A policy may alter measurable behavior without succeeding as an institution: the best available evidence indicates alcohol consumption declined during Prohibition, even though drinking persisted and the national regime is widely judged a failure.
Illegal production prevents a complete consumption count, and each proxy is affected by enforcement, reporting, health lags, and population assumptions. Reduced consumption does not determine whether Prohibition's benefits outweighed its costs.
Lewis triangulates legal-consumption estimates with cirrhosis and drunkenness-arrest proxies. The synthesis separates magnitude of consumption change from overall evaluation of enforcement, legitimacy, corruption, crime, constitutional effects, and long-run durability.
Created from the complete 244-sheet textual and visual review of LIT-111. Developing until independent sources corroborate the chapter-level synthesis.
Developing interpretive note derived from Michael Lewis's measurement chapter and verified statistical-table evidence.
Atomic idea
Behavioral change and institutional success are different questions: alcohol consumption appears to have declined during Prohibition even though drinking persisted and the regime failed as a durable political institution.
Reasoning
Illegal production prevents a complete count. Lewis therefore triangulates legal-consumption estimates with cirrhosis mortality and drunkenness arrests and records the limitations of each. The combined direction supports decline, but not elimination, precise effect size, or a complete policy verdict.
Implications
Website language should avoid both myths: Prohibition neither stopped drinking nor left consumption wholly unchanged. Claims about effectiveness must specify which outcome is being measured.
Connections
This note supports the linked methodological Claim and informs the assessment of enforcement, illicit markets, public legitimacy, and repeal.
Counterpoints and maturity
Reporting, enforcement, health lags, population assumptions, and unmeasured illicit supply affect every series. The original datasets and independent scholarship should be checked before publishing numbers. Maturity remains Developing.
Fresh table audit and teaching application
Lewis PDF97–98 was enlarged and fully read. Cirrhosis deaths per100,000 extend1914–1935; arrests per1,000 extend1914–1922 only. Blank later arrests are missing, not zero. Large declines predate1920; shared police inputs limit independence between opposing organizations’ series. Quantity does not establish prevalence of drinkers. Proposed exercise: label measure, denominator, time window and confounders before assessing effectiveness. Developing maturity retained.
Full audit: 