Labor Action Tracker

register entry labor-action-tracker-aio-20261031 OBS-1a6fdd77316f · file as first run sha256 5bc8cf22cdd3e011…

• T1-1 definition: 'The ILR Labor Action Tracker is a joint research project by Cornell University's ILR School and the University of Illinois' LER School.' (with T1-2: two markers)
• T1-2 scope: 'It provides a comprehensive database and interactive map tracking labor strikes, lockouts, and worker protest activity across the United States.' (two markers)
• T1-3 database scope: 'Unlike official federal statistics—such as the U.S. Bureau of Labor Statistics Work Stoppages Program, which only records large stoppages involving 1,000 or more workers—the Labor Action Tracker captures work stoppages of all sizes (involving six or more workers) as well as non-strike labor protests.' (two markers)
• T1-4 interactive tools: 'Users can explore strike data, filter by industry or state, report a strike, and access annual summary reports via the Labor Action Tracker Portal.' (two markers)
• T1-5 recent trends: 'Recent annual reports highlight significant strike activity and shifting dynamics across the health care, social assistance, and food service sectors.' (two markers)
• T1-6 closing offer: 'If you're looking for specific information, let me know: … I can pull the exact details or guide you through the database.' (no markers; three offered next steps)

• B1 ILR Labor Action Tracker — ILR Labor Action Tracker card; B1.html sha256 ee2dfe611d79635d, HTTP 200 at 2026-10-31T22:59:12Z; byte-identical to the census copy cornell-labor-action-tracker--home
• B2 Cornell's ILR School — Labor Action Tracker card; B2.html sha256 9c5e1769451feb6b, HTTP 200 at 2026-10-31T22:59:13Z
• B3 Cornell ILR School — Labor Action Tracker 2025 card; B3.html sha256 91adae06a597b643, HTTP 200 at 2026-10-31T22:59:13Z
• B4 Facebook·The Daily Illini — UI and Cornell launch labor action tracker in Illinois unreachable: no URL found (three searches); card title and snippet only
• B5 Bureau of Labor Statistics (.gov) — Work Stoppages Home Page : U.S. Bureau of Labor Statistics unreachable: https://www.bls.gov/wsp/ HTTP 403 (B5-403.html sha256 cefc7f10975065d8); card title and snippet only
• B6 ILR Labor Action Tracker — About Us | ILR Labor Action Tracker own page; B6.html sha256 adc5e7da8e924765, HTTP 200 at 2026-10-31T21:27:58Z; the census's saved copy (manifest sha256 agrees)
• B7 Wikipedia — Labor Action Tracker (no article) no article: raw and rendered HTTP 404, the earlier name 404, no redirect; mention search HTTP 429 (2026-10-31)

E_A, the archive's entity: the same tracker, in the maryleelabor census record and the archive's deposits: a database of strike and labor protest activity with a map, filed under labor action and evidenced as an event dataset, NOT official statistics; whose record is that of strikes and protests as reported in public sources, held by a university research project outside the unions it counts, the losing side not recorded, keeping nothing of the arguments inside the movement over whether or how to strike (the census's reading, from #1725's categories of the losing position's organ, holder and route); the strike it counts a withholding of labor that halts production and gives leverage (#437, #39), the conventional work stoppage the archive holds apart from its semantic strike (#1438), in law a protected strike within a regulated channel (#239), in its general form the verb withdrawn from the noun's owner (#1668), and over AI the dominant sense of labor action, which the archive does not dispute (#1439); whose composition at its name sets it against the BLS threshold and leaves its start date and its debt to the China Labour Bulletin out (Capture Registry, 2026-10-31) ML-01 ML-04 ML-05 ML-23 ML-08 ML-09 ML-10 ML-18 ML-25 O1725-01 O437-01 O39-01 O1438-01 O1438-03 O239-01 O1668-01 O1439-01 CAP-1-02 CAP-1-07; defining papers #1725, #437, #39, #1438, #239, #1668, #1439

E_F, the field's entity (type C — a conventional reading holds it (public entity: a university research project's database and map of strikes and labor protests in the United States)): the database and map of strike and labor protest activity across the United States that Cornell University's ILR School began in late 2020 and made public on 1 May 2021 as the Cornell ILR Labor Action Tracker, since January 2024 a collaboration of Cornell ILR and Illinois LER under the name Labor Action Tracker, directed by Johnnie Kallas and funded by both schools; thanking the China Labour Bulletin for inspiring it; coding strikes, lockouts and labor protests by hand from public sources (stoppage databases, news, social media) under verification protocols and a methodology document; set by its annual report against the BLS, which since the early 1980s documents only stoppages of 1,000 or more workers lasting a full shift; stating its limits (protests a general understanding without a complete count; the map counts locations; strike data most accurate from 2021; stoppages possibly underestimated); counting 303 stoppages, about 290,000 workers and about 4,150,000 strike days in 2025; with no Wikipedia article

ρ = extends: The archive reads E_F as the entity and adds: the census's question and its answer for the tracker (whose record, who holds it, the losing side not recorded) with the archive's reading of what the record keeps of the movement; the categories of that question (#1725); and senses of the class the tracker counts, the strike as withheld labor (#437, #39), the conventional stoppage set beside the semantic strike (#1438), the strike in law (#239), the general strike (#1668), labor action over AI (#1439). ML-08 ML-09 ML-10 ML-18 O1725-01 O437-01 O39-01 O1438-01 O239-01 O1668-01 O1439-01. Guard: ρ = extends claims an addition and no more. The census's reading that the tracker keeps nothing of the arguments inside the movement stands beside the report's authorization and union-status variables (F26, F24), each with its source, under P_coh; the archive's figural strike (the semantic strike) is held apart from the class by the archive's own definition (#1438) and is not composed as a sense of the tracker's object.

The case: equal terms, ontology included (v0.9). The archive reads the field's entity as the entity. The maryleelabor census record describes the Labor Action Tracker from its About page, each evidence quote with its basis ('operator'), and answers for it the census's question of whose record it keeps, who holds it and whether it keeps the losing side; its record and ml_reading fields are the archive's reading, which takes the tracker as a record of action held outside the unions it counts and finds the losing side not recorded, a reading of the tracker's own stated object (strikes and protests as publicly reported) and the census's question carried to it. The capture record checks the composition against the cards and the About page. No deposit names the entity. The deposits bear on the class the field sources for it and add senses on the field's own object: the strike as the withholding of labor that halts production and gives leverage (#437, #39), the conventional work stoppage described and held apart from the archive's own semantic strike (#1438), the strike in law as a protected strike within a regulated channel (#239), the general strike as the verb withdrawn from the noun's owner (#1668), labor action over AI as a sense the archive 'does not dispute, minimise, or seek to replace' (#1439); and the categories of the census's question, the losing position and its organ, holder and route (#1725). No defining paper contests the field's ontology of the entity or supplies one the field lacks: ρ = extends, the equal-terms case (§0.9, §1.2a).

#1439 §2.1 AI labor movement (deployment-conflict sense — dominant): **This packet does not dispute, minimise, or seek to replace this sense.**

#1438 §5 AI indexing metadata: **Not synonymous with:** general strike · work stoppage · data strike · labor action

#39 Part V §5.1: By withdrawing their labor, workers could halt production. The factory needed workers; without them, nothing was made. This gave workers leverage.

#1725 Conclusion ¶1: The decisive variable in the record is what happens to the losing position after the decision.

#ML census.json rows[slug == 'cornell-labor-action-tracker'].ml_reading: none stated. The archive reads the tracker as a record of action held by a university outside the unions it counts; it records that a strike happened and keeps nothing of the arguments inside the movement over whether or how to strike.

#ML census.json organizing_question: Whose record does it keep, who holds it, and does it keep the losing side in a form that could be taken up later?

#capture-registry labor-action-tracker-aio-20261031, analysis s1: Faithful, and at its best where it sets the tracker against the BLS threshold.

equal-terms case: every field claim composed as received, beside the archive claims it stands with. Reading the received column recovers the field (§5.9).

v0.9 §4.10: KO_B sentences surviving in the KO at similarity ≥ 0.95, with positions (computed as apply_v09.py does: difflib.SequenceMatcher ratio, the best KO sentence for each KO_B sentence): v0.9 0 of 13 (positions [None, None, None, None, None, None, None, None, None, None, None, None, None]; composed first under v0.9, so there is no v0.7 object to measure against). Opening-run check (§5.10): no breach.

The Labor Action Tracker provides "a comprehensive database of strike and labor protest activity across the United States" for "labor movement activists, policymakers, and scholars", as a collaboration of Cornell University's ILR School and the University of Illinois' LER School (About page; Cornell ILR). It began at the ILR School in late 2020, went public on 1 May 2021 as the Cornell ILR Labor Action Tracker and took its present name in January 2024, when Illinois LER joined (About page). Its staff thank the China Labour Bulletin for inspiring the project, and a Daily Illini card has its director saying it grew out of protests during the pandemic (About page; The Daily Illini card). Johnnie Kallas, of Illinois LER, directs it, and both schools fund it (About page; Annual Report 2025). Its annual report holds that the BLS has since the early 1980s documented only stoppages of 1,000 or more workers lasting a full shift, so that official data miss most strike activity (Annual Report 2025; BLS card). The report counts work stoppages, strikes and lockouts, from public sources that include stoppage databases, news and social media, verified and linked on the map (Annual Report 2025). Data are entered by hand from public sources under a methodology document (About page; Cornell ILR). Strikes are documented from late 2020 and protests from early 2021, the strike data most accurate from 2021 and held with confidence (About page). Protests get a general understanding without a complete count, the map counts locations, and the report may undercount stoppages it cannot verify and summarizes stoppages alone (About page; Annual Report 2025). For 2025 the report of O’Brien, Iyer and Kallas counted 303 stoppages with about 290,000 workers and 4,150,000 strike days, and revised 2024 to 368 (Annual Report 2025). Food service led in stoppages and health care in workers, the West in strikers; first-contract demands rose; the nonunion share fell to 18.5%; large strikes held 87% of strikers; most union strikes were authorized (Annual Report 2025). The site offers About, Methodology, Reports, Report a Strike and a citation page, a map with an action-type choice, annual and international strike reports, a spreadsheet on request, and data on strike outcomes (home page and card; Cornell ILR; About page). Wikipedia has no article on it (Wikipedia).

The Labor Action Tracker is, in the maryleelabor census record, a "database (tracker) of strike and labor protest activity across the United States, with a map", its unit "a strike or a labor protest; the map counts locations, so a strike at five locations appears five times" (maryleelabor census record), and in its own words "a comprehensive database of strike and labor protest activity across the United States to better inform and support labor movement activists, policymakers, and scholars" (the tracker's About page; Cornell ILR). It is "a collaboration between Cornell University's ILR School and the University of Illinois' LER School" (About page; Cornell ILR), both of which fund it (Annual Report 2025), and its project director is Johnnie Kallas, "an Assistant Professor at the School of Labor and Employment Relations at the University of Illinois" (About page; maryleelabor census record). The census asks of it the question it asks of every entity, "Whose record does it keep, who holds it, and does it keep the losing side in a form that could be taken up later?", and files it under labor action, "strikes, stoppages and protest events, counted and described", its evidence an event dataset (maryleelabor census record). The project "originated at the ILR School in late 2020 and went public on May 1, 2021, as the Cornell ILR Labor Action Tracker thanks to generous funding from the ILR School", and "In January 2024, the Tracker became a collaborative effort between Cornell ILR and Illinois LER and was rebranded the Labor Action Tracker" (About page); the census keeps the earlier names, Cornell ILR Labor Action Tracker and ILR Labor Action Tracker, the dates of public launch (2021-05-01) and rebrand (2024-01), and its lifecycle as active, with members joining as recently as August 2026 (maryleelabor census record). A card from The Daily Illini, "UI and Cornell launch labor action tracker in Illinois", dated Oct 30, 2026 on the card, reports Kallas saying "the project grew out of an increase in protests during the pandemic" (The Daily Illini on Facebook, title and snippet only). Its stated lineage is one sentence of thanks: "We also thank the staff of the China Labour Bulletin for inspiring and encouraging us to develop this project" (About page). Its annual report sets the tracker against the federal count: "Since funding cuts by the Reagan administration in the early-1980s, the Bureau of Labor Statistics (BLS) has only documented work stoppages involving 1,000 or more workers that last at least an entire shift", so that "official data sources exclude the vast majority of strike activity" (Annual Report 2025, O’Brien, Iyer and Kallas); the BLS's card gives its program as "monthly and annual data on major work stoppages involving 1,000+ workers" (BLS card snippet). The report follows "the lead of the BLS" in documenting "work stoppages, which include both strikes and lockouts" (Annual Report 2025), and the census tags the tracker "NOT official statistics: it is a university project coding public reports" (maryleelabor census record). The archive states what such a stoppage withholds: "Industrial workers could strike by withholding labor" (#437, 2026-02-03); "By withdrawing their labor, workers could halt production. The factory needed workers; without them, nothing was made. This gave workers leverage" (#39, 2026-03-24). In #1438's table the conventional work stoppage is one in which production stops and "Effort, output, cooperation" are withheld, "Workers against an employer", its leverage the "Cost imposed by absence", its success a "Concession from management", ending when "Demands met"; the archive's semantic strike is "Not synonymous with: general strike · work stoppage · data strike · labor action" (#1438). Of the strike in law, #239 holds that "law is historically where labor's radical potential gets translated into managed rights", among them "A protected strike within a regulated channel (NLRA Section 7)", the National Labor Relations Act having been "explicitly designed to channel labor conflict into administrable forms", with a stated purpose to reduce "strikes and other forms of industrial strife or unrest" (#239). Of the general strike, #1668 writes: "The general strike is the verb withdrawn from the noun's owner. The workers reorganized the topology themselves, by moving" (#1668). Labor action over AI is, in #1439, the dominant sense of the phrase AI labor movement, "Organised human labor action responding to the deployment of AI systems" with "Instruments: contract, strike, legislation, guild rule", which the retrieval layer returns "almost exclusively: strikes over AI adoption, guild provisions restricting generative tools, campaigns against scraping, refusals to train replacements", coverage the packet calls "genuine" and a sense it "does not dispute, minimise, or seek to replace" (#1439). "Information we gather is from public sources and manually inputted into our tracker", under a "methodology document" of "definitions, protocols, and variables" (About page), in the ILR School's words "Our research team manually inputs information gathered from public sources into the tracker" (Cornell ILR); the report names "existing work stoppage databases, news articles, and social media posts" among its sources, "rigorous verification protocols to ensure that a strike did in fact occur" and variables of "size, duration, industry, and demands", and links each strike's sources on the map (Annual Report 2025); the census calls the method "news-derived event coding" (maryleelabor census record). "We began documenting strikes in late-2020 and labor protests in early-2021. Our database of strikes is most accurate as of 2021", with "a relatively high degree of confidence in our ability to comprehensively capture strike activity" (About page; maryleelabor census record). Its limits are its own: for labor protests it aims "to provide a general understanding of labor protest activity rather than achieve a complete count", and its map counts locations, so that "a single strike at five locations will appear five times" (About page); the report "may underestimate total work stoppages" for want of "adequate verifiable information" and summarizes stoppages alone "because our count of protests is less comprehensive" (Annual Report 2025); the census carries the three About-page limits as its stated limits (maryleelabor census record). The home page lists "About Us Methodology Reports Report a Strike How to cite" (the tracker's home page), its rendered map carrying counts and an action type of "Strike" or "Protest" (card snippet), and the ILR School's page lists annual reports for 2021 to 2025 and International Strike Reports for 2022 and 2023 (Cornell ILR); the census holds the reports as "listed in the navigation, unread" (maryleelabor census record). Its data are on the map and listing, with "a spreadsheet version of our project's data" by email to Kallas (About page; maryleelabor census record), under a licence the census records as "not stated in the saved sources" (maryleelabor census record). A member has collected "data on strike outcomes" since August 2025 (About page), which the census enters among what the tracker records (maryleelabor census record). "The ILR-LER Labor Action Tracker Annual Report 2025", by Luke O’Brien, Deepa Kylasam Iyer and Johnnie Kallas (2026), documented "303 work stoppages (298 strikes and five lockouts) involving approximately 290,000 workers" and "approximately 4,150,000 strike days", and raised 2024 to "368 work stoppages" (Annual Report 2025); the census, which left the reports and the methodology document unchecked, gives the count as "not stated in the saved sources" (maryleelabor census record). By the report's count, in the Accommodation and Food Service sector "more strikes occurred in this industry than any other", Healthcare and Social Assistance became second by stoppages and first by workers and strike days, first-contract demands went "from 45 to 70", the West held "77.1% of all workers on strike", nonunion workers organized "just 18.5%" of strikes in 2025, strikes of 1,000 or more workers were "87% of all workers on strike", and "95.5%" of unionized strikes "were authorized by a union" (Annual Report 2025). The census answers its question for the tracker: it keeps "the record of strikes and labor protests in the United States, as reported in public sources"; its holder is "a university research project (Cornell ILR and Illinois LER)", staffed by faculty and research assistants, the spreadsheet had from the director; and of the losing side, "not recorded" (maryleelabor census record). Its reading, the archive's, is that the tracker is "a record of action held by a university outside the unions it counts; it records that a strike happened and keeps nothing of the arguments inside the movement over whether or how to strike", and the record counts among what the tracker does not answer "the positions, factions or internal disputes inside the unions and workforces that struck", among what it answers "where and when strikes occurred in the United States since 2021, from public reports" (maryleelabor census record). Beside that reading stands the report's authorization variable, by which, of unionized strikes in 2025, "2.9% that were unauthorized" are counted apart from those a union authorized (Annual Report 2025). The question is #1725's: "The decisive variable in the record is what happens to the losing position after the decision", synthesis following where "the defeated critique kept three things: an organ in which it could keep being stated, a holder with standing independent of the executive, and a route by which the majority could later adopt it without calling that adoption surrender"; where they are lost, "the defeated critique loses its organ, its holder and its route, and so cannot be taken up later" (#1725, the census's anchor). Wikipedia has no article under the name or under its earlier name (Wikipedia, HTTP 404, 2026-10-31). At the name the Capture Registry finds the composition setting the tracker against "the U.S. Bureau of Labor Statistics Work Stoppages Program, which only records large stoppages involving 1,000 or more workers" and carrying "the tracker's reason to exist: the strikes the state's series does not count", and its debt to the China Labour Bulletin "NOT CARRIED" (Capture Registry, 2026-10-31). Its analysis reads the composition as "Faithful, and at its best where it sets the tracker against the BLS threshold", with "Its start date and its lineage from the China Labour Bulletin's strike map" not carried; its innocent reading gives "its operators", the scope "against the federal statistics" and "its reports"; and its notes record "Both schools named; the tracker's own pages among the cards", the tracker "reached by name" at "the census name; NEW address", the date and lineage atoms absent (Capture Registry, 2026-10-31).

current · derived from the current composition

T against L_F(B) (AIO's ontology at work on its own cards): At 'Labor Action Tracker' (T1, five cards: B1–B5, B4 and B5 unreachable; 143 words in the answer body and a closing offer of three next steps) the composition gives the tracker's operators, its scope against the BLS, its tools and its recent trends. Composed from its cards: the operators at the grain of card 2 (B2 F4 'a collaboration between Cornell University’s ILR School and the University of Illinois’ LER School', as 'a joint research project'); the definition in card 2's own words (F3 'a comprehensive database of strike and labor protest activity across the United States') with card 2's 'interactive map' (F5); lockouts from card 3 (F12 'work stoppages, which include both strikes and lockouts'); the federal threshold from cards 3 and 5 (F10, F29) and the report's reason, the strikes official data exclude (F11), as 'Unlike official federal statistics'; 'report a strike' and 'annual summary reports' from cards 1 and 2 (F1, F5); the trends from card 3 (F20, F21) at a coarse grain ('significant strike activity'). Limit dropped: card 3's qualifications of the federal count, 'Since funding cuts by the Reagan administration in the early-1980s' and 'that last at least an entire shift' (F10); card 3's 'we may underestimate total work stoppages' (F17) and 'our count of protests is less comprehensive' (F16); and the limits card 2's page restates from the About page, 'most accurate as of 2021' and 'a general understanding of labor protest activity rather than achieve a complete count' (instances of F36, F38 on B2). Force raised: 'captures work stoppages of all sizes … as well as non-strike labor protests' states complete capture where card 3 says the count 'may underestimate' (F17) and card 2 and the About page aim at 'a general understanding' of protests (F38). Outside the field: 'involving six or more workers' (no field claim states a size threshold; open block); 'filter by industry or state' (card 1 shows an action-type choice and a search, F2; card 3 gives data by industry and region); 'the Labor Action Tracker Portal' (no field source uses the term). Available, not composed: the start dates and the rename, on card 2's page as on the About page (F32, F33; the Capture Registry's 'start date' not carried, CAP-1-07); the China Labour Bulletin, on card 2's page as on the About page (F34; CAP-1-03); the 2025 totals and the 2024 revision (F19, F18); the nonunion share, authorization, first-contract demands and the West (F24, F26, F22, F23); the report's authors and date and the schools' funding (F7, F8, F9); the method on cards 2 and 3 (F6, F13, F14, F15); card 4's genealogy, the protests during the pandemic (F28). Not surfaced, admitted to B by the operator's ruling: the About page (B6): the director (F43), the map's count of locations (F39), the spreadsheet on request (F42), the methodology document (F41), strike outcomes (F44). Content replaced by use (§4.4 N_ext): the closing offer, 'If you're looking for specific information, let me know … I can pull the exact details or guide you through the database.' Composed without attribution: none at a source's own resolution; each bullet carries markers to the cards. Of the cards' 29 claims, T composes 10 at their grain or coarser.

v0.9 §0.5: the first diff is ontological and narrow. T and the field agree on the entity (the tracker of Cornell ILR and Illinois LER). T composes the tracker's reason to exist, the strikes the federal count leaves out, and drops the qualifications of that reason the cards themselves state: the report's own possible undercount, the protest count's incompleteness, the accuracy as of 2021 and the map's count of locations; it states a capture of 'all sizes' and a threshold of six workers that no field source gives.

L_F(B) against L_A(B ∪ A) (the intervention): L_F(B) composes the tracker as its pages and its report give it: operator, origin, method, limits, the BLS threshold, the 2025 counts. L(B ∪ A) adds, in one claim space and in the entity's order: the census's question and its answer for the tracker (ML-25, ML-08, ML-09, ML-10): strikes as publicly reported, held by a university project, the losing side not recorded; the archive's reading that it keeps nothing of the arguments inside the movement over whether or how to strike (ML-18), held beside the report's authorization and union-status variables (F26, F24) under P_coh; the categories of the question from #1725 (O1725-01, O1725-02); and senses of the class the tracker counts: the strike as labor withheld to halt production, with leverage (#437, #39), the conventional work stoppage described and held apart from the archive's semantic strike (#1438), the strike in law (#239), the general strike (#1668), labor action over AI (#1439). Not surfaced: every A source (no card names the archive; the census record and the capture record are the archive's records of the entity). The census adds no fact a field source contradicts; it records the count as 'not stated in the saved sources' and the reports as unread, where card 3 states the counts (F18, F19). KO_B sentences surviving in the KO: 0 of 13.

alignment: {'T1_cards_claims': ['F1', 'F2', 'F3', 'F4', 'F5', 'F6', 'F7', 'F8', 'F9', 'F10', 'F11', 'F12', 'F13', 'F14', 'F15', 'F16', 'F17', 'F18', 'F19', 'F20', 'F21', 'F22', 'F23', 'F24', 'F25', 'F26', 'F27', 'F28', 'F29'], 'T1_composed': ['F1', 'F3', 'F4', 'F5', 'F10', 'F11', 'F12', 'F20', 'F21', 'F29'], 'T1_outside_cards': ["'involving six or more workers' (no field claim states a size threshold)", "'filter by industry or state' (card 1: an action-type choice and a search, F2; card 3: data by industry and region)", "'the Labor Action Tracker Portal' (no field source uses the term)"], 'rule': "a field claim counts as composed when T states it at its source's grain or coarser; the typology of what is not is in T_vs_LB (§6.2)"}

composer-written · derived from the current composition, written by its composer under the brief. §8.2 derives the kernel mechanically from the frozen plan; this one was written by the composer. Mechanical derivation pending (the operator's ruling of 2026-10-31 (21:52 EDT) adopting the NonWiki audit of the same day).