Freedom of Contract and Contract Costs

by Tim Lee on November 3, 2008 · Comments

I made the point last week that freedom of contract includes the right not to be party to contracts without your consent, and that consent has to involve some sort of affirmative action. Shrink-wrap contracts that are presented only after a sale is complete don’t cut it in my view. If the contract wasn’t available for review at the point of sale, then I don’t care what might be in the box. You didn’t consent to the contract.

Now, some people might (and in the comments to previous posts, did) claim that this is just nitpicking, and that the really important thing is to promote economic efficiency by making contract formation easier. On this theory, contract negotiations are a transaction cost, and it’s economically beneficial to lower transaction costs as much as possible. So even if you haven’t technically agreed to the shrink-wrap EULA before you leave Best Buy, some people might argue that you “should have known” there would be a EULA in the box, and therefore it’s economically efficient to bind you to the contract unless you return the product to the store.

The problem with this argument is that it focuses myopically on the costs of contract negotiation to the exclusion of other costs that in many cases are much more important. It should be remembered that every contract signed is a prelude to possible state coercion if the contract is broken. Like all other kinds of coercion, the possibility of contract-related litigation creates uncertainty and other deadweight costs. In addition, the act of offering contracts imposes a deadweight cost. Every time I’m presented with a contract, I have to at least skim through it to make sure that the terms are acceptable. A society in which contract formation is extremely cheap for one party will be a society in which other people have to spend a lot of time scrutinizing the contracts they offer. Finally, contracts impose costs on the court system. A legal system that makes contracts to cheap to create will lead to too much taxpayer money being wasted on contract litigation.

All of which is to say that economic efficiency is not promoted by making contract formation as cheap as possible. Rather, the goal should be to align incentives so that a party only offers a contract if its benefits to all parties outweigh its expected costs. One thing that we should particularly try to avoid is a situation in which offering contracts is almost costless to one party, but reviewing them is expensive for the other party. Because then the offering party will offer inefficiently many contracts favorable to itself, and its counterparties will accept inefficiently many contracts because the costs of scrutinizing them individually is too high.

In contrast, if things are structured so that each party bears roughly half the costs of contract negotiation, then each party is only going to propose a formal, written contract if he believes that the benefits of doing so will outweigh the costs to both parties. This is one of the good things about paper contract negotiations between flesh-and-blood people: If you give me a long contract to sign, you’re going to have to stand there and wait while I read the contract and decide if I want to sign it. Since standing around is a waste of your time, you’re only going to do that if you believe the transaction can’t happen without it. And you’re going to try to make the contract as short as possible so you don’t have to stand around too long.

In the vast majority of business transactions, the default UCC terms work just fine. Grocery stores don’t try to attach contracts to the items they sell because it would slow down the checkout line too much if customers had to stand around reading their Banana Licensing Agreements before they were allowed to take their groceries home.

The reason shrinkwrap licenses on software are more popular than supermarket checkout contracts is not because software sales are some kind of exotic financial transaction that require special contractual terms. The UCC defaults can and do work just fine for software. Rather, the reason software firms make extensive use of EULAs is because they’ve found a clever gimmick that allows them to use copyright law as a way to bypass the ordinary rules of contract law and foist almost all the costs of contract negotiation onto the customer.

If a grocery store checkout clerk slipped a Banana Licensing Agreement into your grocery bag, no court of law would regard that as an enforceable contract. But software vendors have been pushing the legal fiction that software is licensed rather than sold. And if software is licensed rather than sold, then using software without accepting the license agreement is copyright infringement, which operates under an entirely different set of rules than ordinary contract law.

This claim is pretty clearly contrary to copyright’s First Sale Doctrine, and some courts have explicitly rejected it in some cases, but other courts have upheld it, and software vendors find the ability to skirt the ordinary rules of contract law so convenient that they keep trying.

But as a policy matter, there isn’t any good reason to let them get away with it. It would be bad policy to allow grocery stores to attach Banana License Agreements to their customers’ banana purchases by putting a BLA in each grocery bag, and it’s equally bad policy to allow software vendors to bind their customers to contracts they’re not able to review until after a sale is made. If software vendors want the software they sell to come with particular contractual restrictions, they should have the clerk at Best Buy provide the customer with a copy of the license agreement and require her to sign it before she can leave the store. If software vendors aren’t willing to put their customers through that hassle (and I’d bet money that they’re not) then it’s obviously not that important to them for their products to come with contractual restrictions attached, in which case the right outcome is for the software to be sold without special contractual restrictions, the same way that movies, music, books, and other creative works do.

Comments Posted in: Uncategorized

  • WilsonF
    Tim, I have very much enjoyed this series of posts from both you and Alex. . .I'm not an expert on these matters, but I find this question really interesting. I basically still don't see the efficiency problem here.

    "Rather, the goal should be to align incentives so that a party only offers a contract if its benefits to all parties outweigh its expected costs."

    I think it's easy to imagine a world where this assumption leads to the conclusion that contracts should be this easy to form. The 'half-the-costs' analysis fails because the items at issue are very small probability, so to any individual consumer the benefits for negotiating on them are vanishingly small, but to a company that sells a million units, the benefits are large.

    The terms do not apply to the vast majority of consumers; either they are uninterested because the terms relate to unlikely situations (forum selection clauses, like in Carnival Cruise) or they relate to conduct the user has no intention of engaging in (such as commercial exploitation, as in Pro-CD). If Pro-CD is not able to prevent Zeidenberg from using the software in a commercial fashion, he can do a lot of damage to their business. They'd just rather not sell it to consumers at all than be unable to impose this contract on Zeidenberg.

    If the alternative is forcing Pro-CD to paste the necessarily complex agreement on the outside of the box, or to make people read it at the door of Best Buy, no one is going to buy their product. However, in a world with shrinkwrap contracts people can happily buy the Pro-CD software, and take it home and use it and never risk violating the license agreement.

    I think (but don't know, an important distinction) that the reason other products aren't sold with special contractual attachments is that the potential benefits are rather small, and the market is keeping them in check rather than a theory of agreement. Maybe I'm wrong, maybe the court has held them unenforceable, but I would think that between Carnival Cruise and Pro-CD if there was something there they'd exist. I could be totally wrong about this stuff, but those are my thoughts.

    As you know, I love your stuff. Keep it coming.

    http://people.hofstra.edu/peter_j_spiro/cyberla...
  • Clearly, what you economist types (and I include Tim in that statement) will think of the contracts in question depends on how good you think their content is. If the contracts all said "...and, further, the customer agrees to paint Bill Gates' house and wash Paul Allen's cat and the customer is deemed to accept this agreement by breathing for thirty seconds after having received these terms," I doubt you would be such a fan. Tim's point is that, in principle, such obviously inefficiently one-sided terms can sneak in when one side can make a "contract" on behalf of both parties.

    I'm a deontologist. I don't care what the terms say. It's just plain unjust to bind me to something I never agreed to.
  • Hey Wilson,

    Thanks for the thoughtful reply. I think this is where we disagree:

    If Pro-CD is not able to prevent Zeidenberg from using the software in a commercial fashion, he can do a lot of damage to their business. They'd just rather not sell it to consumers at all than be unable to impose this contract on Zeidenberg.

    The Pro-CD case is a great example, and I'm glad you brought it up. However, I don't think we ought to start with the assumption that ProCD needs to be able to employ a particular business model and then reason backwards to the legal system required to support that business model. I think it's way too pessimistic to assume that in the absence of this particular enforcement mechanism, that they'd be unable to engage in price discrimination and therefore unable to produce their product at all.

    More to the point, if you think that the kind of information ProCD produced is worthy of copyright-like protection, the right way to deal with that is to amend copyright to include some kind of right database protection. Congress and the courts have specifically considered and rejected the idea that this kind of factual information should be covered by copyright law. It strikes me as perverse to use a contract-copyright hybrid to allow copyright holders to restrict the use of their works in ways that neither contract law nor copyright law, by themselves, permit.

    Now, you want to know what the big deal is. A key point is that the fact that some activity is only performed by a small fraction of customers doesn't necessarily mean that it's not important. As a computer geek, my personal hobby horse is reverse engineering. Software EULAs routinely have clauses forbidding reverse engineering, which makes it very hard for engineers to examine software, learn how it works, and produce interoperable software or hardware devices.

    Why does that matter? Well, consider Open Office. It exists because some determined free software developers reverse engineered the Microsoft Office file formats and implemented competitive versions. Open Office produces immense social benefits. Its users obviously benefit from it, but users of Microsoft Office also benefit, because it gives Microsoft some competition and saves customers money. So all things being equal, it's better to have a policy that makes projects like Open Office easier to create.

    So lots of people benefit from Open Office and therefore rely indirectly on the benefits of reverse engineering, but only a tiny fraction of those people actually want to engage in reverse engineering themselves. So it's to everyone's benefit to have contract that don't make it too easy to restrict reverse engineering, even though only a few people will actually take advantage of that privilege.

    To return to my earlier point, copyright law protects reverse engineering as fair use for precisely the reasons I described above. So software developers' efforts to write "no reverse engineering" rules into their EULAs amounts to an effort to expand their monopolies beyond the bounds permitted by ordinary copyright law. If Congress decides that reverse engineering isn't beneficial after all, it can amend copyright law to say that reverse engineering is fair use. But as long as copyright law permits reverse engineering, I think contract law should look very skeptically at novel interpretations of contract law that effectively nullify the traditional limits on the scope of copyright protection.
  • WilsonF
    Tim;

    Thanks for giving me a good take at what benefits can come from unlicensed software. It's something for me to chew on.

    I still think it's possible that this type of offer and acceptance can be efficiency increasing. I wonder then if the small number of users who want to reverse engineer the product can't simply get together and sign separate agreements with the software companies? Clearly still we have a situation where the company derives a benefit from making a contract with all users that only a small number of users would have rejected. It's just a less obviously efficiency-enhancing provision. This reverse engineering issue is really interesting.

    It's hard for my to deny the enormous benefits of open office. I will ponder...
  • The problem is that the no-reverse-engineering clause is specifically targeted at the small minority of users who want to reverse engineer. The firm isn't going to sign separate agreements with those people because they don't want people reverse engineering their software and producing compatible versions that could compete with their own. See, for example, the BNetD case.
  • WilsonF
    To clarify, I would say that it's OK to impose a reviewing cost in this situation since that cost is likely to be very very low--people won't review because they have no incentive to, and the benefits that construe to the software company are large.
  • Congress and the courts have specifically considered and rejected the idea that this kind of factual information should be covered by copyright law.
  • Adaku praise
    How can we discuss the concept of freedom of contract
blog comments powered by Disqus

Previous post:

Next post: