Ayres y Gertner para Traducir

Filling Gaps in Incomplete Contracts: An Economic Theory of Default Rules
Author(s): Ian Ayres and Robert Gertner

Source: The Yale Law Journal , Oct., 1989, Vol. 99, No. 1 (Oct., 1989), pp. 87-130
Published by: The Yale Law Journal Company, Inc.
Stable URL: https://www.jstor.org/stable/796722
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
is collaborating with JSTOR to digitize, preserve and extend access to The Yale Law Journal
This content downloaded from

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
All use subject to https://about.jstor.org/terms
Filling Gaps in Incomplete Contracts: An
Economic Theory of Default Rules
Ian Ayrest and Robert Gertnertt
INTRODUCTION
The legal rules of contracts and corporations can be divided into two
distinct classes. The larger class consists of "default" rules that parties can
contract around by prior agreement, while the smaller, but important,
class consists of "immutable" rules that parties cannot change by contrac-
tual agreement.' Default rules fill the gaps in incomplete contracts; they
govern unless the parties contract around them. Immutable rules cannot
be contracted around; they govern even if the parties attempt to contract
around them. For example, under the Uniform Commercial Code
(U.C.C.) the duty to act in good faith is an immutable part of any con-
tract,2 while the warranty of merchantability is simply a default rule tha
parties can waive by agreement.3 Similarly, most corporate statutes
t Assistant Professor, Northwestern University School of Law; Research Fellow, American Bar
Foundation. B.A., Yale University; J.D., Yale Law School; Ph.D. (Economics), Massachusetts Insti-
tute of Technology. The support of Northwestern's Corporate Counsel Center is gratefully
acknowledged.
fl Assistant Professor, University of Chicago, Graduate School of Business. A.B., Princeton Uni-
versity; Ph.D. (Economics), Massachusetts Institute of Technology. The authors would like to thank
Douglas Baird, Randy Barnett, Richard Craswell, John Donohue, Frank Easterbrook, Mark Grady,
David Haddock, Steve Harris, Oliver Hart, Ian MacNeil, Joel Rogers, Steve Salop, David
Scharfstein, David Van Zandt, Robert Vishny and seminar participants at the University of Chicago,
Georgetown, Michigan, Northwestern, and Illinois law schools, and at the U.S. Department of
Justice for helpful comments. Rebecca Mitchells and Jerry Richman provided excellent research
assistance.
1. See B. Black, Corporate Law As Neutral Mutation, (Nov. 1988) (unpublished manuscript on
file with authors) (arguing that few corporate laws are immutable). Immutable rules are similar to
what Calabresi and Melamed call "inalienable" rules, Calabresi & Melamed, Property Rules, Liabil-
ity Rules, and Inalienability: One View of the Cathedral, 85 HARV. L. REV. 1089, 1093 (1972),
except that immutable entitlements are created by and conditioned upon contract, while inalienable
entitlements exist outside of contract. See Schwab, A Coasean Experiment on Contract Presumptions,
17 J. LEGAL STUD. 237, 239 n.6 (1988) (distinguishing between inalienable and immutable rules).
2. U.C.C. ? 1-203 (1976); see Morin Bldg. Prod. Co. v. Baystone Constr., Inc., 717 F.2d 413,
414-15 (7th Cir. 1983); RESTATEMENT (SECOND) OF CONTRACTS ? 205 (1979); R. POSNER, Eco-
NOMIC ANALYSIS OF LAW 81 (3d ed. 1986).
3. U.C.C. ? 2-314 (1976). U.C.C. ? 1-102 (1976) distinguishes between default and immutable
rules and states its preference for the former:
(3) The effect of provisions of this Act may be varied by agreement, except as otherwise pro-
vided in this Act and except that the obligations of good faith, diligence, reasonableness and
care prescribed by this Act may not be disclaimed by agreement but the parties may by agree-
ment determine the standards by which the performance of such obligations is to be measured
if such standards are not manifestly unreasonable.
(4) The presence in certain provisions of this Act of the words "unless otherwise agreed" or
87

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
88 The Yale Law Journal [Vol. 99: 87
require that stockholders elect directors ann

incorporation to contract around the defau
tory language such as "[u]nless otherwise provided in the certificate of
incorporation"' or unlesses otherwise unambiguously indicated"7 makes
it easy to identify statutory default, but common-law precedents can also
be divided into the default and immutable camps. For example, the com-
mon-law holding of Peevyhouse v. Garland Coal & Mining Co.,8 which
limited damages to diminution in value, could be contractually reversed by
prospective parties.9 In contrast, the common law prerequisite of consider-
ation is largely an immutable rule that parties cannot contractually
abrogates
There is surprising consensus among academics at an abstract level on
two normative bases for immutability. Put most simply, immutable rules
are justifiable if society wants to protect (1) parties within the contract, o
(2) parties outside the contract.11 The former justification turns on
parentalism; the latter on externalities. Immutable rules displace freedom
of contract. Immutability is justified only if unregulated contracting woul
be socially deleterious because parties internal or external to the contract
cannot adequately protect themselves.12 With regard to immutable rules,
the disagreement among academics is not over this abstract theory, but
words of similar import does not imply that the effect of other provisions may not be varied by
agreement under subsection (3).
U.C.C. ? 1-102 (1976).
4. See, e.g., DEL. CODE ANN. tit. 8, ? 211(c) (1974).
5. See, e.g., DEL. CODE ANN. tit. 8, ? 214 (1974).
6. DEL. CODE ANN. tit. 8, ? 223 (1974).
7. U.C.C. ? 2-206 (1976).
8. 382 P.2d 109 (Okla. 1962), cert. denied, 375 U.S. 906 (1963).
9. Whether the Peevyhouse majority actually intended for prospective parties to be able to
the "cost of performance measure" is discussed more fully infra text accompanying notes 15
10. The consideration requirement is not immutable if written agreements "under seal" serv
contractual substitute for consideration. See U.C.C. ? 2-203 (1976) (making inoperative "the law
respect to sealed instruments").
This default rule/immutable rule dichotomy also pervades other areas of the law that have con
tual components. In the law of divorce, for example, wealth accrued before marriage is alloc
according to default rules that can be altered in pre-nuptial agreements, while income earned
marriage is immutably divided. Similarly, the repayment priorities set by state debtor-credit
can, like default rules, be reordered through private contract. The laws of intestacy are also d
rules: they fill any testamentary gap, but can be contracted around. As discussed below, infr
accompanying notes 83-91, one can distinguish between defaults that must be bilaterally cont
around and defaults that may be unilaterally overcome.
11. See I. MACNEIL, CONTRACTS: EXCHANGE TRANSACTIONS AND RELATIONS 346-47 (
Calabresi & Melamed, supra note 1; Easterbrook & Fischel, The Economic Structure of Cor
Law, 89 COLUM. L. REV. (forthcoming 1989) (manuscript at 21-30; on file with authors).
12. Note that even when there are negative externalities, third parties may be able to pro
themselves without immutable rules. One implication of the Coase theorem is that in a world wit
transaction costs, third parties will have an incentive to contract to reduce externalities to an eff
level. See Coase, The Problem of Social Cost, 3 J. LAW & ECON. 1 (1960). There are no extern
if the class of parties to the potential contract is defined broadly enough. G. Priest, Internalizing
(Yale Law School Program in Civil Liability 1988; working paper no. 93) (explicating pervasiv
of private incentives to internalize costs).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
1989] Default Rules 89
whether in particular contexts parentalistic concerns or externalities are

sufficiently great to justify the use of immutable rules."3
When the preconditions for immutability are not present, the normative
legal analysis devolves to the choice of a default rule. Yet academics have
paid little attention about how to choose among possible default rules."4
The law-and-economics movement has fought long and hard to convince
courts to restrict the use of immutable rules,"5 but has lost most of its
normative energy in constructing a theory of default choice.16 Economists
seem to believe that, even if lawmakers choose the wrong default, at worst
there will be increased transaction costs of a second order of magnitude."
Few academics have gone beyond one-sentence theories stipulating that
default terms should be set at what the parties would have wanted.18
13. A recent conference (Dec. 9-10, 1988) on "Contractual Freedom in Corporate Law" at Co-
lumbia's Center for Law and Economic Studies focused directly on the appropriate application of
immutable rules. For example, Jeffrey Gordon argued that having multiple precedents that construe
single legal standard produces positive externalities that might justify imposing an immutable rule. Se
J. Gordon, The Mandatory Structure of Corporate Law (Dec. 2, 1988) (unpublished manuscript on
file with authors).
14. Courts or legislatures are inevitably forced to set defaults, because contracts with gaps need t
be interpreted. Courts must do something-even if that something is non-enforcement. As discusse
infra text accompanying notes 51-52, defaults of non-enforcement can play an important role in
efficient law.
15. For instance, Anthony Kronman has written:
[E]x ante arguments for the efficiency of a particular legal rule assume that individuals remain
free to contract around that rule, and a legal system that denies private parties the right to
vary rules in this way will tend to be less efficient than a system that adopts the same rules but
permits contractual variation.
Kronman, Specific Performance, 45 U. CHI. L. REV. 351, 370 (1978). See Haddock, Macey &
McChesney, Property Rights in Assets and Resistance to Tender Offers, 73 VA. L. REV. 701, 736
(1987) ("The ability of firms to contract around costly legal rules when lower-cost private alternatives
are available must be a feature of any efficient standard-form contract."); see also Goetz & Scott,
Liquidated Damages, Penalties and the Just Compensation Principle: Some Notes on an Enforce-
ment Model and a Theory of Efficient Breach, 77 COLUM. L. REV. 554 (1977) (arguing that "immu-
table" standards for determining enforceability of liquidated damages clauses should be relaxed). But
see Clarkson, Miller & Muris, Liquidated Damages v. Penalties: Sense or Nonsense?, 1978 Wis. L.
REV. 351 (providing efficiency rationale for immutable liquidated-damages rules).
16. For example, Haddock and Macey have suggested that immutable rules against insider trad-
ing are inefficient but have remained agnostic about whether corporations wishing to allow their
insiders to trade should be forced to "opt out" of an insider trading prohibition, or whether corpora-
tions wishing to prohibit insider trading should be forced to "opt in" to such a system. Haddock &
Macey, A Coasian Model of Insider Trading, 80 Nw. U.L. REV. 1449 (1987).
17. See, e.g., Easterbrook & Fischel, Limited Liability and the Corporation, 52 U. CHI. L. REV.
89, 102 (1985) ("In light of the ability of firms to duplicate or at least approximate either limited or
unlimited liability by contract, does the legal rule of limited liability matter? The answer is yes, but
probably not much.").
18. Looking backward to what the present litigants "would have wanted" is analytically analo-
gous to looking forward to what prospective contractors will want. It is to ask (as Lea Brilmayer often
does) "who are the prospective parties rooting for?" In both cases the court examines ex ante incen-
tives. While ex post each party will have economic incentives to shift costs to the other side, ex ante
the parties have an incentive to place the risks on the least-cost avoider. Kronman, Mistake, Disclo-
sure, Information and the Law of Contracts, 7 J. LEGAL STUD. 1 (1978). If a court can identify that
ex ante the parties to the contract had identical interests in allocating a certain risk or duty of per-
formance, then it can, in a sense, pierce the ex post adversarial veil. Thus, for example, even if ex post
a particular tenant wants to avoid the risk of fire damage, ex ante both landlords and tenants may
have preferred to have tenants bear this risk as the least-cost avoider. Thus, the fact that after a fire a
tenant tries to avoid liability is not dispositive of what prospective tenants would contract for or,

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
Frank Easterbrook and Daniel Fischel have championed the "would have
wanted" theory in a number of articles suggesting that "corporate law
should contain the [defaults] people would have negotiated, were the costs
of negotiating at arms'-length for every contingency sufficiently low.""9
Similarly, Richard Posner has argued that default rules should "econo-
mize on transaction costs by supplying standard contract terms that the
parties would otherwise have to adopt by express agreement." Douglas
Baird and Thomas Jackson have argued that the default rules governing
the debtor-creditor relationship "should provide all the parties with the
type of contract that they would have agreed to if they had had the time
and money to bargain over all aspects of their deal."'21 While this litera-
ture has vigorously examined what particular parties would have con-
tracted for in particular contractual settings,22 it has failed to question
whether the "would have wanted" standard is conceptually sound.23
Thus, although the academy recognizes the analytic difference between
indeed, of what this particular tenant would have contracted for.

19. Easterbrook & Fischel, supra note 11 at 14-15; see also id. at 20-21 (default term should be
"the term that the parties would have selected with full information and costless contracting"); Easter-
brook & Fischel, Corporate Control Transactions, 91 YALE L.J. 698, 702 (1982) (default fiduciary
duties are derived from a hypothetical contract, imagined by judges, between investors and managers
dickering with each other free of bargaining costs); Easterbrook & Fischel, The Proper Role of a
Target's Management in Responding to a Tender Offer, 94 HARV. L. REV. 1161, 1182 (1981)
(corporate law should supply "standard form 'contracts' of the sort shareholders would be likely to
choose ...."). Calabresi and Melamed's analysis may be an early antecedent of the "would have
wanted" analysis. They argue that efficiency-minded law would establish default entitlements as the
parties would allocate them in a world without transaction costs. Calabresi & Melamed, supra note 1,
at 1093-98.
20. R. POSNER, supra note 2, at 372; see also Schwartz, Proposals for Products Liability R
form: A Theoretical Synthesis, 97 YALE L.J. 353, 361 (1988) (offering as default rule "the contra
that most well-informed persons would have adopted if they were to bargain about the matter").
21. Baird & Jackson, Fraudulent Conveyance Law and Its Proper Domain, 38 VAND. L. REV.
829, 835-36 (1985); see also Goetz & Scott, The Mitigation Principle: Toward a General Theory of
Contractual Obligation, 69 VA. L. REV. 967, 971 (1983) ("Ideally, the preformulated rules supplied
by the state should mimic the agreements contracting parties would reach were they costlessly to
bargain out each detail of the transaction.").
In Lewis v. Benedict Coal Corp., 361 U.S. 459 (1960), the Supreme Court applied this default-
setting standard in deciding whether, in the absence of explicit contractual language, payments to a
third-party beneficiary pension fund should be subject to setoff if the union breaches the underlying
labor contract. The Court stated, "[i]t may be fair to assume that had the parties anticipated the
possibility of a breach by the promisee they would have provided that the promisor might protect
himself by such means as would be available against the promisee under a two-party contract." Id. at
468 (emphasis added). The Court later distinguished what parties to a collective bargaining agree-
ment might have wanted and accordingly established a no-setoff default. Id. at 469.
22. See, e.g., Craswell, Contract Remedies, Renegotiation, and the Theory of Efficient Breach, 61
S. CAL. L. REV. 629 (1988); see also infra text accompanying notes 136-39 (discussing judicial disa-
greement of Judges Easterbrook and Posner).
23. Charles Goetz and Robert Scott have written one of the more thoughtful examinations of
default choice. Goetz & Scott, The Limits of Expanded Choice: An Analysis of the Interactions Be-
tween Express and Implied Contract Terms, 73 CALIF. L. REV. 261 (1985). Their theory of default
rules is based on the idea that a large body of precedent will develop regarding the interpretation and
application of a standard-form clause. Parties who contract around a standard-form clause will face
the prospect that courts will interpret their contract in a manner that is inconsistent with the parties'
initial intentions. Thus, parties who prefer an alternative to the standard-form may accept the latter
for fear of misinterpretation.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
default and immutable rules, a detailed theory of how defaults should be

set has yet to be proposed. Indeed, the lack of agreement over even what
to call the "default" concept is evidence of the underdeveloped state of
default theory.24 Default rules have alternatively been termed background,
backstop, enabling, fallback, gap-filling, off-the-rack, opt-in, opt-out,
preformulated, preset, presumptive, standby, standard-form and supple-
tory rules.25
This Article provides a theory of how courts and legislatures should set
default rules. We suggest that efficient defaults would take a variety of
forms that at times would diverge from the "what the parties would have
contracted for" principle. To this end, we introduce the concept of "pen-
alty defaults." Penalty defaults are designed to give at least one party to
the contract an incentive to contract around the default rule and therefore
to choose affirmatively the contract provision they prefer. In contrast to
the received wisdom, penalty defaults are purposefully set at what the
parties would not want-in order to encourage the parties to reveal infor-
mation to each other or to third parties (especially the courts).
This Article also distinguishes between tailored and untailored defaults.
A "tailored default" attempts to provide a contract's parties with precisely
"what they would have contracted for." An "untailored default," true to
its etymology, provides the parties to all contracts with a single, off-the-
rack standard that in some sense represents what the majority of con-
tracting parties would want. The Restatement (Second) of Contracts' ap-
proach to filling gaps, for example, provides tailored defaults that are
"reasonable in the circumstances."" "Reasonable" defaults usually entail
a tailored determination of what the individual contracting parties would
have wanted because courts evaluate reasonableness in relation to the "cir-
24. The "default" characterization seems currently in vogue. Professor Robert Clark explains its
etymology:
For those who haven't been exposed to this jargon from the world of computers, "default
rules" are the rules that a program follows in "default" of an explicit choice by the user to
have some other principle apply. For example, your word processing program may set paper
margins of 1 inch on all sides unless you take the trouble to learn the relevant commands and
set the margins otherwise.
Clark, Contracts, Elites, and Traditions in the Making of Corporate Law, 89 COLUM. L. REV.
(forthcoming 1989) (manuscript at 3 n.9; on file with authors).
25. See Bebchuk, Limiting Contractual Freedom in Corporate Law: The Desirable Constraints
on Charter Amendments, 102 HARV. L. REV. 1820 (1989) (using "opt-out" and "opt-in"); Coffee,
The Mandatory/ Enabling Balance in Corporate Law: An Essay on the Judicial Role, 89 COLUM. L.
REV. (forthcoming 1989) (using "enabling"); Easterbrook & Fischel, supra note 11 (using "standby,"
"enabling," "presets," and "fallback"); Eisenberg, The Foundational Model of the Corporation, 89
COLUM. L. REV. (forthcoming 1989) (using "enabling" and "suppletory" terms); Goetz & Scott,
supra note 21, at 971 (using "preformulated"); Haddock, Macey & McChesney, supra note 15, at
736 (using "default" and "standard-form"); Schwab, supra note 1, at 237 (using "presumptive" and
"off-the-rack"); Speidel, Restatement Second: Omitted Terms and Contract Method, 67 CORNELL L.
REV. 785 n.2 (1982) (using "gap-filling").
26. RESTATEMENT (SECOND) OF CONTRACTS ? 204 ("Supplying an Omitted Essential Term")
(setting default for missing term to be "a term which is reasonable in the circumstances"). See Speidel,
supra note 25, at 792-809.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
cumstances" of the individual contracting parties.27 In contrast, Charles

Goetz and Robert Scott have proposed that courts should set untailored
default rules by asking "what arrangements would most bargainers
prefer?"28
This Article provides a general theory of when efficiency-minded courts
or legislatures should set penalty defaults and how they should choose
between tailored and untailored default rules. Some common-law and
statutory defaults are flatly at odds with the "would have wanted" princi-
ple. Although this Article does not make the full-blown positivist claim
that current default rules are efficient, it does offer a more complete ex-
planation of the current diversity of defaults.
An essential component of our theory of default rules is our explicit
consideration of the sources of contractual incompleteness. We distinguish
between two basic reasons for incompleteness.29 Scholars have primarily
attributed incompleteness to the costs of contracting. Contracts may be in-
complete because the transaction costs of explicitly contracting for a given
contingency are greater than the benefits.30 These transaction costs may
27. For example, U.C.C. ? 2-306, governing output and requirement contracts, establishes as a
default that: In the absence of a stated estimate, "no quantity unreasonably disproportionate . . . to
any normal or otherwise comparable prior output or requirements may be tendered or demanded."
U.C.C. ? 2-306(1) (1976). In determining reasonableness, courts are expressly asked by the U.C.C. to
look at specific characteristics of the contracting parties. For example, "[a] shut-down by a require-
ments buyer for lack of orders might be permissible when a shut-down merely to curtail losses would
not." U.C.C. ? 2-306(1) comment 2 (1976).
28. Goetz & Scott, supra note 21, at 971 (emphasis in original).
29. There are two distinct ways for a contract to be incomplete. First, a contract may fail to
specify the parties' duties for specific future contingencies. For example, a contract for the construc-
tion of a third floor to a house may not state the parties' respective rights and responsibilities should
the entire house burn down before construction is started. Since construction of a third floor is impos-
sible (without the lower two floors), the contract does not cover the contingency of the house burning
down.
The second form of contractual incompleteness is more subtle. A contract may also be incomplete in
that it is insensitive to relevant future contingencies. Under this second form of contractual incom-
pleteness, parties' duties are fully specified, but the contracts are incomplete because those specified
duties are not tailored to economically relevant future events. See K. Spier, Incomplete Contracts in a
Model with Adverse Selection and Exogenous Costs of Enforcement (Dec. 1988) (unpublished manu-
script on file with authors) (discussing causes for such incompleteness). For example, consider a con-
tract that simply obligates one party to construct a garage adjacent to a house. On the face this
contract imposes a duty to build a garage whether or not the adjacent house burns down before
construction of the garage is complete. The contract is incomplete in this second sense, however, be-
cause the duty to build a garage is not sufficiently dependent on future contingencies. If the adjacent
house burns down, the parties probably would want to adjust the terms of contract. Such contracts we
call insufficiently state-contingent.
Courts recognize the first form of incompleteness and know they must decide how to fill the gap.
For instance, non-enforcement is one way courts can fill the gap. Courts seldomly recognize the sec-
ond form of contractual incompleteness. That is, they are generally unwilling to alter (they strictly
enforce) the terms of a contract that is insufficiently state-contingent. The main exception to strict
enforcement is the doctrine of impossibility (or economic impracticability) with which courts some-
times refuse enforcement when performance, although literally possible, is not ex post efficient. See
Posner & Rosenfield, Impossibility and Related Doctrines in Contract Law: An Economic Analysis, 6
J. LEGAL STUD. 83, 87 (1977) (discussing legal contours of impossibility doctrine).
30. See 0. WILLIAMSON, THE ECONOMIC INSTITUTIONS OF CAPITALISM 70 (1985); MacNeil,
Contracts: Adjustment of Long-Term Economic Relations Under Classical, Neoclassical and Rela-
tional Contract Law, 72 Nw. U.L. REV. 854, 871-73 (1978); Shavell, Damage Measures for Breach

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
include legal fees, negotiat

of researching the effects
the parties and the courts
Rational parties will weigh
addressing a particular contingency. If either the magnitude or the
probability of a contingency is sufficiently low, a contract may be insensi-
tive to that contingency even if transaction costs are quite low.
The "would have wanted" approach to gap filling is a natural out-
growth of the transaction cost explanation of contractual incompleteness.
Lawmakers can minimize the costs of contracting by choosing the default
that most parties would have wanted. If there are transaction costs of ex-
plicitly contracting on a contingency, the parties may prefer to leave the
contract incomplete. Indeed, as transaction costs increase, so does the par-
ties' willingness to accept a default that is not exactly what they would
have contracted for. Scholars who attribute contractual incompleteness to
transaction costs are naturally drawn toward choosing defaults that the
majority of contracting parties "would have wanted" because these
majoritarian defaults seem to minimize the costs of contracting.
We show, however, that this majoritarian "would have wanted" ap-
proach to default selection is, for several reasons, incomplete. First, the
majoritarian approach fails to account for the possibly disparate costs of
contracting and of failing to contract around different defaults.3" For ex-
ample, if the majority is more likely to contract around the minority's
preferred default rule (than the minority is to contract around the major-
ity's rule), then choosing the minority's default may lead to a larger set of
efficient contracts. Second, the received wisdom provides little guidance
about how tailored or particularized the "would have wanted" analysis
should be.32 Finally, the very costs of ex ante bargaining may encourage
parties to inefficiently shift the process of gap filling to ex post court de-
termination.33 If it is costly for the courts to determine what the parties
would have wanted, it may be efficient to choose a default rule that in-
duces the parties to contract explicitly. In other words, penalty defaults
are appropriate when it is cheaper for the parties to negotiate a term ex
ante than for the courts to estimate ex post what the parties would have
wanted. Courts, which are publicly subsidized, should give parties incen-
tives to negotiate ex ante by penalizing them for inefficient gaps.
of Contract, 11 BELL J. ECON. 466, 468 (1980) ("[Bjecause of the costs involved in enumerating and
bargaining over contractual obligations under the full range of relevant contingencies, it is normally
impractical to make contracts which approach completeness.").
31. See infra Section II.
32. Id.
33. Jeffrey Gordon argues that the "would have wanted" approach is also flawed b
risks to the ex ante least-cost avoider but, as applied by most courts, does not in fact co
ex post risk bearer. Gordon, supra note 13, at 61 (problem with hypothetical bargain
other Kaldor-Hicks arguments, is that it doesn't guarantee that each party will in fact r
slice, or a slice of the right size." (citation omitted)).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
This Article also proposes a second source

that is the focus of much of our analysis.
pleteness as strategic.34 One party might
tion that would increase the total gains fr
pie") in order to increase her private sha
(her "share of the pie"). By attempting to contract around a certain de-
fault, one party might reveal information to the other party that affects
how the contractual pie is split. Thus, for example, the more informed
party may prefer to have inefficient precaution rather than pay a higher
price for the good.3" While analysts have previously explained incomplete
contracting solely in terms of the costs of writing additional provisions, we
argue that contractual gaps can also result from strategic behavior by rela-
tively informed parties. By changing the default rules of the game,
lawmakers can importantly reduce the opportunities for this rent-seeking,
strategic behavior.36 In particular, the possibility of strategic incomplete-
ness leads us to suggest that efficiency-minded lawmakers should some-
times choose penalty defaults that induce knowledgeable parties to reveal
information by contracting around the default penalty. The strategic be-
havior of the parties in forming the contract can justify strategic contrac-
tual interpretations by courts.37
Our analysis therefore moves beyond the received wisdom that default
rules should simply be what the majority of contracting parties would
have wanted. In choosing among default rules, lawmakers should be sensi-
tive to the costs of contracting around, and the costs of failing to contract
around, particular defaults. We show that different defaults may lead to
different degrees of "separating" and "pooling."38 In "separating" equi-
libria, the different types of contracting parties, by bearing the costs of
contracting around unwanted defaults, separate themselves into distinct
34. See infra Section II for examples of strategic incompleteness. A third reason for contractual
incompleteness is that some contingencies may simply be unforeseen by all contracting parties. In this
case, the default rule will not affect the actions of the parties since by definition they do not consider
the contingency in deciding what to do. There will normally be no reason to consider the rule's ex
ante effect because it will have none.
There is one caveat to this statement: Behavior may be affected if parties are aware that unforeseen
contingencies exist but are unable to ascertain the nature of these contingencies. For example, parties
who are aware that a variety of unforeseen contingencies may affect the price at which they should
transact may choose a contract that includes a reasonable price clause rather than fixing a price or a
price rule as a function of foreseen contingencies. In this way the price can respond to unforeseen
contingencies. See D. Kreps, Static Choice in the Presence of Unforeseen Contingencies (Aug. 1988)
(unpublished manuscript on file with authors) for a utility-theoretic characterization of behavior in
the presence of unforeseen contingencies. The choice of optimal defaults for such unforeseen contin-
gencies is beyond the scope of this article. For a discussion of the appropriate default, see Posner &
Rosenfield, supra note 29.
35. See K. Spier, supra note 29 (formalizing this strategic reluctance to reveal information).
36. See infra Section I-C.
37. Inefficient strategic behavior will often induce efficiency-promoting counterstrategies by other
economic actors. See Easterbrook, Predatory Strategies and Counterstrategies, 48 U. CHI. L. REV.
263 (1981).
38. See infra text accompanying notes 113-16.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
contractual relationships. In "pooling" equilibria, different types of con-

tracting parties fail to contract around defaults, thus avoiding transaction
costs but bearing the inefficiencies of the substantive default provisions.
In contrast to the majoritarian analysis, our analysis shows that it may
be efficient to choose a rule that a majority of people actually disfavor. To
set defaults efficiently, lawmakers must not only know what contracting
parties want, but how many are likely to get it and at what cost. We
recommend a greater and more explicit legal sensitivity toward the ways
in which different defaults will affect the resulting contractual
"equilibrium."39
Finally, before deciding how to fill gaps, courts must decide whether
the contract even has a gap. In other words, courts must decide whether
the contract already allocates a particular risk or duty. We show that this
issue of whether a gap exists is identical to the issue of what is sufficient
to contract around a particular default.40 While the received wisdom is
that lawmakers should minimize the costs of contracting around default
rules,41 we suggest that efficiency-minded courts and legislatures may
want to intentionally increase these transaction costs to discourage parties
from contracting around certain defaults.
The Article has three sections. Section I discusses the possible efficiency
of penalty defaults. Section II embeds penalty defaults in a more general
model of default choice, a model which suggests when penalty, tailored, or
untailored defaults will be efficient. Section III then develops a theory of
gap definition that determines what should be sufficient to contract
around a given default.
I. PENALTY DEFAULT RULES
A. The Zero-Quantity Default
The diversity of default standards can even be seen in contrasting the

law's treatment of the two most basic contractual terms: price and quan-
tity. Although price and quantity are probably the two most essential is-
sues on which to reach agreement, the U.C.C. establishes radically differ-
ent defaults. If the parties leave out the price, the U.C.C. fills the gap
with "a reasonable price."42 If the parties leave out the quantity, the
39. We will sometimes find it useful to distinguish between situations in which the parties negoti-
ate in ignorance of the default rule and situations in which the parties negotiate in the shadow of the
default rule. In the former case, the parties do not know how the courts will decide if the contingency
in question occurs, while in the latter case the parties know the legal default (but may not know
certain information about the other party). See infra text accompanying notes 58-60.
40. See infra Section III.
41. See Easterbrook & Fischel, supra note 11, at 21-30.
42. U.C.C. ? 2-305(1) reads: "The parties if they so intend can conclude a contract for sale even
though the price is not settled. In such a case the price is a reasonable price at the time for deliv-
ery...." U.C.C. ? 2-305(1) (1976).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
U.C.C. refuses to enforce the contract.43

that the default quantity should be zero.
How can this be? The U.C.C.'s reasonable-price standard can be partly
reconciled with the received wisdom that defaults should be set at what
the parties would have contracted for." But why doesn't the U.C.C. treat
a missing quantity term analogously by filling the gap with the reasonable
quantity that the parties would have have wanted? Obviously, the parties
would not have gone to the expense of contracting with the intention that
nothing be exchanged.
We suggest that the zero-quantity default cannot be explained by a
"what the parties would have wanted" principle. Instead, a rationale for
the rule can be found by comparing the cost of ex ante contracting to the
cost of ex post litigation. The zero-quantity rule can be justified because it
is cheaper for the parties to establish the quantity term beforehand than
for the courts to determine after the fact what the parties would have
wanted.
It is not systematically easier for parties to figure out the quantity than
the price ex ante, but it is systematically harder for the courts to figure
out the quantity than the price ex post. To estimate a reasonable price,
courts can largely rely on market information of the type "How much
were rutabagas selling for on July 3 ?"4 But to estimate a reasonable
quantity, courts would need to undertake a more costly analysis of the
individual litigants of the type "How much did the buyer and seller value
the marginal rutabagas?"
43. U.C.C. ? 2-201 states that a "contract ... is not enforceable under this [provision] beyond the
quantity of goods shown...." U.C.C. ? 2-201(1) (1976). The official comment adds that "[tlhe only
term which must appear is the quantity term which need not be accurately stated but recovery is
limited to the amount stated." U.C.C. ? 2-201(1) (official comment) (1978). In some cases, lack of a
quantity term will merely be evidence that the parties did not have a meeting of the minds. But even
if there are sufficient objective indicia of an intent to contract (and even if the statute of frauds is not
raised as an affirmative defense), courts may refuse to enforce the contract because it is indefinite. See
Jessen Bros. v. Ashland Recreation Ass'n., 204 Neb. 19, 281 N.W.2d 210 (1979) (contract for sod
unenforceable for lack of specific quantity term); Burke v. Campbell, 258 Mass. 153, 154 N.E. 759
(1927) (contract unenforceable in part because contract did not state how much stock defendant would
receive in exchange for financing corporation); King v. Krischer Mfg. Co., 220 A.D. 584, 222 N.Y.S.
66 (1927) (contract unenforceable because "a quantity of merchandise" too indefinite).
44. A simple "what parties would have wanted" approach has trouble explaining why the parties
would choose reasonable price at time of delivery instead of at the time of contracting. There is no
reason to think that parties would systematically prefer one risk allocation to the other. However, one
can determine the efficient default rule by asking the question, "why didn't the parties explicitly
contract for price?" Those parties who wish to allocate the risk of cost fluctuations to the seller will
most likely contract for a price at the time of contracting. Those who wish to allocate the risk to the
buyer will attempt to contract for a time-of-delivery, cost-based price. Such a clause may be costly to
write into the contract explicitly because of the difficulty in measuring the seller's cost exactly. The
parties may instead prefer to rely on reaching an agreement in the shadow of the court's reasonable-
price default rule.
45. This analysis suggests that courts would be less likely to enforce contracts in "thin" markets
in which the market price is not readily ascertainable. See Haddock, McChesney & Spiegel, An Ordi-
nary Economic Rationale for Extraordinary Legal Sanctions, 78 CALIF. L. REV. (forthcoming 1990)
(discussing how efficient legal rules will turn on "thickness" of market).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
The U.C.C.'s zero-quantity default is what we term a "penalty de-

fault." Because ex ante neither party would want a zero-quantity con-
tract, such a rule penalizes the parties should they fail to affirmatively
specify their desired quantity. Because the non-enforcement default poten-
tially penalizes both parties, it encourages both of them to include a quan-
tity term.46
B. Toward a More General Theory of Penalty Defaults
Penalty defaults, by definition, give at least one party to the contract an

incentive to contract around the default. From an efficiency perspective,
penalty default rules can be justified as a way to encourage the production
of information.47 The very process of "contracting around" can reveal in-
formation to parties inside or outside the contract. Penalty defaults may be
justified as 1) giving both contracting parties incentives to reveal informa-
tion to third parties, especially courts,48 or 2) giving a more informed con-
tracting party incentives to reveal information to a less informed party.
The zero-quantity default, for instance, gives both contracting parties
incentives to reveal their contractual intentions when it would be costly for
a court to discover that information ex post. This justification-that ex
ante contracting can be cheaper than ex post litigation-can also explain
the common law's broader rule that "for a contract to be binding the
terms of the contract must be reasonably certain and definite."49 Simi-
larly, this rationale can explain corporate statutes that give incorporators
an incentive to affirmatively declare the number of authorized shares, the
address of the corporation for legal process and, indeed, the state of incor-
46. Even if the judicial system were not subsidized, the zero-quantity default might be justified on
parentalistic concerns. For example, if private parties are uninformed or systematically underestimate
the costs of ex post judicial determination of a "reasonable" quantity, it might be in society's interests
to dissuade parties from mistakenly failing to negotiate the contract quantity ex ante.
This rationale for this penalty default depends on the assumption that the private parties pay less
than the full costs of their ex post litigation. The parties may lower their transaction costs by shifting
the privately funded ex ante negotiations to publicly subsidized ex post litigation. If parties fully
internalized ex post litigation costs, at first cut they should be able to choose the cheaper type of
negotiation.
47. In encouraging the revelation of information, lawmakers should be sensitive to the influence
that defaults can have on the incentives for private parties to acquire information in the first place. See
infra notes 93-94 and accompanying text.
48. Penalty defaults may be established to provide information to third parties other than the
courts. For example, in corporate law certain alterations to the default corporate governance can be
accomplished through a by-law amendment, while other alterations can only be made by changing the
articles of incorporation. See infra note 148 and accompanying text. Requiring certain amendments in
the articles of incorporation reveals information to interested third parties, such as creditors, because
these articles are publicly filed with the Secretary of State. See, e.g., REVISED MODEL BUSINESS
CORP. ACT ANN. ? 2.01 (1987) (requiring filing of articles of incorporation).
49. Steinberg v. Chicago Medical School, 41 Ill. App. 3d 804, 807, 354 N.E.2d 586, 589 (1976);
see also Parks v. Atlanta News Agency, 115 Ga. App. 842, 156 S.E.2d 137 (1967); 1 A. CORBIN,
CORBIN ON CONTRACTS ? 95 (1952 & Supp. 1989); 1 S. WILLISTON, A TREATISE ON THE LAW OF
CONTRACTS ? 37 (3d ed. 1957 & Supp. 1978).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
poration.6? Statutes that refuse to enforce

provisions create incentives similar to those
refusal to enforce vague or indefinite cont
can make these contractual choices more ef
Lawmakers should select the rule that det
social cost. When the rationale is to provid
non-enforcement default is likely to be eff
are likely to provide least-cost deterrence b
enforce and give each party incentives to contract around the rule. It
might seem that a penalty default set solely against one side of a contract
would be sufficient to get both sides to reveal information. For example, a
penalty default that makes the seller sell at one-tenth the market price
would certainly encourage sellers to affirmatively fill any price gaps. But
one side's penalty may be the other side's windfall. One-sided penalties
can create incentives for opportunism.51 The non-penalized buyer in the
above example would have incentives to induce sellers to enter indefinite
contracts in order to extract the penalty rent.62 By taking each party back
to her ex ante welfare, the non-enforcement default eliminates this poten-
tial for opportunism.
In contrast, when the rationale is to inform the relatively uninformed
contracting party, the penalty default should be against the relatively in-
formed party.63 This is especially true when the uninformed party is also
uninformed about the default rule itself. If the uninformed party does not
know that there is a penalty default, she will have no opportunistic
incentives.
In some situations it is reasonable to expect one party to the contract to
be systematically informed about the default rule and the probability of
the relevant contingency arising. If one side is repeatedly in the relevant
contractual setting while the other side rarely is, it is a sensible presump-
tion that the former is better informed than the latter. Consider, for exam-
ple, the treatment of real estate brokerage commissions when a buyer
breaches a purchase contract. Such contracts typically include a clause
which obligates the purchaser to forfeit some given amount of "earnest"
money if she breaches the agreement. How should the earnest money be
50. See REVISED MODEL BUSINESS CORP. ACT ANN. ? 2.02 (1987).
51. Posner defines opportunism (in relation to the common-law contractual duty to act in good
faith) as "taking advantage of the vulnerability of the other party to a contract . . . that is due to the
sequential character of performance." R. POSNER, supra note 2, at 81.
52. Similar opportunistic incentives have been analyzed in other areas of contract law. See, e.g.,
Clarkson, Miller & Muris, Liquidated Damages Versus Penalties: Sense or Nonsense?, 1978 WIS. L.
REV. 351 (non-enforcement of penalty clauses prevents opportunistic breach inducement); Goetz &
Scott, supra note 15, at 586 (mitigation requirement eliminates incentive for opportunism by obligee
in case of breach). More generally, the inefficiency of excessive penalties has been detailed in the
economics-of-crime literature. See, e.g., R. POSNER, supra note 2, at 205-12; Becker, Crime and
Punishment: An Economic Approach, 76 J. POL. ECON. 169 (1968).
53. See infra text accompanying notes 57-73.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
split between the seller and the broker if their agency contract does not
address this contingency? Some courts have adopted a "what the parties
would have wanted" approach and have awarded all the earnest money to
the seller."4 We agree with this outcome, but for different reasons. The
real estate broker will more likely be informed about the default rule than
the seller. Indeed, the seller may not even consider the issue of how to
split the earnest money in case of default.56 Therefore, if the efficient con-
tract would allocate some of the earnest money to the seller, the default
rule should be set against the broker to induce her to raise the issue. Oth-
erwise, if the default rule is set to favor the broker, a seller may not raise
the issue, and the broker will be happy to take advantage of the seller's
ignorance. By setting the default rule in favor of the uninformed party,
the courts induce the informed party to reveal information, and, conse-
quently, the efficient contract results.
Although social welfare may be enhanced by forcing parties to reveal
information to a subsidized judicial system, it is more problematic to un-
derstand why society would have an efficiency interest in inducing a rela-
tively informed party to a transaction to reveal information to the rela-
tively uninformed party. After all, if revealing information is efficient
because it increases the value created by the contract, one might initially
expect that the informed party will have a sufficient private incentive to
reveal information-the incentive of splitting a bigger pie. This argument
ignores the possibility, however, that revealing information might simulta-
neously increase the total size of the pie and decrease the share of the pie
that the relatively informed party receives. If the "share-of-the-pie effect"
dominates the "size-of-the-pie effect," informed parties might rationally
choose to withhold relevant information.57
Parties may behave strategically not only because they have superior
54. See, e.g., Dennis Reed, Ltd. v. Goody, 2 K.B. 277 (1950); see also J. DUKEMINIER & J.
KRIER, PROPERTY 554-55 (2d ed. 1988) (discussing evolution of common-law rule).
55. Earnest money is used to force the purchaser to internalize the cost to the seller of taking the
property off the market during the time from the signing of the sale agreement to the closing or, in
this case, the breach. Since these costs are largely incurred by the seller, she should receive the com-
pensation. Furthermore, the seller may wish to give the broker incentives to find a buyer who will not
default. Allowing the broker to share in the earnest money will lower or eliminate this incentive. One
reason, however, that the broker and the seller would ex ante contract for the broker to receive some
of the earnest money is that the breach by the initial buyer may necessitate a quick sale which may
cause the seller to lower the selling price. This, in turn, would lower the broker's commission. Thus, a
"what the parties would have wanted" approach might yield a default in which a risk-averse broker
receives a portion of the earnest money.
56. Of course, people hire lawyers in part to ascertain the relevant negotiation issues, contingen-
cies, and default rules. Our argument is therefore most applicable in contractual settings in which
lawyers are not employed.
57. Withholding socially valuable private information to obtain private gains is common. Compa-
nies may withhold information about innovations from competitors to increase profits; car buyers may
withhold information about particular options or accessories that they value if this information signals
to car dealers a greater willingness to pay for the underlying automobile; and professional athletes
may withhold information about injuries to increase their salaries, even though as a result their team
may inefficiently hire reserves.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
information about the default, but also bec

mation about other aspects of the contract.
knows that a particular default rule is inef
tiate to change it. The knowledgeable party may not wish to reveal her
information in negotiations if the information would give a bargaining
advantage to the other side.
How can it be that by increasing the total gains from contracting (the
size-of-the-pie effect) the informed party can end up with a smaller share
of the gains (the share-of-the-pie effect) ?58 This Article demonstrates how
relatively informed parties can sometimes benefit by strategically with-
holding information that, if revealed, would increase the size of the pie. A
knowledgeable buyer, for example, may prefer to remain indistinguishable
from what the seller wrongly perceives to be the class of similarly situated
buyers. By blending in with the larger class of contractors, a buyer or a
seller may receive a cross-subsidized price because the other side will bar-
gain as if she is dealing with the average member of the class. A knowl-
edgeable party may prefer to remain in this inefficient, but cross-
subsidized, contractual pool rather than move to an efficient, but unsub-
sidized, pool. If contracting around the default sufficiently reduces this
cross-subsidization, the share-of-the-pie effect can exceed the size-of-the-
pie effect because the informed party's share of the default pie was in a
sense being artificially cross-subsidized by other members of the contrac-
tual class. Under this scenario, withholding information appears as a kind
of rent-seeking59 in which the informed party foregoes the additional
value attending the revealed information to get a larger piece of the con-
tractual pie.60
58. If, under a given set of default rules, a seller wants to sell a sweater that she values at $50 to a
buyer who values it at $150, then without contracting around any of the defaults the parties' agree-
ment will create $100 of value. The total gain from contracting, in other words, will be $100. The
parties will split this gain in value between themselves by bargaining for a price between $50 and
$150. Suppose, however, that the buyer (and only the buyer) has information that would make the
sweater worth $200 to him if the seller would take on a duty that is outside of the default provisions
and that would cost the seller $10. The total gains from this non-default exchange would be $140
($200 - $50 - $10). How could the buyer lose by revealing information that increased the size of the
pie by $40? If the parties accept the default provisions and negotiate a $100 price (implying that each
party receives a $50 share of the total gains), how can it be that revealing the value-enhancing infor-
mation-by contracting for the non-default duty-would reduce the buyer's share to less than $50
(implying that the negotiated price would exceed $150 and that the seller's share would exceed $90)?
59. Most broadly, rent-seeking "arises wherever parties have an incentive to expend real resources
to capture something of value." V. GOLDBERG, READINGS IN THE ECONOMICS OF CONTRACT LAW
49 (1989). Strategic withholding represents a species of rent-seeking because the relatively informed
party commits the real resources of an inefficient contract to capture the cross-subsidization.
60. An equity-minded court might encourage information revelation to foster an equitable distri-
bution of the gains from contracting, even if doing so reduces those gains.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
C. Penalty Defaults in Action
1. Uncompensated Damages in Hadley v. Baxendale
An example of how a penalty default can restrict rent-seeking behavior

can be seen in the venerable case, Hadley v. Baxendale.61 In Hadley a
miller in Gloucester contracted with a carrier to have a broken crank shaft
transported to Greenwich. The shipment was delayed, and the miller sued
the carrier for consequential damages of the profits lost while the mill was
inoperative. The court, holding that only foreseeable consequential dam-
ages should be awarded, reversed a damage award and remanded for a
new trial.2
The holding in Hadley operates as a penalty default. The miller could
have informed the carrier of the potential consequential damages and con-
tracted for full damage insurance.63 The Hadley default of denying un-
foreseeable damages may not be consistent with what fully informed par-
ties would have wanted. The miller's consequential damages were real
and the carrier may have been the more efficient bearer of this risk. As a
general matter, millers may want carriers to compensate them for conse-
quential damages that carriers can prevent at lower cost." The default
can instead be understood as a purposeful inducement to the miller as the
more informed party to reveal that information to the carrier. Informing
the carrier creates value because if the carrier foresees the loss, he will be
able to prevent it more efficiently.65 At the same time, however, revealing
61. 9 Ex. 341, 156 Eng. Rep. 145 (1854). William Bishop has also discussed how the Hadley rule
could promote efficient revelation of information. See Bishop, The Contract-Tort Boundary and the
Economics of Insurance, 12 J. LEGAL STUD. 241, 254 (1983); L. Bebchuk & S. Shavell, Information
and the Scope of Liability for Unusual Damages from Breach of Contract, (March 1983) (unpub-
lished manuscript on file with authors).
62. Hadley, 9 Ex. at 356, 156 Eng. Rep. at 151-52.
63. There is some evidence that the miller in fact attempted to inform the carrier of the probable
damages. Thus, it may be difficult for parties to contract around the Hadley default. For a discussion
of how courts can alter the mutability of a default rule by varying the requirements for contracting
around it, see infra Section III.
64. The efficient risk-sharing agreement between symmetrically-informed shippers (e.g., the
miller in this case) and carriers will depend upon their relative attitudes toward risk, the ability of the
carrier to prevent the damages, and the ability of the shipper to mitigate damages in case of breach.
Richard Posner and Richard Epstein argue that in many situations there are actions the shipper
could take to reduce consequential damages. See R. POSNER, supra note 2, at 114-15; Epstein, Be-
yond Foreseeability: Consequential Damages in the Law of Contract, 18 J. LEGAL STUD. 105,
121-25 (1989). In such situations the parties would choose to share consequential damages through
appropriate liquidated damage clauses. See Cooter, Unity in Tort, Contract, and Property: The
Model of Precaution, 73 CALIF. L. REV. 1, 15 (1985). In Evra Corp. v. Swiss Bank Corp., 673 F.2d
951, 955-59 (7th Cir. 1982), Judge Posner attempts to determine what insurance arrangement the
parties would have contracted for. Epstein prefers a less tailored rule that does not require the courts
to analyze the relative abilities of both sides to reduce damages.
We proceed under the assumptions that the miller had no economically practical means of reducing
the losses from delay (e.g., keeping a spare crankshaft in inventory would have been too expensive)
and that both parties were risk-neutral. Under these assumptions the efficient contract is for the
carrier who is the least-cost avoider to bear the costs of delay.
65. This is one lesson from Goetz and Scott's analysis of the anxious alumnus: A bus company
driving Dean Smith to the Final Four would probably take more efficient precautions if it were

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
the information to the carrier will undoub

ping. Nonetheless, so long as transaction co
with high consequential damages will gain from revealing this informa-
tion and contracting for greater insurance from the carrier because the
carrier is the least-cost avoider.
This is not to say that there could not be an equally efficient market
response if Hadley had gone the other way. If the default required carri-
ers to compensate for unforeseeable consequential damages, low-damage
millers would have the incentive to raise the issue of consequential dam-
ages. In a competitive industry, the uninformed carrier, in effect, assumes
she is facing an average-damage miller and charges a price accordingly.
The market price will reflect the expected cost of insuring high-damage
millers. A low-damage miller will want to contract for less-than-average
insurance and, therefore, a lower price. But the gains from contracting
around the default may be insignificant if the proportion of millers with
high damages is small. Furthermore, it may be very difficult for a low-
damage miller to determine how much of the price is an implicit insur-
ance premium for millers with higher damages.
Thus, there may be situations in which the low-damage millers fail to
contract for the low-insurance/low-price contract. In the resulting equilib-
rium, carriers may charge a price representing their average cost of serv-
ing both high- and low-damage millers and take an average amount of
precaution (which will be relatively low if there are relatively few high-
damage millers)." Richard Posner suggests a similar result: If the damage
default changed so that manufacturers of photographic film were liable
for unforeseen consequential damages, "[t]he manufacturer of the film
will probably take no additional precautions . . . because he cannot iden-
tify the films whose loss would be extremely costly, and unless there are
many of [the high-damage photographers] it may not pay to take addi-
tional precautions on all the films he develops."67 Section II formalizes
this discussion of Hadley to show that low-damage millers might fail to
contract around a default that awarded unforeseeable damages while
high-damage millers will contract around the Hadley rule."8
warned of how large Dean Smith's consequential damages would be were he to miss the game. Goetz
& Scott, supra note 15, at 578.
66. The uninformed parties, the carriers, could simply exclude unforeseeable consequential dam-
ages from their standard-form contract (thereby contracting around the default at very low cost), and
high-damage millers, if they want insurance, could simply contract around the contract default. In
other words, if the legal default rule is inefficient, contracting parties may have ways of supplanting it
with a default of their own. (This point can also be made with regard to Posner's story about film
development; see infra text accompanying note 67.) The limits to this contractual response to the
contra-Hadley default are discussed infra notes 70-71 and text accompanying notes 69-71.
67. R. POSNER, supra note 2, at 114 (emphasis in original).
68. In the economics literature several articles examine situations in which asymmetric informa-
tion induces inefficient contracting. See, e.g., Akerlof, The Market for "Lemons": Quality Uncertainty
and the Market Mechanism, 84 Q.J. ECON. 488 (1970); Myerson, Mechanism Design by an In-
formed Principal, 51 ECONOMETRICA 1767 (1983).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
It is important to consider another channel for information revelation.

The uninformed party, the carrier, may attempt to learn the expected
damages of the informed parties, the millers, by offering a menu of insur-
ance contracts.69 The millers might then be induced to self-select the in-
surance contract that is optimal for their expected damages.70 The prob-
lem faced by uninformed parties trying to induce information revelation,
however, is that in many situations the necessary menu is more compli-
cated than in the common carrier example. Devising a menu that induces
information revelation may require a great deal of sophistication by the
uninformed party and may entail large transaction costs.7"
The main lesson to draw from our discussion of Hadley is that there
may be strategic reasons for parties' choosing not to reveal information. If
the default rule awarded all consequential damages, to be sure, the low-
damage millers would want to distinguish themselves from the high-
damage millers. But the high-damage millers may intentionally choose to
withhold information that would make their contracts more efficient. An
informed party may not realize the full social value of revealing the infor-
mation and, hence, her private benefits from revealing may diverge from
the social benefits of having the information revealed. As elaborated be-
low, by not distinguishing themselves, informed parties may be able to
free-ride on the lower-cost qualities of others and thereby contract at a
subsidized price. To counteract this strategic behavior, courts should
choose defaults that are different from what the parties would have
wanted.
Easterbrook and Fischel have argued that courts should choose defaults
that "the parties would have selected with full information and costless
contracting."72 Their standard fittingly tracks the two reasons for contrac-
tual incompleteness that we have identified-strategic withholding and
transaction costs. In a sense, their standard seeks what the parties would
have selected if there were no barriers to contractual completeness.73 We
disagree, however, with their conclusion that courts should choose the de-
fault that the parties would have selected with "full information." When
69. See supra note 66.

70. This is in fact done by many common carriers. The Federal Express standard contract, for
example, limits consequential damages to $100 but permits the shipper to buy greater insurance at
stated rates. See Epstein, supra note 64, at 120.
71. See, e.g., infra text accompanying note 81 (consumers trying to ascertain information about
dealer's expected lost-profit damages may encounter significant resistance in devising menu that en-
courages dealers to reveal information). For an example of a menu that would require large transac-
tion costs and sophistication on the part of the uninformed party, see Note, Imperfect Information,
The Pricing Mechanism, and Products Liability, 88 COLUM. L. REV. 1057 (1988) (suggesting that
manufacturers should be required to provide menus of warranties so that consumers can judge quality
and safety of products).
72. Easterbrook & Fischel, supra note 11, at 20-21.
73. Full information by both parties would eliminate any opportunity for strategically withhold-
ing information, and costless contracting obviously eliminates transactions costs as an impediment t
complete contracting.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
relatively informed parties strategically wi

promote information revelation, should choose a default that the informed
party does not want. Imposing ex post what completely informed parties
would have contracted for may not result in the ex ante revelation neces-
sary for efficient reliance or precaution. For example, we have shown that
in Hadley fully informed parties may have wanted the carrier to be fully
liable for consequential damages. Yet choosing full liability as the damage
default may lead carriers to invest suboptimally in accident prevention.
Hadley is inconsistent with a full-information, "what the parties would
have wanted" standard. Instead, Hadley penalizes high-damage millers
for withholding information that would allow carriers to take efficient
precautions.
2. Goldberg's Solution to the Lost-Profits Puzzle
Victor Goldberg's analysis of the lost-volume retail seller74 comports

with our discussion of penalty defaults. Goldberg examines what a re-
tailer's damages should be when a customer breaches a contract to buy.
The U.C.C. mandates, and other commentators have suggested, that the
retailer should receive the lost profits on the good." Goldberg's own anal-
ysis suggests that the lost profits approximate the real loss borne by the
retailer. But Goldberg instead proposes that courts deny "recovery for lost
profits in the absence of explicit contract language to the contrary."76 In
other words, Goldberg proposes that the default damages be zero.
His rationale parallels our rationale for penalty defaults. The zero-
damage default is intended to give retailers an incentive to come forward
and contract for a nonrefundable deposit or for a liquidated damages
clause. Again, the process of contracting around the default apprises the
customer of her potential liability. A sales contract gives the consumer an
option either to purchase the product or to pay damages. Goldberg's zero-
damage default encourages retailers to inform consumers about the price
of exercising the option to breach. Consumers will therefore internalize
the cost to the retailer of breach and more efficiently take precautions
against breach.77 Furthermore, the consumer may not know the default
rule for breach.78 If a consumer is unaware that the default rule makes
74. Goldberg, An Economic Analysis of the Lost-Volume Retail Seller, 57 S. CAL. L. REV. 283
(1984).
75. See U.C.C. ? 2-708(2) (1976); Goetz & Scott, Measuring Sellers' Damages: The Lost-Profits
Puzzle, 31 STAN. L. REV. 323, 323 (1979); Speidel & Clay, Seller's Recovery of Overhead Under
UCC Section 2-708(2): Economic Cost Theory and Contract Remedial Policy, 57 CORNELL L. REV.
681, 694 (1972).
76. Goldberg, supra note 74, at 291.
77. In this situation the uninformed party, the consumer, probably could not offer a menu of
contracts to the informed retailer to induce revelation of the markup. See supra text accompanying
notes 69-71. The information requirements, complexity, and transaction costs of such a scheme would
be prohibitive.
78. This is similar to our earlier discussion of what the real estate default rule should be for

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
her liable for lost profits, the seller will have little incentive to bargain
about damages.79 Consumer liability for lost profits can lead to efficient
breach and precaution only if consumers know the amount of their liabil-
ity. The zero-damages penalty default encourages the retailer to reveal her
markup.80
Although consumers would value this markup information, retailers
still have incentives to withhold it.81 By revealing their profits, retailers
may simultaneously reduce their bargaining power. Even Goldberg's pen-
alty default, therefore, could be too weak to induce information disclosure.
Retailers may sometimes prefer to take their chances with zero damages
for breach rather than disclose a high markup. If the zero-damage penalty
default is insufficient to induce information revelation, a stiffer penalty
may be necessary to induce the parties to contract for liquidated
damages.
3. Unilateral Defaults as Penalties: The Perverse Incentives of

Lefkowitz
Some contractual rules establish defaults that individual parties may

unilaterally change. Consider, for example, the legal effect of offers. The
U.C.C.'s mandate that an offer intends any reasonable form of acceptance
is a unilateral default that obtains "unless otherwise unambiguously indi-
cated."83 Consistent with the proceeding analysis, lawmakers may want to
set penalty defaults that encourage offerors to reveal information to offer-
ees. Thus, courts may want to choose the default that the offeror does not
want.
In Le/kowitz v. Great Minneapolis Surplus Store, Inc.,84 the Minne-

sota Supreme Court rejected the plaintiff's claim that the defendant, a
retailer, had failed to honor two advertised offers that the plaintiff had
accepted.85 In the first offer the defendant had advertised the sale of three
splitting earnest money between the seller and the broker when a buyer breaches a purchase contract.
See supra text accompanying notes 54-56.
79. If the seller raises the issue of damages, thereby revealing her markup, consumers can more
efficiently take precautions to avoid breach. Since the consumer will learn she is liable for damages,
however, she will insist on a lower price for the good.
80. Similarly, the common-law doctrine of construing ambiguities in contracts against the drafter,
see I. MACNEIL, supra note 11, at 372, can be viewed as a penalty default. The doctrine is not based
on the judgment that the parties would have wanted the anti-drafter provision, but that such a penalty
encourages drafters to draft more precise contracts.
81. For a fuller discussion of this point, see Ayres & Miller, "I'll Sell It To You At Cost": Legal
Methods to Promote Honest Retail Markup Disclosure, 84 Nw. U.L. REV. (forthcoming 1990) (man-
uscript on file with authors).
82. It is not necessarily true, however, that retailers will reveal their markups via their choice of
liquidated damages. All retailers may uniformly negotiate for minimal liquidated damages that pro-
vide no more information than the zero-damage default.
83. U.C.C. ? 2-206(1) (1976).
84. 251 Minn. 188, 86 N.W.2d 689 (1957).
85. The court concluded that the newspaper advertisements constituted offers and not merely
tations to make an offer. Id. at 192, 86 N.W.2d at 691.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC2:34:56 UTC
fur coats " 'Worth to $100.00' " for " '$1 Ea

defendant advertised the sale of one black lapin stole " 'worth $139.50' "
for "'$1.00' "86 The court, following the common-law rule that indefinite
contracts are not enforceable, refused to enforce the contract arising out of
the first offer but awarded the plaintiff damages for his attempted accept-
ance of the second.
The court reasoned that the first offer's estimate that the coat was
'worth to $100'" was too speculative and uncertain to award damages.87
But in applying the common-law standard that indefinite contracts are
unenforceable, the court ignored the likely market response to the non-
enforceability default. As argued above, non-enforceability can be viewed
as a penalty default that encourages both parties to come forward and fill
in the gap;88 that is, refusing to enforce indefinite contracts drives out
indefinite contracts. In Le/kowitz, however, the court's refusal to enforce
the indefinite offer leads to exactly the opposite result.
Ask yourself the simple question: What kind of ad is the Great Minne-
apolis Surplus Store going to run the week following the court's decision?
By lending its imprimatur to the indefinite ad, the court allows retailers to
induce inefficient consumer reliance with impunity.89 The Le/kowitz case
dramatically illustrates that only by enforcing indefinite offers against the
offeror can one drive out indefinite offers.
Le/kowitz was wrongly decided. The defendant's offer was intentionally
vague to induce inefficient reliance on the part of the buyer (Lefkowitz
incurred the "shoe-leather" costs of traveling to the store). Courts can re-
tain the common law's general reluctance to enforce indefinite contracts so
that both parties will have an incentive to make the contracts more defi-
nite.9O But Le/kowitz illustrates an exception to this general rule. When
the indefiniteness is clearly attributable to one party and induces ineffi-
cient reliance from the other party, punitive enforcement may be efficient
to drive out inefficient indefinite offers.
D. Summary
Penalty defaults stand as stark counter-examples to the proposition th

courts should simply choose defaults that the parties would have wanted
Particularly when individual parties have private incentives to withhold
86. Id. at 189, 86 N.W.2d at 690.

87. Id. at 190, 86 N.W.2d at 690.
88. See supra text accompanying notes 49-52.
89. Our argument that consumers will be "suckered" into the store by non-enforceable, indefinite
offers assumes that consumers are unable to distinguish between enforceable and unenforceable offers.
90. See supra text accompanying note 88.
91. It could be argued that rational contractual parties behind a "veil of ignorance" would want
penalty defaults that increase the total gains to contracting. Cf. J. RAWLS, A THEORY OF JUSTICE
24-38 (1974) (discussing preferences under utilitarian assumptions of persons behind "veil of
ignorance").

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
information, it may be desirable for the law to give them a nudge. The
possibility that efficient defaults will at times be used to reflect what most
people want while at other times be used to encourage the revelation of
information is analogous to the disparate uses of presumptions in the laws
of evidence."' In both instances, the law is sometimes chosen to promote
the revelation of information.
Finally, having shown that lawmakers will sometimes want to set de-
faults that encourage one or both parties to reveal information, we now
should warn lawmakers that they should sometimes protect the private
incentives to become informed. In some instances forcing parties to reveal
information will undermine their incentives to obtain the information in
the first place.9" Lawmakers therefore should not impose penalty defaults
that have a net effect of reducing the amount of socially useful informa-
tion. But in some instances, a particular party may need to acquire certain
types of information before contracting, so that forcing disclosure would
have minimal disincentive effects. For instance, in Goldberg's example of
lost-profit damages, retailers naturally knew the profits from a sale. It is
hard, therefore, to conceive how forcing retailers to disclose their profits
would undermine their private incentives to calculate the profitability of a
sale.94
92. The list of penalty defaults analyzed above is far from exhaustive. For example, the U.C.C.
sections which establish implied or default warranties, U.C.C. ?? 2-314 and 2-315, cannot easily be
justified as "what the parties would have contracted for." Instead, the defaults, consistent with the
foregoing analysis, force sellers to reveal information to consumers about the extent of their coverage.
Indeed, one way to identify penalty defaults is to investigate the pervasiveness with which parties
contract around them, as is done with the seemingly ubiquitous use of limited warranty disclaimers.
U.C.C. ? 2-207 is also inconsistent with "would have wanted" default analysis. This section sup-
plants the common-law mirror-image rule with the default that additional terms in an acceptance that
do not materially alter the terms of the offer become part of a contract between merchants. This
default cannot be reconciled with a "what the parties would have contracted for" analysis, because
there is no reason to think that the merchants would have wanted to include the additional terms of
their contract. Instead, the rule places an informational burden on the party with the last clear chance
to come forward and notify the other side if the additional terms are objectionable.
Evidentiary presumptions in litigation are sometimes used to reflect a relationship between facts
and at other times to place the burden of producing evidence on the party who is more likely to be
informed. See Allen, Presumptions in Civil Actions Reconsidered, 66 IOWA L. REV. 843, 845 (1981)
(suggesting that presumptions are used "to construct rules of decision to avoid factual impasse at trial;
to allocate burdens of persuasion; to instruct the jury on the relationship between facts; and to allocate
burdens of production").
93. See R. POSNER, supra note 2, at 115; Easterbrook, Insider Trading, Secret Agents, Eviden-
tiary Privileges, and the Production of Information, 1981 SuP. CT. REV. 309, 359; R. Allen, M.
Grady, D. Polsby & M. Yashko, Confidentiality of Legal Affairs (1988) (unpublished manuscript on
file with authors); cf Laidlaw v. Organ, 15 U.S. (2 Wheat.) 178 (1817) (permitting relatively in-
formed contracting party to profit by keeping information to himself). Of course, much of patent law
is justified as a means of encouraging the private production of information.
94. Anthony Kronman distinguishes "deliberately acquired information" from "casually acquired
information." For example, he suggests that if one side to a contract is aware that the other has
entered a mistaken bid, the special knowledge of the non-mistaken party "is unlikely to be the fruit of
a deliberate search." Requiring disclosure by the knowledgeable party of this casually acquired infor-
mation will accordingly not undermine incentives to become informed. Kronman, Mistake, Disclosure,
Information, and the Law of Contracts, 7 J. LEGAL STUD. 1, 32 (1978) (revealing lost profits from
contract can undermine incentives for parties to search for undervalued assets). There may be a trade-
off between inducing efficient search by one party and efficient breach by the other. Retailers do not

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
II. A GENERAL THEORY OF DEFAULT CHOICE
The prior analysis of Hadley can be embedded in a more general model

of default choice. To formalize our ideas, we begin by introducing some
notation. Let the percentage of millers who had high (unforeseeable) dam-
ages be a? and the percentage of millers who had low (foreseeable) dam-
ages be aL.96 Let DH and DL be the monetary value of damages for the
high- and low-damage millers, respectively. The court chooses between
two possible defaults: denying or awarding the high, unforeseeable dam-
ages. Let the costs of contracting around these defaults be CH and CL,
where:
cH = cost of contracting around the Hadley default rule that denies

awards for unforeseeable consequential damages; and
CL = cost of contracting around the default rule that awards unfore-
seeable consequential damages.
We assume that all parties are risk neutral, so the only goal of the
insurance aspect of the contract is to induce the carrier to take efficient
precaution. As argued above, in a world with full information, both high-
and low-damage millers would contract for the carrier to pay for their
consequential damages.96 If the carrier knows the miller's type and bears
all damages, she will choose the optimal level of precaution.97 Let KH and
KL equal the carrier's optimal investment in precaution for the two types
of millers, respectively. Let qH and qL equal the probability of damages
for each type of miller given the optimal level of precaution for the class.98
Before analyzing the likely equilibrium associated with each default, we
must say a bit more about the carrier's price. Assume for the moment that
the carrier is in a competitive market in which, by definition, there are
zero economic profits.99 Let the competitive price of shipping crankshafts
for known high-damage millers be PH and for known low-damage millers
be PL With this notation, if MC is the marginal cost of shipping with no
precaution, then the following equations will hold:
(1) PH MC + KH + qHDH'
(2) PL =MC + KL + qLDL
search for their markup information-in Kronman's terminology it is n

that information revelation will enhance efficiency.
95. We assume that these are the only types of millers so that
mathematical model of Hadley, see L. Bebchuk & S. Shavell, supra no
97. We assume that the contract cannot mandate a specific level of pre
miller nor a court can observe the level of precaution. Otherwise, carriers
ing different levels of precaution.
98. Thus, KH is the level of K that minimizes the expression K + q(K)
K that minimizes the expression K + q(K)DL, where the function q(K
damages from delay for a given level of precaution, K.
99. Economic profits are the residual earnings after all implicit and o
counted for.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
Because of competition, these prices reflect the carrier's marginal cost of

shipping, precaution, and expected damages.
What will the market equilibrium look like under the alternative de-
faults? First consider the Hadley low-damage default. Since the carrier is
liable only for foreseeable damages, DL, she will choose a precaution in-
vestment of KL. The probability of damages will therefore be qL, and the
price will be PL* Now we can ask whether the high-damage millers will
contract around the low-damage default for higher insurance. If the high-
damage millers fail to contract around the low-damage default, they will
pay PL and, in addition, bear the probability of damage, qL, multiplied by
the uninsured damages (DH - DL). Thus, the total expected cost, pHin' to
the high-damage millers if they inefficiently fail to contract around the
low-damage default is:
(3) PH PL + qL(DH - DL).

But because the carrier invests suboptimally in precaution, the millers'
total expected cost in equation (3) must be greater than if the carrier had
taken the efficient level of precaution and passed the cost on in the price,
PH:
(4) PHin > PH.100
This difference (PHin - PH) represents the inefficiency (fH) associated
with failing to contract for the efficient level of precaution:
(5) in - =
"fH'" then, repres
fail to contract for
pay:
(6) PHin PH + fH
when they fail to contract around the Hadley low-damage default.
If the high-damage millers contract around this default, they should
expect to bear the costs of the increased precaution as reflected in a higher
shipping price (PH rather than PL) as well as the additional transaction
costs of contracting around the default (cH).'02 This total cost would be:
100. From equations (1), (2) and (3):

PH PL + qL(DH - DL) = MC + KL + qLDH, and
PH= MC + KH + qHDH
So that:
PH - PH= KL + qL(KL)DH - (KH + qH(KH)DH)-

Because KH minimizes K + q(K)DH, see supra note 98, KL + qL(KL)DH must be larger
KH + qH(KH)DH. Therefore, the right side of this equation must be positive, and inequality (4
must hold.
101. From the previous note we can calculate the inefficiency cost:
fH= KL + qLDH - (KH + qHDH) = (qL - qH)DH - (KH - KL)-
This latter expression illustrates that the inefficiency of low precaution derives from the fact that the
costs of the higher probability of damage, (qL - qH)DH, outweigh the lower costs of precaution, (KH -
KL)-
102. The high-damage millers bear all of these additional transaction and precaution costs be-
cause competition will keep the carrier's price at marginal cost.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
(7) PH + CH.
Thus, the high-damage mille
fault if the total cost of failin
costs of contracting (PH + CH):
(8) PH + fH > PH + CH,
or more simply if:
(9) fH > CH.
High-damage millers will c
fault when the cost of ineff
contracting around the def
around the Hadley default a
will contract for the efficie
ers will effectively pay PH
ers will pay PL
Now consider the non-Had
age of high-risk millers, a H
produce an equilibrium in which no one contracts around the default and
the carriers take only low levels of precaution, KL.104 If the carriers can-
not distinguish between high- and low-damage millers, competition would
cause the shipping price to become:
(1 0) P = aLPL + aHPH
Since in a competitive market, carriers
expected average cost of serving both h
represents a weighted average of these
the carriers PL' while the high-damage
inefficient, low level of precaution is ta
This will be the equilibrium if neithe
millers have an incentive to reveal to t
might think that the low-damage mille
forward and reveal their status in orde
But they will do so only if:
(11) cL < P* - PL- 06
Contracting around the high-damage default
savings from the reduced shipping price (P*
103. We assume that the costs to the court of determinin

quential damages-i.e., determining whether a given miller
104. If precaution investment is a continuous choice vari
spent training carriers how to drive safely), its optimal lev
aH the investment will be close to KL. If there are discrete
example, use two or three horses) and aH is small enough,
105. Substituting equations (2) and (3) into equation (10) yields:
P* = MC + KL + qLDL + alqL(DH - DL),

which, again substituting equation (2), simplifies to:
P= PL + alqL(DH - DL)-
106. From the prior footnote, this inequality can be expressed as:
CL < alqL(DH - DL)

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
additional transaction cost, CL. But from equation (10) we see that the
equilibrium price, P*, is a positive function of aH, the percentage of high-
damage millers.107 If this percentage (aH) is sufficiently low, the transac-
tion costs (CL) will keep the low-damage millers from contracting around
the high-damage default. Note that this is unlike the low-damage default
in which the incentives to contract around the default rule are based on
the gains to the high-damage millers and are independent of the percent-
age of the population that has high damages (aH).
The high-damage millers will not reveal their true status to the carriers
because they would be forced to pay more (PH - P*) but would gain no
additional coverage. The high-damage millers fail to distinguish them-
selves not because of transaction costs, but because they prefer to withhold
this information strategically and to receive the subsidized shipping
price.108 The shipping price is subsidized because transaction costs preve
low-damage millers from contracting around the default. Even though the
information is socially valuable because it leads to more efficient precau-
tion and even though this value exceeds the transaction costs, the high-
damage millers prefer to remain undistinguished from their low-risk
counterparts. The high-damage millers do not mind that carriers take in-
efficiently low levels of precaution because, like all shippers, high-damage
millers are fully insured. The low-damage millers bear the costs of this
inefficiency, but are not hurt enough individually to distinguish them-
selves contractually.
Prior analyses of incomplete contracts have suggested that parties fail to
contract around inefficient defaults because of transaction costs.109 Our
analysis is striking because it demonstrates the possibility that parties may
fail to contract around defaults for strategic reasons.110 A relatively in-
formed party may strategically withhold information that would increase
the joint gains from trade.111 Moreover, the example illustrates two ex-
treme forms of default equilibria. The Hadley, low-damage default caused
all high-damage millers to contract around the rule and thus engendered
107. One could imagine situations in which the costs for low-damage millers of contracting
around a high-damage default would be small. For example, the Federal Express standard-form con-
tract limits consequential damages if the sender does not contract for more insurance. See supra note
70. Such a standard form would allow low-damage millers to cheaply opt for low-damage protection
and consequently a lower price.
108. In other words, even if the costs of contracting for higher precaution were zero, the high-
damage millers would not reveal their status.
110. In our simplified model, low-damage millers failed to contract around the pooling equil
rium because of transaction costs. See supra inequality (11). In a more general model, however, e
without transactions costs the informed cross-subsidizing parties (such as the low-damage mil
may fail to contract around pooling equilibria if in doing so they reveal information which redu
their bargaining power. See K. Spier, supra note 29 for a more formal demonstration of why sim
strategic concerns can keep contracts from being efficiently state-contingent.
111. This, then, is an example in which the "share-of-the-pie effect" exceeds the "size-of-the-
effect. See supra text accompanying note 57.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
what economists sometimes call a "separating

ferent types of contracting parties sort them
process into different groups at different prices. In contrast, the non-
Hadley, high-damage default created a "pooling" equilibrium in which
the high- and low-damage millers failed to distinguish themselves to the
carriers.112
The tension between "pooling" and "separating" equilibria is crucial in
choosing the efficient default. Separating equilibria entail the additional
transaction costs of causing some parties to contract around the rule; pool-
ing equilibria entail the costs of inefficient reliance or precaution for fail-
ing to contract around the rule.1"3 In the Hadley example the transaction
cost of the high-damage millers' contracting around the low-damage de-
fault is acHcH. This is the cost of separating equilibria. The cost of the
high-damage default stems from the inefficiently low precaution that car-
riers take with regard to the high-damage millers, aHfH. This is the cost
of pooling equilibria. If inequality (9) holds (i.e., if fH > CH), then the
costs of pooling exceed the costs of separating, and the Hadley default is
efficient even though it is not what the high-damage millers would have
wanted.
Although the prior section described the Hadley rule as a penalty de-
fault, it can be alternatively conceived as an untailored default rule that
provides what the majority of the parties would want (since under our
assumptions, aL > aH). This reconception suggests that untailored de-
faults that apply a single rule to different types of contracting parties act
as penalty defaults with regard to those parties who disfavor them. Be-
cause a majority of the millers had low damages (aL > aH) and would
only want to contract for low damages, one might deduce that the un-
tailored Hadley rule is efficient because a majority of the parties prefers
it.
But the commonly accepted notion that untailored defaults should be set
at what the majority of parties wants does not hold in a more general
model of default choice in which the pooling and separating equilibria are
not extreme. To extend the Hadley model, consider the choice between
112. This analysis is consonant with the economics of insurance literature. See, e.g., Rothschild &
Stiglitz, Equilibrium in Competitive Insurance Markets: An Essay on the Economics of Imperfect
Information, 90 Q.J. ECON. 629, 634-37 (1976) (discussing "separating" and "pooling" equilibria).
Low-risk insureds will have incentives to drop out of (or separate from) pools in which they cross-
subsidize the premiums of high-risk insureds. See generally Priest, The Current Insurance Crisis and
Modern Tort Law, 96 YALE L.J. 1521 (1987) (arguing that this sort of separation has occurred in
third-party insurance pools and is largely responsible for recent "insurance crisis").
113. In this model all millers continue to ship crankshafts regardless of the default rule. If, how-
ever, the cross-subsidization of the non-Hadley default reduced the consumption of the low-damage
millers (who must pay P* - PL more) or increased the consumption of the high-damage millers (who
have to pay PH - P* less), then the pooling equilibria would additionally entail the dead-weight losses
associated with inefficient amounts of contracting. See Note, Contract Damages and Cross-
Subsidization, 61 S. CAL. L. REV. 1125 (1988).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
two defaults: one and two.

the population of contractin
ferred, al and a2, respectiv
around each default, cl an
default two to the default one rule (and define c2 analogously). The Had-
ley model showed that failing to contract around a default could result in
inefficient levels of precaution.115 Accordingly, let f1 equal the inefficiency
generated if a type one contracting party fails to contract for the type one
rule (and analogously define f2). Most important, in a more general model
there may be intermediate amounts of pooling and separation for the dif-
ferent defaults. Accordingly, let:
1= the percentage of type one contracting parties who in equilib-

rium would actually contract around default two, and
02 = the percentage of type two contracting parties who in equilib-
rium would actually contract around default one.
Thus, while a, percent of the contracting parties prefer default one, it is

possible that if default two is the rule, a smaller percent (#1 < a,) would
contract around it for the efficient amount of precaution. And analogously,
some type two contracting parties might not contract around default one if
it is the rule (02 < a2).116
The completeness of the Hadley pooling equilibrium implies that OH
aH,5while the completeness of the non-Hadley separating equilibrium im-
plies that OL = 0. These extreme results turn on the homogeneity of, for
example, fH and CH. More generally, if high-damage millers have hetero-
geneous costs of contracting or of failing to contract, then different de-
faults may engender intermediate forms of pooling and separating.
In this more general model, each default can engender both costs of
contracting and costs of failing to contract. For example, the costs of de-
fault one will be:
c232 + f2(a2 - 02)5
which equal the transaction costs of those type two contractors who con-
tract around default one (c2g2) and the inefficient reliance or precaution
costs for those type two parties who fail to contract around default one
(f2(a2 - f2)). Default one will be optimal only if:
(12) cfld + f1(a1 - j1) > C2g2 + f2(a2 - 02)
This expression establishes the optimal condition for choosing between
two untailored defaults. In equilibrium each default may cause a portion
of the population to incur the expense of contracting around the default
114. a1 +a2 = 1.
115. See supra text accompanying note 100.
116. Because we assume that parties would never bargain for a less efficient contract, a, and a2
will always be greater than or equal to ,1 and ,2 respectively.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
(chilo), and each rule may create costs (f

inefficiently fail to contract around it. The
sum of these two costs, which, as stated bef
separation and pooling, respectively:
min [cifi + fi(ai -
i =1,2 11 1
In contrast to the majoritarian analysis, condition (12) does
that efficiency-minded lawmakers should choose the default ru
of the parties would want. In other words, condition (12) doe
that:
al > a2.
Indeed, exploiting the fact that a2 1 - a1,117 one can rearrange condi-
tion (12) in terms of a1. Default one will be efficient if and only if:
(13) a> f2 + 32(C2 - f2) - #3(C1 - fl)

f1 +f2.
The right-hand side of this cumbersome inequality implicitly defines a
critical value of a1 * for choosing between defaults one and two. When a1
is less than a1*, default two is efficient; when a1 is greater than a1*,
default one is efficient.
The crude majoritarian criterion (a1,* = /2) only emerges from highly
constrained assumptions about #1, 02, c1, c2, f1, and f2. For example, the
majoritarian default analysis in the existing contracts literature seems to
derive from two divergent sets of assumptions about these crucial vari-
ables.118 One set of majoritarians seems to assume that transaction costs
are small enough that no one fails to contract around inefficient defaults
(a1 = f1 and a2 = 1B2) and that the costs of contracting around each
default are the same (c1 = c2).119 Under these assumptions, the second set
of terms in condition (12) drop out of the analysis (fi(ai - 13) = 0), so that
the costs of failing to contract are irrelevant. Inequality (13) can then be
simplified to:
a1 > C2/(c1 + C2) 1?2,

which implies the majoritarian result that default one should be chosen
only if a majority of the contracting parties prefers it.
A second set of majoritarians seems to assume that the transaction costs
of contracting around a default are so great that no parties will contract
around inefficient defaults (l1 = =2 = 0) and that the inefficiencies of
failing to contract for the right rule are the same (f1 = f2). Under these
assumptions the first terms of condition (12) drop out of the analysis (cifi
0), so that inequality (13) can be simplified to:
a1 > f2/(f1 + f2) = 1/2,

118. See Craswell, supra note 22, at 632-39 (discussing two approaches).
119. Id. at 633.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
which again implies that default one will be efficient only if a majority
prefers it. Majoritarians are forced to make highly restrictive (and some-
times contradictory) assumptions to produce their desired rule. Most fun-
damentally, the majoritarian analysis errs by looking at only one of the
relevant variables, a.
As an alternative, some commentators have suggested that courts fill
gaps with the provisions that most parties bargain for in actual con-
tracts.120 Some academics have labelled this style of gap-filling as "mim-
icking-the-market."'121 The "mimic-the-market" approach to default rules
ignores the fact that the type of parties who contract around a given rule
depends upon the rule itself. Parties who dislike a given default rule will
contract around it; if we change the default rule to mimic the contracts
these parties write, other types of parties may contract around the new
default back to the original rule. This process could cycle forever.122 Set-
120. See Epstein, In Defense of the Contract at Will, 51 U. CHI. L. REV. 947, 951 (1984) (a
default rule "is normally chosen because it reflects the dominant practice in a given class of cases and
because that practice is itself regarded as making good sense for the standard transactions it gov-
erns."). Epstein argues that the default for consequential damages should be limited because this "is
what the express contracts have typically provided." Epstein, supra note 64, at 118. Frank Easter-
brook orally suggested a similar standard for choosing corporate default rules at the Columbia confer-
ence. See supra note 13.
The NLRB has looked at actual collective bargaining agreements in deciding whether there should
be a default limiting management's right to transfer work. See Milwaukee II, 268 N.L.R.B. Dec.
(CCH) 601, 603 (1984) (quoting Ozark Trailers, Inc., 161 N.L.R.B. Dec. (CCH) 561, 570 (1966)
(citing M. CHANDLER, MANAGEMENT RIGHTS AND UNION INTERESTS 217 (1964))). The Supreme
Court has at times looked to actual contracting practices to determine whether a particular issue
should be a mandatory subject of bargaining. See, e.g., First Nat'l Maintenance Corp. v. NLRB, 452
U.S. 666, 684 (1981); see also Alchian, Decision Sharing and Expropriable Specific Quasi-Rents: A
Theory of First National Maintenance Corporation v. NLRB, 1 SuP. CT. ECON. REV. 235 (1982)
(suggesting how "decision sharing" default should be set); Wachter & Cohen, The Law and Econom-
ics of Collective Bargaining: An Introduction and Application to the Problems of Subcontracting,
Partial Closure, and Relocation, 136 U. PA. L. REV. 1349, 1364-77 (1988) (suggesting how "de-
fault entitlements" in labor market should be set). However, in Lewis v. Benedict Coal Corp., 361
U.S. 459 (1960), discussed supra note 21, the Court eschewed any empirical analysis of the private
reaction to a particular contract.
121. See Schwab, Collective Bargaining and the Coase Theorem, 72 CORNELL L. REV. 245,
286-87 (1987).
122. This is not to say that the mimic-the-market rule would never be desirable. If, for instan
filling a particular type of contractual gap is an issue of first impression, it may be reasonable f
court to look at existing contracts as a guide to what the parties would have wanted. If parties w
unaware of the default rule when they were contracting, the cycling problem would not arise. If
well-established default exists, many contracting parties may explicitly contract for what they wan
order to avoid the penalty of ex post uncertainty. In this case existing contracts provide evidence f
what the parties would have done, so mimicking the market may be justified. For example, Epste
uses this approach to argue that one can ascertain the efficient form for workers' compensation leg
tion by observing the contractual insurance provisions that existed prior to legislation. Epstein, sup
note 64, at 118-19.
Mimicking the market may also make sense when parties have failed to record any contract. For
example, in devising their estates, many parties may not go to the trouble of writing a will, but those
that do may restate even well-established intestacy defaults. Thus, looking at actual wills can give
some guide to what the general populace wants. But even this argument fails if the contracting sample
misrepresents the intestate class. If the reason certain parties fail to contract is related to the substan-
tive provision that those parties want, then the inference between actual and hypothetical contracts is
attenuated. For example, if only upper-class people can afford to write wills, then an upper-class
preference for children over parents may not be relevant in determining the intestate preferences of

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
ting defaults that mimic the market therefore will not assure efficiency. A
slightly more sophisticated version of mimicking the market would set the
default at what most people would contract around another default for.
This approach would focus solely on the 3's. For example, if a larger
percentage of parties were willing to contract around the first default than
the second default ((2 > (1), then rule two would be chosen as the de-
fault. But maximizing the 3's suffers the same flaw as maximizing the
a's-neither choice rule conforms with the efficiency criterion in inequal-
ity (12).123
Implementing a complete theory of default choice requires attention to:
1) what the parties want (the a's),
2) whether they will get it (the d's), and
3) the costs associated with getting it (the c's) or not getting it (the f's).
It is especially important that lawmakers ascertain the degree of separat-
ing and pooling that each default engenders. For in determining the equi-
librium levels of the A associated with each default, the court must esti-
mate the importance of transactional and strategic barriers to contracting
around particular default rules, as well as understand the costs associated
with failing to strike the efficient contract.
We have shown that at times the efficient default will be one that a
majority of contracting parties disfavors. As the number of different types
of preferred contracts (and consequently the number of possible defaults)
increases, any untailored default is likely to be disfavored by the majority
of contractors.124 Untailored defaults act as penalty defaults with regard to
the lower class.

123. Still another partial and inefficient criterion for default choice would attempt to minimize the
number of people who failed to contract to their efficient rule. In the model's terminology, this would
be equivalent to minimizing ai - i.
124. The results of this dichotomous model can be extended to a situation in which courts are
choosing among more than two default alternatives. In many situations, there will be more than two
possible default choices from which the lawmaker may choose. Consider a class of contracts in which
tailoring is prohibitively expensive, so that the lawmaker must choose among N (N > 2) possible
defaults. Assume that there are different classes of contracting parties for which different rules would
be efficient. For example, heterogeneity across buyer risk preferences could lead to a range of optimal
risk-sharing rules for different contracts. We assume that when parties contract around a default, they
contract for the rule that is efficient for their class. But see Schwab, supra note 1, at 251 (students
sometimes contracted around default to less efficient outcome). A contracting party will be called a
type j contracting party if default rule j would be efficient for it. Let:
ai = the percent of contracting parties who want default i (where Xj ca = 1);
flij = the percent of type j contracting parties who will contract around default i (where flij = 0);
cij = the cost for type j parties of contracting around default i; and
fij = the cost for type j parties of failing to contract around default i (where fij = 0).
Extending the earlier model, we can say that the cost of any default i will be:
j=1,N [c.W. + f..(a. A-

1N iJ1 i 1Jl J iJ
which represents for all contracting types the total costs of contracting around default i or of fai
contract around it. A lawmaker choosing the least-cost default will accordingly want to minim
mmni Z ( c.. + f.(a.-
i1,N I j-=I,N [ 'j ij j J I
As in the dichotomous model, a penalty default may be the least costly. Even a de
wants, ax = 0, may be optimal if most people contract around the default (tx; '.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
the classes of contracting p

contracting types grows, any untailored defaults start to resemble the
types of penalty defaults we described in Section L."25 As the number of
possible defaults expands, courts must choose between a penalty default
that is efficient within the class of untailored defaults or a tailored default
that requires the court to ascertain what individual parties would have
wanted. We have analyzed how courts should find the optimal rule within
the class of untailored defaults; we now examine whether a tailored de-
fault is superior to an untailored default.
With a tailored default the court attempts to determine the default for
which the particular parties would have contracted. A major cost of tai-
lored defaults, then, is the cost of this ex post determination.'28 Instead of
having the different types of parties separating themselves into different
contractual groups ex ante, the court attempts the separation ex post. In-
stead of contracting costs, the costs of ex post tailoring are the costs of
distinguishing between types of contractual parties, where each type
would have contracted for a different rule.
These costs of determining what the particular parties would have con-
tracted for can be significant. For example, in Jordan v. Duff & Phelps,
Inc. ,127 two leading law-and-economics practitioners, Judges Easterbrook
and Posner, disagreed about what the particular parties would have
wanted. In Jordan the Seventh Circuit was called on to fill a gap in a
shareholder/employment agreement. The court needed to determine
whether the defendant corporation had a duty to disclose material infor-
mation about a merger to an employee who was about to quit and who
was, according to the agreement, thereby required to sell his shares. Pre-
dictably, both judges agreed that contractual gaps should be filled with
terms that the parties would have wanted.'28
The judges, however, strongly disagreed about what these terms would
have been. Judge Easterbrook, authoring the majority opinion, found it
"unwarranted to say that the implicit understanding between Jordan and
Duff & Phelps should be treated as if it had such a no-duty clause; we are
the costs of contracting around it are low (Cj 0, for all j).
For example, consider our earlier discussion of the zero-quantity default, supra text accompanying
notes 42-46. The default choice is non-dichotomous. Numerous quantities could be chosen as the
default. Although contracting parties would not contract to exchange zero quantity (x. = 0), most
people contract around this default, and the costs of contracting around it are low.
125. Any untailored quantity default would be, except for the smallest proportion of transactions,
a penalty default. For example, only the smallest percentage of contracting parties would actually
want a default quantity of some randomly chosen number such as, say, 39 or 2003.
126. Richard Epstein criticizes Judge Posner's analysis in EVRA Corp. v. Swiss Bank Corp., 673
F.2d 951, 954 (7th Cir. 1982), because contributory negligence requires costly ex post tailoring by
courts. Epstein, supra note 64, at 134.
127. 815 F.2d 429 (7th Cir. 1987).
128. Id. at 436; id. at 446-47 (Posner, J., dissenting).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
not confident that this is the clause firms

would prefer.""'
Judge Posner, in dissent, argued that the parties would have waived
any duty of disclosure because waiving such a duty would have "aligned
their respective self-interests better than the legal protection that the court
devises today."'30 This disagreement between sophisticated jurists suggests
that tailoring costs include not only the out-of-pocket litigation expenses
but also the potential costs of judicial error. Indeed, if the expected costs of
ex post tailoring are sufficiently large, even a tailored rule may give the
parties an incentive to contract explicitly ex ante. Ex post tailoring may be
more expensive to contracting parties than ex ante contracting. Even if
judicial tailoring is accomplished without error, parties may prefer to con-
tract for the same terms that the courts would provide at a higher cost ex
post. 131
But these costs of distinguishing between different types of parties are

not the only costs of tailored rules. Because courts do their tailoring after
the fact, tailored rules can actually exacerbate the inefficiency of strategic
incompleteness. With a tailored rule the relatively informed
party-assured that the court will provide the terms that fully-inform
parties would want-may have no incentive to reveal her private infor
tion to the other party. For example, in our Hadley model the high-
damage rule was a tailored default that gave each contracting type what
they would have contracted for were they fully informed-damages of DH
or DL, depending on the miller's type. Yet, we have shown conditions
under which this tailored high-damage default gives rise to the costs of
inefficient reliance or precaution.'32 By not explicitly contracting for dam-
ages ex ante, the carrier cannot know if the miller is high-damage or low-
damage and therefore may not choose the correct level of precaution. Even
if it is costless ex post for the court to determine exactly what two specific
parties would have agreed to ex ante, tailored defaults may not be
optimal.
In sum, finding the efficient default can involve a complicated inquiry.
Knowledgeable parties can leave holes in contracts for strategic rea-
sons-they might prefer to remain in an undifferentiated pool than pay
their full freight in an efficient but unsubsidized equilibrium. Efficiency-
minded lawmakers must therefore be attuned to the sources of contractual
incompleteness and to the attendant costs of pooling and separating associ-
ated with their default choice.
129. Id. at 436.

130. Id. at 448 (Posner, J., dissenting).
131. This point inverts Posner's observation that it may sometimes be "cheaper for the court to
'draft' the contractual term necessary to deal with the contingency if and when it occurs." R. POSNER,
supra note 2, at 82.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
III. A THEORY OF GAPS
The preceding sections have attempted to develop a theory for choosing

default rules. But before implementing any default standard, courts need
to establish, as a logically prior matter, rules for deciding when a contract
is incomplete.'33 Indeed, the litigants in many cases will argue not only
about how the gap should be filled but also about whether there is a gap
at all.'34
For example, returning again to Jordan v. Duff & Phelps,'35 Judges
Easterbrook and Posner disagree not only on how to fill a gap in the
employment-stockholding contract but also, and more fundamentally, on
whether there was a gap to be filled.'36 Easterbrook's majority opinion
concludes that although the fiduciary duty to disclose material information
could be contracted around, the parties' agreement did not sufficiently ne-
gate this duty: "[Tihe possibility that a firm could negotiate around the
fiduciary duty does not assist Duff & Phelps; it did not obtain such an
agreement, express or implied."5137 Easterbrook's decision then proceeds to
fill this disclosure gap with the fiduciary duty default. Posner, in dissent,
finds that the "stockholder agreement that defined [the employee's] rights
as a shareholder 'with greater specificity' "'38 sufficiently contracted
around any default fiduciary duty: "The arrangement that resulted (call it
'shareholder at will') is incompatible with an inference that Duff and
Phelps undertook to keep him abreast of developments affecting the value
of the firm."139 In Jordan, then, two prominent legal economists dis-
agreed not only about how the gap should be filled but also about what
constitutes a gap.
133. At an even more basic level, courts will not be able to determine whether a contract has gaps
without a prior theory of contract formation. In general, courts will need to determine:
1) whether the parties have formed a contract,
2) whether the contract has gaps, and
3) how the gaps should be filled.
We implicitly assumed in Sections I and II that there were sufficient objective indicia of the parties
meeting of the minds to infer contractual formation. See Barnett, A Consent Theory of Contract, 86
COLUM. L. REV. 269 (1986) (discussing competing theories of contractual formation). This Section's
focus on legal formalities should inform courts' theories of contractual formation and contractual gaps.
134. This point is also made in M. Freed & D. Polsby, Hard Cases Make Bad Law: Employ-
ment at Will at the Edge (1988) (unpublished manuscript on file with authors). For example, Freed
and Polsby criticize Richard Epstein's analysis, see Epstein, supra note 120, at 980, that the court
supplied the correct "gap-filling" default in Coleman v. Graybar Elec. Co., 195 F.2d 374 (5th Cir.
1952), an employment at will/wrongful termination case: "The contract in Graybar was anything but
silent . . . . Such an agreement could hardly 'provide otherwise' with more clarity, short of specifi-
cally reserving the employer's right to act in bad faith." M. Freed & D. Polsby, supra, at 2.
135. 815 F.2d 429 (7th Cir. 1987); see supra text accompanying notes 127-31.
136. Judge Posner argued in the alternative. Although he concluded that the contract did not have
a gap, id. at 449 (Posner, J., dissenting) ("the parties were not silent"), he alternatively found that
even if there were a gap, it should be filled differently than the way Easterbrook recommended. See
supra text accompanying note 129.
137. 815 F.2d at 436.
138. Id. at 447 (Posner, J., dissenting) (quoting majority).
139. Id.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
This question of when a contract is incom

tion of what is sufficient to contract arou
that the parties' attempt to contract aroun
is identical to a holding that there is still a gap in the contract. By an-
swering one question, the courts necessarily answer the other. Courts need
to develop a theory of gaps before they can address how best to fill them.
If the academy has been remiss in developing a theory of default choice,
then to an even greater degree it has failed to address what the necessary
and sufficient conditions for contracting around defaults should be. In de-
termining these conditions, courts are determining the costs of contracting
around a given default.140 The received wisdom that transaction costs are
responsible for contractual incompleteness implicitly suggests that
lawmakers should minimize the costs of contracting around defaults so
that if any contracting parties do not like the off-the-rack standard, they
can inexpensively tailor their corporate or contractual structure to suit
themselves.
But legal rules evince a greater diversity than this simple theory sug-
gests."4' Parties wishing to contract around both statutory and common-
law defaults will encounter varying degrees of difficulty. For example,
many sections of the U.C.C. establish defaults that are in force "unless
otherwise agreed,''l42 but ? 2-206 provides that contractual offers shall be
construed as inviting any reasonable manner of acceptance "[u]nless other-
wise unambiguously indicated,"'143 and the Official Comment punctuates
this point by saying that this default obtains "unless the offeror has made
quite clear that it will not be acceptable."1144
Similarly, in the corporate context the common law of promoter liabil-
ity requires extraordinary efforts to negate the joint and several liability
default. For example, in Stanley J. How & Assocs. v. Boss,145 a promoter
was held liable under a contract even though he signed the pre-
incorporation agreement: "By: Edwin A. Boss, Agent for a Minnesota
Corporation to be formed, who will be the Obligor."1146 The court held
that "[w]hile the agreement was not completely clear, the words 'who will
be the obligor' are not enough to offset the rule that the person signing for
the nonexistent corporation is normally to be personally liable."1147 Con-
140. These costs were represented as ci in the early model of default choice. See supra Section II.
141. But see Black, supra note 1 (arguing that corporate default rules are trivial). Black's argu-
ments imply that corporate formalities do not impede corporations from establishing tailored forms of
corporate governance. But the costs of contracting around some "strong" defaults (that is, fulfilling
certain corporate formalities) are likely to give some formalities a more substantive or nontrivial na-
ture. See, e.g., infra notes 148-50.
142. See, e.g., U.C.C. ? 2-303 (1976).
143. U.C.C. ? 2-206 (1976) (emphasis added).
144. Id. at Official Comment 1 (emphasis added).
145. 222 F. Supp. 936 (S.D. Iowa 1963).
146. Id. at 939.
147. Id. at 942. The holding is consistent with the common-law doctrine of construing contr

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
tracting around corporate statutes especially exhibits different degrees of

difficulty. For example, certain default rules of corporate governance can
be changed only in the articles of incorporation, while others can be over-
come through less costly bylaw amendments.148
At times even seemingly immutable rules can be circumvented at a cost.
To be sure, some methods of contracting around legal rules are better
described as examples of abuse of the law.149 But lawmakers at other
times seem to allow ostensibly immutable rules to be negated if the private
parties structure the transaction properly. For example, the seemingly im-
mutable rule that limited partnerships must have at least one general
partner (that is, one partner who retains unlimited personal liability) can
be contractually negated in jurisdictions that allow corporate persons to
serve as the sole general partner.150
A. Peevyhouse and the Default-Immutable Spectrum
As the cost of contracting around a default rule becomes extremely

large, the default starts to look like an immutable rule.151 The Oklahom
Supreme Court's classic decision in Peevyhouse v. Garland Coal & Min-
ing Co.152 exemplifies this insight. The Peevyhouses leased their land con-
templating that it would be strip-mined. The contract included specific
covenants that Garland would restore the land at the end of the lease
ambiguities against the drafter. See supra note 82.

148. Defaults that may be contracted around only in the articles of incorporation include: Board
of directors may be dispensed with entirely in limited circumstances or its functions may be restricted,
REVISED MODEL BUSINESS CORP. ACT ANN. ? 8.01 (1984); power to compensate directors may be
restricted or eliminated, REVISED MODEL BUSINESS CORP. ACT ANN. ? 8.11 (1984); special voting
groups of shareholders may be authorized, REVISED MODEL BUSINESS CORP. ACT ANN. ? 7.25
(1987); and classes of shares may be given more or less than one vote per share, REVISED MODEL
BUSINESS CORP. ACT ANN. ? 7.21 (1987). Defaults that may be contracted around in either the
articles of incorporation or the bylaws include: Number of directors may be fixed or changed within
limits, REVISED MODEL BUSINESS CORP. ACT ANN. ? 8.03 (1987); qualifications for directors may
be prescribed, REVISED MODEL BUSINESS CORP. ACT ANN. ? 8.02 (1987); shares may be issued
without certificates, REVISED MODEL BUSINESS CORP. ACT ANN. ? 6.26 (1987); power of board of
directors to amend bylaws may be restricted, REVISED MODEL BUSINESS CORP. ACT ANN. ?? 10.20,
10.22 (1987). See REVISED MODEL BUSINESS CORP. ACT ANN. ? 2.02 official comment.
149. The field of tax abounds with instances in which clever citizens abusively attempt to contract
around tax liabilities in ways that Congress did not foresee or intend. See, e.g., Heintz v. Comm'r, 25
T.C. 132 (1955) (stockholder of merging corporation contractually abrogates seemingly immutable
rule for capital gains realization rules); Zuckman v. United States, 524 F.2d 729 (Ct. Cl. 1975)
(stockholders of corporation contractually abrogate seemingly immutable rules to carry through
losses).
150. In Delaney v. Fidelity Lease Ltd., 526 S.W.2d 543 (Tex. 1975), the Texas Supreme Court
voided this contractual attempt to circumvent the general-partner personal-liability rule. The Texas
legislature, however, amended its corporations statute to permit a corporation "[t]o be an organizer,
partner, member, associate, or manager of any partnership. TEX. BUS. CORP. ACT ANN. art.
2.02A(18) (Vernon 1989).
151. Conversely, if we admit the possibility of lobbying to change current law, even immutable
rules are really defaults where the costs of lobbying to change them are prohibitively expensive. This
point was made by Joseph Grundfest at the Columbia conference on December 9-10, 1988, supra
note 13.
152. 382 P.2d 109 (Okla. 1962), cert. denied, 375 U.S. 906 (1963).

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
period. As it turned out, the restoration "would involve the moving of

many thousands of cubic yards of dirt, at a cost estimated by expert wit-
nesses at about $29,000.00.''153 Garland introduced evidence that such
restoration would increase the market value of the land by only $300.151
The court was asked to choose the proper measure of damages: cost of
performance ($29,000) or diminution in value ($300).
The majority opinion, which limited the Peevyhouses' damages to dimi-
nution in value, clearly claimed that the court's ruling could be contracted
around.
It should be noted that the rule as stated does not interfere with the
property owner's right to "do what he will with his own," or his
right, if he chooses, to contract for "improvements" which will actu-
ally have the effect of reducing his property's value. Where such re-
sult is in fact contemplated by the parties, and is a main or principal
purpose of those contracting, it would seem that the measure of dam-
ages for breach would ordinarily be the cost of performance.155
This passage seems to emphasize that, notwithstanding the holding in this

particular case, parties retain the right to contract around the diminution-
in-value default standard.
The dissenting opinion, however, points to the specific provisions of the
contract that had been added to the usual covenants of the defendant's coal
mining leases and that seemed to place the duty of restoration on the de-
fendant. In the face of these contractual provisions, the reader of
Peevyhouse is left with two alternative interpretations. Either the major
opinion is 1) disingenuously creating an immutable rule; or 2) creating a
"strong" default rule (that the Peevyhouses' amendments did not suffi-
ciently override).156 The most straightforward way to distinguish between
these competing interpretations is to ask what extra words the
Peevyhouses could have added to the contract to "reverse" the majority's
153. Id. at 111.

154. Many commentators have criticized the majority's opinion for ignoring the subjective
market) value of the land to the Peevyhouses. See, e.g., Muris, Cost of Completion or Diminut
Market Value: The Relevance of Subjective Value, 12 J. LEGAL STUD. 379 (1983) (arguing
courts must recognize subjective value to fill compensatory goal of contract law). However, the c
reliance on market value is probably more a function of the plaintiffs' trial strategy. The Peevy
in trying to focus the jury's attention on cost of performance, may have strategically decided n
introduce any evidence of subjective value. And the court's opinion seems to indicate that eviden
subjective value should be included in diminution damages: "After a careful search of the reco
have found no evidence of a higher figure [than the $300 market diminution], and plaintiffs
argue in their briefs that a greater diminution in value was sustained." Peevyhouse, 382 P.2d
One would predict that in subsequent litigation plaintiffs would begin to introduce subjectiv
evidence that would be included in diminution damages.
155. Peevyhouse, 382 P.2d at 114 (citations omitted).
156. The majority's own use of the word "ordinarily" in the last quoted sentence seems to add
some support for the immutability interpretation. After all, is the court suggesting that even if the
parties explicitly attempt to contract for cost of performance, there still may be "extraordinary" situa-
tions in which the diminution in value standard will apply?

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
default. If the majority is really doing what it says, we should be able to

write such a contract. If we cannot confidently determine what such a
contract would need to say, the alternative hypothesis-that Peevyhouse
really is creating an immutable rule-becomes more compelling.157
While the Peevyhouse opinion has been criticized for reaching the
wrong result,158 this analysis suggests that the majority also erred by not
more clearly establishing what words would be sufficient to contract for
the cost of performance damages. Even though the decision (rightly or
wrongly) resolved uncertainty about what the default damages would be,
it did little to resolve the uncertainty about how one could contract around
this default. Even prospective parties who had read Peevyhouse and had
known that it "does not interfere with the property owner's right to 'do
what he will with his own' "159 would still face considerable uncertainty
about how to exercise that right.
This discussion suggests that in many instances decisions should do
more than merely decide what the efficient default should be; they should
establish "safeharbors" of contractual language that will be sufficient to
reach other contractual outcomes.160 As the Peevyhouse case illustrates, by
giving prospective parties examples of explicit language, courts can dra-
matically reduce the uncertainty and costs of contracting around the spe-
cific default.""1
B. Determining the Efficient Level of Legal Formalities
There can be good economic reasons for "strong" defaults in which the
courts intentionally increase the procedural costs to parties of contracting
around the default. These reasons parallel the rationales given in one of
the first intuitively economic analyses of contract law, Lon Fuller's classic,
Consideration and Form.'62 Fuller suggested that legal formalities serve
157. Tracking the majority's language, we suggest that prospective parties attempting to overcome
the diminution-in-value standard include the following covenant:
The parties specifically intend and contemplate that the lessee shall restore the land, even if
the costs of performance are grossly disproportionate to the diminution in value from failing to
restore the land. This is a main and principal purpose of the lease. In writing this provision,
we are explicitly contracting around the holding of Peevyhouse v. Garland Coal.
If this provision were not sufficient, the Peevyhouse rule would, in fact, be immutable.
158. See, e.g., Birmingham, Damage Measures and Economic Rationality: The Geometry of Con-
tract Law, 1969 DUKE L.J. 49.
159. Peevyhouse, 382 P.2d at 114.
160. Courts may face the constitutional limitation of resolving only "cases or controversies" at
issue. U.S. CONST. art. III, ? 2, cl. 1. But other decisions have established prospective safeharbors.
See, e.g., Miranda v. Arizona, 384 U.S. 436 (1966) (establishing fair and effective warning require-
ment to persons in police custody).
161. The purpose of such contractual safeharbors should not be to preclude the parties from tai-
loring other standards. The safeharbor alternatives might additionally provide benefits as common-law
interpretations would more fully specify their meanings. See J. Gordon, supra note 13
162. Fuller, Consideration and Form, 41 COLUM. L. REV. 799 (1941). In truth, the fact that the
article is claimed by economists (it is, for example, reprinted in A. KRONMAN & R. POSNER, THE
ECONOMICS OF CONTRACT LAW 40 (1979)) is as much a tribute to the soundness of its insights as to
its nexus with economics.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
evidentiary, cautionary and channelling fun

ficient conditions for contracting around a
ities that courts require. Our discussion accordingly tracks Fuller's
categories.
In Fuller's analysis, the evidentiary function of legal formalities is to
provide information to courts in order to lower the costs of subsequent
decision making. This analysis is quite resonant with our earlier conclu-
sion that penalty defaults could be justified on efficiency grounds as ways
of encouraging the revelation of information.163 In choosing the necessary
conditions (the formalities) for contracting around a default, a court
should consider similar informational aspects. We go beyond Fuller
though and argue that the evidentiary function of legal formalities can
serve to inform not only the courts but the parties within the contract as
well.'64 For example, if a penalty default is chosen to encourage one party
to reveal information to another, the court may want to regulate the pro-
cess of contracting around the default so that meaningful information is
conveyed.
Structuring formalities to facilitate internal information flows is also
related to the cautionary function of legal formalities. Fuller argues that
some formalities, such as certain writing requirements, serve a cautionary
function by forcing the parties to undertake a minimal amount of reflec-
tion before being bound by a contract. But, as already discussed, when the
parties to a contract have disparate amounts of information, lawmakers
may want to establish formalities that are more directed toward protecting
the relatively uninformed. For example, a holding contrary to Peevyhouse,
in which the courts not only choose cost-of-performance damages as the
default but also require extremely explicit language to opt for value dimi-
nution damages, might be justified to caution the relatively uninformed
landholders of their legal rights.'65 To caution is to give information. At-
tention to how legal formalities will affect the flow of information should
inform lawmakers' theory of gaps as well as their theory of default choice.
Finally, Fuller suggested that legal formalities could serve a chanelling
function which, for example, would allow parties to channel their contrac-
tual agreements toward legal or non-legal enforcement. Fuller's theory of
channelling is highly analogous to our earlier discussion of pooling and
especially separating equilibria.'66 As we discussed in our Hadley model,

165. Similarly, if strategic withholding of information explains why employers do not reveal to
employees the true probability of future termination, employees may make inefficient investments in
goods such as housing and human capital. In such situations it may be efficient for courts not only to
make a default rule against termination without cause but also to require employers to state specifi-
cally that they wish to be able to fire for even arbitrary reasons. But see Epstein, supra note 120
(arguing against such a rule); M. Freed & D. Polsby, supra note 134

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
different defaults can lead

chanelling, the separating effect of legal defaults may allow parties to sort
themselves into different groups.
There may be situations in which courts should increase the costs of
contracting around defaults to force the majority of parties into a particu-
lar channel. For example, if a certain type of contract generates a mild
externality, we may want to discourage most people from entering this
type of contract. But if there is a small class of contracting parties that
will be damaged more than society gains if they fail to contract, then
courts may want to consider a method of "chanelling" or "separating" the
socially beneficial from the deleterious classes of transactions.
One way to do this would be by direct fiat, but there may be ways of
encouraging the parties to sort themselves into the correct categories
through legal rules. For example, by artificially increasing the costs of
contracting around a no-contract default, courts may discourage the low-
benefit contractors but not the high-benefit contractors. Although such le-
gal formalisms may engender the efficient types of contracting, they may
still be inefficient because the additional transaction costs that the formali-
ties engender are a social cost that could be avoided by simply imposing a
tax of an equal magnitude. The tax would be a transfer to the government
instead of a dead-weight loss. If courts are constrained from imposing
taxes,'68 however, intentionally costly formalisms that increase the cost of
contracting around "strong" defaults may be the most efficient rules
within their choice set. The channelling function of legal formalities can
also be given an informational justification because the very process of
self-sorting reveals information about the parties' preferences.'69
Such intentionally costly formalism would not be used in conjunction
with a penalty default. Since the whole point of penalty defaults is to
encourage parties to contract around them, formalisms that increase the
cost of doing so would be counterproductive. In essence, penalty defaults
encourage, and "strong" defaults discourage, contractual mutation.
C. Legal Responses to Contracting Around Immutable Rules
An important difference between default and immutable rules is that if

parties attempt to contract around a default rule and fail, they will simply

168. See, e.g., I. Ayres & P. Schmitt, Court-Ordered Funding of School Desegregation Remedies:
Federalism Versus Minority Rights ? 3 (1989) (unpublished manuscript on file with authors) (dis-
cussing constitutional limits on judicially imposed taxation).
169. The possibility that courts may intentionally want to increase the costs of contracting around
a default might potentially explain their refusal to articulate alternative safeharbors. If there is uncer-
tainty about what is sufficient to overcome a particular default, the parties will be discouraged ex ante
from even trying. This argument might provide an apology for the Peevyhouse holding. However, in
Peevyhouse the societal externalities of strip-mining militate toward a cost-of-performance default.
This channeling explanation is therefore inapposite to the Peevyhouse fact pattern.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
be bound by the default, whereas if partie

immutable rule and fail, the law may choose to penalize the attempt by
imposing a penalty different from (and, from the parties' ex ante perspec-
tives, worse than) the immutable standard. From an ex ante perspective
the possibility of receiving this ex post penalty is just another expected
cost of contracting around the default rule.
The legal response to parties who try to contract around immutable
rules can also be given a default interpretation. In these cases the courts
remove the clauses that transgress the immutable rule and then choose a
default to fill the gap. For example, in Frostifresh Corp. v. Reynoso,'70
the trial court established an immutable rule that retailers cannot sell at
unconscionably high markups. Accordingly, the court struck out the con-
tract's price provision (which was more than three times the retailer's
cost), but it then had to decide how to fill the gap. The trial court decided
that the reconstructed contract should have a price which only covered the
seller's cost. The appellate court reversed and awarded the seller "a rea-
sonable profit" instead of zero profit.'7' The difference in these holdings is
very similar to the tension between a tailored default and penalty default.
Reconstructing the contract to give the seller a reasonable profit is identi-
cal to the U.C.C.'s reasonable-price standard. The zero-profit standard is
analogous to a penalty default, in this case imposing the penalty on the
seller for trying to contract around a rule, while the penalty defaults dis-
cussed earlier penalize parties for not contracting around the rule.
This same tension arises with covenants not to compete. Courts have
established an immutable rule that parties cannot make covenants of un-
reasonable duration. In Fullerton Lumber Co. v. Torborg,'72 for example,
the Supreme Court of Wisconsin held that a ten-year covenant was unrea-
sonable but then struggled in deciding whether it should reformulate the
contract to impose a duty not to compete for a reasonable period or penal-
ize the employer for transgressing the immutable limitation by allowing
the former employee to compete immediately.'73
If the goal of an immutable rule is to discourage people from even at-
tempting to contract around a provision, then it would seem that the pen-
alty reconstruction would be the favored result.'74 This is again analogous
170. 52 Misc. 2d 26, 274 N.Y.S.2d 757 (Dist. Ct. 1966), rev'd 54 Misc. 2d 119, 281 N.Y.S.2d
964 (App. Term 1967).
171. Frostifresh Corp. v. Reynoso, 54 Misc. 2d 119, 281 N.Y.S.2d 964 (App. Term 1967).
172. 270 Wis. 133, 70 N.W.2d 585 (1955).
173. The majority reconstructed the contract to have a covenant of "reasonable" duration. 270
Wis. at 146-47, 70 N.W.2d at 592. The dissent would have penalized the employer by allowing the
employee to compete immediately. Id. at 148-52, 70 N.W.2d at 593-94 (Gehl, J., dissenting).
174. A penalty for even attempting to contract around a legally immutable provision may b
necessary to deter attempts in a world where some disputes are not litigated. If all disputes were
litigated, attempts to contract around an immutable rule would never succeed. But if some contractual
parties fail to challenge unenforceable immutable rules, then imposing a penalty whenever a case does
come to court might be justified to prevent underdeterrence. For example, it is widely believed tha

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
to our earlier discussion of penalty defaults, in which we suggested that

courts should choose the penalty that provided "least cost deterrence.""17
The difference is that in the earlier discussion the penalties were attempt-
ing to deter gaps and here the penalties are attempting to promote gaps
(that is, deter contracting around). The preference for penalty defaults to
fill the gaps left in unconscionable contracts may justify the common law
"blue pencil" test which simply enforces contracts after the offending pro-
vision has been struck (with a blue pencil).'76
CONCLUSION
This article has suggested how efficiency-minded lawmakers should go

about filling gaps in contracts. The reasons that parties leave gaps in con-
tracts should strongly inform this decision. Our theory of defaults should,
in a sense, be guided by our theories of why there are contractual gaps.
Prior theorists have argued that parties leave gaps in contracts because the
cost of writing additional terms outweighs the benefit. Accordingly, they
have suggested that courts should simply fill in the gap with the term the
parties "would have wanted."
This article, however, has articulated a second cause of contractual in-
completeness. We have shown that when one party to a contract knows
more than another, the knowledgeable party may strategically decide not
to contract around even an inefficient default. Because the process of con-
tracting around a default can reveal information, the knowledgeable party
may purposefully withhold information to get a larger piece of the smaller
contractual pie. This possibility of strategic incompleteness leads us to em-
brace more diverse forms of default rules. In particular, lawmakers may
be able to undercut the incentives for this strategic rent-seeking by estab-
lishing penalty defaults that encourage the better informed parties to re-
veal their information by contracting around the default.
Our analysis does not imply that penalty defaults should be used in all
contractual settings. The decision by efficiency-minded courts or legisla-
tures to impose a penalty, untailored majoritarian, or tailored "what the
parties would have wanted" default in a particular setting is not a trivial
one. The first step in the decision-making process should be to ask "why
does the gap exist?" We have suggested that parties may fail to contract
around inefficient defaults for strategic as well as transaction cost reasons.
When parties fail to contract because they want to shift the ex ante trans-
creditors and residential landlords make a practice of including all sorts of illegal clauses in their
contracts not because they think the clauses will stand up in court, but because they know that most
debtors and tenants have more respect for written contracts than most courts do. The authors are
indebted to conversations with Richard Craswell for this point. See Craswell, supra note 22, at 64-65.
175. See supra Section I.B.
176. See Fullerton, 270 Wis. at 43, 70 N.W.2d at 590. The "blue pencil" test is the traditional
test of severability. If, after removing the unconscionable terms, the contract is still comprehensible in
that the parties might have still entered into it, the court will enforce it.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
action cost to a subsidized ex post court det

of non-enforcement may be appropriate. When strategic considerations
cause a more knowledgeable party not to raise issues that could improve
contractual efficiency, a default that penalizes the more informed party
may encourage the revelation of information.'77
Lawmakers should not, however, impose penalty defaults indiscrimi-
nately. Even if strategic considerations are established as a significant
source of contractual incompleteness, courts or legislatures should consider
the costs and benefits of penalty defaults themselves. Since the goal of a
penalty default is to induce information revelation, lawmakers should con-
sider the likelihood that the penalty will in fact result in information be-
ing revealed,'78 the benefit (in more efficient reliance or precaution) of the
revealed information, and the costs of explicitly contracting around the
default. If the private information is acquired with economic resources,
the value of information revelation must also be weighed against the pri-
vate incentives to acquire it. Penalty defaults are therefore more likely to
be efficient if the private information is acquired passively. In sum, a pen-
alty default should be used if it results in valuable information revelation
with low transaction costs.'79
Throughout the discussion we have cited examples of common-law and
statutory rules that are broadly consistent with our theory's categories of
penalty, tailored and untailored defaults. But we believe that both courts
and legislatures should be more sensitive to the process of contracting.
Lawmakers should consider explicitly the informational as well as the
contractual equilibria generated by alternative default rules. Different de-
faults may generate different degrees of pooling and separating. When
heterogeneous contracting parties separate themselves into different con-
177. In many settings, such as simple contracts for small transactions, it is easy to point to trans-
action costs as the source of incompleteness. In other settings, however, parties write complicated,
lengthy contracts that are carefully considered by both parties and their lawyers. In these situations a
contingency that is not contracted for may be sufficiently important that it is unreasonable to ascribe
incompleteness to transaction costs. Some omitted clauses, for example, may be very inexpensive to
include from a transaction cost perspective, such as liquidated damage clauses or non-refundable de-
posits. In such cases courts will find it difficult to point to transaction costs as the source of incom-
pleteness, so penalty defaults should be considered.
178. For example, we suggested that Goldberg's zero-damage penalty default would not suffi-
ciently encourage retailers to contract for liquidated damages, and even if a larger penalty would, the
amount of liquidated damages in equilibrium might "pool" at a non-informative low level. See supra
Section I.C.2. Courts should also consider the possibility that some parties will fail to contract around
penalty defaults out of ignorance or oversight.
179. The Hadley rule is an example of a penalty default that 1) is cheap to contract around by
including a liquidated damage clause, 2) will likely cause damage information to be revealed, and 3)
will facilitate more efficient precaution.
We must add an important transitional caveat. A legal change from one default to another can be
costly-especially if the move is to a penalty default. Until parties become informed about the new
default, there may be transitional costs as the parties continue to bargain in the shadow of an invalid
law.
Finally, we note that in many contractual settings, both parties may have private information which
they choose to withhold. In such settings of "dual asymmetric information," penalty defaults may not
be sufficient to ensure that all private information is revealed.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
tractual agreements, they r

at the cost of increased con
of pooling and separating in
tally, they need to be able
different rules are likely to
The analysis of the article
range of legal issues.'80 For
often called upon to fill ga
tion-as-contract" approach to resolve such questions.'8' Jon Macey has
proposed a type of penalty default by suggesting that statutory gaps be
filled with public-regarding legislation.'82 Macey's default is not what the
legislature would have contracted for and thus forces the legislature to
make their pork-barrel deals public. Justice Scalia has gone the furthest,
however, in articulating a penalty default standard of statutory interpreta-
tion.'83 In Agency Holding Corp. v. Malley-Duff & Assoc. ,84 the Su-
preme Court was asked to supply a default statute of limitations for a civil
RICO186 action because the legislation was silent on the issue. While the
majority established a four-year limitation as what Congress would have
wanted, Justice Scalia argued that there should be no statute of limitation.
Scalia was "unmoved by the fear that [this] conclusion might prove 're-
pugnant to the genius of our laws' "186 and even suggested that this pen-
alty default might encourage Congress to contract around it: "Indeed, it
might even prompt Congress to enact a limitations period that it believes
'appropriate,' a judgment far more within its competence than ours.''l87
Ironically, Justice Scalia is arguing for an extreme form of judicial activ-
ism. Only by choosing a default that is wholly at odds with Congressional
intent can the Court ensure that Congress will reveal its intent. Scalia's
ultimate message then is that extreme activism can become a form of dis-
empowering judicial restraint. Examples such as these underscore our pri-
mary thesis that efficient defaults may at times deviate from what the
parties would have wanted. More importantly, they demonstrate that our
180. See discussion supra note 10.

181. See Easterbrook, The Supreme Court, 1983 Term-Foreword: The Court and the Eco-
nomic System, 98 HARV. L. REV. 4, 15-16 (1984).
182. Macey, Promoting Public-Regarding Legislation through Statutory Interpretation: An In-
terest Group Model, 86 COLUM. L. REV. 223, 236-40 (1986).
183. For a more general discussion of using penalty defaults to force Congress to enunciate the
law, see L. Marshall, Let Congress Do It: The Case For an Absolute Rule of Statutory Stare Decisis
(1989) (unpublished manuscript on file with authors).
184. 483 U.S. 143, 157 (1987) (Scalia, J., concurring).
185. Racketeer Influenced and Corrupt Organizations Act (RICO), 18 U.S.C. ? 1964 (1982 &
Supp. V 1987).
186. 483 U.S. at 170 (Scalia, J., concurring) (quoting Occidental Life Ins. Co. v. EEOC, 432
U.S. 355 (1977) quoting Adams v. Woods, 6 U.S. (2 Cranch) 336, 342 (1805)).
187. 483 U.S. at 170.

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC
choice of default must be informed by an understanding of why con-

tracting parties-such as legislatures-leave gaps in their texts.'88
Finally, we suggest that teachers focus more explicitly on the difference
between default and immutable rules. Often students are able to recite the
various rules learned in a course without knowing whether they are obli-
gatory or not. Teaching which rules are defaults is not a mere pedagogical
conceit. To represent their clients effectively, attorneys need to know not
only what the legal rules are, but how, if at all, they can be abrogated to
further their clients' interests.'89 A descriptive knowledge of defaults and
how to contract around them is a prerequisite of effective advocacy.
188. For example, the statute of limitation gap of RICO in Malley-Duff is hard to explain as
hidden pork-barrel legislation.
189. In corporate law especially, the mechanism for contracting around the default may be ob-
lique. For example, while corporate statutes do not give the board of directors the right to disapprove
a merger, certain "poison pill" plans have the effect of making the board sign off on any hostile bid.
See Dynamics Corp. of Am. v. CTS Corp., 805 F.2d 705 (7th Cir. 1986) (analyzing economic impact
of "poison pill").

200.89.140.130 on Mon, 12 Oct 2020 12:51:27 UTC

Ayres y Gertner para Traducir

Uploaded by

Copyright:

Available Formats

Ayres y Gertner para Traducir

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ayres y Gertner para Traducir

Uploaded by

Copyright:

Available Formats

Filling Gaps in Incomplete Contracts: An Economic Theory of Default Rules

Author(s): Ian Ayres and Robert Gertner

Stable URL: https://www.jstor.org/stable/796722

This content downloaded from

Ian Ayrest and Robert Gertnertt

This content downloaded from

require that stockholders elect directors ann

This content downloaded from

whether in particular contexts parentalistic concerns or externalities are

This content downloaded from

indeed, of what this particular tenant would have contracted for.

This content downloaded from

default and immutable rules, a detailed theory of how defaults should be

This content downloaded from

cumstances" of the individual contracting parties.27 In contrast, Charles

This content downloaded from

include legal fees, negotiat

This content downloaded from

This Article also proposes a second source

This content downloaded from

contractual relationships. In "pooling" equilibria, different types of con-

I. PENALTY DEFAULT RULES

A. The Zero-Quantity Default

The diversity of default standards can even be seen in contrasting the

This content downloaded from

U.C.C. refuses to enforce the contract.43

This content downloaded from

The U.C.C.'s zero-quantity default is what we term a "penalty de-

B. Toward a More General Theory of Penalty Defaults

Penalty defaults, by definition, give at least one party to the contract an

This content downloaded from

poration.6? Statutes that refuse to enforce

This content downloaded from

This content downloaded from

information about the default, but also bec

This content downloaded from

C. Penalty Defaults in Action

1. Uncompensated Damages in Hadley v. Baxendale

An example of how a penalty default can restrict rent-seeking behavior

This content downloaded from

the information to the carrier will undoub

This content downloaded from

It is important to consider another channel for information revelation.

69. See supra note 66.

This content downloaded from

relatively informed parties strategically wi

2. Goldberg's Solution to the Lost-Profits Puzzle

Victor Goldberg's analysis of the lost-volume retail seller74 comports

This content downloaded from

3. Unilateral Defaults as Penalties: The Perverse Incentives of

Some contractual rules establish defaults that individual parties may

In Le/kowitz v. Great Minneapolis Surplus Store, Inc.,84 the Minne-

This content downloaded from

fur coats " 'Worth to $100.00' " for " '$1 Ea

Penalty defaults stand as stark counter-examples to the proposition th

86. Id. at 189, 86 N.W.2d at 690.

This content downloaded from

This content downloaded from

II. A GENERAL THEORY OF DEFAULT CHOICE