Essay: A Bargaining Power Theory of Default Rules

\\server05\productn\C\COL\109-2\COL204.
txt unknown Seq: 1 23-FEB-09 11:10
ESSAY
A BARGAINING POWER THEORY OF DEFAULT RULES
Omri Ben-Shahar*
This Essay explores the merits of a new criterion for default rules in
incomplete contracts: filling gaps with terms that are favorable to the party
with the greater bargaining power. It argues that some of the more common
gaps in contracts involve purely distributive issues, such as the contract
price, for which it is impossible to choose a unique, joint-maximizing, “most
efficient” term. Instead, the term that mimics the hypothetical bargain in
these settings must be sensitive to the bargaining power of the parties—the
term they would have chosen to divide the surplus in light of their relative
bargaining strengths. This Essay explores the justifications for such a bar-
gain-mimicking principle, the ways in which it could be implemented by
courts, and the subtle ways it is already in place.
INTRODUCTION
How to fill gaps in incomplete agreements is perhaps the most im-
portant question in contract law. It is important both because courts
often interpret and supplement contracts and because the default rules
set by law determine how contracts will be written. One of the greater
successes of the economic approach to contracts is the development of
systematic ways to think about gap filling.1 The most broadly accepted
principle of gap filling is that courts should “mimic the parties’ will.”2
Under this principle, only gap fillers that mimic what the parties them-
selves would have chosen are allowed to remain in place and survive opt-
out, thereby eliminating unnecessary drafting costs. Of course, the no-
tion of the parties’ will is hypothetical. Because the contract contains a
gap, we do not know what they would have consented to. Here, the eco-
nomic approach provides another powerful insight: The parties’ will is to
have the most efficient arrangement. That is, they are best served by de-
* Frank and Bernice Greenberg Professor of Law, University of Chicago Law School;
omri@uchicago.edu. I am grateful to Ian Ayres, Richard Brooks, Clay Gillette, Bob
Hillman, Ariel Porat, Peter Siegelman, Kathy Zeiler, and workshop participants at the
Universities of Amsterdam, Chicago, Cornell, Duke, Michigan, and NYU for helpful
suggestions. Financial support from the Olin Center at the University of Michigan Law
School is gratefully acknowledged.
1. See Ian Ayres, Valuing Modern Contract Scholarship, 112 Yale L.J. 881, 890–92
(2003) (summarizing contribution of economic analysis to theory of default rules).
2. Richard Craswell, Contract Law: General Theories, in 3 Encyclopedia of Law and
Economics 1, 3–4 (Boudewijn Bouckaert & Gerrit De Geest eds., 2000); see Ian Ayres,
Default Rules for Incomplete Contracts, in 1 The New Palgrave Dictionary of Economics
and the Law 585, 586 (Peter Newman ed., 1998).
396
\\server05\productn\C\COL\109-2\COL204.txt unknown Seq: 2 23-FEB-09 11:10
2009] BARGAINING POWER THEORY OF DEFAULT RULES 397
fault rules that maximize the contractual surplus.3 The idea that gap fill-
ers should maximize the contractual surplus is based on the following
well-known logic. Assuming parties are rational, they would have agreed
upon terms that maximize their joint surplus, irrespective of the distribu-
tive impact of such terms. True, such terms might be more favorable to
one side, but the parties would have corrected for any distributive effects
by appropriately adjusting the contractual price or another purely distrib-
utive term.4 But for this theory to be valid, it must assume that there is at
least one contract term that the parties use to make the appropriate dis-
tributive adjustments—usually the price term—to which the theory does
not apply. The content of the purely distributive terms is not determined
by the surplus-maximizing criterion; it is surplus neutral. Rather, the con-
tent of the purely distributive term is determined by the bargaining
power of the parties. In other words, the surplus-maximizing conception
of gap filling is, by definition, insufficient to resolve all gaps because it
does not resolve gaps in the price term or in any other contract term that
is purely distributive. Thus, there is a troubling paradox surrounding the
basic criterion of gap filling. It assumes that the parties’ joint will exists—
that there is a single term such that, if only the parties spent the time and
attention dealing with the gap, they would have jointly desired the sur-
plus-maximizing term. Yet the existence of a gap in a contract is often an
indication that a consensus could not be reached because a single jointly
preferable term does not exist.5 If the parties’ interests had coincided,
they would have been able to agree on a term. But when the issue is
distributive, the parties’ interests are in conflict, and it is this divergence
of interests that leads to the gap. Ironically, as I will show later, many of
the cases used in contracts casebooks to introduce the topic of indefinite-
ness and gap filling involve purely distributive gaps over issues such as
3. See E. Allen Farnsworth, Contracts 486 (4th ed. 2004) (noting courts may provide
terms “that an economist would describe as maximizing the expected value of the
transaction”); Richard A. Posner, Economic Analysis of Law 99 (7th ed. 2007) (“[C]ontract
law cannot readily be used to achieve goals other than efficiency, as a ruling that fails to
interpolate the efficient term will be reversed by the parties in their subsequent dealings.”);
see also Frank H. Easterbrook & Daniel R. Fischel, The Economic Structure of Corporate
Law 21–22 (1991) (stating that gap fillers must duplicate terms that optimally promote
parties’ interests); Mark P. Gergen, The Use of Open Terms in Contract, 92 Colum. L. Rev.
997, 1064–72 (1992) (asserting that default rule should be a joint maximization rule); Alan
Schwartz & Robert E. Scott, Contract Theory and the Limits of Contract Law, 113 Yale L.J.
541, 554 (2003) [hereinafter Schwartz & Scott, Contract Theory] (“Parties jointly choose
the contract terms so as to maximize the surplus, which the price may then divide
unequally.”).
4. See, e.g., George L. Priest, A Theory of the Consumer Product Warranty, 90 Yale
L.J. 1297, 1313 (1981) (pointing out that disclaimers of warranty result in price
adjustments); Schwartz & Scott, Contract Theory, supra note 3, at 554 (“Bargaining power R
instead is exercised in the division of the surplus, which is determined by the price term.”).
5. See, e.g., Omri Ben-Shahar, “Agreeing to Disagree”: Filling Gaps in Deliberately
Incomplete Contracts, 2004 Wis. L. Rev. 389, 399–405 [hereinafter Ben-Shahar, Agreeing
to Disagree] (arguing that gaps in contracts are often created deliberately when parties fail
to agree on a negotiated provision).
398 COLUMBIA LAW REVIEW [Vol. 109:396
price, for which the prescription “choose the terms that maximize the
total surplus” does not provide a definite solution. For example, in
Oglebay Norton v. Armco, two large companies had a long-term relational
contract for transportation of iron ore, but ended up in a bitter legal
dispute about the price.6 Their agreement originally had a price
formula, but over time this formula failed and needed to be revised.
When the parties turned to the court to help fill the price gap, there was
no single term that reflected the “market price” which the court could
invoke.7 Indeed, if there were such a price, the parties would not have
needed the court to supply it. Of course, there was no “surplus-maximiz-
ing” price to fill the gap because price is surplus-neutral. Instead, the
court had to supply a “reasonable” gap filler that was purely distributive.
It ended up doing so by splitting the difference in a creative and unortho-
dox manner, forcing the companies’ CEOs to meet and mediate the fu-
ture price.8 And yet, the difficulty the court encountered and the ad hoc
solution it found merely emphasize the absence of a systematic and ra-
tional criterion for filling such gaps. The purpose of this Essay is to begin
developing a systematic new gap-filling criterion for these distributive
price gaps.
The proposed criterion, which I label the “bargain-mimicking” gap
filler, is consistent with the fundamental norms of mimicking the parties’
will. In cases of purely distributive terms there is no joint will. Instead,
each party’s will is to have a term at the more favorable end of a range of
reasonable terms. What courts need, then, is more information about
how parties would have resolved their a priori conflict of wills. Specifi-
cally, a court needs information that would help it mimic the bargain: the
division of surplus that the parties would have struck given their relative
bargaining powers. Since the division of bargaining power between two
parties may be uneven, the gap filler in these situations would be differ-
ent than the midrange “market” term. When one party has greater bar-
gaining power, the gap filler should tilt to favor this party, because this is
the party whose will would have more likely prevailed if an explicit bar-
gain were struck. Purely distributive gaps, therefore, would be filled with
terms more favorable to the party with greater bargaining power.
At first blush this criterion might seem unfair. As a normative crite-
rion, it is counterintuitive. Uneven bargaining power is hardly a desirable
phenomenon; why then should it be mimicked? I will defend its more
subtle appeal later in this Essay, but the core claim is perhaps less objec-
tionable than it initially seems and can be illuminated with a nonlegal
example. Take, for example, an incomplete command issued by a parent
to a child to “mow the lawn.” If imperfectly specified, it needs supple-
mentation—“only the front yard” or “both the front yard and the back
6. 556 N.E.2d 515, 518–20 (Ohio 1990).

7. The court found that any price between $5.00 and $7.44 per gross ton was a
reasonable rate. Id. at 520.
8. Id. at 518.
yard”—and more than one reasonable version could be offered. Still, if

the parent has the “bargaining power”—the power to dictate the exact
scope of the command—then the precise term that ought to be followed
is the one that is consistent with the meaning intended by this stronger
party. The parent can reasonably say to the child: “You knew or should
have known what I meant,” implying that the parent’s will, by virtue of
her greater bargaining power, is the controlling source of interpretation.
In fact, such interpretive method would render it unnecessary for the par-
ent to be more explicit in her command, and there are benefits to using
minimal language, such as saved transaction costs and the ability to utilize
simple, generic templates of transactions. To be sure, this example is not
about a legal gap filler but rather about an informal norm that governs
intrafamily communications. Nonetheless, their functions are the same.
These norms supply a default content to an otherwise ambiguous provi-
sion, and they do so in ways that mimic the will of the party with the
power to dictate.
It is not always clear that courts can figure out, ex post, how bargain-
ing power was divided before a contract was concluded. While the par-
ent/child example I provided is misleadingly easy, determining relative
bargaining strength in commercial relations is more elusive. This Essay
explores and identifies what courts would need to know in order to fill
contract gaps and discusses whether they have the institutional capacity to
do so. It also argues that in a subtle way courts, when filling price gaps,
are already sensitive to the division of bargaining power. For example,
when courts need to determine what constitutes a “reasonable” price
under section 2-305 of the Uniform Commercial Code, they can let one
party have more influence in choosing where, within a broad range, this
price would lie.9 In doing so, courts often acknowledge that the choosing
party is the one with the greater bargaining power.10
Part I of this Essay introduces the idea of bargain-mimicking gap fill-
ers. It explores the conceptual basis for this idea, how it relates to other
criteria of gap filling, and when it may be regarded as the natural substi-
tute for the otherwise compelling, but indeterminate principle of maxi-
mize-the-joint-surplus. Part II explores the normative grounding for this
regime. Admittedly, there is something objectionable about a legal rule
that favors the party with the greater bargaining power. Bargaining
power is hardly a compelling conception of distributive fairness. Legal
rules that favor the weaker party and level the playing field are more com-
monly defended.11 But, as I demonstrate in Part II using actual contracts
9. U.C.C. § 2-305(1)–(2) & cmt. 3 (2004) (establishing that one party may be
accorded power to set price, but must do so in good faith).
10. See, e.g., D.R. Curtis, Co. v. Mathews, 653 P.2d 1188, 1189, 1191 (Idaho Ct. App.
1982) (using original price set by middleman in damages calculation, even though final
price was not set by contract).
11. For example, the doctrine of duress is often justified as redressing the disparity of
bargaining power. See, e.g., John P. Dawson, Economic Duress—An Essay in Perspective,
as examples, default rules that try to upset the potential bargaining out-
come are undesirable because they are a futile effort—the parties can
always opt out of them, and strong parties likely will. Moreover, the anal-
ysis in Part II illustrates that the traditional justifications for default
rules—saved transactions costs, facilitating entry into desirable forms of
relationship, and inducing optimal reliance—also apply to the bargain-
mimicking conception of gap filling.
Part III of this Essay identifies the existence of bargain-mimicking
gap fillers in current contract law. It demonstrates how this idea was im-
plemented in leading cases, although without courts always recognizing
that their decisions relied on—or at least conformed to—a bargain-mim-
icking principle. In illustrating that courts already do this, I suggest that
it is not institutionally impossible to base a legal rule on a criterion as
elusive as relative bargaining power. This Part also highlights situations
in which the bargain-mimicking idea was rejected, thereby recognizing
that in certain situations the bargain-mimicking idea conflicts with other,
deep-rooted principles of contract law. Finally, Part IV extends the analy-
sis by introducing a problem posed by excessive terms, which go beyond
the threshold of permissible contracting. When such excessive terms are
struck down and need to be replaced, courts are faced with a problem of
gap filling. The original, excessive, term is no longer valid; what should
be put in its place? Here, it is generally clear that one party holds greater
bargaining power, which it employed to dictate the excessive term.12 A
bargain-mimicking term would maintain maximal loyalty to the bargain
struck between the parties by filling the gap with the maximally tolerable
term that remains one-sided and favorable to the party who dictated the
original one-sided term, but which is moderated sufficiently so that it is
tolerable. Instead of substituting the offensive term with the most bal-
anced “majoritarian” term, the court would reduce it only enough to fit
within the range that is considered legitimate. In so doing, the court
would mimic the hypothetical bargain that parties negotiating over a
truncated domain would reach.
I. BARGAIN-MIMICKING TERMS
A. No Joint Will
There is a troubling paradox surrounding one of the most basic ten-
ets of contract law that gaps in contracts should be filled with terms that
mimic the will of the parties—terms that most parties would have jointly
45 Mich. L. Rev. 253, 282–88 (1947) (discussing work of duress doctrine in context of
economic bargaining power). Likewise, weak bargaining power can support a claim of
unconscionability. See, e.g., Shell Oil Co. v. Marinello, 307 A.2d 598, 601 (N.J. 1973)
(declaring that courts may find void a “grossly unfair contractual provision[ ]” when “there
is grossly disproportionate bargaining power”).
12. Many cases of intervention in unconscionable contracts explicitly recognize the
presence of superior bargaining power. See Farnsworth, supra note 3, at 301–02. R
chosen. On the one hand, this conception of gap filling makes basic
sense: It minimizes the need of the parties to contract around the default
rule, and it spells out performance provisions that maximize the parties’
joint well-being. But on the other hand, the mimic-the-parties’-will prin-
ciple assumes that the parties’ joint will exists. It assumes that there is a
single term such that, if only the parties spent the time and attention
dealing with the gap, they would have jointly supported the drafting of
this term. Yet the existence of a gap in a contract is often an indication
that a consensus could not be reached—that a single jointly preferable
term does not exist. The claim from which the analysis in this paper be-
gins is that there are situations in which more than one term satisfies the
standard conception of the joint will of the parties to a contract. Absent a
more powerful prescription, then, the will-mimicking principle would be
indeterminate and too amorphous to fill the gap.
Put differently, contract design involves two tasks: creating the pie
and dividing it, with many terms affecting both aspects. Principles of sur-
plus maximization are synonymous with the creation of the pie. Once
the maximal pie is created, through a combination of express terms and
surplus-enhancing gap fillers, it has to be divided. But the term that ac-
complishes this aspect has no bearing on the size of the pie. If one of the
distributive terms is missing from the agreement, the surplus-maximizing
conception of gap filling would, by definition, be indeterminate in sup-
plementing it. So what do we do if the gap involves one of these distribu-
tive aspects?
The fundamental reason to doubt whether there is a single joint will
that could be mimicked when the gap involves a distributive issue is that
the parties have opposite interests in resolving this issue. In these set-
tings, it is impossible to articulate solely on the basis of economic effi-
ciency what term the parties would have chosen. The process of reaching
agreement over distributive elements is resolved by bargaining, and is
thus determined by ad hoc factors that affect the parties’ bargaining
power.13 Filling distributive gaps, then, is not an exercise in surplus max-
imization or in figuring out the optimal transaction design, but in guess-
ing how the surplus would have been divided.
Consider, for example, a sales contract that does not specify payment
terms. There are many ways to supplement this gap, but it can hardly be
said that the different modalities for payment affect the size of the pie. In
many cases, whether payment is made before, during, or after delivery, is
merely a matter of the time value of money, and it would affect the well-
being of the parties in a zero-sum fashion. There is no more or no less
efficient arrangement; the only effect is distributive. As a result, there is
no joint will to mimic. Earlier payment is usually preferable to the seller
13. See, e.g., Martin J. Osborne & Ariel Rubinstein, Bargaining and Markets 50–55
(1990) (showing that strategic bargaining power depends on bargaining procedure;
parties’ relative costs of delay and relative patience; outside options; and more).
to the same extent that it is detrimental to the buyer. The seller, there-
fore, has one will, while the buyer has another.
How, you might wonder, could parties enter a binding contract with-
out specifying the surplus division and leaving the price term out? Is this
scenario realistic? Not only is this scenario possible, but some of the most
prominent cases on contractual indefiniteness involve gaps in the price
term, the one term that by definition has a purely distributive effect. For
example, one of the leading cases, Joseph Martin, Jr., Delicatessen, Inc. v.
Schumacher, involved a lease of commercial property with an option to
renew under an indefinite price, the contract stating that the renewal
price needed “to be agreed upon.”14 Another classic case, Sun Printing &
Publishing Ass’n v. Remington Paper & Power Co., involved a sales contract
that contained an indefinite price formula.15 Furthermore, sales law
casebooks typically devote a chapter to the case law of commercial con-
tracts with missing price terms—a scenario that is fully anticipated by sec-
tion 2-305 of the Uniform Commercial Code and other international
sales law provisions.16
Courts and commentators may disagree whether such distributive
gaps render contracts too indefinite to be enforced.17 The missing price,
it is sometimes argued, is a conclusive indication that the parties have not
yet intended to be bound, since they left the most essential term for fur-
ther assent.18 And yet, modern contract law tends to conclude that a
14. 417 N.E.2d 541, 542 (N.Y. 1981). In that case, the court refused to fill the gap and
held that the contract was too indefinite to be enforced. Id. at 543–44. But the growing
trend is to enforce such contracts. See, e.g., Daniel E. Feld, Annotation, Validity and
Enforceability of Provision for Renewal of Lease at Rental to Be Fixed by Subsequent
Agreement of the Parties, 58 A.L.R.3d 500, 503–06 (1974) (surveying case law on lease
renewals subject to price agreement).
15. 139 N.E. 470, 470 (N.Y. 1923) (examining role of mutual assent in interpreting
open terms and agreements to agree). See generally Sw. Eng’g Co. v. Martin Tractor Co.,
473 P.2d 18 (Kan. 1970) (addressing gap filling where payment and credit terms—
elements that are purely distributive—are not fully specified); Mantell v. Int’l Plastic
Harmonica Corp., 55 A.2d 250 (N.J. 1947) (addressing gap filling in contract in which
price was deliberately left out and yet court was more than ready to supply it).
16. U.C.C. § 2-305 (2004) (“The parties if they so intend may conclude a contract for
sale even if the price is not settled.”); see also United Nations Convention on Contracts for
the International Sale of Goods (CISG) art. 55, Apr. 11, 1980, S. Treaty Doc. No. 98-9
(1986), 1489 U.N.T.S. 59, 69 (implying price where contract was validly concluded but
without a price); Int’l Inst. for the Unification of Private Law, UNIDROIT Principles of
International Commercial Contracts art. 5.1.7 (2004) [hereinafter UNIDROIT Principles]
(same); Bruce W. Frier & James J. White, The Modern Law of Contracts 256–89 (2005)
(discussing interpretation of contract terms); Richard E. Speidel & Linda J. Rusch,
Commercial Transactions: Sales, Leases and Licenses 174–203 (2d ed. 2001) (discussing
effect of open contract terms).
17. See Feld, supra note 14, at 503–06 (surveying different approaches taken by R
courts); see also sources cited in Ben-Shahar, Agreeing to Disagree, supra note 5, at 395–96 R
nn.15–19 (citing sources that describe some reasons why case outcomes differ).
18. See, e.g., Walker v. Keith, 382 S.W.2d 198, 203–04 (Ky. 1964) (finding that missing
price indicates lack of mutual assent); U.C.C. § 2-305 cmt. 2 (“Under some circumstances
the postponement of agreement on price will mean that no deal has really been
missing price does not render a contract unenforceable, so long as there

are independent indications of intent to be bound.19 These situations
present a binding contract with a substantial gap, and the gap filler can-
not be determined by reference to the term that maximizes the total sur-
plus—the price—because every price, at least within a fairly broad inter-
val, satisfies this criterion.
It is true that many aspects of a contract that are primarily distribu-
tive may also affect the total surplus. For example, payment terms can be
more or less efficient because the seller may or may not be the efficient
supplier. But it is false to conclude that gap fillers for all these aspects
can be set to maximize the surplus. As I noted at the outset, the only
reason a subset of the gap fillers can be indisputably surplus maximizing
is that there is at least one other aspect of the deal that is purely distribu-
tive and that can be used to achieve the bargained-for distribution of
value.
The observation that the surplus-maximization conception is poten-
tially indeterminate is reinforced by an account of why contracts are in-
definite. Negotiations—the bargaining and haggling over terms—re-
quire time, effort, and strategy, and they often fail, not because parties
are undertrained in maximization exercises, or because of limits on the
parties’ ability to foresee and imagine contingencies. Rather, negotia-
tions are hard precisely where the issue is distributive and there is no
single maximizing term over which agreement would naturally arise.
Workers go on strikes because of disagreements over zero-sum wage
terms; nations go to wars because of disputes over zero-sum boundary
lines; merger agreements fail when the price offered by the buyer is re-
garded by the shareholders as too low. Negotiations are hardest and
most likely to fail specifically when the issues are purely distributive, be-
cause these are the areas for which the “engine” of increased surplus can-
not provide a focal point for agreement.
Still, even when parties fail to resolve a distributive term, they may
nevertheless choose to enter a binding contract, leaving the distributive
term open or subject to an agreement to agree. They may do this be-
cause they believe agreement to this distributive term will be more likely
at a future date, and it would therefore be inefficient for them to walk
away from the remainder of the deal.20 Or, they may leave price gaps
because they expect some contingency to materialize that will make the
distributive issue moot or easier to resolve in reference to market indi-
concluded . . . .”); Farnsworth, supra note 3, at 202–03, 210–11 (distinguishing R

indefiniteness that is due to lack of assent).
19. See, e.g., U.C.C. § 2-305 (rejecting principle that indefinite agreements, or
agreements to agree, are unenforceable); Farnsworth, supra note 3, at 207–11 (noting that R
U.C.C. § 2-305 rejects traditional common law nonenforceability rule).
20. See Ben-Shahar, Agreeing to Disagree, supra note 5, at 403 (discussing R
negotiation practice of avoiding contentious issues that may make agreement impossible).
ces.21 Or, they may expect that, if needed, a third party can arbitrate or
split their difference.22 In all these cases, if the price resolution mecha-
nism subsequently fails, courts are presented with the reality that disput-
ing parties entered a binding contract but left out a crucial distributive
term.
B. Mimic One Party’s Will

As a mechanism for gap filling, the surplus-maximization principle is
easy to justify. It improves both parties’ well-being and gives them what,
ex ante, they would rationally have chosen.23 If a distributive gap cannot
be filled with a surplus-maximizing term, it may nevertheless be possible
to provide a similarly justified gap filler that solves the problem of what
the parties would have chosen ex ante.
In the case of distributive terms, the parties do not have a joint inter-
est, ex ante. To be sure, consensus over distributive issues can emerge,
but it would be a result of bargaining and maneuvering in the shadow of
market conditions. The argument, therefore, is that the central concep-
tion of what the joint will is must be supplemented by a criterion that
would apply to settings that are purely distributive. Fortunately, courts
often have information that can help them tease out what the parties
would have agreed upon: information about the parties’ relative bargain-
ing power.
When the parties’ interests concerning a particular term conflict, the
term the parties would have agreed upon depends on the allocation of
bargaining power. Having the greater bargaining power means that a
party can exert more influence in the design of the terms. A gap filler
that mimics the division of bargaining power would then favor that party.
If, instead, the parties have equal bargaining power, the gap filler should
resemble the split-the-difference, midrange term. Generally, a gap filler
that depends on information the court has regarding the parties’ relative
bargaining power at the time of the contract is a superior proxy for the
missing term. I refer to this gap filler as a “bargain-mimicking” default
rule.
Thus, for example, in the context of a missing payment term, the
bargain-mimicking gap filler would potentially favor the party that was in
a bargaining position to force the other to acquiesce and surrender to
her dictates. Unlike midrange, split-the-difference default terms that re-
flect the average interest rate or the most common credit arrangement,
the bargain-mimicking term could fall anywhere within a broad interval
and could be significantly different than the midrange solution. The
21. This is the typical situation in lease agreements with a tenant option to renew
upon its expiration.
22. See U.C.C. § 2-305(1)(c) & cmt. 4 (recognizing situation in which third party’s
judgment as to price is used to fill gap).
23. Craswell, supra note 2, at 3–4. R
greater the seller’s bargaining power, the higher the interest rate that the
gap filler would supply. And conversely, the greater the buyer’s bargain-
ing power, the more lenient the credit terms.
Another example of a gap filler that would tilt in favor of one party
arises in auto manufacturing contracts. Sellers, known as “tier-1” suppli-
ers, compete through a bidding process to produce auto parts to be as-
sembled into a car model manufactured by an automaker.24 Because
there are only a few automakers but many suppliers, the buyer in this
setting has much of the bargaining power. And indeed, once the sup-
plier is selected and the price is set, the buyer dictates all the remaining
terms of the contract, including price adjustments over time.25 The stan-
dard form contracts utilized in the auto industry, however, are short and
contain many gaps.26 For example, they often leave the price under
which “service parts” will be sold unspecified.27 Service parts, which are
sold to dealers and car owners in the retail market for a substantial pre-
mium, are a significant source of profits, but how should this surplus be
divided between the automaker and the parts supplier in the absence of a
specific agreement? There is no “market” term to refer to because there
is no competitive market. There is only a single seller who sets different
prices for different buyers. A midrange, split-the-profit price is one way
to fill the gap. Of course, it would not reflect the parties’ relative bargain-
ing powers. Nor would it come close to mimicking the express deal they
would have reached—that is, the deal the buyer would have dictated, and
that some automakers do in fact dictate.28 The bargain-mimicking gap
filler, by contrast, would supply a price that accords the greater share of
the premium to the buyer.
The content of a bargain-mimicking gap filler is fact dependent and
specifically tailored to the contracting parties. The same contract, with
the same gap, can be filled with a pro-seller term in one case and a pro-
buyer term in another, depending on the parties’ relative bargaining
power in each case. For example, a lease with an option to renew under a
24. See Omri Ben-Shahar & James J. White, Boilerplate and Economic Power in Auto
Manufacturing Contracts, 104 Mich. L. Rev. 953, 961–63 (2006) (describing contracting
process in auto manufacturing).
25. Id. at 963–64.
26. For example, General Motors, who in 2004 entered into close to one million
procurement transactions for a total volume of $80 billion, used for all these contracts a
short, thirty-one paragraph, standard form. Id. at 957.
27. Id. at 961; see, e.g., Toyota Motors Mfg. N. Am., Inc., Terms and Conditions § 4.2
(Oct. 1, 1998) (leaving price for service parts to be determined later). Due to the
confidential nature of this contract, the Columbia Law Review does not have a copy of it on
file.
28. For example, Nissan’s contract forces suppliers to commit to selling the service
parts for fifteen years at the price negotiated during the production phase, which is
typically the lowest possible price and the one that accords the entire surplus from the
service parts market to the buyer. Nissan N. Am., Inc., Master Purchase Agreement art. 19
(2003), available at http://www.butzel.com/AutoIndustry/080907tcNissan_AI.pdf (on file
with the Columbia Law Review).
price to be agreed upon would be supplemented with a high, pro-land-

lord price if the landlord happens to enjoy greater bargaining power be-
cause of a migration of many tenants into the region. The same lease
would be supplemented with a low, pro-tenant price if the tenant has the
greater bargaining power because there are many vacant sites in the area.
In an interesting way, bargain-mimicking gap fillers share the same
empirical premise, but can also be contrasted with, the contra proferen-
tum principle. This is the principle under which ambiguous language in
the contract is interpreted against the drafter.29 Both principles—bar-
gain-mimicking and contra proferentum—envision situations in which
one of the two parties has the bargaining power to dictate the language of
the terms and yet this party left some element ambiguous or unspecified,
such that a court needs to fill the language gap. Both principles can only
be applied after a court makes a determination of relative negotiation
power.30 But, relative to the contra proferentum principle, the bargain-
mimicking principle provides the opposite prescription. When a contract
is ambiguous or indefinite, the contra proferentum principle prescribes
gap filling with the term that is least favorable, within reason, to the party
who drafted the contract. By contrast, the bargain-mimicking principle
supplies a term that is most favorable to the drafter and that most closely
resembles the deal the drafter would have been able to dictate. While the
contra proferentum doctrine relies on the notion that the strong party
should be “punished” for leaving ambiguity or indefiniteness in the con-
tract,31 the bargain-mimicking principle gives the strong party what she
could have gotten explicitly through bargaining.
C. Majoritarian Versus Bargain-Mimicking Terms

The bargain-mimicking conception of gap filling breaks a disconti-
nuity that is otherwise created by midrange, majoritarian gap fillers. If all
gaps are filled with midrange terms, a decision by the stronger party to
leave a gap in the contract would result in an expected forfeiture of a
discrete chunk of the private payoff. For this party, the choice to leave
the contract gap might save some transaction costs, but it would simulta-
neously cost her the opportunity to exploit her bargaining advantage and
29. See 5 Margaret N. Kniffin, Corbin on Contracts: Interpretation of Contracts

§ 24.27 (Joseph M. Perillo ed., rev. ed. 1998) (explaining that contra proferentum is
“technique” in which courts “adopt the meaning that is less favorable in its legal effect to
the party who chose the words”).
30. See, e.g., Gaunt v. John Hancock Mut. Life Ins. Co., 160 F.2d 599, 601–02 (2d Cir.
1947) (“[C]ontra proferentum is more rigorously applied in insurance than in other
contracts, in recognition of the difference between the parties in their acquaintance with
the subject matter.”).
31. See, e.g., Kenneth S. Abraham, A Theory of Insurance Policy Interpretation, 95
Mich. L. Rev. 531, 534 (1996) (noting that typical justification for contra proferentum is
liability for negligent drafting); Ian Ayres & Robert Gertner, Filling Gaps in Incomplete
Contracts: An Economic Theory of Default Rules, 99 Yale L.J. 87, 91 (1989) (arguing that
penalty default rules provide incentives for drafting explicit terms).
appropriate a term more favorable than the midrange default rule. The
agreement that results from midrange gap filling is distinctly different
from that which she would have expressly negotiated. The more gaps she
leaves, the greater the wedge between the hypothetical agreement and
the legally supplemented contract. It is this discontinuity—the diver-
gence between the hypothetically negotiated deal and the legally-implied
deal—that the proposed conception of gap filling resolves. The closer
the gap fillers are to the hypothetical bargain, the smaller the divergence.
In many distributive contexts, the bargain-mimicking criterion would
prescribe terms that are significantly different from the midrange, rea-
sonable terms that would be prescribed by the majoritarian criterion.
Specifically, the majoritarian criterion recognizes that different parties
could have reached different reasonable terms and that there is a distri-
bution of bargained-for terms. Accordingly, it fills contract gaps with a
term that measures a “center” of this distribution.32 Unlike the
majoritarian criterion, the bargain-mimicking criterion relies on specific
information that reflects the division of bargaining power, and conse-
quently the deal, that these particular parties would have come to.
Still, it would be a mistake to conclude from this discussion that the
bargain-mimicking gap fillers would always diverge from the majoritarian,
midrange gap fillers. The two gap-filling criteria may prescribe the same
content of gap filler in situations where the bargaining positions of the
parties are relatively equal. In these situations, information about the
specific parties’ bargaining does not change the inference about the hy-
pothetical bargain. For example, if the parties are price-takers, dealing in
matters for which there is a thick market and neither is uniquely posi-
tioned within this market, it is likely that the term they would have agreed
upon is the same term that most parties in the market adopt.33 Similarly,
if bargaining power is determined by outside options, and if there is a
thick market of alternative partners for each party, the terms of that bar-
gain will necessarily be influenced by the terms in that market.34 In these
situations, the bargain-mimicking principle would prescribe a term that
reflects the market term. But unlike the majoritarian principle, it would
do so not because this term best reflects some statistical regularity regard-
ing the market, but rather because it is the best guess as to each party’s
relative purchasing power. Put differently, majoritarian gap fillers that
refer to “reasonable market prices” will be consistent with the bargain-
32. Craswell, supra note 2, at 4–5. R

33. See 1 James J. White & Robert S. Summers, Uniform Commercial Code § 3–8 (5th
ed. 2006) (describing use of market prices as gap fillers).
34. See, e.g., Oglebay Norton Co. v. Armco, Inc., 556 N.E.2d 515, 519–20 (Ohio 1990)
(affirming use of market shipping rates in thick market for shipping, when contract
pricing mechanism failed).
mimicking criterion when applied to specific situations where the parties’

bargaining power is determined by the market.35
To illustrate, consider the well-known case Mantell v. International
Plastic Harmonica Corp., in which a wholesaler and distributor arranged a
long-term distribution agreement that did not fix the price term.36 The
agreement instead referred to the prices charged to other distributors,
but, as it turned out, there were no other distributors.37 The court de-
cided to fill the gap with a reasonable price term, but explained that what
constituted a reasonable price depended, inter alia, on the price the
seller could get from other dealers, competition between wholesalers as
well as between dealers, the uniqueness of the product, and the quantity
produced.38 These factors are precisely those that determine parties’ rel-
ative bargaining powers.
D. Can Courts Identify the Bargain-Mimicking Terms?

There is something admittedly deceptive about the idea of a bargain-
mimicking gap filler. It assumes that the division of bargaining power
between parties is measurable and verifiable by a court. Bargaining
power is, of course, a real factor in negotiations, and economic theory
demonstrates that it depends on relative risk preferences, outside op-
tions, discount factors, negotiation protocol, and the like.39 It reflects, in
short, the relative facility of each party to refuse the deal. But it is one
thing to recognize the theoretical existence and role of this parameter; it
is quite another to actually measure relative bargaining power and to base
legal decisions on these measurements.
It would be naı̈ve to expect that courts will be able to measure bar-
gaining power with complete precision. Still, implementing a regime
with error, or only in those cases where the parameter is verifiable, is
better than nothing. Moreover, in some situations crude approximations
of relative bargaining power are likely to be correct, even if imperfect.
For example, someone who sells a good for which demand is inelastic
undeniably possesses greater bargaining power than those with whom she
is negotiating.40 Similarly, when many bidders compete for a single job,
the party inviting the bids has greater bargaining power. While it is diffi-
cult even in these situations to quantify a party’s bargaining strength on a
scale of one to ten, is it any more difficult than weighing the other param-
eters that courts ordinarily assess, such as comparative fault in a tort
action?
35. See James Gordley, Foundations of Private Law 363 (2006) (“[T]he market price
preserves (so far as possible) each party’s share of purchasing power.”).
36. 55 A.2d 250, 254 (N.J. 1947).
37. Id. at 254–55.
38. Id. at 256.
39. See, e.g., Osborne & Rubinstein, supra note 13, at 29–65 (analyzing factors that R
affect bargaining outcome).
40. Robert Pindyck & Daniel Rubinfeld, Microeconomics 357–58 (6th ed. 2005).
In fact, courts already quite regularly refer to bargaining power as a

factor that justifies case outcomes. Under the unconscionability doctrine,
the presence of one-sided bargaining power is often identified and in-
voked for the purpose of reforming some explicit term.41 Under the du-
ress doctrine, weak bargaining power is often identified and invoked for
the purpose of relieving the weaker party from a coerced deal.42 Under
the contra proferentum doctrine courts have to figure out which party
had the power to dictate a term, and they then rule against this party. As
the Restatement recognizes, contra proferentum is “often invoked in
cases of standardized contracts and in cases where the drafting party has
the stronger bargaining position.”43 While courts may at times misjudge
relative bargaining positions, particularly because there is a misguided
tendency to view a take-it-or-leave-it offer as a sign of bargaining power,44
a substantial doctrinal tradition is nevertheless founded on the belief that
courts can identify bargaining power and determine legal consequences
based on this identification.
Yet, unlike the approach presented in this Essay, existing doctrines
that refer to bargaining power typically favor the weaker bargainer. This
distinction is crucial. For a weak party, there is no danger in arguing in
court that the other side had all the bargaining power. In contrast, for a
strong party, arguing that she had all the bargaining power may be risky,
lest existing contract doctrines and laws equalize her bargained-for gains.
For example, if she were to argue that her bargaining power stems from a
monopoly position, she could face antitrust consequences. If she were to
argue that her bargaining power is due to information advantage, she
could face heightened disclosure requirements, or simply lose the sympa-
thy of the court and the jury. In other words, one’s superior bargaining
power might be a trait that one prefers to keep secret, rather than prove
in court.45 Accordingly, any legal regime that relies on litigants demon-
strating their own superior bargaining strength faces this obstacle.
Moreover (and adding to the difficulty) even in the clear presence of
verifiably uneven bargaining power, identifying a bargain-mimicking
term may be tricky. Gaps in the contract may result from the parties’
41. See, e.g., Carboni v. Arrospide, 2 Cal. Rptr. 2d 845, 850 (Ct. App. 1991) (“[T]here
was an inequality of bargaining power which effectively robbed [promisor] of any
meaningful choice.”); see also UNIDROIT Principles, supra note 16, art. 3.10(1) (listing R
“lack of bargaining skill” as factor relevant to determination of unconscionability).
42. See, e.g., Restatement (Second) of Contracts § 176 cmt. f (1979) (discussing
improper threats in bargaining process).
43. Id. § 206 cmt. a.
44. See, e.g., Circuit City Stores, Inc. v. Adams, 279 F.3d 889, 893 (9th Cir. 2002)
(reasoning that employer “possesse[d] considerably more bargaining power than . . . its
employees” such that employees had to “take the contract or leave it”).
45. See generally Omri Ben-Shahar & Lisa Bernstein, The Secrecy Interest in Contract
Law, 109 Yale L.J. 1885 (2000) (discussing interest of contracting parties in concealing
information for strategic purposes).
inability to agree.46 Or, they may result from the stronger party’s strate-
gic calculation to leave an issue open, recognizing that on this specific
issue the weaker party would not acquiesce to a one-sided term or would
be alerted to some hidden unfavorable aspect of the deal. If the stronger
party suppressed a specific issue and deliberately left a gap, it could actu-
ally be an indication of the limits of her bargaining power. In other
words, it could be a reflection of the fact that in an explicit agreement
she could not extract the one-sided term she coveted. Here, the bargain-
mimicking term would not necessarily favor the stronger party, and grant-
ing her a favorable gap filler would encourage her to leave gaps in all
areas in which she cannot bargain for an advantage. Instead of mimick-
ing the bargain, then, this regime could distort it.
Thus, the craft of filling gaps with bargain-mimicking terms is more
nuanced than merely identifying the party with the greater overall bar-
gaining power. It requires attention to the specific issue left open and
the parties’ special concerns regarding this issue. Recognizing that a bar-
gain usually involves some concessions even by the overall stronger party,
and recognizing that some leverage has already been spent on other, ex-
pressly drafted terms, the court must figure out how the parties would
have used their remaining bargaining power over this specific term.
Daunting as this task might first appear, it is probably not more com-
plicated than other gap-filling principles. For example, under the sur-
plus-maximizing principle, figuring out which term is most efficient re-
quires a sophisticated account of costs and benefits, an understanding of
how different terms and issues interact, and a perception of what each
party values more—all with an eye to idiosyncratic preferences.47 Here,
too, some terms cannot be supplemented without careful attention to
other aspects of the deal. Still, the surplus-maximization criterion has
broad appeal because it makes normative sense, despite the fact that it is
harder to implement than other, simpler default rules. It is socially desir-
able to instruct courts to make the effort to apply the surplus-maximiza-
tion criterion, even crudely, because the benefit arising from more effi-
cient obligations is worth the adjudicative cost. In the next section, I
propose a normative defense of the bargain-mimicking criterion, sug-
gesting that here too it is a worthy effort to trace the bargain that parties
would have struck. And, at the very least, when courts do have accurate
information about the bargain-mimicking term, it ought not be ignored.
46. See Ben-Shahar, Agreeing to Disagree, supra note 5, at 402–05 (arguing that gaps R
in contracts often result from failed attempts to agree on negotiated provision).
47. See, e.g., Mkt. St. Assocs. Ltd. P’ship v. Frey, 941 F.2d 588, 593–95 (7th Cir. 1991)
(demonstrating various factors that need to be evaluated in figuring out how to interpret
gap in contract).
II. ARE BARGAIN-MIMICKING TERMS DESIRABLE?

The purpose of this Essay is not to advocate for the general use of
bargain-mimicking gap fillers, but to identify their use as a conceptual
and practical possibility and to explore arguments in support of such a
regime. Before turning, in Part III, to examine instances where courts
have actually implemented this regime, let us explore some reasons why a
bargain-mimicking regime is normatively desirable.
A. Transaction Costs
When a contract gap involves an issue affecting the size of the sur-
plus, a well-rehearsed argument explains why the gap filler ought to be a
surplus-maximizing provision.48 It is an argument of exceptional appeal
because it sidesteps any distributive implications. Surplus-maximizing
gap fillers indiscriminately increase the well-being of both parties to the
contract. If the law were to provide off-the-shelf terms that were anything
but surplus-maximizing, it would have the effect of inducing parties to
write explicit provisions which, other than occasional indirect benefits, as,
for example the exposure of private information, would merely increase
transaction costs.49 But once the price or other distributive term is ad-
justed appropriately to divide the savings in transaction costs, each party
ends up with a greater net payoff.
One might assume that in the context of bargain-mimicking terms
this same distributive-neutral defense is inapplicable. In other words, if
the law provides a gap filler that is more favorable to one of the parties,
without affecting the size of the surplus, how can it be said that this term
accords both parties a greater surplus to divide? If it is a term that mimics
one party’s will, against the will of the other party, how could the other
party benefit from it?
Moreover, upon first reflection, bargain-mimicking terms might
seem to encounter an objection that surplus-maximizing terms avoid—
namely, that they conflict with social concerns and intuitions regarding
the fairness of distribution. While surplus-maximizing terms need not
have any distributive effect—they merely secure more value to divide—
bargain-mimicking terms do not create a greater surplus and do have a
clear distributive effect in favor of the stronger party. Why, one might
ask, should it be the law’s objective to resolve distributive ambiguities and
gaps in favor of the stronger party when the overall welfare of the parties
is not enhanced? Surely, this party can take good care of herself and
48. See, e.g., Robert E. Scott, Rethinking the Default Rule Project, 6 Va. J. 84, 94 n.4
(2003) (“[C]hoosing a default rule on the basis of some normative conception of fairness
would be wrong, in the sense that it would not increase the amount of fair contracts in the
world, but it would increase the amount of contracting costs . . . .”); see also sources cited
supra note 3. R
49. Posner, supra note 3, at 96–99 (“[C]ontract law cannot readily be used to achieve R
goals other than efficiency, as a ruling that fails to interpolate the efficient term will be
reversed by the parties in their subsequent dealings.”).
secure her own advantages through bargaining. If anything, it is the weak

transactor that should be protected by the law and enjoy a distributive
bias. A prescription of distributive fairness, so goes the objection, can
hardly be based on bargaining power as the conception of merit. It
should aim to undo the unfairness that unfettered bargaining might gen-
erate, not mimic it.
Compelling as this argument might be, it is beside the point. The
benchmark argument in favor of bargain-mimicking terms is not that
these terms are fair or that they otherwise conform to an attractive con-
ception of distributive desert. They probably do not. Bargain-mimicking
is a principle of gap filling, not of redistribution. The reason why bar-
gain-mimicking terms may be desirable as gap fillers is that, very much
like surplus-maximizing terms, they save transaction costs. If the law ac-
cords a party the same terms that she could secure by explicit (and harsh)
bargaining, the party with the bargaining power need not expend the
costs of explicitly specifying these same terms. If gap fillers do anything
other than mimic the term this party could have dictated herself, they will
have the ex ante effect of inducing this party to dictate the term in order
to preempt any adverse allocation that would otherwise result from the
gap filler. Perhaps even more than in other contexts, when the distribu-
tion is at stake it is likely that the stronger party will insist on contracting
around a nonmimicking gap filler.
In the context of surplus-maximizing gap fillers, it is commonly
noted that both parties enjoy the saved transaction costs afforded by such
terms.50 By similar logic, it must also be true that when one party has the
bargaining advantage, both parties enjoy the saved transaction costs
achieved by bargain-mimicking terms and would therefore prefer them to
majoritarian gap fillers. The only difference in the current context is that
the savings achieved by the mimicking terms are, like other sources of
value in the contract, enjoyed disproportionately by the party with the
greater bargaining power. Her leverage enables her to dictate a division
of the salvaged transaction costs that is favorable to her.
What, exactly, are these transaction costs that are saved by a bargain-
mimicking term? Beyond the obvious category of drafting costs, in the
context of unequal bargaining power, there might be additional psychic
burdens that the parties are spared. In certain settings, for example, we
can imagine that weaker parties endure humiliation when the stronger
party openly dictates a one-sided term. While the cost of punctuating
one’s powerlessness is emotional and cannot be measured in monetary
terms, it is nonetheless recognized as an important cost in negotiation
literature.51 In this vein, stronger parties are often advised to impart to
50. See, e.g., id. at 95–96 (explaining mutual benefit to parties of leaving gaps in
contract for contingencies that are unlikely to occur).
51. See, e.g., David A. Lax & James K. Sebenius, The Manager as Negotiator 112–13
(1986) (highlighting cooperative bargainer’s concern for preservation of “self-esteem” and
“helping counterparts to save face when necessary”); Robert H. Mnookin, Scott R. Peppet
their counterparts a sense that the pie is equally divided, even when it is
not, to make it easier for their opponents to acquiesce.52 A default rule
that eases the need for stronger parties to openly “stick it” to weaker par-
ties has this cost-mitigating effect.
B. Flexibility in Contract Performance

Another reason why a bargain-mimicking regime is normatively de-
sirable has to do with the parties’ need to leave issues unresolved in order
to allow flexible adjustments down the road. In long-term contracts, gaps
result not only from transaction costs (i.e., the difficulty of foreseeing and
stipulating for all future contingencies), but also from deliberate drafting
decisions to leave room for more flexibility when new contingencies
arise.53 Parties recognize that conditions may change and special needs
or priorities may arise, such that it would be mutually beneficial to allow
for future adjustments in their respective obligations. It is, of course, pos-
sible to dictate rigid terms that apply to future contingencies and then
later, if flexibility is needed, to accord waivers and accommodations
through the course of performance. But doing so means the stronger
party would “waste” bargaining power to secure terms and privileges that
he would be willing to waive and that, ex post, might not matter all that
much. Alternatively, recognizing the advantages of flexibility, the
stronger party may choose to leave some issues in the contract un-
resolved, with the expectation that she will nail them down if and when
they become relevant. A bargain-mimicking, gap-filling regime would
render the open term strategy safer for the stronger party whenever it is
the cheaper method for drafting the contract.
This technique of flexible drafting is used in various ways. At the
extreme, a contract might stipulate that the parties will agree upon a par-
ticular provision at a later stage.54 This is not an agreement to agree;
there is enough definiteness in the remainder of the contract, and there
& Andrew S. Tulumello, Beyond Winning: Negotiating to Create Value in Deals and
Disputes 44–68 (2000) (asserting that expressing “concern and respect” during
negotiations “tend[s] to defuse anger and mistrust, especially where these emotions stem
from feeling unappreciated or exploited”).
52. See Richard H. Thaler, The Winner’s Curse: Paradoxes and Anomalies of
Economic Life 21–35 (1992) (surveying experimental research that shows that parties with
less bargaining power will nevertheless refuse to accept deals in which they are treated
unequally).
53. See, e.g., Alan Schwartz, Relational Contracts in the Courts: An Analysis of
Incomplete Agreements and Judicial Strategies, 21 J. Legal Stud. 271, 284–90 (1992)
(analyzing reasons for incompleteness of long-term contracts).
54. This technique is common in auto manufacturing contracts. The big auto
manufacturers stipulate in long-term contracts with their suppliers that replacement parts
will be sold at a price that will be agreed upon later. See, e.g., Gen. Motors, General Terms
and Conditions § 20 (rev. Sept. 2004) (“[T]he price(s) during the first 3 years of this
period shall be those in effect at the conclusion of current model purchases. For the
remainder of this period, the price(s) for goods shall be agreed to by the parties.”); Toyota
Motors Mfg. N. Am., Inc., supra note 27, § 4.2(d) (“[Toyota] will establish, after good faith
is a clear statement of the parties’ intent to be bound, thereby making the

entire agreement enforceable. Rather, it is an agreement with a specific
methodology for subsequent, contingent, gap filling. The gap is ex-
pressly recognized by the parties, and the methodology to resolve it is set
in the contract. When this methodology fails—when the parties do not
manage to agree at a later stage on the to-be-agreed-upon issue—the
court must utilize a different methodology to fill the gap. In choosing a
gap filler, the bargain-mimicking principle instructs a court to lean to-
ward the will of the party who would have had more bargaining power at
the subsequent negotiation stage. Indeed, at the ex post stage in which
the term would have been settled, bargaining power may have shifted and
the party that originally had more power may now have less leverage.
The parties chose the flexible drafting technique recognizing this possi-
ble shift in bargaining power. The parties anticipated and agreed to fol-
low the terms that would emerge from subsequent bargaining—terms
that would reflect bargaining power at the subsequent round. Thus, the
bargain-mimicking term should trace the relative bargaining power at
this subsequent stage.
A common technique in flexible drafting is to allocate to one party
the power to determine, ex post, the content of the term and to change it
as circumstances change.55 The party with this power is not always the
stronger party. For example, in output contracts the seller is entitled to
set the quantity, but it is hardly the seller who has the bargaining power.56
Thus, when a farmer sells his small crop to a large distributor/buyer, the
farmer sets the quantity ex post, but has very little bargaining strength ex
ante. In some contexts, however, the parties use this technique of one-
sided, ex post control over a term to create what is effectively a bargain-
mimicking, gap-filling regime. When a seller has the power to set the
price and vary it throughout the duration of the contract, the seller is
translating her bargaining strength ex ante into a scheme that supplies
terms that are favorable to her ex post. This is how oil companies deal
with their local distributors.57 The role of courts here is to police over-
reaching—to determine if a party acted in bad faith and used its power to
set terms that are unreasonable or intolerable.58 If courts were unwilling
negotiations with Supplier, a price for Service Parts.”). Due to the confidential nature of
these contracts, the Columbia Law Review does not have copies of them on file.
55. White & Summers, supra note 33, at 231. R
56. For example, in the case law favorite Feld v. Henry S. Levy & Sons, Inc., 335
N.E.2d 320, 321 (N.Y. 1975), the parties had an output contract for breadcrumbs. The
seller was entitled to set the quantity but did not have much bargaining power and indeed
failed to induce the buyer to agree to pay for a cost increase of one cent per pound. Id.
57. See, e.g., Mathis v. Exxon Corp., 302 F.3d 448, 452–54 (5th Cir. 2002) (illustrating
that oil companies follow one-sided pricing practices); Shell Oil Co. v. HRN Inc., 144
S.W.3d 429, 432–33 (Tex. 2004) (same).
58. U.C.C. § 2-305(2) & cmt. 3 (2004); White & Summers, supra note 33, at 226–34 R
(surveying cases in which courts applied good faith limitation to scrutinize price
adjustment).
to allow stronger parties to utilize this self-promoting technique, flexible

drafting of long-term contracts would be undermined.
Another technique of flexible drafting attaches the meaning of a par-
ticular term to some objectively observable index. For example, a sup-
plier in a long-term sales contract may demand that price will equal some-
one else’s posted price at the time of delivery.59 Disputes may arise if the
contractually selected index ceases to exist in the midst of the contractual
period and can no longer be referenced. Which price should be used in
its place? Here again, the stronger party has a sensible claim that the
supplemented price ought to mimic the bargain the parties would have
struck had they expressly negotiated over a substitute index. Or, more
directly, the price ought to mimic the division of bargaining power be-
tween the parties. If the failed index was a pro-supplier price, located in
the upper range of the market prices, the gap ought to be supplemented
with a comparable price. Any other choice would force parties to choose
an alternate pricing methodology, perhaps sacrificing some of the
flexibility.
Termination terms also illustrate the benefit of bargain-mimicking
defaults. A party enjoys stronger bargaining power when there are many
potential partners who bid to be chosen by her. Once a bidder is chosen
and awarded the contract, however, the one-on-one relationship no
longer preserves the asymmetric bargaining power. Nonetheless the pos-
sibility of termination allows the stronger party to maintain a bargaining
advantage throughout the relationship because she can credibly threaten
to choose another bidder. Thus, the duration gap filler—the rule that
allows parties to terminate an open-ended contract at will60—is effectively
a bargain-mimicking default rule. True, under this rule both parties have
a symmetric right to terminate.61 But often it is only one party—the ini-
tially stronger party with many bidders—who might potentially want to
terminate.62 The other party has too little choice to go elsewhere, or has
already sunk too much into the relationship.63 If the weaker party
chooses to terminate the agreement, she is unlikely (or at least less likely)
to find other business. The right to terminate effectively mimics the will
59. U.C.C. § 2-305(1)(c) & cmt. 4 (applying to cases in which a particular person is
chosen to set price); White & Summers, supra note 33, at 232 (providing examples for R
such formulae).
60. U.C.C. § 2-309(2).
61. Id. (“[U]nless otherwise agreed [the contract] may be terminated at any time by
either party.” (emphasis added)).
62. See, e.g., Corenswet, Inc. v. Amana Refrigeration, Inc., 594 F.2d 129, 131 (5th Cir.
1979) (describing manufacturer’s termination of distribution contract and holding
“arbitrary termination . . . permissible under both the contract and the law of Iowa”).
63. Id. at 132 (detailing distributor’s claim that it made investment in relationship
that would be squandered if contract was terminated). Indeed, this nowhere-to-go
problem is often the case in termination of franchise contracts. See, e.g., Gillian K.
Hadfield, Problematic Relations: Franchising and the Law of Incomplete Contracts, 42
Stan. L. Rev. 927, 951–53 (1990) (discussing problem of relationship-specific investment).
of the stronger party. With this safeguard in place, the stronger party is
free to use a variety of terms she could not otherwise use. First, she is free
to use an open-ended duration. Second, she can give the other party
leeway and control over aspects of performance without specifying them
in the contract, since she knows that if these privileges are abused she can
simply terminate the contract. Finally, she can choose a lesser-known
bidder.
A final example where a bargain-mimicking criterion might affect
the way parties design their bargain is a prenuptial agreement. Imagine a
situation in which a billionaire is about to marry a person with no assets.
It is plausible to suggest that the billionaire has greater bargaining power
regarding the financial consequences of divorce. In the event of divorce,
a gap filler that tracks this bargaining advantage would differ significantly
from one that provides a more generous distribution to the less affluent
spouse. It is true that in noncommercial settings bargaining power may
be particularly elusive because bargaining power is not simply equivalent
to financial prowess. There are obvious factors other than wealth that
affect each party’s relative eagerness to enter the relationship and thus
his or her relative power to say “no” to versions of the prenuptial agree-
ment proposed by the other party. And yet, bargaining power surely ex-
ists—it is often played out in express prenuptial bargains. Allowing the
parties to leave some aspects vague, by assuring them that their relative
bargaining power will be reflected ex post, might relieve them of the
costly and often damaging need to punctuate who has the upper hand ex
ante.
C. Ex Ante Investment
The analysis so far has assumed that the relative bargaining power of
the parties is an exogenous factor, determined before the parties enter
the negotiations. In reality, many features may affect bargaining power:
outside options, impatience to reach a deal, reputation, financial distress,
negotiation savvy, and more.64 The implicit assumption so far was that
none of these factors depend on the gap-filling methodology. Thus, the
premise was that gap fillers could be a function of the relative bargaining
power of the parties, but not vice versa. But can the cause-and-effect be
reversed? Is it possible that the gap-filling rule would induce parties to
make investments in increasing their bargaining power?
Theoretically, a bargain-mimicking regime could create incentives
for parties to make investments that affect their bargaining power. Of
course, parties already have a motive to invest in strengthening their bar-
gaining power because such actions will help them secure better express
terms in the deal. But in the shadow of bargain-mimicking gap fillers, the
incentive to manipulate bargaining positions would be bolstered. Invest-
64. See supra note 13 and accompanying text (exploring factors that affect bargaining R
outcomes).
ing in stronger bargaining power would now affect not only the explicit
provisions, but also the gap fillers.
It is not clear what to make of these potential effects. Prima facie,
much of the investment in bargaining leverage is a social waste—it is a
social cost that redistributes value without creating a corresponding social
benefit.65 This could suggest that a bargain-mimicking regime would
have the undesirable effect of further distorting already excessive
investments.
But the picture is more complex. There are other types of precon-
tractual investments—such as those having an effect on the total surplus
of the potential bargain, and not on relative bargaining power—that are
often set too low. Specifically, where parties are exposed to the holdup
problem, the anticipation that some of the fruit of this investment will be
appropriated by the other party may induce them to set investment too
low.66 If the party who makes the surplus-enhancing investment is also
the one who is in a position to make the bargaining-leverage investment,
it is no longer clear that the latter investment is a social waste. Investing
in greater bargaining leverage would have an indirect positive effect: It
would diminish the other party’s ability to engage in hold up and would
thus lead to a more efficient level of surplus-creating investments. For
example, a builder who successfully acquires an exclusive position in a
specific market would be able, in the course of negotiating a project with
a client, to make precontractual investments in plans and materials since
he knows his bargaining leverage will shield him from hold up.
Moreover, sometimes the same investment has both a surplus-en-
hancing effect and a bargaining-leverage effect. For example, a potential
employee who invests in learning a specialized skill increases the overall
surplus from the employment arrangement, but at the same time gains
more leverage in negotiating her wages and securing a bigger slice of the
surplus for herself. The incentive to invest too much to enhance bargain-
ing leverage is at least partially offset by the incentive to invest too little
because of the holdup problem. The bargain-mimicking legal regime,
which amplifies the “too much” side of this trade off, is not necessarily
bad.
In the end, though, whatever effect the bargain-mimicking gap-fill-
ing regime has on ex ante investment, one should doubt whether this
effect is significant. Parties have strong incentives to make investments
that increase their bargaining power even in the absence of this gap-fill-
ing regime. Such investments secure greater payoffs through the more
65. See, e.g., David M. Frankel, Creative Bargaining, 23 Games & Econ. Beh. 43,
49–50 (1998) (proving parties overinvest in creative strategies that improve their
bargaining position).
66. See, e.g., Lucian Arye Bebchuk & Omri Ben-Shahar, Precontractual Reliance, 30 J.
Legal Stud. 423, 431 (2001) (showing parties will make insufficient investments in
relationship in absence of precontractual liability).
favorable express terms that the investing party can draft. Ex ante, the
incremental effect of a gap-filling regime would probably be negligible.
III. BARGAIN-MIMICKING TERMS IN ACTION

A. Interpretive Practices
To be sure, a general principle of bargain-mimicking gap fillers has
not been explicitly endorsed by the law. Notably, section 204 of the
Restatement instructs courts to supply gap fillers that are “reasonable in
the circumstances,” expressly rejecting the bargain-mimicking approach:
“[W]here there is in fact no agreement, the court should supply a term
which comports with community standards of fairness and policy rather
than analyze a hypothetical model of the bargaining process.”67 Despite this
sweeping mandate, I hope to illustrate that in many subtle ways courts’
current interpretation of existing contract law applies gap fillers that re-
flect relative bargaining power.
It is important to note that principles of gap filling could be enacted
in contracts themselves. Bargain-mimicking gap fillers could emerge in
practice as a result of contractual drafting instructing courts to apply such
a criterion, effectively telling courts to fill any particular gap with terms
more favorable to one of the parties. How is this done? For example,
typical boilerplate contracts include severability or “savings” clauses that
instruct courts to enforce the contract to the maximum extent permitted
by law.68 If a provision that is otherwise drafted vaguely is appended to
this maximum extent boilerplate, the ambiguity will be resolved in a one-
sided manner. Such a drafting technique may apply to a single provision,
as in warranty disclaimer clauses,69 or it may apply to the entire contract
any time a term is unenforceable.70 Effectively, by including such provi-
sions, the drafting party opts out of the “fair community standards” gap-
filling approach of section 204 and opts into a bargain-mimicking, one-
sided gap-filling regime. The parties are effectively telling a future court
that if a specific term will be deemed unenforceable as too one-sided, it
ought to be filled with a substitute term that achieves the desired effect to
the maximum permissible extent. The incentive to draft such terms is
67. Restatement (Second) of Contracts § 204 cmt. d (1979) (emphasis added).

68. See, e.g., Charles A. Sennewald, Security Consulting 151 (3d ed. 2006) (“If the
scope of any of the provisions of the Agreement is too broad in any respect whatsoever to
permit enforcement to its full extent, then such provisions shall be enforced to the
maximum extent permitted by law . . . .”).
69. See, e.g., RealNetworks, Inc., Realplayer Distribution Agreement for Educational
Institutions § 11, at http://forms.real.com/rnforms/products/tools/rpedist/index.html
(last visited Nov. 16, 2008) (on file with the Columbia Law Review) (“To the maximum
extent permitted by applicable law, RealNetworks further disclaims all warranties . . . .”).
70. C. James Levin & Avery R. Brown, Severability, in Negotiating and Drafting
Contract Boilerplate 539, 547–48 (Tina L. Stark ed., 2003) (noting that severability clauses
can apply to provisions such as indemnity and exculpation, noncompete, acceleration,
damages and penalties, interest rates, and more).
particularly strong when applied to distributive issues, where the one-

sidedness does not come at the expense of the overall surplus.
Before exploring instances where courts have already indicated
openness to a bargain-mimicking principle, it is important to highlight
one major area of contracting in which the principle is, or appears to be,
regularly rejected. In insurance contracts, courts consistently fill gaps
with provisions that favor the weaker party—the insured. Through doc-
trines like reasonable expectations, contra proferentum, and implied war-
ranty of fitness for intended purpose,71 courts supply meaning to ambigu-
ous terms that increase rather than reduce coverage, thereby favoring the
party with the lesser bargaining power. There are good reasons to use
these techniques and to give primacy to the insured’s expectations. But is
this practice truly in conflict with the bargain-mimicking principle? It
may be argued that despite the fact that the insureds have no bargaining
power to change the policy terms, they do have the power to walk away
from policies that provide insufficient coverage, or that cost more than
they are worth. Through these interpretive practices, courts merely give
insureds the coverage they originally sought (and paid for): the terms
without which they would not have signed onto the contract. It is this ex
ante power to say no to the policy that is effectively mimicked. Moreover,
the pro-insured gap filling is consistent with the bargain-mimicking crite-
rion in another way. It is often the case that, despite having the power to
prospectively revise the interpreted language and redraft it in a self-serv-
ing way, insurance companies do not pursue this strategy. When a court
interprets the meaning of an insurance policy in a pro-coverage manner,
insurance companies often adjust the premium, rather than the lan-
guage.72 They are happy to maintain the broader coverage as long as
they can charge for it in a profitable way. Put differently, insurers who
have bargaining power are not interested in selling less coverage. On the
contrary, their interest is in selling more coverage and raising prices.
Viewed in this light, the pro-coverage gap fillers are essentially subtle bar-
gain-mimicking terms. Insureds get the coverage they would have in-
sisted on, and the insurer adjusts the premium to fit the actual coverage.
The insureds have the bargaining power to choose the scope of coverage;
the insurer has the bargaining power to price it; and the gap fillers reflect
this allocation of powers.
71. See, e.g., C & J Fertilizer, Inc. v. Allied Mut. Ins. Co., 227 N.W.2d 169, 176–81
(Iowa 1975) (analyzing interpretation doctrines for insurance contracts); Kenneth S.
Abraham, The Expectations Principle as a Regulative Ideal, 5 Conn. Ins. L.J. 59, 63–64
(1998) (exploring foundations of reasonable expectation doctrine).
72. See, e.g., Kenneth S. Abraham, Insurance Law and Regulation 33–37 (3d ed.
2000) (discussing value of interpreted terms); Michelle E. Boardman, Contra Proferentem:
The Allure of Ambiguous Boilerplate, in Boilerplate: The Foundation of Market Contracts
176, 180 (Omri Ben-Shahar ed., 2007) (discussing reasons why insurance companies do
not redraft).
B. Examples
This section demonstrates instances in which courts explicitly recog-
nize a bargain-mimicking criterion for gap filling and reach decisions in
line with it. It does not argue that the bargain-mimicking criterion is uni-
versally applied. Instead, the argument—by way of examples—is more
modest, asserting that the bargain-mimicking criterion is not as alien to
the task of judicial gap filling as might otherwise seem.
1. Termination Terms. — When parties have an open-ended contract
duration, section 2-309 of the U.C.C. allows each party to terminate at
will.73 Even when the contract guarantees a minimum duration, termina-
tion can occur prior to the expiration of this period if there is miscon-
duct by one of the parties, even if this misconduct does not rise to the
level of total breach.74 In this context, courts are often asked to deter-
mine whether a particular event or misconduct by the franchisee provides
legitimate grounds for termination by the franchisor. A powerful exam-
ple of the application of the bargain-mimicking principle comes up in a
case that called for interpretation of a termination clause in a franchise
contract. In the casebook favorite, The Original Great American Chocolate
Chip Cookie Co. v. River Valley Cookies, Ltd., 75 Judge Posner discusses the
power of the franchisor to terminate the franchise. He rejects the claim
that “in a dispute between franchisee and franchisor the judicial thumb
should be on the franchisee’s pan of the balance.”76 He made this deter-
mination despite the fact that the franchisee was clearly the party with the
weaker bargaining power and should therefore have been the natural re-
cipient of any redistributive sentiment. Such a tilt, he explains, will not
help franchisees as a group: “The more difficult it is to cancel a
franchise, the higher the price that franchisors will charge for franchises.
So in the end the franchisees will pay for judicial liberality . . . .”77 Posner
continues, invoking the logic underlying bargain-mimicking terms:
The idea that favoring one side or the other in a class of con-
tract disputes can redistribute wealth is one of the most persis-
tent illusions of judicial power. It comes from failing to con-
sider the full consequences of legal decisions. Courts deciding
contract cases cannot durably shift the balance of advantages to
the weaker side of the market; they can only make contracts
more costly to that side in the future, because franchisors will
demand compensation for bearing onerous terms.78
73. U.C.C. § 2-309(2) (2004) (noting that contract with indefinite duration “may be
terminated at any time by either party”).
74. Antony W. Dnes, A Case-Study Analysis of Franchise Contracts, 22 J. Legal Stud.
367, 370–74 (1993) (analyzing termination of franchise agreements).
75. 970 F.2d 273 (7th Cir. 1992).
76. Id. at 282.
77. Id.
78. Id. In a somewhat mocking dissent, Judge Cudahy agrees that franchisees have
less bargaining power than franchisors but responds to Judge Posner’s bargain-mimicking
default rule by saying:
Posner’s decision in the Original Great American Chocolate Chip Cookie

Co. case is an illustration of a bargain-mimicking term because the con-
tract at issue presented a vague termination clause that needed interpre-
tation: Was breach by the franchisee “material”? Did it constitute “good
cause” for termination?79 The court’s answer had nothing to do with sur-
plus maximization. The court did not focus on the franchisee’s reliance
or its interest in recouping its investment, as courts sometimes do. Nor
did the court focus on the franchisor’s need to efficiently protect its
brand and provide adequate incentives for management of its franchised
stores. Moreover, the court did not invoke notions of hypothetical con-
sent—terms or meanings that both parties would have willingly chosen, if
only they drafted terms with increased resolution, or terms that are stan-
dard in the industry. Instead, the court viewed the problem as distribu-
tive in nature, but rejected a solution that would be redistributive in favor
of the weak party. It examined the parties’ relative bargaining powers
and held that the franchisor’s superior economic position would make it
futile for courts to interpret the contract in any way that did not mimic
the franchisor’s bargaining strength. In essence, Judge Posner asserted
that any judicial favoring of the weaker party for redistributive reasons
would fail because it would be undone through overriding provisions dic-
tated by the stronger party in the contract.
2. Force Majeure Terms. — Another interesting illustration of the bar-
gain-mimicking principle arises in the context of force majeure clauses,
which expand the scope of excuse otherwise available under the doctrine
of impracticability.80 Any party can use its bargaining power to secure a
favorable list of excuses by drafting a self-interested force majeure clause.
But even when the express force majeure clause is clear, questions arise
that require gap filling and interpretation. For example, if a seller is ex-
Apparently, the legislators had not read enough scholarly musings to realize that
any efforts to protect the weak against the strong would, through the exhilarating
alchemy of economic theory, increase rather than diminish the burden upon the
powerless. I agree that the thumb of judges ought not be placed on the scales of
justice. But judges have no obligation to ignore the numerous thumbs already
put down on the side of economic power . . . .
Id. at 283 (Cudahy, J., dissenting).
79. “Material breach” was defined in the contract to include, among other things:
“failing to maintain and operate the Cookie System Facility in a good, clean,
wholesome manner and in strict compliance with the standards then and from
time to time prescribed by” the Cookie Company; selling any product not
authorized by the Cookie Company; failing to pay any service fee within 10 days
after it is due; failing to pay any of the company’s invoices within that period;
underreporting gross sales (on which the Cookie Company’s royalty from its
franchisees royalty is based) by 1 percent or more; or failing to maintain certain
insurance coverage. . . . Any three breaches, whether or not material, entitle the
company to terminate the franchise within a 12-month period without giving the
franchisee notice or an opportunity to cure.
Id. at 278 (majority opinion).
80. White & Summers, supra note 33, § 3-10 (analyzing U.C.C.’s force majeure R
jurisprudence).
cused against the buyer when the seller’s source of supply defaults, must
the seller assign its remedial rights against its own defaulting supplier to
the disappointed buyer? Usually, when the grounds for excuse are in the
gap-filling list of section 2-615 of the U.C.C., the answer is yes: The disap-
pointed buyer, while unable to get redress from the excused seller, can
instead step into the seller’s shoes and recover from the interfering party
(the defaulting upstream supplier).81 Put differently, the U.C.C. attaches
to its excuse gap fillers an assignment gap filler: Unless stated otherwise,
the rights against the interfering party are automatically assigned from
the excused seller to the buyer.
But what if the grounds for excuse are not in section 2-615 and in-
stead appear in the expressly drafted force majeure clause? Are the rights
of the excused party against the interfering party assigned here too? Is
the disappointed buyer entitled to any recovery rights against the default-
ing upstream supplier? Or does the seller get to keep the right to recover
against his own defaulting supplier, despite being excused against the
buyer? The U.C.C.’s assignment gap filler does not speak to this situation
and so the answer is less clear and must be provided by courts.
In a leading case, Interpetrol Bermuda Ltd. v. Kaiser Aluminum
International Corp., the Ninth Circuit decided that the seller does not have
to assign the recovery rights against the supplier to the disappointed
buyer.82 The court based its decision on a bargain-mimicking principle.
It noted that the seller used its bargaining power to extract a force
majeure clause from the buyer, and that the seller’s supplier was unable,
“because of market forces,” to require a similar excuse provision against
the seller.83 Accordingly, the seller was excused even though his supplier
was not. The court held that it saw
no reason to award the windfall of recovery against the supplier
to the buyer, who agreed to excuse the seller, instead of the
seller, who was able to insist on better protections . . . . We find
no reason to transfer the benefit of [the seller’s] superior nego-
tiating position to [the buyer] by giving [the buyer] rights
against [the defaulting supplier]. We do find that it serves the
forces of natural market adjustments not to transfer [the
seller’s] rights.84
The court’s decision in Interpetrol hinged on mimicking the parties’
likely bargaining outcome. While the parties did not stipulate who, in the
event of excuse, would be entitled to recover from the defaulting sup-
plier, only one party could recover this right, and the court chose to
award it to the party with the greater bargaining power. This decision did
81. U.C.C. § 2-615 cmt. 5 (2004) (“[E]xcuse should not result in relieving the
defaulting supplier from liability nor in dropping into the seller’s lap an unearned bonus
of damages over.”)
82. 719 F.2d 992, 999–1001 (9th Cir. 1983).
83. Id. at 1000.
84. Id. at 1000–01.
not hinge on an efficiency analysis, or on the assumption of a hypotheti-

cal joint will. In fact, the court even speculated that its holding, which
obviously favored the stronger party, might affect the incentive of future
parties to enter contracts with such unfavorable terms.85 It nevertheless
found that the gap filler in this case, with its distributive effect, needed to
mimic the division of bargaining power.
3. Dividing a Windfall. — Current case law dealing with missing dis-
tributive terms may suggest that courts are up to the task of ascertaining
bargain-mimicking terms. A recent Sixth Circuit decision dealing with
the division of a pot of money that the parties had not expected illustrates
this nicely.86 In this case, a pension plan that was originally set up to
provide employees with fixed benefits had to be redesigned under ERISA
and ended up being managed in a way that created unanticipated finan-
cial benefits. The employer and the employees disputed how to divide
this surplus.87 The court turned to the “community standards of fairness”
principle of section 204 of the Restatement and held that the party who
bore the risk was the one entitled to the unforeseen proceeds.88 Interest-
ingly, though, the court recognized that the same result would be
achieved by applying a “hypothetical model of bargaining” approach,
which the court noted was the Restatement’s less favored mode of analy-
sis.89 The court reasoned that the party responsible for any downside in
the case the funding source defaulted or became insolvent would have
demanded that any unanticipated proceeds from this source inure to it.
Put differently, the court did not need to ask who had more bargaining
power in the abstract. Its reasoning was founded on the bargain-mimick-
ing observation that with respect to this specific term, one party would
have naturally prevailed had it been the subject of an explicit agreement.
C. Peevyhouse
The bargain-mimicking idea can also help explain case outcomes in
another important area: the selective application of the cost-of-comple-
tion damages in cases of defective performance. In the classic case
Peevyhouse v. Garland Coal & Mining Co., the court had to determine what
85. Id. at 1000 (stating that future parties might hesitate to move into the “more
contractually secure part of the market”).
86. See Bank of N.Y. v. Janowick, 470 F.3d 264 (6th Cir. 2006).
87. Id. at 267–68.
88. Id. at 272.
89. The court stated that:
Were we to attempt to discern the term to which the parties to the annuity
contracts would have agreed (the less-favored mode of analysis under § 204’s
comment d), we would reach the same conclusion. . . . [U]nder the “hypothetical
model of bargaining” approach, [the employees’ trustee] would have demanded
that any unanticipated proceeds . . . inure to the Employees to compensate them
for this additional risk. Prudential would not have been in a position to favor
either the Employees or Southwire, and would not have objected to this term.
Id. at 272 n.7.
damages applied when a stripmining company breached its promise to

restore mined farm land to its original condition.90 The case provides a
dramatic illustration of the choice between two measures of expectation
damages: cost-of-completion, which would have been $29,000, and dimi-
nution-in-market-value, which was only $300. The Supreme Court of
Oklahoma ruled five to four in favor of the diminution-in-value mea-
sure.91 Other state courts have held differently, and authorities remain
split on the appropriate resolution of this issue.92
For the purpose of our discussion here, it is interesting to analyze
what the trial court did in Peevyhouse. Rather than grant one of the two
competing pure measures of recovery, the trial court awarded the plain-
tiffs $5000.93 This solution, it turns out, was not merely a split-the-differ-
ence compromise, as it might be perceived; it was a remedy that closely
resembled the bargain-mimicking outcome. It is well documented that
when the contract between the plaintiffs and Garland Coal was signed,
the plaintiffs wielded strong bargaining leverage.94 They were not partic-
ularly eager to enter the contract, and when they eventually agreed, they
leveraged their bargaining strength and insisted on including a restora-
tion clause.95 In fact, they waived their right to receive the customary
upfront restoration allowance of $3,000—close to the entire value of the
farm—in order to secure that restoration clause.96 Thus, if instead of a
restoration clause the plaintiffs had bargained for an explicit liquidated
allowance to fund self-managed restoration, it would have been roughly
$3,000—the sum they traded away for the restoration clause, not $29,000,
or $300. The jury award of $5,000, therefore, came close to mimicking
the bargained-for remedy of $3,000, augmented by lost royalties and inci-
dental costs arising from breach, delay, and trial.
The choice of cost-of-completion versus diminution-in-value is a fun-
damental and controversial one, leading to seemingly conflicting out-
comes across cases.97 It is an ongoing struggle for contracts scholars to
provide a descriptive theory of the result reached by courts. Why do
90. 382 P.2d 109, 111 (Okla. 1962).

91. Id. at 111–12, 120.
92. See, e.g., Timothy J. Muris, Cost of Completion or Diminution in Market Value:
The Relevance of Subjective Value, 12 J. Legal Stud. 379, 384–92 (1983) (exploring a
systematic understanding of case outcomes); Alan Schwartz & Robert E. Scott, Market
Damages, Efficient Contracting, and the Economic Waste Fallacy, 108 Colum. L. Rev. 1610,
1620–33 (2008) (surveying prevalence of economic waste notion in case law).
93. Peevyhouse, 382 P.2d at 111.
94. See Judith L. Maute, Peevyhouse v. Garland Coal & Mining Co. Revisited: The
Ballad of Willie and Lucille, 89 Nw. U. L. Rev. 1341, 1358–64 (1995) (describing
negotiations by Peevyhouses to protect their interests by insisting on terms that went
beyond standard industry provisions).
95. Id. at 1365–66.
96. Id. at 1358, 1363.
97. See Patricia H. Marschall, Willfulness: A Crucial Factor in Choosing Remedies for
Breach of Contract, 24 Ariz. L. Rev. 733, 734–36 (1982) (arguing willfulness is a factor
explaining choice of remedy); see also sources cited supra note 92. R
some plaintiffs get the former, usually higher, measure, while others re-
ceive the latter, stingier recovery? Criteria such as the willfulness of the
breach and the disproportionality of the cost of completion are only par-
tial and ad hoc organizing factors. The bargain-mimicking principle, I
argue, can bolster our understanding of case outcomes. Promisees who
had the superior bargaining power to insist on a completed performance,
like the Peevyhouses, should be entitled to the more generous measure.
Providing higher damages to parties with superior bargaining powers
mimics the high-end liquidated damages clauses they would have bar-
gained for.
This idea, focusing on the ex ante bargaining power of the transac-
tors, underlies Cardozo’s famous but cryptic distinction between “com-
mon chattel” and “a mansion or a skyscraper.”98 Why did Cardozo think
courts should allow the margin of noncompletion to be greater (and the
remedy smaller) in the case of common chattels or, as understood by a
later court, when the client purchased a stock floor plan house,99 but
require stricter compliance and award the higher cost-of-completion
measure for mansions? Plausibly, clients who purchase common chattel
and stock floor plan homes have less bargaining power against sellers and
little ex ante leverage to demand strict adherence to detailed specifica-
tions or the cost-of-completion remedy when tender is less than perfect.
But when mansions and skyscrapers are designed, the client is often in a
stronger bargaining position. Hence, when the aggrieved party had the
ex ante bargaining power to insist on precise tender of performance,
courts award the more generous measure.100
Remedies for breach of contract are not the sort of gap fillers that
have solely distributive effects. A long and distinguished literature has
shown that, through their effect on performance and reliance decisions,
remedies significantly influence the overall surplus.101 Thus, there is a
strong argument that contract gaps concerning remedies ought to be fil-
led with surplus-maximizing—rather than bargain-mimicking—terms.
Indeed, in the context of Peevyhouse, commentators have expressed con-
cerns about how the cost-of-completion measure would influence incen-
98. Jacob & Youngs, Inc. v. Kent, 129 N.E. 889, 890 (N.Y. 1921).
99. Plante v. Jacobs, 103 N.W.2d 296, 298–99 (Wis. 1960) (holding that in stock floor
plan house, a small shift of a wall does not entitle buyer to cost of repair measure).
100. See, e.g., Groves v. John Wunder Co., 286 N.W. 235, 238 (Minn. 1939) (finding
cost of performance to be appropriate measure of damages even though this cost was
much higher than decrease in value caused by breach); O.W. Grun Roofing & Constr. Co.
v. Cope, 529 S.W.2d 258, 262–63 (Tex. Civ. App. 1975) (noting that homeowners
contracting for amenities can insist on perfect tender to their specifications); see also
Marvin A. Chirelstein, Concepts and Case Analysis in the Law of Contracts 174 (4th ed.
2001) (arguing that recovery should equal amount promisee could have bargained for at
agreement stage); Robert A. Hillman, Principles of Contract Law 140 (2004) (explaining
that court should have considered “nature of the parties’ bargaining over the restoration
clause at the time of contracting” since this “would have shed light on Groves’ motives”).
101. See, e.g., Steven Shavell, Foundations of Economic Analysis of Law 304–12
(2004) (analyzing effect of remedies on overall surplus).
tives to breach or perform.102 And yet, despite this fundamental concern,

it is quite plausible that the damage measure chosen in cases like
Peevyhouse would have only a distributive effect. If the damage measure
were too high it would induce inefficient performance and the parties
would likely renegotiate. And if the damage measure is too low it would
induce inefficient breach, and again the parties would have an incentive
to renegotiate.103 Thus, to the extent that the damages merely affect the
parties’ bargaining positions in the renegotiation phase, but not the per-
formance outcome, the argument that such damages should reflect the
ex ante bargaining power of the parties—thereby saving them the trouble
of explicitly stipulating these damages in the contract—is all the more
compelling.
IV. MAXIMALLY TOLERABLE TERMS104

When bargaining power is unevenly distributed, a stronger party
would naturally use its bargaining leverage to draft one-sided, self-serving
terms. But the doctrines of unconscionability and duress (as well as other
rules) grant courts the power to invalidate excessively one-sided terms,
thereby preventing stronger parties from overreaching. Once an exces-
sively one-sided term is vacated, the court needs to fill the gap that is
created with an alternative provision.
There are several possible heuristics that can shape the way in which
this sort of gap is filled. First, a court could replace the excessive term
with a midrange, majoritarian term that it deems reasonable. Alterna-
tively, a court could plug in a term that is least favorable to the overreach-
ing party, as a penalty for overreaching and as an incentive not to over-
reach. Finally, a court could fill the gap with a bargain-mimicking term
that is still one-sided and favorable to the strong party. Admittedly, there
is something paradoxical about a bargain-mimicking principle in this
context. The term that best reflects the division of bargaining power was
in fact written in the contract, and yet it was found unenforceable under
a policy aimed at limiting the reach of bargaining power. A pure bargain-
mimicking term would be equivalent to that offensive term; surely, the
court would not reinstate the same term it has just struck down. The
practical effect of the bargain-mimicking principle in this setting would
be to prescribe a “maximally tolerable” term—a term that is still one-
102. See, e.g., Posner, supra note 3, at 121 (suggesting that overcompensatory R
remedies would make efficient breach more costly).
103. Ian Ayres & Kristin Madison, Threatening Inefficient Performance of Injunctions
and Contracts, 148 U. Pa. L. Rev. 45, 95–98 (1999) (discussing effect of damage measure
on subsequent rounds of bargaining over release from inefficient performance and
inefficient breach).
104. For background on the concept of maximally tolerable terms, see generally Omri
Ben-Shahar, How to Repair Unconscionable Contracts (Univ. of Chi. Law Sch., John M.
Olin Law & Economics, Working Paper No. 417, 2008), available at http://ssrn.com/
abstract=1082926 (on file with the Columbia Law Review).
sided, still favorable to the drafting party, but which falls just within the
range that is considered tolerable. Once the offensive term is replaced by
a term that is within the tolerable range, even if only barely so, there is no
remaining justification for intervention.
Thus, for example, if the excessive term were an intolerably high
price, the legal intervention would be to lower the price. If we analogize
the process of judicial intervention in the contract to a force that pulls
the price from its currently intolerable level toward the permissible re-
gion, the force gradually weakens as the price gets closer to the tolerable
level, and vanishes entirely once this level is hit. The point at which this
adjustment process is no longer justified is not the midrange,
majoritarian, most balanced term. Instead, the justification disappears at
the maximally tolerable price, which, though it remains one-sided, is not
as intolerable as the original term. Once this term is set, the weaker party
no longer has a reasonable basis for demanding additional redress.105
There are numerous instances in which courts apply maximally toler-
able terms. The doctrine of partial enforcement is one such example.106
Under this doctrine, a court is authorized to reform an unreasonable
term in a contract and enforce it to the extent necessary to avoid the
unreasonableness.107 The most common application of this technique
involves noncompete clauses that are excessive in either duration or geo-
graphic scope. In most states, courts repair excessive noncompete terms
by reducing them to the maximally tolerable level.108
At times, the maximally tolerable level is defined explicitly by statute.
Some states have enacted bright line rules stating the maximal duration
of noncompete clauses in employment contracts.109 There, only the in-
crement of the restraint that is socially intolerable is eliminated; the rest
stands.110 In other states there is no bright line statute. There too, courts
reduce the noncompete term, bringing it down to a level that is maxi-
mally tolerable. The restraint “is not enforceable beyond the time or area
considered reasonable by the [c]ourt.”111
105. This rationale is recognized by Corbin: “[T]he line [representing the

enforceable term] must be drawn somewhere, and it is drawn at the point where the
protection to which the buyer is justly entitled ends.” Arthur L. Corbin, A Comment on
Beit v. Beit, 23 Conn. B.J. 43, 46 (1949).
106. 15 Grace McLane Geisel, Corbin on Contracts § 89.4–.5, at 626–31 (Joseph M.
Perillo ed., rev. ed. 2003) (discussing practice of severing offensive portion of contract and
enforcing remainder).
107. Id. at 629 (surveying cases in which courts partially enforced an overreaching
term).
108. See, e.g., Harlan M. Blake, Employee Agreements Not to Compete, 73 Harv. L.
Rev. 625, 646–51 (1960) (exploring common law treatment of noncompete clauses).
109. See, e.g., Fla. Stat. Ann. § 542.335(1)(d)1 (West 2007) (“[A] court . . . shall
presume unreasonable in time any restraint more than 2 years in duration.”).
110. See, e.g., Flickenger v. R.J. Fitzgerald & Co., 732 So. 2d 33, 34–35 (Fla. Dist. Ct.
App. 1999) (holding that three year duration is excessive and reducing it to maximally
tolerable term of two years).
111. Justin Belt Co. v. Yost, 502 S.W.2d 681, 685 (Tex. 1974).
Another example of the use of maximally tolerable terms relates to

the doctrine of unconscionability. Corbin noted in the context of a
money loan that “a contract that requires the payment of a very high rate
of interest will be enforced, up to the point at which ‘unconscionability’
becomes an operative factor.”112 In the context of unconscionable arbi-
tration clauses, courts can replace an arbitration mandate that includes
unreasonable terms with one that is tolerable. For example, in Brower v.
Gateway 2000, Inc.—a leading New York unconscionability decision—the
court held that a term mandating arbitration in the International
Chamber of Commerce forum was unconscionable because of an exces-
sive $4,000 filing fee.113 It also held, though, that Gateway could cure this
defect by agreeing to arbitrate in another forum, if it entailed filing fees
that were not unconscionable.114 The consumer still viewed this as a one-
sided provision, but once the provision was tolerable, the court found no
grounds for further intervention.115
Maximally tolerable terms can also be used in the context of the le-
gal control over liquidated damages clauses. Courts do not enforce liqui-
dated damage terms that are clearly excessive and punitive,116 but what is
the damage term that courts supply instead? Does the court replace the
liquidated damages term with its own ex post estimate of the loss, equal
to the average or most reasonable compensatory measure? Or does it use
a bargain-mimicking term, somewhere at the high-end estimate of expec-
tation damages—the level that most closely reflects the division of bar-
gaining power?
While there are statements throughout American case law that reject
the use of a bargain-mimicking approach in this context,117 there are
many other legal traditions that appear to directly endorse this approach.
Under Israeli contract law, for example, courts are instructed to reduce
excessive damages to a level that reflects the magnitude of loss reasonably
expected at the time of contracting. In one case where an Israeli court
reduced liquidated damages, it explicitly set the damages above actual
harm, at a level equaling “the maximal amount the parties could have
anticipated as possible harm from delay.”118 Furthermore, a leading
Israeli commentary has stated that excessive liquidated damages should
be reduced
to the highest level the court regards as reasonably related to
the harm anticipated at the time of contracting . . . that is, re-
duced to the measure closest to the agreed sum, such that if that
112. 1 Arthur L. Corbin, Contracts § 129, at 402 (1964).

113. 676 N.Y.S.2d 569, 573–74 (App. Div. 1998).
114. Id. at 575.
115. Id.
116. Farnsworth, supra note 3, at 811–12. R
117. For an explicit rejection of the reduce-and-enforce methodology in penalty
clauses, see Cad Cam, Inc. v. Underwood, 521 N.E.2d 498, 503 (Ohio Ct. App. 1987).
118. CA 539/92 Zaken v. Ziva [1994] 1, 4 (Isr.) (on file with the Columbia Law Review).
measure were the one agreed upon in the first place, the court
would not have been justified in reducing it.119
This, in other words, is the maximally tolerable level, the term within
the reasonable range that comes closest to mimicking the parties’ bar-
gaining power.
CONCLUSION
When we think about courts interpreting contracts and supplying
missing terms, we do not usually regard it as a distributive task. Unless
there is clear evidence about actual but imperfectly specified intent,
courts are instructed to identify the “reasonable” term, the most efficient
term, the majoritarian term, or some other uniquely distinguished con-
tent. Yet none of these criteria are defined with respect to their distribu-
tive effect. This Essay suggests that in fact the gap-filling task of courts is
often purely distributive. In many contexts—as in the case of a missing
price—there are several potential provisions courts can choose from, all
of which satisfy the reasonable or efficient criteria, but differ in their dis-
tributive allocation. Thus, contract law needs to provide an appropriate
criterion for choosing the default rule in these instances of distributive
gaps.
It is one thing to argue that we are faced with a recurring contractual
gap that needs a systematic solution, for which existing gap-filling princi-
ples do not apply. It is quite another to propose a bargain-mimicking
solution that favors the stronger party. Some readers may object to this
proposed criterion as morally unjust. If there is a distributive aspect to
gap filling, they would argue, why not turn it into an opportunity to favor
the weaker party? Should not the law reverse the outcomes of unfettered,
unequal bargaining power?120 The answer I provide in this Essay is that
distributive gaps cannot effectively become an occasion for redistribution.
I argue that gap fillers cannot effectively favor the weak party. Gap fillers
that go against the background allocation of bargaining power will only
induce stronger parties to exert express, self-favoring terms. The more
redistributive gap fillers are, the less likely it is they will come into play.
Identifying the division of bargaining power between parties can be
difficult and error prone. In this Essay I suggest that the problem might
not be as challenging as it initially appears, but nonetheless concede that
119. Uri Yadin, Hok Hahozim: Terufot Beshel Hafarat Hozeh 1970 [Contract Law:
Remedies for Breach of Contract 1970] 132 (2d ed. 1979) (Hebrew text); see also Eyal
Zamir et al., Haperush Hakatsar Lehukim Bamishpat Haprati [Brief Commentary on Law
Relating to Private Law] 302 (2d ed. 1996) (Hebrew text) (stating, as translated, that “the
measure of reduction of liquidated damages ought to be to the level for which the element
of excessiveness no longer applies . . . [such that] if that level was set in the first place, it
would not have been reduced by the court”).
120. See, e.g., Symposium, Power, Inequality and the Bargain: The Role of
Bargaining Power in the Law of Contract, 2006 Mich. St. L. Rev. 841 (describing various
approaches to dealing with bargaining power asymmetries).
it poses a significant obstacle to the implementation of a bargain-mimick-

ing regime. Still, no matter how daunting this judicial task may be, the
difficulty alone cannot justify utilizing a different gap-filling criterion.
There may well be easier-to-apply criteria that are straightforward and re-
quire very little case-specific information. The problem, however, is that
if parties expect courts to supply gap fillers other than the bargain-mim-
icking ones, they will be forced to draft contracts with fewer gaps and less
flexibility.
The concept of bargain-mimicking default rules is quite general and
can apply beyond the standard contractual context explored in this Essay,
anytime a contest occurs over distribution. It is applicable, for example,
as a gap-filling principle for unclear legislation, by justifying statutory in-
terpretation that mimics the preferences, or the division, of power within
the enacting polity.121
The idea of the law mimicking the unequal division of bargaining
power may seem to conflict with notions of fairness.122 In the end,
though, it is merely a sober descriptive principle, which can be harnessed
to advance any normative agenda, including redistributive ones. It is a
reminder that gap-filling regimes cannot effectively level the playing field
or advance any significant redistribution. Even if default rules are sticky,
parties that have greater bargaining power are likely to opt out of redis-
tributive gap fillers. Still, this is not a manifesto against redistribution.
Laws can surely accomplish redistribution in effective ways. If we are seri-
ous about helping weaker parties secure better deals through contract
law, we probably need to get our hands dirty and engage in more aggres-
sive forms of market regulation including mandatory quality rules and
regulation of the bargaining process. Using redistributive gap fillers
might create the ad hoc impression that contract law is sensitive to the
plight of weaker parties, but it would have very little, if any, systematic
effect.
121. See Einer Elhauge, Statutory Default Rules: How to Interpret Unclear
Legislation 5–12 (2008).
122. Ocean Accident & Guarantee Corp. v. Indus. Comm’n of Ariz., 257 P. 644, 645
(Ariz. 1927) (“Our enlightened modern thought realizes that an equality of bargaining
power between two such unequal parties is impossible, and has attempted to equalize the
balance . . . .”).

Essay: A Bargaining Power Theory of Default Rules

Uploaded by

Copyright:

Available Formats

Essay: A Bargaining Power Theory of Default Rules

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

Essay: A Bargaining Power Theory of Default Rules

Uploaded by

Copyright:

Available Formats

\\server05\productn\C\COL\109-2\COL204.

txt unknown Seq: 1 23-FEB-09 11:10

A BARGAINING POWER THEORY OF DEFAULT RULES

2009] BARGAINING POWER THEORY OF DEFAULT RULES 397

398 COLUMBIA LAW REVIEW [Vol. 109:396

6. 556 N.E.2d 515, 518–20 (Ohio 1990).

2009] BARGAINING POWER THEORY OF DEFAULT RULES 399

yard”—and more than one reasonable version could be offered. Still, if

400 COLUMBIA LAW REVIEW [Vol. 109:396

2009] BARGAINING POWER THEORY OF DEFAULT RULES 401

402 COLUMBIA LAW REVIEW [Vol. 109:396

2009] BARGAINING POWER THEORY OF DEFAULT RULES 403

missing price does not render a contract unenforceable, so long as there

concluded . . . .”); Farnsworth, supra note 3, at 202–03, 210–11 (distinguishing R

404 COLUMBIA LAW REVIEW [Vol. 109:396

B. Mimic One Party’s Will

2009] BARGAINING POWER THEORY OF DEFAULT RULES 405

406 COLUMBIA LAW REVIEW [Vol. 109:396

price to be agreed upon would be supplemented with a high, pro-land-

C. Majoritarian Versus Bargain-Mimicking Terms

29. See 5 Margaret N. Kniffin, Corbin on Contracts: Interpretation of Contracts

2009] BARGAINING POWER THEORY OF DEFAULT RULES 407

32. Craswell, supra note 2, at 4–5. R

408 COLUMBIA LAW REVIEW [Vol. 109:396

mimicking criterion when applied to specific situations where the parties’

D. Can Courts Identify the Bargain-Mimicking Terms?

2009] BARGAINING POWER THEORY OF DEFAULT RULES 409

In fact, courts already quite regularly refer to bargaining power as a

410 COLUMBIA LAW REVIEW [Vol. 109:396

2009] BARGAINING POWER THEORY OF DEFAULT RULES 411

II. ARE BARGAIN-MIMICKING TERMS DESIRABLE?

412 COLUMBIA LAW REVIEW [Vol. 109:396

secure her own advantages through bargaining. If anything, it is the weak

2009] BARGAINING POWER THEORY OF DEFAULT RULES 413

B. Flexibility in Contract Performance

414 COLUMBIA LAW REVIEW [Vol. 109:396

is a clear statement of the parties’ intent to be bound, thereby making the

2009] BARGAINING POWER THEORY OF DEFAULT RULES 415

to allow stronger parties to utilize this self-promoting technique, flexible

416 COLUMBIA LAW REVIEW [Vol. 109:396

2009] BARGAINING POWER THEORY OF DEFAULT RULES 417

418 COLUMBIA LAW REVIEW [Vol. 109:396

III. BARGAIN-MIMICKING TERMS IN ACTION

67. Restatement (Second) of Contracts § 204 cmt. d (1979) (emphasis added).

2009] BARGAINING POWER THEORY OF DEFAULT RULES 419

particularly strong when applied to distributive issues, where the one-

420 COLUMBIA LAW REVIEW [Vol. 109:396

2009] BARGAINING POWER THEORY OF DEFAULT RULES 421

Posner’s decision in the Original Great American Chocolate Chip Cookie

422 COLUMBIA LAW REVIEW [Vol. 109:396

2009] BARGAINING POWER THEORY OF DEFAULT RULES 423

not hinge on an efficiency analysis, or on the assumption of a hypotheti-

424 COLUMBIA LAW REVIEW [Vol. 109:396

damages applied when a stripmining company breached its promise to

90. 382 P.2d 109, 111 (Okla. 1962).

2009] BARGAINING POWER THEORY OF DEFAULT RULES 425

426 COLUMBIA LAW REVIEW [Vol. 109:396

tives to breach or perform.102 And yet, despite this fundamental concern,

IV. MAXIMALLY TOLERABLE TERMS104