Law students, I have found, often have a hard time seeing how the Coase Theorem applies outside the context of land use conflicts. They also tend to think Coase’s insight is not so important because, they recite (parroting some of their professors), “Transactions costs are always present.” This saddens me, for the more I look at the law, the more relevant the theorem becomes.
My Business Organizations class recently had occasion to consider the Coase Theorem in connection with our discussion of agency law. I thought some TOTM readers might be interested in how the discussion progressed.
We were discussing the general rule that an agent, unless otherwise agreed, is not to earn profits in competition with his principal and must disgorge to the principal any profits he so earns. (See Restatement (2d) of Agency Section 393.) To illustrate the rule, we considered the case of General Automotive v. Singer, in which the court ruled that a star employee (Singer) of a small machine shop (Automotive) had to disgorge to the shop the profits he had made by brokering jobs that the shop was not equipped to handle.
Singer, it seems, was an expert in estimating the cost of machine shop jobs and in determining which particular shops could do the work most cheaply. Accordingly, when a customer would place an order Automotive couldn’t fill, Singer would find another shop to do the work and would earn a brokerage fee from the customer. Since Automotive couldn’t do the work itself and didn’t engage in brokering services, Singer’s sideline business did not compete directly with Automotive. Nonetheless, the court held that Automotive was entitled to the brokering fees Singer earned. In so holding, the court effectively gave Automotive a property right in all machine shop work and machine shop brokering services that came Singer’s way. Automotive had “dibs” on all that work and could prevent Singer from doing any of it for his own profit.
The students don’t have much problem with Singer’s liability. He had signed a contract in which he promised to devote all his time and attention to Automotive’s business, and his sideline job would seem to breach that contract. The remedy, though, seems odd. Since contract damages are measured by the amount needed to put the promisee in the position it would have occupied but for the breach, it would seem Automotive’s damages would be near zero. After all, Automotive couldn’t have done the brokered jobs and didn’t engage in any brokering services itself, so it wouldn’t have gained anything had Singer kept his “no moonlighting” promise. The court, though, awarded Automotive an amount equal to the profits Singer had earned off his brokering activities. Such an award ($64,000 in the case) would seem to violate the bedrock principle that punitive damages are not available for breach of contract.
Of course, the legal nuance that explains the case is that Automotive’s claim was for breach of fiduciary duty rather than breach of contract. But that just raises the question, why have the remedy differ so dramatically based on how the claim is phrased? Most actions against moonlighting employees could be construed as either a breach of contract (implied, perhaps) or a breach of fiduciary duty. Why have the remedy differ so dramatically if the action is cast in terms of fiduciary duty?
Here’s where the good Professor Coase makes his entry.
I ask the students what would have happened had Singer approached Automotive and asked for permission to engage in brokering. As noted, Singer was a star employee, so Automotive wouldn’t want to lose him. It might therefore permit him to engage in the sideline business, though it would likely monitor his independent brokering to ensure he wasn’t diverting jobs that Automotive could actually handle. On the other hand, Automotive might want a share of the brokering business itself and might therefore demand that Singer provide the brokering services on Automotive’s behalf (i.e., it might expand the firm’s services to include brokering of complicated jobs). If it took that tack, it would have to increase Singer’s compensation.
While we can’t be sure what precise bargain Singer and Automotive would strike, we can confidently predict that the parties would bargain for an outcome in which Singer was permitted to provide the services — either for himself (with Automotive’s monitoring) or for Automotive (with additional compensation to Singer). The parties would not bargain for an outcome in which Singer was precluded from engaging in brokering, for such an outcome would forego wealth that could otherwise be created and distributed among the parties. Thus, even if the law gives Automotive the right to ban Singer’s brokering, the parties will eventually strike a bargain allowing Singer to engage in the work.
This, of course, is the Coase Theorem: It matters not to whom the property right in brokering services is initially allocated (Singer or Automotive); if bargaining costs are sufficiently low (and they would seem to be so here), the parties will strike a bargain in which the party who can use the right to create the most wealth gets it.
Yeah, but if the parties would eventually bargain for Singer to have the right to broker, why wouldn’t the law just give him that right from the outset? Wouldn’t that be more efficient, since it would eliminate the need to engage in bargaining?
Here’s where the penalty default makes its entry.
When we were playing out our hypothetical bargain between Automotive and Singer, we could predict with confidence that Singer would ultimately be allowed to engage in brokering, but we couldn’t confidently predict the terms of the deal. Would Automotive expand its business to include brokering services and pay Singer extra for providing them? Or would Automotive permit Singer to have his own sideline business but monitor him more closely? We can’t say ex ante. The answer turns on the parties’ preferences, to which we are not privy. We’re pretty sure, though, that if the parties were to sit down and talk through all this, they’d strike a mutually beneficial deal. We therefore want to encourage a face-to-face bargain.
How might we do this? Well, if we set the default rule — the outcome that will govern in the absence of a contrary agreement — to be something we’re pretty sure the parties wouldn’t like, we can encourage them to come to the bargaining table.
And that’s what the law does: It sets the default rule to be the opposite of what the parties would likely select (i.e., star employee is not allowed to exploit wealth-creating opportunity) in order to generate the sort of Coasean bargain that ultimately maximizes wealth.
Maybe George Costanza was on to something.
[Professor Bainbridge provides much more (including an audio lecture) on the fiduciary duties of agents.]