By lowering fares to that point, UA (or any carrier) is already assuming that they will sell lower fares and not higher ones. It's a gamble, yes... but IMHO a worthy one.

It's not a new concept, either. Many retailers offer price-protection, whether as an explicit policy or simply implicit in their return policy. If I buy a widget for $100 and next week it goes on sale for $80, in almost all cases I can either get the $20 back directly (using the retailer's price-protection policy) or I can return the widget I bought for a full refund and buy it again for the lower price.

So, a retailer puts a widget on sale to attract more customers... but what if the retailer fails to attract more customers and instead loses some of their already-earned revenue from people returning their full-price widgets and buying them on sale? It's exactly the same chance that UA is taking. UA's refare policy is nothing more than the airline version of the above, although it works out even better for UA because they usually issue vouchers which have significant breakage and which encourage future bookings, compared to giving cash back (as retailers do) which does neither one.

Yes, because they (whether by choice or necessity) waited longer to buy. People who book early not only guarantee a sale to the airline (refundable fares notwithstanding) but also provide the airline with revenue earlier, allowing them to "ride the float" on that revenue compared to people who book later.

Of course, the whole discussion of "fair" or not is irrelevant. It would be equally fair for the airline not to have a refare policy, but the fact that they have one doesn't make anything less fair. Regardless, as you can see the refare/price-protection concept isn't new, and almost every retailer implements some form of it. The only reason it's surprising here is because it's not a concept usually applied to airfare, but I don't see why it's any more risky or harmful to apply it to airfare than to any tangible widget. If it works for retailers, it should work equally well for airlines (or in UA's case, perhaps even better since they give vouchers back rather than actual cash).

I don't, either. The only thing I can think of is that it somehow reduces their back-end costs since they don't have to pay for printing and postage for the vouchers, not to mention the manpower required to reissue the vouchers when the originals never arrive. Overall, though, I'd think they could save more money (and still keep breakage relatively high) by simply issuing electronic credit like WN does.