Replication package for "What public ticket data can and cannot establish about algorithmic airline pricing: distributional evidence from the U.S. OD40 sample"

Processed public U.S. airline ticket cells, complete analysis code, numericalvalidation reports, publication tables and figure source data for anobservational study of carrier-relative fare-distribution trends duringpublicly disclosed AI-assisted revenue-management experimentation. The archive regenerates all reported calculations from deposited analyticalcells and distinguishes independently target-matched values fromreconstructed pipeline outputs. It does not contain personal data,proprietary airline data, pilot assignment or displayed-but-unpurchasedoffers, and it does not identify an AI treatment effect or individual-levelpersonalisation. Data are derived from the public DB1C and DB1B Origin and Destination Surveyproducts and the T-100 Domestic Segment and Domestic Market files publishedby the U.S. Bureau of Transportation Statistics. Raw BTS archives arereferenced and hashed rather than redistributed. Mixed licensing: original code is MIT; original documentation and compileddata products are CC BY 4.0.