I haven't looked it up, but my guess is such a change, were it to happen, could possibly cut the other way. When you go to Home Depot and accept 10% off your bill for applying for their store credit card that is probably not taxable income. When you take advantage of a "buy three, get one free" sale at the tire store that is probably not taxable income My understanding is those examples are not taxable income because they are treated as a purchase price adjustment between the buyer and the seller. The same reasoning might extend to a case where an airline award program was structured so as to effectively yield a nearly-flat percentage "rebate" to the passenger.

Moving on to other issues discussed in this thread, several posts have mentioned profitability and the idea that the most profitable passengers should be rewarded the most. I fear that in this regard revenue and profit are not being adequately viewed as distinct concepts. Profit for an enterprise is a concept that is easy to understand and a number that is calculated fairly easily. Allocating total profit amongst individual passengers is far more complicated than just looking at how much each passenger paid, how far they flew, or any combination of those two. Let's look at some numbers from the financial statements and then examine some hypothetical scenarios that should demonstrate why.

For FY 2007 the WN load factor was reported (pdf) to be 72.6%; with that load factor, operating income was $791 million. Over 27.3 billion seat miles were not filled by revenue butts. At average fares, that represents missed revenue opportunities of $3.573 billion. Incremental costs would be minimal (a bit more fuel, snacks, paper & ink for boarding passes not printed at home, labor at check-in, baggage handling, and the gates, etc.). For simplicity suppose those costs would have been $0.573 billion. That means those additional passengers would have bumped profit up from $791 million to $3.791 billion. Stated another way, the revised profit would be 80% attributable to those new-found passengers. Of course wasted seats are always a big issue and we can't presume we could sell them all for the average ticket price. Suppose instead that WN had a system where it could sell every empty seat on a space-available basis for $12 over direct variable costs (DVC). The assumption above about $0.573 billion DVC is totally made up but it implies $17 DVC for each additional PAX, so the fare would be $29). Ignoring practicalities and issues like cannibalism of the existing revenue stream, filling those empty seats with (DVC + $12)/pop butts would boost system-wide profit by over 50%. Under a concept whereby award points are issued in proportion to profit, the folks who paid (DVC + $12) would "deserve" to be awarded a full 1/3 share of all award points given out for all flights. If you wouldn't like a system like that, you shouldn't be in favor of a system that attempts to award points to the most "profitable" passengers, and you shouldn't argue that points should be awarded in proportion to price from a fairness perspective. Of course the average passenger doesn't view the issues in this way, so marketing issues will of necessity outweigh fairness issues.

We can look at another example that is not so outlandish. Suppose each flight carries 80 PAX who paid the average $100 fare, 10 PAX who bought too late and had to pay $140; and 10 PAX who scored a DING! for $60., for total revenue of $10,000. Further suppose profit was $600. If the PAX who scored the great DING! are discretionary travelers who would not have flown if they had to pay $100, then the flight would have earned zero profit without the 10 DING! flyers. It wouldn't be practical to only issue award points to those 10 passengers, so it seems futile to attempt to base an award system on how much profit each passenger contributes to the mix.

The short-haul versus long-haul debate probably gets even more complicated. Costs and revenue are always reported per mile, but you can't just say the passengers who pay more per mile are more profitable. Some costs are fixed, some vary with miles, some vary with time. It takes more hours per mile to fly short-haul routes, so fares per mile have to be higher to cover the costs. But as noted in another post, short-haul flyers pay dramatically more per mile for their air travel, quite possibly more so than is justified by the operating expense difference. I have nowhere near enough data (or time) to try to calculate that.

As a partial substitute for identifying the most profitable passengers, WN could award bonus points to folks on sold-out flights. To be fair and/or as a token of goodwill, they could also try to give bonus points to those who paid an average or even above-average fare but but received a below-average service in return. Taking a middle between two strangers would seem to qualify. Again, practicalities would interfere with implementation. Awarding bonus points to "C" group boarders would be another way to go, and that would take some of the pressure off the rush to be first for OLCI. So again once we try to throw concepts of fairness into the awarding of "loyalty points, things can get awfully messy.