Ticket prices move most with market concentration at small hubs; at large hubs, medium hubs and non-hub airports the link cannot be told apart from zero. It is a pattern in fares, not a causal price effect.
What this shows: The chart plots Elasticity of fare to carrier concentration (HHI), 95% interval lower (clustered by airport) and 95% interval upper (clustered by airport) on a vertical axis of Elasticity of fare to concentration (Δlog fare / Δlog HHI).
Conditions and method
Status: descriptive (a pattern in the data, not a causal estimate).
Each bar is one hub tier’s slope of the log of an airport’s average fare on the log of its carrier concentration (the Herfindahl-Hirschman index, HHI), across airports and quarters of 2020 to 2025. The most negative slope is at small hubs, where the interval lies below zero. At large hubs, medium hubs and non-hubs the slope is not distinguishable from zero. HHI and fares are determined together, so each slope is a descriptive association, not a causal price effect.
Each airport is in the tier it held in 2019, the last complete year before the window, so a tier is one set of airports throughout (FAA share thresholds applied to T-100 domestic passengers). 9 airports with no tier in 2019 are left out.