this particular form of price discrimination [...] is done to compete with airlines offering direct flights between two cities, A and B. The airline offering the deal seeks to serve A<->B passengers without adding a direct flight, by using excess capacity that’s available on the A<->C and C<->B routes. The airline can offer these seats below the A<->C ECA, because it expects not to fill them with other A<->C or A<->Wherever passengers anyway.
So if the the airline has free seats on the A-C flight, it seems it could do to things; either lower the price on the A-C flight until the plane fills up, or use the spare capacity to serve A-B passengers as you describe.
If they didn’t have the price discrimination option (e.g. because there was a law against it, or because everyone is using hidden city ticketing), then they would be forced to take the first option. If we then change the situation so that price discrimination is allowed, then A-B passengers are better off (the get cheaper tickets) and the airline is better off (evidently)---but A-C passengers are worse off (the price of their tickets increased).
So unlike price discrimination between tourists and business travelers, which you described in your first paragraph, this type of price discrimination does not seem to be a Pareto improvement?
So if the the airline has free seats on the A-C flight, it seems it could do to things; either lower the price on the A-C flight until the plane fills up, or use the spare capacity to serve A-B passengers as you describe.
If they didn’t have the price discrimination option (e.g. because there was a law against it, or because everyone is using hidden city ticketing), then they would be forced to take the first option. If we then change the situation so that price discrimination is allowed, then A-B passengers are better off (the get cheaper tickets) and the airline is better off (evidently)---but A-C passengers are worse off (the price of their tickets increased).
So unlike price discrimination between tourists and business travelers, which you described in your first paragraph, this type of price discrimination does not seem to be a Pareto improvement?