: Alphabet Inc. — Annual report and investor communications
What was read: 3.3 million is a single year’s count, not an annual rate: Google gave it to Axios for 2019, and has published no comparable figure since. Printed as a recurring per-year volume it states something the publisher never measured.
Restated as: 3.3 million job applications in 2019 — citation [6]
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
The document does not existWithdrawn
r=0.58, vs. r=0.31 for positional authority
Said to measure: the correlation between emergent leadership and team performance, against positional authority
Attributed to: Journal of Organizational Behavior — 2023 study of emergent leadership
What was read: No 2023 article in the Journal of Organizational Behavior prints these coefficients. The journal’s 2023 paper on the subject — Galvin, Badura, LePine and LePine, “A theoretical integration of leader emergence and leadership effectiveness”, 10 June 2023 (doi:10.1002/job.2724) — is a theoretical integration and reports no correlation of either size.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
rapid learners outperformed experienced hires by 23% at the 12-month mark
Said to measure: a Google internal result comparing fast learners with experienced hires after a year
Attributed to: Google re:Work — Google internal people-analytics research
What was read: Neither re:Work nor any other Google publication prints a 23% performance gap between fast learners and experienced hires. The internal research Google has published — Project Oxygen and Project Aristotle, both cited by this dossier — measures manager and team behaviour, not learning speed against tenure.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
2.5x more likely to be rated "transformative" after two years
Said to measure: the odds ratio said to attach to a high “Googleyness” score two years after hire
Attributed to: Google re:Work — Google internal people-analytics research
What was read: “Transformative” is not a rating category Google publishes, and no Google document prints a 2.5× multiplier for any attribute score. The attribute itself is documented in Google’s structured-interviewing guidance; the multiplier is not.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
15% of Google employees on some teams have no college degree
Said to measure: the share of Google staff on some teams said to hold no degree, as of 2024
Attributed to: Google — Google workforce composition (unnamed)
What was read: Google publishes workforce composition by gender, race and region in its diversity reporting, and has never published a share of employees without a degree — for the company, for a team, or for any year. There is no document behind the 15%.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The figure measures something elseWithdrawn
0.51 vs. 0.20 for unstructured (Huffcutt & Arthur, 1994)
Said to measure: the validity of structured against unstructured interviews, credited to Huffcutt and Arthur
Attributed to: Journal of Applied Psychology — Huffcutt & Arthur (1994), Hunter and Hunter (1984) Revisited: Interview Validity for Entry-Level Jobs
What was read: Huffcutt and Arthur (1994), Journal of Applied Psychology 79(2), 184–190, band interviews by degree of structure and report a corrected validity near 0.37 for the most structured band; the 0.51 / 0.20 pair is not their published pairing. The document this dossier already cites for the comparison is Schmidt and Hunter (1998), whose table prints 0.51 for structured and 0.38 for unstructured interviews.
Restated as: structured interviews at 0.51 against 0.38 for unstructured — citation [5]
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
The document does not print the figureWithdrawn
76% of talent professionals prioritize skills-based hiring
Said to measure: the share of talent professionals said to prioritise skills-based hiring
Attributed to: LinkedIn — Global Talent Trends 2024
What was read: LinkedIn’s talent research for 2024 — Global Talent Trends (30 October 2024) and The Future of Recruiting 2024 — was located and read: neither prints this share. The citation also names a rolling resource page rather than an edition, so there was no dated document to check it against in the first place.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
60% better quality of hire with skills-based approaches
Said to measure: the improvement in quality of hire attributed to skills-based approaches
Attributed to: LinkedIn — Global Talent Trends 2024
What was read: The same two LinkedIn publications print no quality-of-hire improvement of any size, and “quality of hire” is not a measured variable in either. A percentage improvement in an unmeasured variable cannot be repaired; it is withdrawn.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
20% drop in degree requirements in tech postings since 2021
Said to measure: the fall in degree requirements in technology job postings since 2021
Attributed to: LinkedIn — Global Talent Trends 2024
What was read: Neither located LinkedIn publication prints a change in degree requirements against a 2021 base. LinkedIn does publish postings analyses, but none this desk could locate carries this measurement, so the trend leaves the dossier with the two shares beside it.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
skills-based organizations are 63% more likely to achieve results
Said to measure: the advantage skills-based organisations are said to hold over traditional hiring
Attributed to: McKinsey Global Institute — Skills-Based Organizations (2024)
What was read: No McKinsey or McKinsey Global Institute publication of that title exists, and none prints this figure. The number circulates in third-party marketing content, where it is attributed to Deloitte rather than McKinsey — a publisher swap that is itself the finding. McKinsey’s own skills work (“Taking a skills-based approach to building the future workforce”, 15 November 2022) prints nothing of the kind.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
73% of hiring managers say a strong GitHub profile outweighs credentials
Said to measure: the share of hiring managers said to value a GitHub profile above credentials
Attributed to: HackerRank — 2024 Developer Skills Report
What was read: HackerRank’s 2024 Developer Skills Report (published 5 February 2024) surveys developers and hiring managers on skills, AI use and interview practice, and prints no comparison between a GitHub profile and credentials. The report is real, dated and does not contain the figure.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
65% of Fortune 500 companies will use simulation-based hiring by 2027
Said to measure: a World Economic Forum prediction about simulation-based hiring among the Fortune 500
Attributed to: World Economic Forum — The Future of Jobs Report (2024)
What was read: The Forum published no Future of Jobs Report in 2024 (the editions are 2023 and 2025), and no WEF publication forecasts simulation-based hiring among the Fortune 500 for 2027. The prediction has no document, and the edition it was credited to does not exist.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
roughly 320,000 people across more than sixty countries
Said to measure: the size and geographic spread of the bank the dossier reads
Attributed to: J.P. Morgan — Annual Report and ESG Report
What was read: The 2024 Form 10-K, filed 14 February 2025, prints the figures exactly: 317,233 employees globally as of 31 December 2024, an increase of 7,307 on the prior year, located in 66 countries, with 59% of employees in the United States, and $4.0 trillion in assets. Neither 320,000 nor "more than sixty countries" is printed anywhere in it; both were rounded away from a disclosure precise enough to quote.
Restated as: 317,233 employees in 66 countries — citation [1]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
Said to measure: the price premium French provenance is said to command over comparable goods
Attributed to: Bain & Company — Global Luxury Market Study
What was read: The Bain & Company–Altagamma Luxury Goods Worldwide Market Study, Fall 2024 (23rd edition, 13 November 2024) was read in full: it names France twice, in market tables, and measures no provenance premium of any size.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
the premium rises to 38%
Said to measure: the premium said to attach to leather goods designed and manufactured in France
Attributed to: Bain & Company — Global Luxury Market Study
What was read: The same edition carries no leather-goods provenance breakdown; its category tables report market size and growth, never a manufacturing-origin premium.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
~25x forward P/E
Said to measure: the forward earnings multiple at which LVMH shares were said to trade
Attributed to: Bloomberg — Bloomberg Terminal — LVMH Financial Data (MC FP Equity)
What was read: A terminal screen is a live product, not an edition: no dated document this dossier cites prints this multiple, and none can be identified after the fact.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
the CAC 40 (14x)
Said to measure: the index earnings multiple LVMH was said to trade at a premium to
Attributed to: Bloomberg — Bloomberg Terminal — LVMH Financial Data (MC FP Equity)
What was read: The index multiple shares the terminal’s provenance and the same defect: there is no dated edition to name, so the comparison cannot be checked by a reader.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
its 5-year average (28x)
Said to measure: LVMH’s own five-year average earnings multiple, said to sit above the current one
Attributed to: Morgan Stanley Research — Global Luxury Outlook 2025
What was read: A sell-side note circulated to clients is not a document a reader can obtain; no public edition of it exists, so the five-year average leaves the dossier with the multiples it framed.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
213,000
Said to measure: the Group headcount at the end of 2023, printed in the metrics table and in a section heading
Attributed to: LVMH — LVMH 2023 Annual Report
What was read: The 2023 Universal Registration Document prints the headcount to the unit: "Total headcount as of December 31, 2023 stood at 213,268 employees", of whom 196,686 were on permanent contracts. A rounded figure standing in front of an exact one loses the reader nothing but our credit for having read the filing.
Restated as: 213,268 — citation [4]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The figure measures something elseWithdrawn
5,664
Said to measure: the number of stores in the Group network in 2023
Attributed to: LVMH — LVMH 2023 Annual Report
What was read: The filing prints both numbers in one sentence: "The total number of stores in the Group’s network rose from 5,664 in 2022 to 6,097 in 2023." The dossier printed the 2022 count in a column headed "2023 Actual" — the number is real, and it measures the year before.
Restated as: 6,097 — citation [4]
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
Said to measure: five-year compound growth rates for each headline metric, credited to a sell-side note
Attributed to: Bernstein Research — The Luxury Sector in 2025: Structural Winners
What was read: Bernstein Research publishes no such note to any reader: sell-side research is distributed to clients and carries no public edition, no page a reader can open and no date a reader can quote. A growth rate a reader cannot re-derive is an assertion, and this dossier now prints only rates it derives from the filings it cites.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
the Industry Rank column: #1 on revenue, operating profit, employees, stores and Maisons
Said to measure: a world ranking of the Group on five separate measures
Attributed to: Bernstein Research — The Luxury Sector in 2025: Structural Winners
What was read: No published league table ranks luxury groups on employees, store count and number of brands, and the note credited with it is not published at all. The revenue lead is visible in the peers’ own filings; the other four ranks were never measured by anybody.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
an estimated net worth of €233 billion
Said to measure: the personal wealth of the Group chairman, in euros, as of March 2025
Attributed to: Forbes — Forbes Real-Time Billionaires List
What was read: The Real-Time Billionaires List is a live screen, not an edition: it recomputes through the trading day, keeps no archive a reader can open at a stated date, and quotes dollars, never euros. There is no dated Forbes document behind a euro figure, and the dossier’s argument never needed one.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
3.2 million job applications in 2023
Said to measure: the volume of applications the Group received in one year
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: LVMH publishes no "Social & Environmental Responsibility Report 2023": its non-financial disclosure is the social and environmental chapter of the Universal Registration Document, which reports hires, departures, turnover and training and prints no application count in any year. The number has no publisher.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
a selectivity rate of approximately 6.6%, comparable to Harvard’s admission rate
Said to measure: how selective the Group is as an employer, expressed as hires over applications
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The rate was computed from two figures the filing does not print together — an application count that appears in no LVMH publication and a hiring total that does not match the filing’s recruitment table. A ratio inherits the status of its worst input, and this one had no numerator at all.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
42,000
Said to measure: new hires made by the Group in 2023
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The 2023 filing prints the recruitment total in its own table: 60,379 employees hired on permanent contracts during 2023, against 43,363 permanent-contract departures. 42,000 appears nowhere as a hiring figure, and it understates the printed one by nearly a third.
Restated as: 60,379 — citation [4]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
LVMH will hire 42,000+ people in 2025
Said to measure: a forward hiring commitment a candidate could plan against
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: LVMH publishes no hiring target for 2025, in the Universal Registration Document or anywhere else; the sentence projected a withdrawn 2023 figure two years forward and printed the projection as a plan. What the filing does support is the size of a normal hiring year, which is what the dossier now prints.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
24%
Said to measure: the share of Group roles filled by internal mobility, set against a 12% industry average
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The filing measures internal mobility twice, and neither measure is 24%: close to 18,000 employees took an internal mobility opportunity in 2023, and 72% of the most strategic roles were filled internally. The mechanism the dossier argued is real and stronger than the number it invented for it.
Restated as: 72% of the most strategic roles were filled internally — citation [4]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
the Industry Average column: 12% internal mobility, 33% women in key positions, 22% under-30 employees, 16h training per employee
Said to measure: sector benchmarks against which the Group talent metrics were read as outperformance
Attributed to: Universum — Universum Global Employer Rankings 2024
What was read: Universum ranks employer attractiveness from student and professional surveys; it publishes no sector averages for internal mobility, gender mix, age structure or training hours, and no other publisher issues a luxury-sector series on those four measures. The comparison column was the argument, and it had no document.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
34%
Said to measure: the share of Group employees under thirty
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The Universal Registration Document breaks headcount down by contract, region, business group and gender, and publishes no share of employees under thirty; the report the figure was credited to is not published at all. Nothing dated carries it.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
28h
Said to measure: average training hours per employee per year at the Group
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The filing reports training in absolute hours delivered by programme and school, never as an average per employee, and the report credited with the average does not exist. A per-head average is a different measure from a programme total, and no publisher computed it.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
a bull case at +12% China growth and +€3.2B of LVMH revenue with 25% probability, a base case at +5% and +€1.1B with 50%, a bear case at -3% and -€1.8B with 25%
Said to measure: probability-weighted 2025 outcomes for LVMH revenue under three China scenarios
Attributed to: Morgan Stanley Research — Global Luxury Outlook 2025
What was read: Morgan Stanley Research distributes to clients and publishes no reader-obtainable "Global Luxury Outlook 2025": there is no edition, no page and no date behind the three scenarios or the probabilities attached to them. A probability nobody published is the least checkable figure a dossier can print, and it was carrying the dossier’s central risk section.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
46% reduction in Scope 1
Said to measure: the Group reduction in energy-related Scope 1 and 2 emissions against its 2019 baseline
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The 2023 filing prints the trajectory in one sentence: "Between 2019 and 2023, Scope 1 and 2 emissions declined by 28% and the proportion of renewable energies rose from 47% to 63%." The 50% figure in the LIFE 360 chapter is the 2026 target, not a result; 46% is neither.
Restated as: Scope 1 and 2 emissions declined by 28% between 2019 and 2023 — citation [4]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
100% renewable electricity
Said to measure: that the Group European operations run entirely on renewable electricity
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The filing prints a Group-wide renewable share of 63% of the energy mix in 2023, up from 47% in 2019, and reports renewable supply site by site rather than as a completed European conversion. A 63% mix and a 100% claim cannot both describe the same year.
Restated as: renewable energy rose from 47% to 63% of the energy mix — citation [4]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
€1.2 billion invested
Said to measure: cumulative Group investment in sustainable innovation since 2021
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: No LVMH publication discloses a cumulative sustainable-innovation spend: the environmental chapter reports targets, results and programme scope, and the only €1.2 billion in the filing is an insurance coverage limit in the risk chapter. The figure was read from the wrong page of a document that never claimed it.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
83% of strategic raw materials
Said to measure: the share of strategic raw materials already certified or traceable
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: The LIFE 360 chapter states the commitment as a deadline, not an achievement: all strategic raw materials are to be certified by 2026, with full supply-chain traceability targeted for 2030. The filing prints no share achieved in 2023, so the dossier now prints the deadline it does print.
Restated as: all strategic raw materials to be certified by 2026 — citation [4]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
€2 Billion LIFE 360 Program
Said to measure: the budget of the Group environmental programme, printed as a section heading
Attributed to: LVMH — LVMH Social & Environmental Responsibility Report 2023
What was read: LIFE 360 is declared by LVMH as a set of dated targets across climate, biodiversity, circular design and traceability, with no budget attached in the Universal Registration Document or in any release. A programme heading that prints a number nobody published makes the number look like the programme’s definition.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
Said to measure: the share of postings for formerly degree-free roles that now list a degree
Attributed to: Harvard Business School — Dismissed by Degrees: Updated Analysis
What was read: Dismissed by Degrees (October 2017; no 2024 update exists) prints 65% about childcare workers who said an associate’s degree was sufficient. Its degree-gap finding is that 67% of production-supervisor postings asked for a degree while 16% of incumbents held one.
Restated as: 67% of postings vs 16% of incumbents — citation [2]
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
The document does not existWithdrawn
50% of tech workers at Apple, Google, and IBM
Said to measure: the share of tech workers at three employers without a four-year degree
Attributed to: Glassdoor — Education Requirements in Tech Hiring
What was read: Glassdoor publishes no document titled “Education Requirements in Tech Hiring”; its 2018 list of employers that dropped degree requirements prints no share of the workforce.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
United States band is $95,000–$130,000
Said to measure: seven United States entry-salary bands set beside the APEC bands
Attributed to: NACE — Starting Salary Data by Role
What was read: NACE publishes no document titled “Starting Salary Data by Role”; its salary survey is a paywalled member product and none of the seven bands appears in any public NACE release.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
Said to measure: Generation Z’s share of the global workforce by 2025
Attributed to: World Economic Forum — The Future of Jobs Report
What was read: Neither the 2023 nor the 2025 Future of Jobs Report prints a Generation Z share of the global workforce; the WEF’s own wording elsewhere is “more than one in four”, a phrase, not a measured 27%.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The publisher revised the figureWithdrawn
87% cite purpose and impact; n=22,856 across 44 countries
Said to measure: the share of Gen Z respondents who say purpose matters, and the survey’s sample
Attributed to: Deloitte — Gen Z and Millennial Survey (2024)
What was read: The 2025 Gen Z and Millennial Survey (the current edition) prints 23,482 respondents across 44 countries and states that 89% of Gen Zs consider a sense of purpose important to job satisfaction and well-being.
The publisher printed this figure and then revised it. The earlier value is withdrawn and the revision is named, because the reason for a revision is itself the finding.
The document does not print the figureWithdrawn
46% of Gen Z rate mental health
Said to measure: the share of Gen Z rating their mental health fair or poor
Attributed to: American Psychological Association — Stress in America: Gen Z Report (2024)
What was read: The APA’s Stress in America: Generation Z release (30 October 2018 — there is no 2024 Gen Z report) prints 27% of Gen Z reporting their mental health as fair or poor, against 15% of Millennials and 13% of Gen X.
Restated as: 27% (2018) — citation [4]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
82% say mental health benefits
Said to measure: three shares on mental-health benefits in job choice (82%, 68%, 50%)
Attributed to: Mind Share Partners — Mental Health at Work (2024)
What was read: Mind Share Partners’ Mental Health at Work reports print none of the three shares; the 82% belongs to a Mindspace survey, a different publisher answering a different question.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
79% prefer hybrid/remote-first
Said to measure: the share of Gen Z preferring hybrid or remote-first work
Attributed to: Microsoft — Work Trend Index (2024)
What was read: The 2024 Work Trend Index measures AI adoption at work and prints no hybrid or remote preference share for Generation Z.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
equivalent to an 8% pay raise
Said to measure: the pay-rise value workers place on remote options
Attributed to: Stanford SIEPR — Bloom, N. (2024)
What was read: “Stanford SIEPR” is an institute, not a document; the working paper the citation pointed to (NBER 30292) is “How Hybrid Working From Home Works Out” and prints no 8% pay-equivalent.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
72% also value in-person mentoring
Said to measure: the share of Gen Z valuing in-person mentoring
Attributed to: None named — No document was cited for this share
What was read: The sentence attributed the figure to nobody, and none of the nineteen documents the dossier carries prints a 72% on in-person mentoring.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
Mental Health Support (82%)
Said to measure: the share of Gen Z ranking mental-health support as a priority
Attributed to: Deloitte — Gen Z and Millennial Survey (2024)
What was read: Neither the 2024 nor the 2025 Gen Z and Millennial Survey prints 82% as a Generation Z priority share for mental-health support; the heading’s figure had no printed basis.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
Flexibility and Autonomy (79%)
Said to measure: the share of Gen Z ranking flexibility and autonomy as a priority
Attributed to: Deloitte — Gen Z and Millennial Survey (2024)
What was read: Neither the 2024 nor the 2025 Gen Z and Millennial Survey prints 79% as a Generation Z priority share for flexibility and autonomy; the heading’s figure had no printed basis.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
Career Development (76%)
Said to measure: the share of Gen Z ranking career development as a priority
Attributed to: Deloitte — Gen Z and Millennial Survey (2024)
What was read: Neither the 2024 nor the 2025 Gen Z and Millennial Survey prints 76% as a Generation Z priority share for career development; the heading’s figure had no printed basis.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
Transparent Compensation (74%)
Said to measure: the share of Gen Z ranking transparent compensation as a priority
Attributed to: Deloitte — Gen Z and Millennial Survey (2024)
What was read: Neither the 2024 nor the 2025 Gen Z and Millennial Survey prints 74% as a Generation Z priority share for transparent compensation; the heading’s figure had no printed basis.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
50% more time on learning platforms
Said to measure: Gen Z learning time relative to other generations
Attributed to: LinkedIn — Workplace Learning Report (2024)
What was read: The 2024 Workplace Learning Report prints no generational comparison of time spent on learning platforms.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
82% more likely to apply
Said to measure: the lift in applications when a salary range is posted
Attributed to: Payscale — Compensation Best Practices (2024)
What was read: The 2024 Compensation Best Practices Report measures employer practice — it prints that 60% of organisations publish pay ranges — and no applicant-side likelihood of applying.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
job-switchers earn 15-20% more
Said to measure: the real wage gain of workers who change jobs
Attributed to: Pew Research Center — Young Workers and Job Mobility (2024)
What was read: Pew’s 28 July 2022 analysis prints that half of workers who changed jobs between April 2021 and March 2022 saw a real wage gain of 9.7% or more; it prints no 15–20% range and issued no 2024 job-mobility report.
Restated as: 9.7% median real gain — citation [5]
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
89% of Gen Z prefer employers
Said to measure: the share of Gen Z preferring “try before you apply” employers
Attributed to: Handshake — Gen Z Career Preferences Survey (2024)
What was read: Handshake’s Network Trends reports for 2024 print no 89% and no question about trying a role before applying.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
average debt for 2024 graduates: $37,840
Said to measure: average student debt of the 2024 graduating class
Attributed to: Federal Reserve Bank of New York — Household Debt and Credit Report (2024)
What was read: The quarterly Household Debt and Credit Report prints aggregate student-loan balances and delinquency rates by age band; it prints no average debt per graduate for any graduating class.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
3.1x more likely to stay 3+ years and 2.4x more productive
Said to measure: retention and productivity multiples for purpose-aligned Gen Z employees
Attributed to: McKinsey & Company — Gen Z at Work (2024)
What was read: McKinsey publishes no 2024 document titled “Gen Z at Work”; the citation row carries no address, and no McKinsey publication located prints a 3.1x retention or 2.4x productivity multiple for purpose alignment.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
41% higher engagement scores
Said to measure: the engagement premium of managers who explain the “why”
Attributed to: Gallup — State of the Global Workplace (2024)
What was read: The 2024 State of the Global Workplace report prints engagement shares by region and manager status and no 41% premium for managers who explain the “why”.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not existWithdrawn
34% lower Gen Z turnover
Said to measure: the turnover advantage of employers with robust learning and development
Attributed to: Josh Bersin — No document was cited; the credit named a person and a year
What was read: The sentence credited a person and a year, not a document; the dossier carries no Bersin row, and no Josh Bersin Company publication located prints a 34% Gen Z turnover differential.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The publisher revised the figureWithdrawn
20-24 is 1.2 years
Said to measure: median tenure of United States workers aged 20 to 24
Attributed to: BLS — Employee Tenure Summary (2024)
What was read: Table 1 of “Employee Tenure in 2024” (USDL-24-1971, 26 September 2024) prints 1.4 years for workers aged 20 to 24 in January 2024; 1.2 years is the January 2022 column.
Restated as: 1.4 years (January 2024) — citation [6]
The publisher printed this figure and then revised it. The earlier value is withdrawn and the revision is named, because the reason for a revision is itself the finding.
The figure measures something elseWithdrawn
weekly meaningful feedback = 3.6x more engagement
Said to measure: the engagement effect of weekly feedback against annual reviews
Attributed to: Gallup — State of the Global Workplace (2024)
What was read: Gallup’s “How Effective Feedback Fuels Performance” prints the 3.6: employees are 3.6 times more likely to strongly agree they are motivated to do outstanding work when their manager gives daily rather than annual feedback — a motivation measure for daily feedback, not an engagement multiple for weekly feedback.
Restated as: 3.6 times more likely — citation [9]
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
The document does not existWithdrawn
56% describe their generation as "resilient."
Said to measure: the share of young Americans calling their generation resilient
Attributed to: Harvard Institute of Politics — No document was cited; the credit named an institute
What was read: The dossier carries no Harvard row; the Institute of Politics Youth Poll releases read for 2021 through 2025 print no question asking respondents whether their generation is resilient.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
40-year-high inflation
Said to measure: the height of the 2022 inflation peak
Attributed to: None named — No document was cited for this comparison
What was read: The sentence named no statistical release; the dossier carries no consumer-price document for any country, so the comparison has nothing to rest on and the sentence keeps its argument without it.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
applications up 300%, time-to-hire down 75%
Said to measure: the effect of Unilever’s game-based screening on applications and hiring time
Attributed to: None named — No document was cited for this case
What was read: The list credited no document; Unilever’s own account of the programme is a corporate case study, not a dated report, and no located Unilever publication prints a 300% rise in Gen Z applications.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
4% turnover vs. 17% industry average
Said to measure: Patagonia’s staff turnover against its industry
Attributed to: None named — No document was cited for this case
What was read: The list credited no document; Patagonia publishes no turnover rate, and the 4%-against-17% pairing circulates only in secondary articles quoting one another.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
Fortune "Best Workplace for Gen Z" 3x
Said to measure: the number of times HubSpot appeared on a Fortune Gen Z workplace list
Attributed to: None named — No document was cited for this count
What was read: The list credited no document and named no editions; a count of list appearances is a claim about three dated publications, none of which the dossier carries.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not existWithdrawn
2.8M participants in 2024
Said to measure: participation in JPMorgan’s virtual job simulations
Attributed to: None named — No document was cited for this count
What was read: The list credited no document; JPMorgan Chase’s 2024 annual report and ESG report print no participant count for its virtual job simulations.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The figure measures something elseWithdrawn
roughly two hundred hours a year
Said to measure: the yearly cost of ninety extra minutes of daily commuting
Attributed to: CareerOn — The dossier’s own arithmetic
What was read: Ninety minutes on each of 220 office days is 330 hours; the sentence printed 200, an arithmetic error of the desk, and now prints the computation with its assumption.
Not restated — no located document prints a value.
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
Said to measure: how much job-performance variance traditional interviews explain
Attributed to: McDaniel, M.A. et al. — The Validity of Employment Interviews (1994)
What was read: McDaniel et al. report validity as a correlation coefficient; .26 is the pooled estimate across all interview types, not a share of variance and not the unstructured-interview subgroup.
Not restated — no located document prints a value.
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
The figure measures something elseWithdrawn
simulations can predict up to 54%
Said to measure: how much performance simulations predict
Attributed to: Schmidt, F.L. & Hunter, J.E. — The Validity and Utility of Selection Methods in Personnel Psychology (1998)
What was read: Schmidt and Hunter print the validity of work-sample tests as a correlation of .54, a coefficient, not a percentage of performance predicted.
The document prints the number. It measures something other than the sentence beside it claimed, so the sentence is withdrawn with the number it carried.
The document does not existWithdrawn
reduce demographic hiring bias by 46%
Said to measure: the bias reduction of structured work samples over unstructured interviews
Attributed to: NBER — Reducing Bias in Hiring Through Structured Assessments (Working Paper 28820)
What was read: NBER Working Paper 28820 is a different paper on a different subject, and no NBER paper bears the title “Reducing Bias in Hiring Through Structured Assessments”.
Not restated — no located document prints a value.
No publisher issues the document the figure was attributed to. A figure of this class is withdrawn and never re-attributed, because there is no document that carried it.
The document does not print the figureWithdrawn
23% higher retention
Said to measure: two-year retention of hires selected through skills-based assessment
Attributed to: Deloitte — Global Human Capital Trends (2024)
What was read: The 2024 Global Human Capital Trends report argues for skills-based organisations and prints no retention differential for skills-assessed hires.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
The document does not print the figureWithdrawn
72% of candidates
Said to measure: the share of candidates preferring employers that offer realistic previews
Attributed to: Talent Board — Candidate Experience Research Report (2024)
What was read: The Candidate Experience (CandE) benchmark reports print no 72% on realistic job previews or simulations.
Not restated — no located document prints a value.
The document exists, is dated, and was read. It does not print this figure, so the figure leaves the dossier rather than acquiring a new source.
How a withdrawal is made
A withdrawn figure is quoted verbatim, in both languages, with the day it was withdrawn and the document it had been attributed to. A withdrawal with no printed quote would be a deletion.
Every entry states in one sentence what was read: the document that does not exist, the sentence the document does print, or the edition that replaced it. Hedging is refused.
The figure leaves the argument. Gate 142 refuses the entry unless the quoted figure is absent from the dossier’s stored prose and exhibits in both languages.
Where a located document does print a figure, the corrected value is declared beside the withdrawn one and cites the dossier’s own citation index. A figure whose document does not exist is never restated.
