« Brad DeLong's lectures | Main | Irrational beliefs I hold about carbon emissions »
Why are there strikes?
The UAW at General Motors has gone on strike. But why are there strikes at all? John Kennan wrote (full text gated here):
There is no commonly accepted economic theory of strikes. The main obstacle is that if one has a theory which predicts when a strike will occur and what the outcome will be, the parties can agree to this outcome in advance, and so avoid the costs of a strike..strikes are apparently not Pareto optimal, since a strike means that the pie shrinks as the employer and the workers argue over how it should be divided. If the parties are rational, it is difficult to see why they would fail to negotiate a Pareto optimal outcome.
Hicks suggested two possible explanations for strikes: either the union is trying to maintain a "reputation for toughness", or there is private information on at least one side of the bargaining table...
The NYT claims that the union had to "draw a line in the sand." More generally, David Card wonders whether striking workers are forcing employers to reveal how profitable they are. The firm, if indeed it is profitable, will come back with a higher wage offer, but only if it's hand is forced. Otherwise the union does not know how much surplus can be grabbed because the firm will not have to reveal it. Or can workers, using institutions and morale, somehow precommit to a resistance curve? Such a precommitment is optimal ex ante (it reaps a greater share of the surplus by conferring a bargaining advantage) but not always best ex post because gains from trade can break down. That is my guess in this case.
Perhaps Dave Ribar has the wisest comment:
The only "good" news (if you can call it that) is that GM's manufacturing workforce has shrunk so much that many fewer workers are involved than in the earlier UAW strikes.
About 73,000 workers have gone on strike. Here is the decline of General Motors.
Posted by Tyler Cowen on September 25, 2007 at 06:41 AM in Current Affairs | Permalink
Comments
But isn't the chance that the union goes on strike one of the variables that determine the Pareto optimum, and so by going on strike the union changes one of the variables in (they hope) their favor?
Posted by: Hei Lun Chan at Sep 25, 2007 8:14:40 AM
'Otherwise the union does not know how much surplus can be grabbed because the firm will not have to reveal it.'
This a revealing statement - the workers are 'grabbing' the surplus of their labor, a surplus I suppose is instead reserved for CEOs, various financial institutions, and those almost mythical individuals (see preceding group), shareholders.
Yet strangely, when a company like Exxonmobil, making a measly 10 billion dollars in profit a quarter is asked by some of its shareholders to fully fund pension commitments to its employees, to the tune of 4 billion dollars, I have yet to hear about how those involved in not distributing 40% of one quarter's profit to fulfill the company's commitments are 'grabbing' the surplus.
Strange how far the idea that labor creates wealth has drifted in America. Especially when workers, bargaining collectively, are now considered to be 'grabbing.' Strangely, though, as management pay has risen at GM, while that same management has been making decisions which have driven the company to the ground, I hear nothing about how they kept grabbing what they could.
Must be a matter of perspective - after all, if workers would just stop demanding higher wages, the factory floor/management level pay disparity could rise with the majesty which so many Americans now accept - except for those nasty unions.
Strangely, though, Porsche still employs union labor, while proudly possessing the highest profit margin in the industry - and a CEO that rejects what the stock market demands of him in terms of managing his business to its standards, and not his. BMW also builds its car with union labor, as does Mercedes. And VW. Possibly this is just something uniquely German - maybe it has something to do with all that vacation time and the idea that employees are the primary asset of a company?
Posted by: grabby at Sep 25, 2007 8:18:17 AM
...i'm not sure that anything happening in europe related to labor is at all a result of the decisions of the company, nor should be attributed to them as a credit, but rather a cost of doing business *in* europe.
Posted by: shawn at Sep 25, 2007 8:34:36 AM
I think the Pareto efficiency argument is a bit of a red herring: both sides are seeking to maximize their own gains, not rationalize the overall system. Besides, in a simple pie-slicing negotiation, ALL outcomes are Pareto efficient. If anything, you see strikes because it's in each sides interest to misrepresent their own utility curve so an apparantly Pareto efficient solution is actually to their advantage, and thus they suspect that the other side is doing the same.
There's something to be said for the signaling argument, since an expensive signal is usually more likely to be believed than a cheap one. After all, you only have a strike because both sides think the other can give somewhere, so there is by definition an information discovery component to it. A strike is, quite literally, each side discovering the other's resistance/concession curve.
But striking might simply be the rational gain-maximizing move. Sure, a strike destroys value, but if the other side is loosing faster than you, you've just shifted the negotiation frontier in a way that might make your side better off in the end. Again, comes down to each side having uncertanty about the other's true utility function.
Why does the partner ever reject the deal in the ultimatum game (http://en.wikipedia.org/wiki/Ultimatum_game)?
Posted by: TomR at Sep 25, 2007 8:51:43 AM
I had always heard that the auto companies were the most unyielding in bad years for all the obvious reasons PLUS they wanted to use non-production-time during the strike period to sell off a backlog of surplus vehicles (or create scarcity) without having to pay the cost of involuntary layoffs. In other words, the union pays the workers for the down time. A layoff costs the auto companies much more, thanks to union contracts already in place. When car demand is good, there are lots of reasons to settle by negotiation, including this. This probably doesn't explain strikes outside the auto industry, but maybe there are similar factors.
Posted by: anon at Sep 25, 2007 9:11:18 AM
Hicks suggested two possible explanations for strikes: either the union is trying to maintain a "reputation for toughness",
Um, duh?
Let me guess, Tyler_Cowen. Do you also stay awake at night pondering these *perplexing* mysteries:
-Why people don't pay kidnappers' ransoms, even when they value their daughter more than $15,000.
-Why in practice people don't pay nuisances to stop throwing off negative externalities.
-Why some people terminate contracts at a loss, even though putting up with more short term dishonesty
from the counterparty might lead to a cheaper solution this time.
I'd love to approach the world with the level of awe that a lot of economists have.
Posted by: Person at Sep 25, 2007 9:31:17 AM
Professor Earl Thompson at UCLA had a theory that collective bargaining was a legal means to by-pass
antitrust laws. Producers could collude at the union bargaining table. The goal of a cartel is
monopoly profits derived from a coordinated reduction in output. By "hanging tough," the producers
could force the unions to strike. That reduces output, concentrated at one time rather than spread
out over the year. In this theory, producers create strikes for their benefit.
I would add that strikes end when producers agree to share monopoly profits the union.
I would further add that fresh competition destroys the cartel profits. If the union has been sharing
the cartel's profits, the union must give back its share, or the cartel will become so unprofitable that
it shrinks and ultimately vanishes. That is my explanation for what has happened to "Detroit."
Posted by: B.H. at Sep 25, 2007 9:56:46 AM
anon's point is a good one here: Given that GM's dealers' lots are full, what cost is it to them? They'll sell of surplus for a bit and get inventory levels more sensible.
The dealers must be cheering the strike.
Posted by: ah at Sep 25, 2007 9:58:15 AM
What does economics say about the game of Chicken?
Posted by: Anderson at Sep 25, 2007 10:13:15 AM
You might also ask why wars occur, since they are clearly not Pareto optimal either.
My suggestion is not that there is private information, but that there is a lack of information. Nobody can really predict with great confidence how the strike will turn out right now, so it's not possible to agree on the ultimate outcome and save the costs of the strike.
Posted by: Foolish Jordan at Sep 25, 2007 10:13:58 AM
Ah:
From the perspective of GM it's not the dealers that benefit from the strike it's the entire company. My only question is that if we can come up with clear ways in which GM benefits from the strike this whole discussion has nothing to do with any Pareto argument. The only remaining question is, why would the UAW choose to strike?
Posted by: Fever at Sep 25, 2007 10:40:40 AM
Could it be that they are upset about the HHS asking its employees to buy Japanese or South Korean cars?
http://www.ft.com/cms/s/456c8478-686a-11dc-b475-0000779fd2ac,Authorised=false.html?_i_location=http%3A%2F%2Fwww.ft.com%2Fcms%2Fs%2F456c8478-686a-11dc-b475-0000779fd2ac.html&_i_referer=
Just a thought.
NB: FT requires registration...
Posted by: Shefaly at Sep 25, 2007 11:23:19 AM
Kennan's theory on why strikes shouldn't exist relies entirely on the existence of strikes. Strikes are actually quite rare, but are a result of asymmetrical information.
Posted by: Willy at Sep 25, 2007 11:45:39 AM
To Fever: Why would the union choose to strike if it's in GM's interest to let them pay for a layoff? If they offer a really disadvantageous contract, the union management has to do something -- calling a strike is a good something -- to keep the confidence of the rank and file. It's been quite a long time since the union and the workers really had indivisible interests, and it's hard to vote in new union officials who really represent the workers. So strikes involve a lot of playing chicken with more than two "sides."
Posted by: anon at Sep 25, 2007 11:48:38 AM
If you want to play a fun game, try explaining a strike to a child. "They just don't go to work? Why don't they get fired?" Why indeed.
Posted by: Noah Yetter at Sep 25, 2007 12:05:39 PM
I was thinking the analogy to war was a good one. In a world of perfect information and no uncertainty, you'd think wars would be quite rare. They would only involve parties with interests that are simply not compatible under any circumstances (i.e. racialist absolutist ethnic-cleansers like the Nazis.) The rest of the time, a clear understanding of the two sides' capabilities, cost structures and utility curves would presumably result in an optimal solution being negotiated prior to hostilities.
But usually neither side has perfect information about the other side's capabilities or utility curve, nor even about their own. Same with unions and managers. Add in that both the managers and union leadership are subject to agency problems.
Posted by: Bernard Guerrero at Sep 25, 2007 12:49:55 PM
Actually, I think there is more "good" news. Given the quality of American vehicles, won't we all benefit if there are fewer GM cars on the road.
Here's to a long strike!
Posted by: Allan at Sep 25, 2007 1:14:09 PM
Looks like Bernard_Guerrero is just as confused as the average economist.
Yes, war is another case where people get confused for the same reason.
Prior to a war, one side could prefer ceding X land, to a war in which many of its people will die and then
it gets occupied and loses 3X land.
However, if it actually did take the first option, tomorrow, a nation -- which could include the one it
just negotiated with, could make the same demand, again: "Give us more land."
Build a rep as a pushover, people will take advantage of that.
Why is this so easy for everyone except economists to understand?
Posted by: Person at Sep 25, 2007 1:57:35 PM
In this coountry, Venezuela, we tenured professors were paid while we were in strikes.You had no choice but join strikers because the University was closed, almost.In 1997, the goverment supended transfer to the Universities, during the strikes.It was a national strike.After 4 months without payment ,the strike finished.It was the last one.Anyway the government paid us the months not worked.
Btw , after every strike we agrred with the initial offer of the goverment.We never won any strike.
Posted by: juancarlos at Sep 25, 2007 4:12:39 PM
I think a part of the issue that has not been mentioned is emotions - you might as well call it irrationality. When I hear workers on strike interviewed, they practically always seem honestly fed up and really cross because they feel that The Big Man is exploiting them. Going on strike is their way of sticking it to him.
Posted by: LemmusLemmus at Sep 25, 2007 4:34:15 PM
Person,
I don't think you're quite getting it, yourself. My entire point was that if everybody was certain of the outcome, exactly what you describe would happen? What would I care for your "rep" if I knew I could beat you regardless? The outcome being essentially pre-ordained, your "rep" would be of no importance.
Introduce uncertainty about your actual capabilities and your "rep" then becomes important. Assuming I give it some credence, my calculation of who would win might be thrown off in the manner of a predator looking at a frilled lizard. At the least it introduces uncertainty. A "rep" is necessary but not sufficient; it can only exist as separate from your actual capabilities in an environment of uncertainty. Perfect information means the strongest win, period.
Posted by: Bernard Guerrero at Sep 25, 2007 5:02:00 PM
Just a thought - why doesn't the UAW hold the strike long enough to severely damage the value of GM, and then just buy it outright? Their 73,000 members could easily leverage a quick 7.3 billion to get started, plus what the union coffers have acumulated.
Then they could run GM whatever way they wanted, since it would be theirs - they could run it on a non-profit basis, solely for their own good, if that's what they wanted. Of course, that would mean taking on all the business risk themselves... But it would be a socialist's dream - the workers in ownership of the means of production. What a great experiment that would be!
Posted by: persiflage at Sep 25, 2007 5:29:54 PM
Why is this so easy for everyone except economists to understand?
Actually, economists understand this quite well. One of the ideas as to why strikes occur is that the union must show that it is willing to damage the company, even at a cost to itself, to get what it wants, so that future strike threats will be taken seriously.
I'm not sure if that's what Tyler means when he talks about "precommitting to a resistance curve," but it is exactly maintaining the "reputuation for toughness" that he cites.
Posted by: Bernard Yomtov at Sep 25, 2007 6:21:15 PM
Is it possible there are strikes for the same reason that people won't sell their home for less than they bought it for?
Posted by: Shane Milburn at Sep 25, 2007 9:08:23 PM
In a world of perfect information and no uncertainty, you'd think wars would be quite rare. ... The rest of the time, a clear understanding of the two sides' capabilities, cost structures and utility curves would presumably result in an optimal solution being negotiated prior to hostilities.
You could make essentially the same remark about a football game.
Posted by: RJ at Sep 25, 2007 11:39:05 PM
