Legal Architecture — Analysis

The Regulatory-Lag Pattern

From the first documented internal evidence of harm to the first binding response — four industries, one recurring delay.

What counts as an endpoint

RATIFIED 2026-06-03 (operator): the STRICT reading, applied uniformly. 'First binding regulation' = the first enforceable government requirement on the product or its marketing — tobacco's 1965 warning-label act (12-yr lag), lead's 1973 EPA phasedown (49 yr), opioids' 2001 FDA black-box label requirement (5 yr). Settlements and criminal pleas — the 1998 tobacco MSA, the 2007 Purdue plea, the opioid bankruptcies — are shown as secondary typed markers, NOT counted as the regulation endpoint. Attention capture has no binding regulation yet (KOSA proposed, not enacted) and is rendered open. The honest pattern the strict reading shows: a first binding regulation has historically arrived within ~5–50 years of internal evidence; for attention capture, none has arrived at all — even as the strong structural accountability (tobacco's 45-yr MSA) shows how much longer real redress can take.

The pattern, at a glance

Internal evidence The industry's own documented evidence of harm.
Binding regulation An enforceable government requirement on the product or its marketing.
Non-binding milestone Advisory report, declined petition, or other non-enforceable step.
Civil settlement A civil settlement (money plus terms) — not itself a regulation.
Criminal plea A criminal plea or conviction — not itself a regulation.
No binding response yet No binding regulation, settlement, or criminal outcome has occurred.

Every endpoint, with its source

This table is the authoritative record: each date is sourced to the corpus page that asserts it and/or the primary source that verifies it. Where the corpus and the primary record diverge, both are shown.

Industry	Endpoint	Date	Type	Event	Source & verification
Tobacco	First internal evidence	Feb 1953	Internal evidence	Teague, “Survey of Cancer Research” — an R.J. Reynolds internal memo citing 40+ sources on the smoking–cancer link	Claude E. Teague, “Survey of Cancer Research,” R.J. Reynolds, Feb 1953 (UCSF Truth Tobacco Industry Documents) /sagas/the-archive/tobacco/what-they-knew/corpus + Wave-2 audit (corrected from a prior mis-attribution to Clarence Cook Little, the industry's public denier)
	Milestone	1964	Non-binding milestone	Surgeon General’s report, Smoking and Health An authoritative federal finding — but advisory, not an enforceable requirement.	Smoking and Health: Report of the Advisory Committee to the Surgeon General (1964) /sagas/the-archive/tobacco/the-tobacco-archive-as-template/corpus
	First binding response	Jul 1965	Binding regulation	Federal Cigarette Labeling and Advertising Act — first federally mandated warning label The first BINDING federal requirement on the product. Weak (a label) but enforceable.	Federal Cigarette Labeling and Advertising Act, signed July 27 1965; labels required from Jan 1 1966 (15 U.S.C. ch. 36)checked against the primary source — the corpus did not assert this endpoint
	Milestone	Nov 1998	Civil settlement	Master Settlement Agreement — 46 states, ~$206B (first 25 years), advertising restrictions The site's prose-headline endpoint (“45 years, 1953→1998”). A civil settlement, not a regulation — typed here so the 45-year span is not misread as the regulatory lag. (The ~$246B total figure includes the four states — Florida, Minnesota, Mississippi, Texas — that settled individually before the MSA; the 46-state MSA itself is ~$206B.)	Master Settlement Agreement, Nov 1998 (~$206B over 25 years, 46 states) /sagas/the-archive/tobacco/the-tobacco-archive-as-template/corpus + primary-source-checked scope split (gate fix)
Lead (tetraethyl gasoline)	First internal evidence	Oct 1924	Internal evidence	Standard Oil Bayway refinery deaths from acute tetraethyl-lead poisoning (industry lab toxicity documented earlier, 1922–23)	Standard Oil / GM Research internal documentation; Bayway, NJ incident, Oct 1924 (UCSF / contemporaneous press) /sagas/the-archive/lead/seventeen-workers/corpus — see audit flags on the casualty count and the earlier GM lab evidence
	Milestone	1925–26	Non-binding milestone	Surgeon General’s TEL conference (May 1925) → special-committee report (Jan 1926) finds “no good grounds” to prohibit Regulation DECLINED, not enacted — sale continued with worker-protection caveats. The diagram shows this as a non-binding deferral, the opposite of a regulation.	U.S. Public Health Service TEL conference, May 1925; committee report Jan 19 1926 /sagas/the-archive/lead/the-1926-conference/checked against the primary source
	First binding response	1973	Binding regulation	EPA begins the leaded-gasoline phasedown — first binding reduction regulation The first enforceable federal limit on lead in gasoline.	EPA lead phasedown regulations, 1973 (Clean Air Act) /sagas/the-archive/lead/clair-patterson/checked against the primary source
	Milestone	Jan 1996	Binding regulation	Clean Air Act ban on leaded gasoline for on-road vehicles takes full effect Completion of the phasedown — the corpus’s “72 years” endpoint.	Prohibition on Gasoline Containing Lead, eff. Jan 1 1996 (61 FR 3832) /sagas/the-archive/lead/seventeen-workers/checked against the primary source
Opioids (OxyContin)	First internal evidence	1996	Internal evidence	OxyContin launched on a low-addiction marketing claim built on the misused 1980 Porter–Jick letter; internal knowledge of addiction risk documented as marketing continued	Purdue internal documents; Porter & Jick, NEJM 1980 (a five-sentence letter, later cited 600+ times as evidence of low addiction risk) /sagas/the-archive/opioids/the-porter-jick-letter/corpus
	First binding response	Jul 2001	Binding regulation	FDA-mandated OxyContin black-box warning + label change — strongest FDA warning; the false “reduced abuse liability” sentence removed from the 1995 label The first BINDING federal regulation on the product — the strict-reading endpoint, parallel to tobacco's 1965 label.	FDA-approved revised OxyContin label adding a boxed warning, July 2001 (the highest FDA warning level)checked against the primary source — sourced under the strict-regulation ruling (the corpus's opioid milestones were criminal/settlement, not a regulation)
	Milestone	May 2007	Criminal plea	Purdue Frederick Co. + 3 executives plead guilty to misbranding OxyContin; $600M⚠ corpus says 2008 · primary says 2007 Criminal accountability (secondary marker). CORPUS DIVERGENCE — the page dates this plea to “2008”; the plea was entered May 10 2007.	U.S. v. Purdue Frederick Co., W.D. Va., guilty plea May 10 2007 ($600M) /sagas/the-archive/opioids/what-the-sacklers-knew/checked against the primary source
	Milestone	Sep 2019	Civil settlement	Purdue files for Chapter 11 bankruptcy Restructuring/accountability vehicle, not a product regulation.	Purdue Pharma Chapter 11 filing, Sept 2019 /sagas/the-archive/opioids/the-settlement-architecture/corpus
	Milestone	2020–25	Civil settlement	Settlement architecture continues (2nd DOJ plea; SCOTUS rejects the Sackler release 2024; revised ~$6.5B+ plan) Ongoing; not a product regulation.	Harrington v. Purdue Pharma, 603 U.S. ___ (June 27 2024); revised settlement reporting 2025 /sagas/the-archive/opioids/the-settlement-architecture/corpus + checked against the primary source (SCOTUS date)
Attention capture (social platforms)	First internal evidence	2019–21	Internal evidence	Facebook/Instagram internal research documents harm to teens (“we make body-image issues worse for 1 in 3 teen girls”); routed internally, disclosed Sep 2021 via the Haugen documents	Internal Facebook research (disclosed via Frances Haugen); Wall Street Journal “Facebook Files,” from Sept 13 2021; Haugen Senate testimony Oct 5 2021 /sagas/the-children/instagram/what-the-internal-research-showed/corpus + public record (disclosure date)
Attention capture (social platforms)	Milestone	ongoing	No binding response yet	No binding federal regulation, settlement, or criminal outcome as of 2026 (KOSA proposed, not enacted; product changes optional; state laws fragmentary) The visual point of the diagram: the only row whose binding-response endpoint is still open. Rendered as an open span to the present, NOT with an invented date.	KOSA legislative history (not enacted as of the build date); the corpus's account of the disclosure producing no structural product change /sagas/the-children/instagram/what-the-internal-research-showed/corpus + public record

Definitions

First internal evidence: The first documented internal or industry evidence of harm — an internal memo, internal study, or industry-documented casualty event. Not the first outside suspicion: the first evidence the industry itself is documented to have held.
First binding response: The first binding, enforceable response to the product or its marketing. We type each response so the reader is never asked to read two different kinds of event as the same thing: a government REGULATION, a civil SETTLEMENT, or a CRIMINAL plea are distinct; advisory reports and declined petitions are NON-BINDING; and where nothing binding has occurred the endpoint is NONE-YET (open).

Date-audit notes

Tobacco

Wave-2: page originally mis-attributed first evidence to a “Clarence Cook Little memo (1953)” — Little was the public denier; corrected to Teague.
Wave-2: the Hill & Knowlton strategy meeting was Dec 1953, not 1954.

Lead (tetraethyl gasoline)

Wave-2: “17 workers” is the cumulative 1924–25 toll across the Ethyl/DuPont/Standard Oil plants; the discrete 1924 Bayway/Standard Oil incident killed 5.
Wave-2: the conference was held May 1925; the committee report and federal go-ahead came Jan 1926 — cite “1925–26,” not “1926 conference.”

Opioids (OxyContin)

Corpus dates the first guilty plea to 2008; the primary record (W.D. Va.) is May 10 2007. The JSON uses 2007 and records the divergence (corpus_says/primary_says).
RESOLVED at the gate (strict-regulation ruling, 2026-06-03): the strict reading needs a binding *regulation*, not the 2007 criminal plea. Sourced the first binding FDA regulation — the July 2001 OxyContin black-box label requirement (primary-source-checked) — as the headline endpoint; the 2007 plea and the bankruptcy/settlements remain as secondary typed markers.

Attention capture (social platforms)

Attention's binding-response endpoint is intentionally OPEN — render ongoing; do not assign a date. KOSA's enacted/not-enacted status is the live variable (see /legal-architecture/the-kids-online-safety-act-record/).

Methodology. Endpoints are harvested from the Institute's own audited Archive-saga records (tobacco, lead, opioids) and the children's-saga Instagram record; dates that are load-bearing for a lag span, or that the wave audits flagged, were additionally checked against primary sources (statutes, agency records, court filings) — no date is asserted from memory. Each endpoint is typed so a regulation, a settlement, and a criminal plea are never read as the same kind of event, and the attention-capture row is rendered open rather than given an invented date. Source of truth: data/precedent-timeline.json. Built 2026-06-03 (D-099).