Time to auction flight slots & clamp down on flight time padding

Last May, the DOT dropped a proposal to auction landing slots at New York’s 3 airports.   I think this was a huge mistake.

If you’ve been unfortunate enough to sit on the tarmac of JFK, Newark, or La Guardia for 90+ minutes waiting to take off on a clear, beautiful day, you can understand my frustration.  Airlines routinely schedule more takeoffs per hour than these airports can accommodate, therefore making it impossible that all scheduled flights can take off on time.  Delays are 100% guaranteed in this case.

According to a 2009 report from the Partnership for NYC, flight delays cost the regional economy more than $2.6 billion in 2008, including $835 million in fuel.  Importantly, this takes into account the padding that airlines put into schedules. (more on that in a minute)

Everyone knows what happens when the pilot comes on to cheerily announce something like “There’s a little traffic jam this morning in NYC!  Looks like we are number 19 for takeoff, which might happen in about 40 minutes.”  There’s an audible grumble, and some people worry if they’re going to make their connections.

But here’s the thing: if you polled every person on that aircraft at that exact moment,  offering them the option to pay $50 to take off right away, there’s no doubt that some would eagerly accept and others would decline, preferring to accept the delay.  Rather than having everyone inconvenienced, the people who value the specific takeoff time the most will pay for it.  Airlines should be figuring that out in advance and pricing seats accordingly. Instead, they are transferring the ‘cost’ of higher tickets into a cost of wasted time (and fuel, and productivity). This is econ 101: price the scarce resource so that demand falls to meet its supply.

Opponents to auction pricing say that it’s antiquated air traffic control technology and a lack of experienced air traffic controllers that’s to blame, not runway congestion.  Other opponents may have a more hidden agenda: it is assumed that if takeoff slots were auctioned/rationed, then airlines would start flying fewer, larger jets and reduce smaller, regional jet service.  Less regional jet service means inconvenience for congress people (and their constituents) who want frequent service to regional airports close to them.

Opponents also point to alternatives to increase capacity: opening up underutilized military airspace, for example, or investing in on-board navigation systems that can give pilots more maneuvering options than ground-based radar allows.

But regardless of the underlying reasons and regardless of the possible long-term solutions, two facts remain clear: 1) flights can’t take off at the rate that airlines schedule them, and 2) airlines aren’t pricing these flights to take this into account, in order to shift some demand to less busy times. Side note: It’s worth noting that Mayor Bloomberg supported auction pricing, to his credit.  Not surprising, given his very sensible approach to congestion pricing for cars in Manhattan (a measure which was also struck down but which I strongly supported).

I have another gripe: the practice of airlines to massively pad their schedules so that their on-time performance record doesn’t suffer from their addiction to over-scheduling.  I routinely fly a 2 hour route from LGA to Charlotte that has at least 40 minutes of padding in the schedule.  I’ve been on JFK-Europe routes that sat on the ground for 90 minutes before taking off, yet arrived well before schedule.

Besides wasting time and fuel, padded schedules are highly misleading and distort on-time performance measures.  It’s akin to reducing a product’s base price but then adding hidden fees to the final price.  The airlines are, in fact, already penalized for this practice since flight crews and cabin crews are often paid based on scheduled, rather than actual flight times.  But clearly they find it more expedient to incur this cost rather than to shift flights to times when they can actually depart shortly after leaving the gate.

Since the DOT started requiring airlines to report their on-time performance, surprise! they’ve tried harder to be on time. (it’s the old “What gets measured, gets done” theory of management) What’s more, the airlines turned this into a marketing opportunity:  You can be sure that whoever is in the top few performers will be talking about that in their ads the next year. So it’s unfair and misleading that airlines can simply ‘lower the bar’ on performance by padding their schedules.

My proposal is that airlines be forced to publish their schedules based on actual flight time + no more than a single digit percentage.  Some allowance is reasonable since there will be variations based on headwinds and unpredictable ground conditions, but consciously overbooked takeoff slots don’t fall into those buckets.  A less favorable alternative approach would be to penalize airlines when on-time tables are calculated, by discounting their reported on-time performance by the relative disparity in their scheduled flight times vs. actual flight times.  Either way, this would put more pressure on airlines to stop scheduling flights when they know the airport is congested and facing almost certain system-wide delays.

It’s high time for the airlines to stop pretending that there is no cost to their over-scheduling of flight slots.  Auction pricing would be an efficient way to take care of this.  It’s also time for airlines to stop artificially manipulating their on-time performance with wasteful and misleading schedule padding.