Author: Richard H. Thaler
Year: 2015
Original file: RAW/books/Misbehaving-The Making of Behavioral Economics.md
Source type: Popular science / disciplinary memoir
Publisher: W. W. Norton
Created retroactively 2026-08-10 during a ghost-link remediation pass. Thirty citing links resolved to nothing before this file existed. The citations were not fabricated — author, title and text verified on disk at the path above. The leaf node was missing, not the source. ARCHIVES/index.md had described this class of link as "fabricated source citations"; for this source that is inaccurate and has been corrected.
Economics built its models on an agent who optimises, and real people do not. Thaler's argument is that the deviations are not random noise that cancels out in aggregate but systematic, predictable and consequential — which means they belong inside economic models rather than in a footnote about human frailty. The book is simultaneously a survey of those deviations and a first-person account of getting a discipline to admit them.
🚩 REPLICATION RISK — and this one is unusually live. A substantial portion of the behavioural-economics literature of this era has faced replication difficulty since 2015. Some findings here are robust and heavily replicated (endowment effect, loss aversion); others are not. A claim's presence in this book is not evidence of its replication status, and pages drawing on it should say which category a specific finding falls into where that is known.
🚩 MOTIVATED REASONING, mild but real. A participant's account of a disciplinary victory, written by the participant. The narrative selects for anomalies that turned out to matter and arguments the author won.
Tag empirical claims [PLAUSIBLE — needs corroboration] unless the specific finding's replication record has been checked. Accounts of disciplinary history are [POPULAR SOURCE].
The ~30 existing citations predate this stub and have not been individually verified against the text. This stub establishes that the source exists and is correctly identified; it does not certify every claim attributed to it.