Fare Changes After the Delta–Northwest Merger: Market Composition and Competitive Exposure
Using approximately 4.18 million Borenstein DB1B market summaries from 2005–2010, I examine relative fare changes around the Delta–Northwest merger. Accounting for baseline route composition reduces the estimated relative fare decline and leaves it statistically indistinguishable from zero.
I use Python for data construction, R for HonestDiD sensitivity analysis, and R and Stata for independent regression checks. Methods include route and quarter fixed effects, event studies, baseline composition controls, and within-route carrier fare comparisons.
Data and methods
I construct passenger-weighted route-quarter fares and evaluate distance, baseline itinerary composition, Southwest exposure, and comparison-route selection. The technical supplement documents data cleaning, full results, raw fare trends, and sensitivity to departures from parallel trends. The replication repository includes code, saved analysis panels, and verified source-download instructions.