The following is an edited transcript of a Facebook conversation between Joseph Itkis (JI), Jonathan Brice (JB), Shaked Sivan (SS) and myself, Rohan Grey (RG) that was inspired by our class yesterday and we thought might be worth sharing. Apologies in advance for any edit-fails.

JI: I agree with I think is Eben’s main point that law isn't the main drive behind people keeping promises, I just don't see how that's a problem.

RG: Isn’t the important question to ask “what is point of contract law”? If society agrees that a contract is supposed to be a "promise" for action, then is there a value-loss in broader society in being morally neutral to whether that contract is enforced or broken with compensatory damages?

JB: Joe, I agreed with your point that one could easily say that people in the diamond trade keep their promises simply because it is inefficient for them not to...that’s perfectly compatible with efficient breach.

RG: I would disagree - the reason it's inefficient there is because the legal system (in that case, a dude with a machete) imposes an implicit punitive cost on failure to break promises. That type of cost is completely absent in efficient breach theory. Moreover, all current intentional breach is "efficient" breach - the question is whether the law should make it inefficient to breach contracts in acknowledgment of the social value of increased trust in contractual integrity.

JB: From my understanding of efficient breach, in breaching, people take into account whether or not they will die as a consequence. To some, the cost of death outweighs the benefits that they would get from breaching.

JI: Rohan - that was Eben’s other main point, which I wasn't really talking about. I think contract law not caring about keeping promises is totally fine. I prefer that than the law following me around making sure I keep my promises. Not sure how that would be different than making me go to church/synagogue on weekends. The morality of keeping promises is people's own business.

JB: Why does it matter if the law makes it efficient or not? I wouldn’t agree that all breach is "efficient" breach.

RG: The difference to making you go to church/synagogue is that a contract is a "legally enforceable promise" by definition. Otherwise what use is contract law at all? All intentional breach is efficient in that the actor breaching determines that the risks of expected costs are worth the likely benefits to itself. The question is whether the legal system should remain agnostic or not to the broader social harms of that breach that may not be internalized currently.

JI: If we follow Holmes, a "contract" is whatever we want it to be.

RG: Eben’s two points are not distinguishable, I think - if "efficient breach" simply means the person who breached thought it would be a good idea and was willing to take the costs on themselves, then that's just a truism and doesn't give us any direction as to the question of how a court should assess damages. What does the legal profession/society say the purpose of enforcing any contract law is at all? What is the difference between a promise to my grandma to go to church and a promise to sell 4000 units to another company written by a team of lawyers?

JI: ‎"contract" is a word that allows courts/the law to step in and control official dealings between people when those dealings break down.

JB: I think the law has a broader view of contracts. A contract is a means to an end, not the end in itself. Most people don’t care if a contract is upheld, they just care about why they got into the contract...often times, damages get them to that point

RG: Exactly - so the real question, and Eben’s point (I think), is what value do we place on social trust? An efficient breach theory says zero.

JI: the difference is that in the first case, you presumably don't intend to give the law authority to resolve the issue when a dispute comes up, as opposed to in the second case. Once the law is involved, I don't see why it should be a given that the law should care more about the morality of promises than about economic efficiency.

RG: I agree a law of contract is means to an end - that end would be social progress, no? Eben’s point as I see it is that economic efficiency has to take into account the value of social trust, which efficient breach assumes away definitionally by being agnostic on whether a promise is enforced or not.

JB: How does efficient breach put zero value on social trust? It simply says that sometimes, it’s better for all parties involved if someone breaks a contract.

JI: I think contract law recognizes that trust plays a huge part in promises and dealings and will continue to do so regardless of the law.

RG: Jonathan, have you ever seen an efficient breach analysis take into account the cost of a legal approach that is neutral on performance and non-performance with compensation on the breakdown of a stable system of legally enforceable promise-keeping between consenting members of society?

JI if contract law has to force people to trust each other, then that's a pretty artificial version of trust.

JB: It depends.

RG: Joe, contract law (or at least, efficient breach) might recognize that trust exists outside of the law, but doesn’t input that value adequately into its economic calculus, in my opinion. Also, we force people to not kill each other (i.e. be peaceful) and that system has tangible benefits to peace, don’t you think? I'm pretty sure a policeman doesn't stand by and agnostically let you kill someone as long as you're willing to cop a ten year sentence. Also, Jonathan, I would love to a link to that efficient breach analysis if you have one.

JB: Well in breaching, the person also takes into account whether people will deal with them in the future and whether future parties will abide by future agreements with them (or try to screw them over vindictively)...isn’t that in a way taking into account the "cost allowing a breach on the breakdown of a stable system of legally enforceable promises"?

RG: Jonathan, I think that analysis suffers a bit from a fallacy of composition - if all businesses assume everyone will breach every time its economically efficient because the law explicitly condones it, then the aggregate trust is gone already at the outset of the analysis as it’s conducted by any particular firm.

JI: I think the problem is that, despite what we talk about in torts, we don't want to start talking about economic efficiency when talking about one person intentionally taking the life of another.

JB: I agree with Joseph – that’s basically what Holmes says as well.

JI: And actually, looking at what we've done in torts, sometimes an "efficient breach" type of question does come into play even when discussing killing people. The reason why it's much more difficult is because assigning monetary value to a life is a complicated/impossible matter.

JB: Or assigning a monetary value to having a free life as opposed to going to prison for life or dying

JI: That's why I think contract law doesn't need to account for trust - because people won't start cheating each other left and right and completely lose all trust, just because contract law doesn't care about keeping promises.

JB: not necessarily. I must distinguish economic efficiency from financial efficiency...often times, it might be financially efficient for someone to breach, but not economically efficient. Economic efficiency takes into account everything including the negative externalities that results from ones actions.

RG: Joe, I agree, and people won’t necessarily start committing crimes left and right simply because the law only imposes a cost-benefit approach to criminal justice. The question is, is criminal law agnostic to the act of a crime (with punishment) and the non-act, if the victim is sufficiently compensated for their loss? If not, there is clearly a prioritization of one option over the other. The question to ask is if we should prioritize breach or non-breach, or remain agnostic? Efficient breach theory says the latter. Additionally, while there might be social reasons not to break every promise if it’s financially efficient, that is exactly what happens with contracts in a perfect efficient breach universe. I also agree that there may still be social trust promises, but that is not the same as trust in the value of contracts per se, merely just trust in other things (handshake, reputation, etc). If that's the case, what's the value of contract law at all?

JB: Isn’t efficient breach telling us to look at the law as the bad man?

RG: Jonathan – my take was that’s true from the perspective of a legal practitioner, not necessarily a jurist. So we should only do that if the judge accepts a theory of efficient breach, which is the crucial question here - I could quite easily advocate a theory of contract law that defines damages as "overall cost of breach + fifty percent penalty for destroying social trust in the value of contractual promises." That is an alternative efficient breach theory that has potentially as much validity as any economic analysis of efficient breach theory I’ve ever seen. And I agree, true economic efficiency takes into account those externalities, but not "efficient breach theory" in any iteration I've ever heard, hence why I said a true "efficient breach theory" is merely a tautology.

JI: I think if you increase the penalty because you value "trust" and non-breach - sure, people will breach less. But will people trust each other more? I doubt it. People will still assume that some parties will breach if non-breach costs more - it's just that now that condition is harder to achieve. So you're decreasing the number of breaches that happen but you're not changing the fabric of society or something like that, you're just shifting the economic equation over in one direction.

RG: By very definition people will place a higher economic value on trust when you shift the economic equation in one direction - people might not trust each other more but they will trust the contract system more.

JI: No they won't. It will be the same thing - people will trust that other people won't breach if they lose money. The only difference will be in how much money people lose. There will be no change in the morality of society or in the general feelings of trust - the motivations and calculations will be the same, just with different numbers.

SS: Isn't there something basic missing from this conversation? Efficient breaches do not negate the value of promises - they just convert promises to perform into monetary damages. The breaching party is still on the hook. Rohan, this is from much earlier, but: "what is the different between a promise to my grandma to go to church and a promise to sell 4000 units to another company written by a team of lawyers?" Contracts are at an arms length, while grandmothers pinch your cheeks. They're different intuitively and legally (putting aside for the moment promissory estoppel / donative promises etc). Also, if punitive damages were allowed in contract law, there may be less flexibility and thus less inclination to contract in the first place. And we like contracts.

JI: I like Shaked's last point a lot, didn't think of it before.

RG: Joe - I agree you can't "generate" trust through monetary sanctions, but you can make it far more valuable in any monetary calculus than the price of $0 it is given in traditional efficient breach theory - hence why I said "economic value" not moral value or "feelings". Even Eben said this today when he said "if you bankrupt everyone who breaks a contract" - the question is whether the economic value should be trivial or non-trivial, not whether you can necessarily inspire metaphysical awakening in people. Shaked, efficient breach theory doesn’t convert promises to perform into monetary damages! That's the exact problem. It transfers monetary damages incurred by one person to the other from non-performance of a promise, but places no inherent value on the non-performance itself. Also, I agree there is a difference with contracts - the difference is they are legally binding. If you want to create a promise you don't plan on having enforced, you can do that in a million ways that don't require a legally binding contract. The difference in a (consensual) contractual case is that you are explicitly affirming a promise, much as one makes a commitment "before God" when they get married - the value of the witness in enforcement is crucial, both legally and spiritually. There may be less likelihood to create contracts, or there may be more liquidated damages clauses, or a whole host of things - maybe more people would make promises that they kept, which would reduce litigation costs and increase business confidence regarding supply chains, disclaimers, warranties, etc.

SS: Well, if a contract is breached the breaching party is obliged to compensate for the expectation damages. You are of course correct that these damages are not the equivalent of turning a promise to perform into monetary damages. Still, promises matter!

RG: Not under efficient breach theory.

SS: Under the efficient breach theory the goal is to maximize total value, but the party suffering the breach is still compensated for the loss caused by the broken promise.

RG: No he's not - he's compensated for damages that can be mathematically calculated and directly attributed under a burden of proof that the judge is satisfied with. Aside from the broader social damage of reduced aggregate trust (of which he suffers like everyone else) he may also directly suffer as a result of loss of trust/reputation from his customers, other business partners, etc, due to problems that arose from the original case that can’t be directly measured.

SS: True, the category of 'incidental loss' does not in fact cover all incidental loss, though there is some latitude for courts (remember "bottles and books" getting a large payout).

RG: Incidental loss still only covers those things that can be directly attributed, which is different to the effect on broader practices/business environment

JI: I just don't think any of this stuff increases trust or anything. It's nice that you're giving an extra penalty for breaking a promise but, at the end of the day, the practical result won't be some substantive change in people's honesty or trust for each other. If, as our readings suggest, we should care about what laws and stuff actually accomplish, and how the bad man sees them, I don't see any reason for contract law to "care".

RG: Let's go extreme here - the law imposes a 1000% penalty on any deliberate breach of a contract. You don't think that will increase trust in the likelihood that contracts will be enforced? I think that would accomplish something significant. Now, perhaps you think that would be less efficient - that's a valid discussion to have (what the value of social trust is vis-à-vis flexibility afforded by allowing parties to voluntarily break contract with no penalty) but I’m hard pressed to see how you don't think it would accomplish anything to the point of contract law being ambivalent to the question.

JI: Of course contracts will be more enforced. I fail to see why that's automatically a good thing though. People won't be "trusting" each other more, they'll just know that other people won't find it efficient to breach anymore, so they won't breach. I don't think we're gaining "morality" here and we're taking a clear economic downturn. I still fail to see how this benefits society in any way.

RG: The argument is that it has a clear economic benefit in the aggregate because it is internalizing an externality, much like a carbon tax. I don't think Eben started his point with a call to morality - he talked about "social value" of trust. People may, for example, make a moral argument for progressive income taxation, but it could be equally possible to make an economic claim about the increased social value of limited inequality (i.e. the conversation being had among economists right now about the validity of Krueger’s Great Gatsby Curve). Unless I misheard Eben, this entire conversation began with him expressing puzzlement that efficient breach advocates never seem to place "any value" on social trust in the integrity of contracts themselves when talking about the goals of contract law. I could analogously say its puzzling to me that opponents of carbon taxes/cap-and-trade seem to never acknowledge the social value of high quality global commons/living conditions when talking about the efficient market use of natural resources. It's an economic argument. Your implicit assumption that there is a clear economic downturn is the problem with your point I think - hence Eben’s example of the diamond industry and it's flourishing due to enforcement of contracts.

JI: I think the diamond industry is a good example of a self-regulating efficient breach system. It's not efficient for parties to breach because #1: they'll probably be killed, and #2: it would put their whole industry in danger. They don't not breach because they trust each other and give each other huggsies, they do it because they realize that they have more to lose than to gain.

RG: Self-regulating? It's only self-regulating in the sense that they collectively determine punitive damages through informal enforcement mechanisms, much in the same way as the courts function in legitimate society. Again, I don't think Eben is trying to say everyone will become care bears - he's talking about financial trust. Taking your two points, I could just as easily say: "under an ideal system of contracts it wouldn’t be efficient for any party to intentionally breach because #1 they'll probably be punitively fined, and #2 it would put the entire economy in danger.

JI: If it puts the entire economy in danger then it's not an efficient breach. And by the way, I think what you just said highlights my problem with Eben’s use of the diamond example, which was what made me raise my hand in class. You said "much in the same way as the courts function in legitimate society". I completely agree, but Eben seemed to be using the diamond industry as a alternative example to how things could function differently than our contract law system, and my point is that the two systems are a lot more similar than he was trying to make them seem - it's just that in the diamond industry, the same rules and motivations that are found stated in contract law are more unwritten and assumed in the diamond industry. But they're not run on qualitatively different mechanisms.

RG: “If it puts the entire economy in danger then it's not an efficient breach" - that's Eben’s exact point about social trust! He says an efficient breach theory that places zero value on social trust potentially does this. The analogy to the diamond example wasn't intended to point out a different system, I think, rather that when we don't have people like Posner rationalizing a limited and narrow view of "economic efficiency" from a borderline autistically mathematical perspective, you're more likely to end up with a contract law system that functions like the diamond industry. So perhaps you're more in agreement with Eben (or at least, my interpretation of him) than you think - you wouldn't have to revolutionize contract law (or even necessarily a conception of efficient breach) theory to move our current system towards one like the diamond industry, merely an internalization of the social value of trust in the likelihood of legal contracts being actually enforced into the economic calculus.

JI: If you want to claim that the social trust value is part of the economic calculus and should be calculated as part of "efficient breach", then maybe we can talk. But to me he seemed to be against the idea of efficient breach in general, not just saying that it's missing one factor in the calculus, but maybe I'm wrong about that. I still have my doubts as to the economic value of the "trust" element you're talking about for the reasons I already explained earlier, but this does bring us closer than before.

RG: Perhaps Eben is against the idea of efficient breach in general - something to ask him after next class perhaps? - but as I said before, I would venture that a "true" efficient breach theory is merely a tautology. For example, it’s akin in my mind to saying "a law shouldn't enforce behavior that has more harmful social effects than positive ones". I doubt Eben would say, for example, that someone should be legally obligated to fulfill a contractual promise if that promise led to the guaranteed destruction of the entire economy.

JI: Yeah, probably not. Maybe the main difference is that I think you're overstating the effect of a "promise-breaking penalty" on actual trust. I think the effect would be more of people adjusting their economic calculations and still going by whether it's more or less profitable to breach, as opposed to caring more about the element of "trust" itself. Although who knows? A lot of people would say that law is where we see the values of society and it can shape those values too, so maybe if such a penalty was instituted, eventually it would infuse these kinds of values into society. But I think that's not at all a given.

RG: Think of trust as legal trust - trust in legal contracts not trust in people in general (although I agree with your idea of law shaping values, and the general practices in the diamond industry would suggest it does end up shaping industry practices). All you're really saying here is that we need to adjust our penalty to the degree to which we actually value social trust. If we decided it was absolutely valuable, we could impose a death-penalty like the diamond trade. Who would find it profitable then?

JI: Figuring out what to degree we actually value social trust is a very difficult question, one that might be impossible to answer. How are you going to figure out how much "breaking a promise" is worth? What separates contracts from, say, killing, is that you have a right to lawful protection of your life. You don't have a lawful protection to not being lied to, at least not outside of contract law. And who's to say that this is a protection we should be creating here, on our own? I just see a lot of problems in instituting this kind of penalty and I'm still not sure that the fact that fewer contracts will be broken because the penalty is higher results in more "trust in legal contracts".

RG: At least not outside of contract law - that’s the crucial point here – Posner, I think, would disagree that you even have a right to not be lied to "within" contract law, wouldn’t he? I.e. "Dear Joe, I promise to buy forty stocks of wood from you" (while in reality, I have no intention of buying them, and simply want to preserve the possibility of doing so if it happens to become economically efficient in the future, and am willing to risk paying the likely penalty for that lie). If fewer contracts will be broken, doesn't that by definition increase the level of trust people will have in legal contracts? I trust that a contract that I make with you will actually be honored more when the economic probability that it will be is 99.8% than when it is 50%, no?

JI: Well, yeah, my point is that if we granted some right of "protection from being lied to" it would be entirely through the law of contract itself. So you can't say we need that protection because it's mandated somewhere else - if we're creating it, it's important to realize that we're doing it completely on our own, not because we already have to. And I think at least when Eben uses the word "trust" he means something more than a pure economic calculation. When the probability shifts upwards, your "trust" is still in an economic formula and in human nature, and that "trust" does not increase, it stays the same, just the specific numbers change. I don't think that kind of "trust" is particularly valuable. I would say that if we could make people trust each other more (for non-economic reasons), that would be an accomplishment worthy of changing contract law for. The kind of "trust" you're talking about though seems pretty empty and I still don't think it's all that valuable on its own.

RG: Did Eben say we had a mandate to create trust beyond the practical effects such trust would have on the functioning of society (and the economy)? Our social value of trust is a decision we have to make collectively, much as our value in "creating a more perfect union" vis-à-vis protecting individual liberty in the state of nature. Aside from your own point that such law can in fact generate value in the type of real trust you're talking about (when people see how the system works, value the effects and want it to continue, etc), I don't know how you can claim that there is no value to such an economic calculation when we can input our collective value of trust into that calculation. It could ultimately increase in importance via a positive feedback loop of increasing the value in the calculation, if the initial choice was found to be quite valuable. The reason I keep saying a perfect economic calculation is a tautology is because all decisions are fundamentally "economic" in the sense that we assess all the things that we value and desire, and then work out whether an action will promote more of those values and desires than it detracts from. The attempt to internalize externalities is simply an attempt to be more clear and complete in articulating that internal value calculation.

-- RohanGrey - 25 Jan 2012

This pretty effectively demonstrates why recorded class audio and transcripts aren't always productive in this context. Did we get 5,000 words worth of insight out of these 4,400 words?

One of you, or someone else, should now "refactor" this page, editing it to take out the repetition and the side notes, distilling out clearly the points of view and their differences, setting forth what we learned from the conversation, and where we might go next. That should require less than 900 words, or somewhere in the neighborhood of an 80% reduction in bulk with a concomitant increase in clarity and usability.

Given we had the opportunity to listen in class to your original point in a highly condensed version and it clearly left ambiguity of meaning to some, and the fact that we did admittedly go around in circles multiple times before finally reaching some level of consensus and progress, I don't necessarily agree that there is no value to be had ever in providing multiple iterations of the same concept/articulating the process of conceptualizing in its original form by actual learners in class. It is possible to be receptive to one formulation of an idea and not another, and moreover jarring or overly aggressive phrasing, an overly distracting example, or an unrealistic assumption of conceptual familiarity can result in even the most condensed and direct prose being less persuasive than something more suited to the style in which the listener is used to communicating. That said, i will happily do a "refactored" version tomorrow morning before class, as time does not permit me to do it tonight. Any guidance on your own views to provide some conclusion on what your actual meaning was at the end would be appreciated.

-- RohanGrey - 25 Jan 2012

My take:

1. What should the function of contract law be?

Contract law should empower judges to adjudicate disputes emerging out of certain forms of commitments between private actors. Remedies available to judges for contractual breaches include a court order for specific performance, as well as an order awarding damages to the party suffering the breach.

2. How should courts determine which outcome to prescribe?

The law should not place any intrinsic value in one remedy over another, but rather should compare the specific and general effects of each and adopt the one deemed most socially optimal given the totality of the circumstances.

The optimal remedy may be defined as the action that results in the largest increase in economic value for society from the same starting conditions. However, attempts by courts to accurately measure economic value are problematic, since existing pricing mechanisms assign value imperfectly and reflect path-dependent effects of previous such attempts to quantify social value. Consequently, an economic calculus solely based on assessment of real and expected monetary damages will tend to reinforce rather than rectify existing market imperfections and pricing failures.

The difficulty of conducting genuine predictive economic analysis can be seen in discussions of so-called “efficient-breach” theory, which advocates intentional breach of contract in situations where the breaching party expects to profit more from breaching and paying compensatory damages than it would from honoring the contract. This approach assumes the economic cost of a breach is identical to the liabilities that can be legally imposed on the breaching party. Hence, this theory only holds if the purported method for identifying, measuring and allocating all costs of breach to the breaching party is as perfect as advertised.

3. How, if at all, should the law value the act of intentional breach in calculating the costs and benefits of ordering specific performance versus awarding damages for non-performance?

If the courts wholly adopted an efficient breach theory that assigned nominal or zero value to specific performance of contractual obligations, then the economy would ideally adjust to eventually place little or no reliance on actual performance in instances where the risk of efficient breach is high. Instead, individuals and firms would attempt to minimize the cost of resulting uncertainty by taking out insurance, developing more comprehensive contingency clauses, and/or assigning a lower value to the contract in the first place. While impossible to predict, it is possible that the costs imposed by such ameliorative measures, as well as the more general uncertainty regarding contractual performance, would end up being higher than the benefits gained from promoting intentional breach when it is deemed profitable to one party.

On the other hand, courts could choose to assign a non-trivial value to specific performance over non-performance with compensation. There are three ways this could be done. Firstly, courts may attempt to calculate the expected marginal cost of greater contractual uncertainty represented by each individual act of intentional breach. Alternatively, courts may decide that such calculation is impossible, and deem the optimal solution to be an approximate or even arbitrarily determined Pigovian tax, in order to disincentivize all intentional breaches. Thirdly, courts may decide to presumptively favor specific performance and forced injunctions, and only carve out specific or general case exceptions when performance is unfeasible or unreasonable. Regardless of which particular approach was adopted, each one would result in a higher degree of certainty in contractual performance than in the same context under an efficient breach theory. However, the benefits gained from greater certainty in judicial enforcement of contractual obligations may be offset by excessive restrictions on the ability of contracting parties to opt-out when such an option has vastly greater benefits to almost all parties involved (including society), but unacceptable costs to the party that would directly suffer the breach.

Ultimately, the decision of which of these approaches to adopt depends on the potential risks and benefits the court chooses to acknowledge in their value calculus, and the particular values they assign to each. Perhaps there is an intrinsic tradeoff between the benefits of market efficiency and higher certainty in contractual enforcement. Perhaps the tradeoff is illusory; a mere failure by efficient breach theorists to fully internalize the external costs associated with encouraging breaches deemed profitable by the breaching party. Regardless of whether the answers can be conclusively determined, it is critical to address these issues when determining the broader effect of a system of permissible intentional breaches, rather than assuming them away at the outset of any economic analysis of contractual disputes.

-- RohanGrey - 1 Feb 2012