Arnold Kling  

The Airline Union Game

Milton Friedman on Social Secu... Economics for Environmentalist...

'Jane Galt' gives an explanation for airline bankruptcies .

In single-union negotiations, money left on the table now can be partially reclaimed by demanding higher wages in the next round of negotiations. Not so with multi-lateral arrangements. Any money left on the table by one union will be picked up by another one, since negotiations with different unions are staggered rather than simultaneous.

The idea is that if there are multiple unions, each union has an incentive to try to extort rather than concede, even though the combined extortions bankrupt the company. This can be illustrated using game theory. Consider the following payoff matrix, where there are two unions, A and B, and the payoffs are for union A.

Union A's StrategyUnion B ConcedesUnion B Extorts

Regardless of which strategy the other union chooses, our union should choose to extort. Assuming that the other union faces the same payoff matrix, that union also should choose to extort, leading to a payoff of -5 for each, as the airline goes bankrupt.

For Discussion. Do you think that this is a good explanation for airline bankruptcies, or do you think that there are other reasons that the airline industry is unstable?

Comments and Sharing


COMMENTS (10 to date)
Bernard Yomtov writes:

You frame this as a Prisoners' Dilemma. But doesn't that depend on the inability of the participants to enter contracts?

What is to stop unions from agreeing not to walk into this self-created trap?

David Thomson writes:

I have always been ambivalent towards unions. They are indeed a major reason why airlines are always in economic trouble. Airline executives are at the mercy of unions possessing an adversarial mindset.

But is this the only reason? Warren Buffett once said that he had no use for airline stocks because the winners are only able to barely stay in business! I suspect that the relentless consumer price shopping dooms the airlines to be at best modestly profitable. The majority of us perceive airline travel as little more than a commodity product. We could care less if we travel on Southwestern, Continental, or Joe Blow’s Shuttle Service.

Jim Glass writes:

Airlines have two other real problems:

Economically, they have the pricing problems of high fixed cost, low marginal cost industries. Planes are expensive but the cost of filling an othewise empty seat on a flight is almost nil. When competition drives the prices they can charge towards a marginal cost lower than their average cost, ouch. (See ISPs, software companies, etc.)

Structurally, the major, older airlines are still carrying forward needlessly expensive cost structures set during the era when the industry was regulated, protected, and run on a cost-plus basis with airlines competing on the basis of service rather than price. (Remember when flying was a luxury -- "I'm Braniff, Fly Me!")

Those were the days when their current labor relations were instutionalized, and institutions count. The labor problems really are a hangover from the old pre-deregulation days.

The latter structural cost problem of course greatly compounds the marginal pricing problem. It's been doing in the great old airlines one after another -- TWA, PanAm, Eastern... United and American are on the same flight path.

OTOH, the newer airlines that have been able to organize in the deregulated era without inheriting that cost structure from the bad old days have been able to profit. Southwestern showing the way, obviously. It has the same marginal cost pricing issues but with a *much* lower average cost and a great deal more operating flexibility it has been able to devour the market operating under the price umberlla set by the majors that still leaves them
hemorrhaging money.

There are a good number of others following the Southwestern model now. IOW, not all airlines are the same.

Zimran writes:

Unions may not be able to contract with each other to fix wages (and so avoid the bad outcome in the prisoner's dilemma) because it would run afoul of anti-collusion laws. Collusive contracts are illegal.

I understand that applying this to unions is weird since they themselves are in essence a carve out to anti-collusion laws, but there you go. They may be allowed to merge into a single big union, but incumbent union leaders would block that.

Apart from a unionized labor force, the airline industry also has perennial over capacity. Companies may enter and exit the market, but the planes never leave. It is this that puts downward competitive pressure on prices, not the high-fixed low-marginal nature of the business (all that does is make them lose alot of money when times are bad, and make alot of money when times are good).

I went to a UAL talk at Chicago with the ex-CEO of Northwestern(?) airlines, a Chicago Law banckrupcy expert, and Peltzman (a U Chicago industrial organization guy). They said we should not overestimate the Southwest point-to-point model for airlines. Given the economics point-to-point is more efficient only in about 30% of existing routes, leaving hub-and-spoke still the dominant way to travel.

All the best,


David Thomson writes:

"Unions may not be able to contract with each other to fix wages (and so avoid the bad outcome in the prisoner's dilemma) because it would run afoul of anti-collusion laws. Collusive contracts are illegal."

I’m not buying this argument. You are conveniently ignoring the tacit signaling between the union leaders. Everything does not have to be explicitly stated to obtain the desired results.

“I went to a UAL talk at Chicago with the ex-CEO of Northwestern(?) airlines, a Chicago Law banckrupcy expert, and Peltzman (a U Chicago industrial organization guy). “

One should take the public utterances of these folks with a huge grain of salt. I prefer to know what’s being said behind closed doors.

Jane Galt writes:

Collusion of the sort you're talking about is very hard to maintain even under ideal situations. I don't know how you'd maintain it where the unions are concerned; most such tacit agreements rest on quick pricing response, and a dominant player with capacity to severely punish infringers. The unions would have a multi-year lag time from cheat to caught, and I don't know how you'd punish cheaters -- what can the pilot's union do to the mechanics except what they're doing already, which is maximize their value claiming? It seems to me that value claiming is a dominant strategy unless you can build a contract with exogenous penalties.

David Thomson writes:

"...what they're doing already, which is maximize their value claiming?"

You are almost certainly correct. The airline unions are indeed analogous to pigs at feeding time: "I'm getting mine and the rest of you should worry about yourselves!"

Bernard Yomtov writes:

But are we sure that these sorts of collusive contracts are illegal? After all, the agreement would be to hold down wage demands, which would serve the interests not only of the unions, but also of the company itself. If they're not illegal then they can be enforced.

What if the unions simply agreed to synchronize their contracts, so that they were always negotiating simultaneously? Would that help?

Jane Galt writes:

Yes, they're illegal. The very existance of unions is a carve-out to the anti-trust laws; letting them all get together and decide how many crumbs to lead the shareholders would cause trouble six ways from sunday. I'll ask the labor guys I know, but I'm pretty sure on this one.

I think synchronizing wouldn't help very much. For one thing, it would be a hell of a burden on management to try to simultaneously manage six negotiations, and for another, unless the unions were colluding, they'd have incentive to be even more intransigent because of the information problems.

RC writes:

I think the mistake being made here is to view this as a single iteration game. A better analogy would be a multiple iteration game where the game stops and everyone loses if the airline goes bankrupt.

So the unions are like pigs feeding at a trough, but with the knowledge that if they are collectively too greedy in this round they will all starve in the next one. And the airline knows that if it gives in to the extortion demands of all the unions, bankruptcy will result.

Another way to look at it: the airlines know that there is a certain amount of money they can spend on total labor costs without going bankrupt. The unions know this too. So essentially the negotiations become about how this fixed pot of money is divided up among the different unions. If one particular union tries to hold out for more money, it runs the risk that the other unions will come to agreement first, and leave less of the pot for it.

Comments for this entry have been closed
Return to top