« Why Air Transportation Delays? Becker | Main | Against Bailouts--Posner's Comment »



Feed You can follow this conversation by subscribing to the comment feed for this post.


An excellent post, but from my own anecdotal personal experience and those of most of my friends and family, I think the substitution of automobiles for airtravel is larger than suggested. Personally, I now do something I never would have a decade ago - which is avoid air travel at all costs if the trip is under 700 miles, mostly because of Judge Posner's well-described frustrations of modern American air-travel plus the annoyances of the modern security requirements.

Except for the fact that I lose two days of productivity by actively driving instead of passively being flown (and perhaps exposing myself to the higher risk of injury of road travel), I find that for these "short" trips cars beat planes in convenience, comfort, flexibility, privacy, cost, and often even time, especially if one has a layover or experiences an even-more-common delay.

The fuel cost-per-mile component of a trip is only lower if one is traveling alone and only slightly at that (depending, of course, on the automobile making the journey). Add a few family members, however, and the trip is much cheaper per person in a car-pooled automobile than on an aircraft. Also, you will be able to carry much more luggage, and carry it more easily, than you can with air-travel.

I could go on, but my point is that there are two groups of marginal consumers. There are those who would fly regardless of quality if only the price were a little lower, and those like myself who would fly (even for a higher price), if the quality of the experience were a little better. Cheaper air-travel, if it reduces quality substantially, makes me less, not more, likely to use it.


"My guess (and that's all it is) is that the..."

Well, at least thanks for warning me so I didn't have to read the next eight paragraphs.

Are you sure you teach at Chicago?

dc user

You've both ignored one factor that some commentators have pointed to. One of the features airlines compete on (because consumers turn out to be quite sensitive to it) is the timing of flights. As a result, airlines now are tending to offer very frequent small-jet service (think: hourly direct flights on an Embraer) rather than less frequent large jets (e.g., bi-hourly direct flights on a 737). But take-off and landing density is largely determined per-aircraft, rather than by aircraft size. This means that when bad weather somewhere forces an increase in spacing between aircraft, the increased number of aircraft causes cascading delays.

This may be a problem of incompletely internalized costs. Consumers might actually be better off with less-frequent departures that result in less-frequent delays. But since all aircraft are delayed equally, no individual consumer has any incentive to choose the 737 flight over the Embraer. The same goes for airlines -- it would probably be far more efficient to operate the larger jets (putting aside union contracts, which artificially separate labor prices between the two), and it would definitely be more efficient to have fewer delays. But since customers go to the most convenient departure time, the airlines couldn't do anything to rectify the situation without (illegal) concerted action.

How could the problem be fixed? The suggestion for congestion pricing is one method. Presumably, if airlines were forced to pay more for peak-hour take-off or landing slots, they would have an incentive to spread those costs over more customers by using larger planes then. (It's not clear, how this would work, however, given that the airline generally has to use the same plane at peak-times as it does for non-peak times.)
Another solution I haven't seen discussed anywhere would be to give larger aircraft should have some sort of take-off priority when there are weather delays. (After all, from a total consumer surplus standpoint, it may make sense to speed 180 passengers on their way while delaying the 30, 60, or 90 on smaller jets.)


This blog has turned into, "I read the Wall Street Journal and I want to share my thoughts on what I think about the articles."

The reason for the majority of air travel delays right now is the (hidden) battle between the air traffic controllers and the FAA, which has refused them a contract.

The administration is playing Reagan-tough, the ATC guys are slowing things down to "show those guys" and we all suffer.


Will give you some clue as to the power struggle between the administration and the controllers. The FAA is pounding on them and they are retiring in droves--but are not being replaced.

So it's not "nframarginal" (a made-up word?) consumers, it's infrastructure and political power struggles. Always look at the easy answer first.


The quality degradation you describe is due more to institutional failures than to the differences between inframarginal and marginal users. Specifically, while service quality has a number of dimensions, the most significant for airlines is schedule reliability-- which, for travelers, amounts to the absence of delay. Yet with very limited exceptions, government, not airlines neither own and control two critical factors required to produce reliable schedules -- airport runway access rights and the air traffic control system. But the FAA treats both these key resources as "commons" -- with the predictable outcome of overuse and delays.


The quality degradation you describe is due more to institutional failures than to the differences between inframarginal and marginal users. Specifically, while service quality has a number of dimensions, the most significant for airlines is schedule reliability-- which, for travelers, amounts to the absence of delay. Yet with very limited exceptions, government, not airlines neither own and control two critical factors required to produce reliable schedules -- airport runway access rights and the air traffic control system. But the FAA treats both these key resources as "commons" -- with the predictable outcome of overuse and delays.

Judge Judy

The previous poster is correct. When you have a "free" resource, too much of it gets used.

Airports and airspace are free, terminal space and fuel are not. Guess which of these four get used with restraint?


I enjoy reading your blog, but I must comment because there is an error in your post. But first, full disclosure, I am an FAA employee, but am making this post apolitical. You claim the following:

"Speaking of which, the airlines argue that the air traffic control system is antiquated and that this is contributing to air-traffic delay. I find this implausible, because the the system, though operated by the government, is 90 percent financed by taxes on aviation fuel. If the airlines want a better system, they should support rather than oppose higher taxes."

But if you look at this link off of the FAA's Airport and Airways Trust Fund (AATF) website, which you will need Excel (http://www.faa.gov/about/office_org/headquarters_offices/aep/aatf/media/AATF_Tax_Receipts_since_1998.xls), you will notice that air fuel taxes do not make up the majority of revenue for the FAA. In fact, it is considered "Transportation of persons," or the ticket tax. Ticket taxes make up about 70% of the AATF, whereas fuel taxes make up 8%.

This is where the problem lies. Talking to Air Traffic Controllers, they complain of two things -- too few runways and an antiquated air traffic system. The AATF, funded primarily through ticket taxes (a percent on the price of the ticket), is currently being reconsidered in Congress as the FAA's reauthorization is up this year. The FAA's position is to tax users of the airspace (i.e. planes) rather than the passengers. This is because a "blip" on the radar, whether it is a small private jet, or a commercial airliner carrying 200 passengers, is the same. Coupled with the fact that ticket prices have decreased and the introduction of Very Light Jets (VLJs), AATF receipts will not grow as fast as air traffic demand.

Now let's look at the current air traffic system. It needs to be modernized. Currently, it is based on a 1960s system, and as such, controllers have to make sure that planes are separated by x miles (arbitrary numbers for argument). A new system, projected to be billions, will allow for a more efficient system so that airplanes can be separated by x-n miles. This will allow more planes to fly in more airspace and land and takeoff more frequently. It will also allow local disruptions, for example thunderstorms, to be mitigated so it does not ripple through the system and cause delays thousands of miles away.

And the other thing is, yes, we need more runways. However, I personally see that as a band-aid fix.

Sorry for the rambling post, but I hope this was helpful, and I welcome comments.


A quick addendum ... If the above URL doesn't work, try the link below (the above URL was clipped in the post). The above has the history of the AATF, including the Excel doc referenced in my previous post.


And, I did not mean that the state of the Air Traffic system is the only problem. It is a piece of the problem. Cheers.

not an faa guy

Here is a message board post from a pilot who is experiencing the delays. Not due to prices charged by the airlines as Posner states, but due to the reasons the FAA employee (above) states:

you are exactly right. American Airlines definitely had a hand in the "share the pain" traffic mismanagement scheme that is messing up the entire ATC system.

It is being exacerbated by the controller issue.........

This is gonna be a bad summer. I've been flying the northeast corridor for over 20 years, and have never seen the kind of delays that I've seen in the last month or so, for no reason. There is definitely weather out there, but before, they would just put some in trail restrictions and let everyone deviate around the problem. Now, everyone gets stopped until the fix is completely clear, again, making the problem far worse. Before, you got slowed down, but airplanes were departing. Now, when the fix does open, they still put the in trail, but you now have 30 airplanes trying to leave through the open window instead of 10, because they all sat on the ground waiting it out.

They've also taken the decision making process away from the local tracons and center controllers. Before, the center guy would tell the tracon that he can take so many airplanes in trail, and worked the guys around the weather. If it got saturated, HE'D call for the stop, and then start it when he could. Now, some EEOC moron in DC is looking at a tv screen with the routes and fixes on it, with radar superimposed over it. If he sees a drop of rain over a fix, he will stop everybody, take an hour to figure out reroutes, another hour to actually get the clearances to the airplanes, by which time, the weather has moved off the original fix, and you end up getting re re routed over the original fix.

It's ridiculous.

The only thing you can do is get off the gate, so you are in sequence when the morons figure it out, and can leave. You are also getting paid at this point, but it is small consolation when you are 5 hours late, for no reason.


It's interesting that the FAA isn't held responsible for delays, airlines are. So delays are free to the FAA but costly to the airlines, but the FAA gets to decide on delays.

From an economic standpoint, that is bad juju.

not an faa guy

From a controller:

There are offically more new hires these days. It takes about 3 years to get fully quallified as a FPL (Full Performance Level) controller. This year alone, 25% of the controllers will be eligible to retire. Approximately 80% of those that I have talked to will leave. The rest have a need for their full paycheck.

Those controllers that are eligible will have what they call a "throw down badge." At anytime, they can throw their badge down and walk out the door. Retired with pay. Not many will do it but it has happened.

The ATC system is not one that overstaffs very much. They hire when they have a position open. They are having a difficult time keeping up with the staffing levels. The starting pay was slashed 30% about a year ago and the job is not as attractive as it has been in the past. The lower pay is not attracting the best people. Higher paying jobs are where the better thinkers are going. The new people are relying on the technology to separate traffic instead of having a plan. They are reactors instead of controllers. The technology they are using is not that great. More needs to be done with the ATC system but that will take time.

Haris H.


According to the article above, it's more profitable for airlines to fly three smaller flights or so than combine them all into one larger flight. This clearly has benefits: more frequent service gives customers more choices. However, since fees are based on weight, the smaller planes are also cheaper to fly out of airports, actually incentivizing more flights, which creates congestion and where any delay cascades to many flights down the line. Instead of weigh pricing, takeoff and landing slots should simply be auctioned off. The most popular slots will cost more, and while there will probably be fewer flights overall, the number of seats might not change as the prime time flights would be handed to the biggest planes.

Also, two things re: weather. First, USA Today published an industry survey a while back, listing how often airline blamed delays on weather [about two-thirds of the time] to how often weather was actually the proximate culprit [17-20%, or one-sixth to one-fifth]. Second, we've had commercial flying for what, five decades? You'd think we'd have weather figured out by now. Can't we fly around storms these days? Unless the storm is sitting on the runway, I don't see how most weather, being very geographically limited in extent, should be a serious issue at this point in the history of flight.

H Hufner

Here's your (and Posner's) fallacy: this type of pricing didn't just happen this summer. The reason for the delays is ATC coming apart at the seams.

And to answer your question, no airplanes can't just go around weather, ATC shuts down entire sectors and takeoff and arrivals at the first sign of a thunderstorm.

Yes, there are a lot of airplanes trying to get into the same place, but the real problem lies with the FAA. At least the real problem that we're seeing this summer.

The FAA is incredibly short-staffed and limits traffic when they are under-manned on any given day.


Lately I'd rather drive than fly anyway. But look at it from a positive standpoint. For instance, because of all the luggage problems people are looking for new ways to track their luggage via GPS or other methods of tracking. Look at Global Bag Tag for instance. If the owner hadn't gotten fed up with the way he and his wife were treated, do you think he would have come up with such a great product and business idea?


I'm aware that the pricing scheme as it stands has been around for a while. But given economic and population growth, demand for flights has gone up, probably leading to a bigger supply of flights, while the FAA probably hasn't increased staff enough to compensate. We could probably alleviate some of the delays with a more rational pricing policy.

Second, about flying around weather: I realize that ATC shuts down entire areas, probably prematurely. My point is that physically, airplanes are capable of getting over and around a lot of storms and that ATC should let them do so.


Everyone that I know flys coach and takes the cheapest flight that they can find. This tells me that the airlines can afford to give even worse service.


"When price is at its competitive level, the sellers...can reduce price further only by increasing the quality of their product..."

I don't follow this.

Freddie Sirmans

Just browsing the internet, your blog is very, very interesting.


If that problem is as you state - it seems that one airline could snatch up all the inframarginal customers by raising its prices and offering super-service air service. If the service issues are with the airlines, as you state, then this airline could fix the problems with extra revenue. If the problems are with the system, as many comments have stated, this method would clearly not work. It just seems like there should be some service standard between regular and owning shares in private jet - unless by definition there are not enough inframarginal custormers to go around .... just a thought...


"When price is at its competitive level, the sellers...can reduce price further only by increasing the quality of their product..."

I don't follow this.

Basically, if the price is at its competitive level, firms are pricing very close to cost, and profit margins are small. If a company wanted to make its product cheaper by lowering price, it would have to charge less than cost and thus lose money, so that is not an option. Instead, a company can improve its quality while keeping price the same; this functions the same as a price cut, since you get more for the same price.
Hope that helps.


To the Controller

If the system is unable to hire enough skilled controllers then the simple solution would be to charge the airlines more for the service. If that improves service and profits, the airlines should be happy to invest in the improved service. It would seem that the airlines do not think this is the core of the problem (or that the benefits are too indirect.)

To the Pilot

If too many flights are placed in a limited window then you need to find a way to allocate the slots. We can agree that if you are so close to capacity all the time, weather can create a cascade of problems that can cripple the system. So the solution is to allocate the landing slots smarter.

To The FAA

More runways are a very expensive solution. But even if we go that way, we still should look for the best way to allocate landing slots. It is not an either or choice.

To Posner

The airlines do try to segment the market and gain some of the consumer surplus. For example, Southwest does offer discount fares for those who travel on Saturday, Weds, or Thursday. I assume that they are off peak days for them. That of course implies that flights on other days have greater demand. Rather then reduce prices on slow days, they should be able to raise prices on busy days and use the additional revenue to buy prime landing slots.

An interesting experiment would be to auction landing rights on Mondays and Fridays at the ten largest airports (by passengers) for two years. For at least two days a week the airlines would not be competing for the marginal flyer but for the inframarginal flyer i.e. the flyer who is willing to pay a higher price for a higher level of service and dependability.

You could also sell insurance on these Monday, Friday flights i.e. the chance to buy insurance against flight delays, lost luggage, etc. You would promise that first class passengers get on a flight that day. If you have enough flex in the system, you could allow the airlines to buy some priority slots in an emergency at a higher price (or even sell options on additional slots for a future potential emergency)

Rather then find some optimal pricing scheme, auction the slots. Potential problems are that small cities may get shut out on some days (or face much higher prices.) Politically this could cause problems. Of course it may become more attractive to fly direct flights between midsize cities (Milwaukee to Pittsburgh to avoid Chicago congestion.) Hub airports on the edges of major cities could become more attractive - high demand consumers could fly to central airport, low demand consumers fly to more distant airport. These fringe airports should be cheaper to build (ignoring the politics of such things) and could serve the marginal consumer who is a burden to flyers like Judge Posner.

Steven Weseman

I am grateful for Professor Posner's comments and explanation on this problem. I enjoyed learning the "economics" explanation for what business travellers seem to understand: intra-U.S. airline flying seems to be substantially driven by the cost-conscious consumer market.

I would be grateful for Professor Posner's further comments on the relative significance of control over airport access rights and gates. Here in Chicago I question whether new airline can competitively gain access to a gate.

A related question I would have is whether market forces in the context of limited airport access could ever lead competing airlines to reduce flight schedules at an overtaxed airport. American Airlines and United Airlines at Chicago's O'Hare airport would be my example.

Although my populist sensibilities have lead me to question whether I am just a whining over-privileged business traveller, I know that our common flight delays in the U.S. are becoming a national competitiveness issue in the minds of foreign business leaders.

Finally, without wishing any further inconvenience on him, I hope Professor Posner avoids the temptation of the private flights I am sure he could obtain. Since I am looking for at least a quasi-government solution here, I believe a key part of the problem is that the very privileged level of U.S. society is not sitting on runways with the rest of us.

Steve W. (I withhold my last name only because the word "populist" could be used against me at the office).

Mr. Econotarian

Reason has a series of articles on this issue:


In particular this article:


Mentions that over 40 countries have self-supporting, commercial ATC operations (such as France, Canada, the UK, and Germany).

I find it funny that even in a country where private health care is basically illegal, Nav Canada is a non-governmental organization. Like the UK NATS, they both depend on per-plane user fees rather than ticket & fuel taxes.


مركز تحميل

The comments to this entry are closed.

Become a Fan

May 2014

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31