Having sifted through airfare studies and the quarterly DOT fare data for newspapers for coming up on eight years, here's my two bits on the topic though no one asked and I probably shouldn't be saying anyway. (But it's Friday and the thread is my guilty pleasure and I can't help myself)

First, the nation's fare data can be manipulated to present nearly any conclusion. Any. It's because of the variables of what day fares were sampled, what time of that day they were sampled, what airline they were sampled from, what month (seasonality), what search engine (they vary considerably) and especially what competitive situation you're pulling from. If you use the precise same variables to compare cost-per-mile walk-ups from CVG and FLL, I think we all know the conclusions because Delta faces practically zero LCC trouble at CVG and FLL is an LCC free-for-all. Plus, CVG is going to have a proportionately larger "business" percentage out of all the tickets sampled than our sunny, leisure-flavored traffic at FLL. Of course average fares will be sky high at CVG, as they are at any single-carrier dominated hub in this country. But think about WHY the "average" fare is that way in how the fares are sampled by the government.

My point is that statistically, the DOT figures that are oft-quoted samples every market the same, pulling one of every ten fares and spitting out the results, which generally show "big hub fare premium" at places such as D/FW and MSP and what have you. Media dutifully report this premium and suddenly it's accepted fact. Now, if any of you deal with statisical sampling you might wonder how the variables of each airport might skew the DOT fare data. I offer for debate the concept that proportionately, big hub cities are going to have a much, much bigger pool of walk-up-type fares to be sampled than you'd see at metro airports that aren't hubs. That's the pure nature of hubs - they create thousands of non-stop seats and the majors set them aside specifically for the last minute traveler so they can extract profit from them. I'm not here to comment on that practice, but the reality of how hubs have this huge inventory of last-minute business seats compared to other cities is that the fine folks at DOT are going to pull proportionately MORE walk-up fares from hubs than they will from nonhubs. We know walk-ups are the Lexuses of fares, and if you compare a sample of ten fares from a non hub you're getting mostly, well, Kias. Take ten mostly Kia fares from a non-hub city (maybe one or two Lexus fares tossed in there) compared with DFW's sample of ten fares that includes probably FIVE Lexus fares, what will the AVERAGE fare be between DFW and a non-hub city in that comparison? Of course! It's substantially higher. It's because the DOT's unsophisticated sampling model doesn't differentiate between hub and non-hub and the fact that hubs are here to extract business fares from the local O&D, while nonhubs are offering fares that, on an adjusted basis, are always going to be cheaper than what we pay for nonstop service out of a hub. And hence you have many business writers dutifully writing about the "hub premium."

But is the cost to fly walk-up from a non-hub city more than what it is from D/FW? I'd offer that for a non LCC-affected city it's mostly a wash route-by-route with a big hub. If you're Sacramento and have Southwest, it's good times on walk-ups. If you're Des Moines, tough luck on that walk-up fare and your RJ flight to a hub... That difference can't be reflected in most fare studies, and certainly not in the DOT data.

In the DOT sampling, the hub premium exists. I'm asking discriminating fare watchers to understand what they're sampling and why it's not particularly illuminating data. Does AA overcharge at D/FW? North Texans certainly seem to keep buying what they're selling, with load factors over 90 percent this summer...

This doesn't absolve AA from charging a lot on an advance-purchase basis to the West Coast, which is my chief gripe, but no LCC there is just a fact of life. Our good friends at AA take great pains to point out that if they're forced to unravel their hub in a post WA scenario, the net amount of service to all points beyond North Texas drops because they lose the ability to feed traffic across their hub to support "thin" markets. Indeed, they've got a point if they chose to intentionally unwind their best money-maker, and I join others here in having some skepticism about that tactic. (That's another rant why AA would do this, and there's been great posts here to accurately question their reasoning. Is it a big bluff?)

I'm not here to disprove hub premiums or prove them, but only to say that it's nearly impossible to generalize fare comparisons because of the complexity of the situation. When I first starting writing about fares back in 1996-1997, Continental Airlines offered about 80 different fare levels on most routes. It's gotten a lot less complex, but even with Delta's "simplifares" mostly matched in the system, there are still dozens of types of fares on most routes, and that variability makes tough sledding in our hopes to model what might happen post Wright amendment.

Beware conclusions from fare "studies," including Southwest's, which AA rightly points out used only AA's priciest, pre-Delta-match walk-ups to achieve maximum economic impact. Is there economic benefit to be had by tossing the WA? Yep. Is it $4.2 billion a year? Not so sure. But we do know that AA faces material dilution of its most profitable hub in almost any WA-lifted scenario, and hence they're making it their business to keep Wright around. That much we do know.

Many many of my bosses past and present wonder why we simply can't reach a definite conclusion about fares in markets, and I try to explain that the variables involved in studying fares from markets would overwhelm my small brain, and I think the brains of my readers.

At the DMN, our last fare story compared Southwest from Houston to six cities on a walk-up basis and American from D/FW to the same cities. Southwest won about four of the six; American won the two where it also had LCC competition forcing it to lower its walk-up fares. We wrote it up as a mixed bag, though the prevailing thinking is that AA is just fleecing us.

End of rant. Please continue the excellent discussion. ET