Available via license: CC BY 4.0
Content may be subject to copyright.
Governing Relationships: Theory, Evidence,
and Opportunities 25
Ricard Gil and Giorgio Zanarone
Contents
25.1 Introduction .............................................................................. 622
25.2 Theoretical Framework .................................................................. 625
25.2.1 The Canonical Model of Relational Incentive Contracts ...................... 625
25.2.2 The Clarity Problem and the Role of Formality ............................... 634
25.3 From Theory to the Data ................................................................. 636
25.3.1 Surplus Increases in the Shadow of the Future (for Given Contract
Provisions) . . . . .................................................................. 637
25.3.2 Under Relational Contracting, Responses to and the Design of Court-
Enforceable Contracts Differ from Arm’s-Length Contracting ............... 639
25.3.3 A Better Contractual Technology May Crowd Out Relational Contracts .... 640
25.3.4 Relational Contracts and Organizational Structure ............................ 642
25.3.5 Future Challenges: Empirical Analyses of the Clarity Problem . . . . . . . . . . . . . . 643
25.4 Concluding Remarks ..................................................................... 645
References ....................................................................................... 646
Abstract
In this chapter, we provide a framework to investigate the interaction between the
“hardware”(formal contracting) and the “software”(relational contracting) of
economic collaborations, both within and between firms. First, we present a
simple formulation of the “canonical model”of relational contracting and high-
light its key empirical implications. Next, we discuss selected empirical works
that shed light on the relevance and testability of the model’s predictions. While
significant progress has been made at measuring the complementarity between
formal and relational contracts, our investigation highlights several implications
R. Gil (*)
IESE Business School, Barcelona, Spain
e-mail: rgils@iese.edu
G. Zanarone
Hautes Etudes Commerciales, University of Lausanne, Lausanne, Switzerland
e-mail: giorgio.zanarone@unil.ch
© The Author(s) 2025
C. Ménard, M. M. Shirley (eds.), Handbook of New Institutional Economics,
https://doi.org/10.1007/978-3-031-50810-3_25
621
of the model that are consistent with anecdotes yet lack systematic empirical
verification. Testing these implications is an exciting opportunity for future
scholars in the field.
Keywords
Governance · Formal contracts · Relational contracts · Enforcement · Evidence
25.1 Introduction
What makes self-interested individuals cooperate, both within and across firm
boundaries? Economists have developed two answers to this question. On the one
hand, incentives to cooperate can be created through carefully designed, farsighted
contracts, which a court (or another third party, such as an arbitrator or, in certain
societies, a community/customary judge) will enforce.
1
This is the perspective
adopted by much of agency theory, contract theory, and transaction cost economics.
On the other hand, people can develop trust and long-term relationships that make
their promises to cooperate credible even when third-party enforcement is absent or
dysfunctional. This perspective, initially championed by legal scholars and sociol-
ogists (e.g., Granovetter, 1985; Macaulay, 1963), has been later incorporated into
models of self-enforcing agreements, which economists have come to label “rela-
tional contracts.”
2
The economic literatures on contracts and relationships as mechanisms to govern
cooperation are vast and have been extensively reviewed elsewhere (Gil &
Zanarone, 2017,2018; Lafontaine & Slade, 2013; Macchiavello, 2022; Malcomson,
2013). A key question that remains empirically understudied, and to which this essay
is devoted, is how these two dimensions of governance interact with each other. This
question is especially important in developed economies, where well-functioning
courts and third-party enforcement systems make it possible for firms to use both
1
Both transaction cost economics and agency theory acknowledge that most contracts are inher-
ently incomplete and thus cannot directly specify desired cooperation under all possible contingen-
cies. Accordingly, these branches of economics emphasize contractual provisions that realign the
parties’incentives, thereby indirectly facilitating or incentivizing cooperation. See Lazear and Oyer
(2013) and Lafontaine and Slade (2013) for a comprehensive review of the empirical evidence,
respectively, within firms and between.
2
While most economists (including ourselves) use the term “relational contracts”as a synonym of
self-enforcing agreements, the term was originally coined by legal scholars to define nonstandard
agreements that are meant to flexibly govern an ongoing and idiosyncratic relationship (MacNeil,
1978). This definition was also adopted by some prominent transaction-cost economists
(Williamson, 1979,1991). According to this second usage of the term, relational contracts may
well be enforced by courts but based on a special body of contract law that aims to fill gaps in the
agreement in line with the parties’intentions (Schwartz, 1992). See Gil and Zanarone (2017) and
Masten (2021) for a deeper discussion of the different definitions of relational contracting in the
literature.
622 R. Gil and G. Zanarone
contractual and relational mechanisms to govern their collaborations. Examples of
the interaction between the two are ubiquitous and diverse. On the one hand, there
are studies showing that even in relationships governed by detailed and court-
enforceable contracts, the extent of cooperation is sensitive to the “shadow of the
future,”that is, to the existence and value of a long-term relationship between the
parties (see, most recently, Barron et al., 2020, on distributor-exhibitor agreements in
the Spanish movie industry, and Gil et al., 2022, on major-regional airline contracts
in the US airline industry, both of which are extensively discussed below). Moving
from interfirm collaborations to employment, leading manufacturers like Lincoln
Electric have become famous for jointly using contractual, court-enforceable piece
rates and discretionary subjective bonuses to incentivize their employees (Bartlett,
1998; Fast & Berg, 1975). Does the presence of court-enforceable contractual pro-
visions in these interfirm and intrafirm relationships strengthen (and complement)
the parties’trust-based, self-enforcing cooperation promises? If so, how?
On the other hand, recent surveys of business-to-business collaborations in the
USA show that even in contexts where litigating disputes in court is impractical, and
hence the enforcement of cooperation promises must fully rely on relational mech-
anisms, firms enter sophisticated formal contracts that are drafted by lawyers and
specify performance targets, monitoring mechanisms, and penalties and rewards in
great detail (Bernstein & Peterson, 2023; Hadfield & Bozovic, 2016). A similar
phenomenon occurs within firms. A company’s chief executive has ultimate author-
ity over all decisions that involve the use of the company’s resources, implying that
he or she cannot contractually commit not to exert such authority (Williamson,
1991). Thus, when chief executives delegate spending and decision authority to
divisional managers—and sometimes, as at Netflix (Hastings & Meyer, 2020) and
Oticon (Foss, 2003), even to middle managers and junior employees—they are
implicitly or explicitly making a promise that the latter need to trust, that is, a
relational promise. Indeed, as discussed below, there is evidence that delegation of
authority from headquarters to divisional managers is more likely in the presence of
mutual trust (Bloom et al., 2012). Despite the intrinsically relational and
non-contractual nature of authority delegation, company executives spend consid-
erable time crafting detailed and highly formal organizational charts that specify the
allocation of decision authority within their organizations. What role do contractual
and organizational formalism play in collaborations where the enforcement of
cooperation promises is purely relational?
While there are theoretical papers addressing the two sets of questions above
(we further elaborate on them below), the number of large-sample empirical studies
is remarkably scarce. This paucity of empirical research is in a way not surprising.
Obtaining data on contracts and formal governance (across relationships or over
time) is difficult due to the private and confidential nature of this information.
Measuring relational promises and the shadow of the future in long-term collabora-
tions is even harder because these variables are almost by definition not reported in
company documents. Thus, constructing a dataset that combines both sets of infor-
mation becomes almost a heroic task. The goal of this chapter is to draw upon
existing theoretical and empirical research to make the case that, albeit not easy,
25 Governing Relationships: Theory, Evidence, and Opportunities 623
investigating the interaction of formal and relational contracts (and governance,
more broadly) is both feasible and worthwhile.
We begin by presenting a simple version of the “canonical model”of relational
contracts, adapted from Gil and Zanarone (2018). While early theoretical analyses of
relational contracts go back to the 1980s (Bull, 1987; Klein, 1980; MacLeod &
Malcomson, 1989; Telser, 1981), the modeling of this phenomenon has later con-
verged toward a standard tractable framework developed by the subsequent papers
of Baker, Gibbons, and Murphy (1994,1999,2002) and Levin (2003). Because most
recent models draw upon and enrich the Baker-Gibbons-Murphy-Levin framework
(sometimes explicitly stating which of its assumptions they modify or relax), we
refer to this framework as the canonical model.
Our model captures the interaction of relational contracts with both formal
contracts (as in Baker et al., 1994)and formal organization design (as in Baker
et al., 1999) and, thus, connects to all the dimensions of our discussion above. The
model suggests three key insights. First, when contractual provisions alone cannot
support efficient cooperation, cooperation levels increase in the expected duration
and value of future interactions between the parties. This value is lost if promises are
broken and/or the relationship terminates so the greater it is, the more the parties can
trust that neither of them has an incentive to defect. Importantly, this insight applies
both to the case in which cooperation is a supplier’s or employee’s promise to exert
productive effort (first part of the model) and to the case in which cooperation is a
chief executive’s promise to respect the organizational chart’s allocation of authority
(second part of the model).
The second insight from the canonical model is that parties may enter imperfect
court-enforceable contracts not because they plan to behave according to their terms
but because, given the amount of cooperation that they target, these contracts reduce
the portion of such cooperation that must be enforced relationally and, thus, enable
the parties to select high levels of cooperation even if the value of their future
relationship and their mutual trust are somewhat limited. Moreover, we show that
when formal contracts are used to support an overarching relational agreement, their
provisions look quite different from those in arm’s-length contracts (i.e., contracts
that are supposed to fully describe and dictate the parties’behavior).
The third insight is that regardless of whether they are court-enforceable or not,
formal contractual provisions play an important role in ensuring that relational
cooperation promises are clear to both parties, thereby reducing the risk that good-
faith misunderstandings are interpreted as breach of trust and lead to the termination
of valuable relationships. That is, in the words of two leading scholars in the field,
formal contracts may play both a “credibility-building”role (our second insight) and
a“clarity-building”role (our third insight) in sustaining superior relational agree-
ments between collaborating parties (Gibbons & Henderson, 2012).
In the second part of the paper, we discuss recent empirical studies in the light of
the canonical model of relational contracting. Our purpose in doing so is not to offer
a detailed and comprehensive review of the relevant literature (which, as mentioned
above, is available elsewhere) but rather to highlight promising data and empirical
strategies that have been developed to address some questions on the interaction
624 R. Gil and G. Zanarone
between formal and relational contracts and how these data and strategies could be
further leveraged to investigate theoretical predictions and questions that are still
awaiting empirical scrutiny.
25.2 Theoretical Framework
In this section, we analyze a simple version of the canonical Baker-Gibbons-Mur-
phy-Levin model of relational incentives, adapted from Gil and Zanarone (2018), to
articulate the literature’s key insights on the interaction between contractual and
relational enforcement. Before we proceed, a few observations are in order. First, we
use a model (as opposed to verbal reasoning) to precisely lay down the economic
mechanisms underlying these insights. The model is therefore directed not so much
to economists already versed in the literature on relational contracts but rather to
readers (economists and not) who are approaching this literature for the first time.
Second, as in any model, there is a trade-off between generality and analytical
tractability. While the canonical model (analyzed in Sects. 25.2.1.1,25.2.1.2, and
25.2.1.3) focuses on a relatively specific contracting problem—namely, incentive
provision through pay-for-performance—the underlying economic forces are quite
general. We discuss generalizability of the model’s insights to other settings in Sect.
25.2.1.4. Lastly, while the canonical model focuses on how formal contractual
provisions complement relational ones through the threat of court-enforcement,
recent research emphasizes a different and complementary role of such formal
provisions—namely, building a shared understanding about the content of the
relational agreement. We discuss this “clarity-enhancing”role of contractual formal-
ism in Sect. 25.2.2.
25.2.1 The Canonical Model of Relational Incentive Contracts
25.2.1.1 Environment
Consider a “principal”(P), who hires an “agent”(A) to perform a task for an
indefinitely long number of periods. Both parties discount next-period payoffs by
the common factor δ∈[0, 1]. A convenient interpretation of δis as the exogenous
probability that, assuming both parties cooperate while in a relationship (in a way
that will be clarified momentarily), their relationship survives from one period to
another—for instance, because neither party dies, retires, or goes bankrupt. P and A
can be equivalently interpreted as parties in an interfirm collaboration (for instance, a
buyer and a supplier) or in an employment relationship (for instance, a manager and
a subordinate).
P hires A because A’s effort e(a real number) produces a valuable contribution
to P, y, which we will call P’s“individual output”(Gibbons, 2005) or, simply, output.
Formally, we assume y¼e, that is, output increases in A’s effort. Exerting effort
costs ceðÞ
1
2e2to A, who must therefore be compensated for doing so. In principle,
25 Governing Relationships: Theory, Evidence, and Opportunities 625
P could agree on a simple contract with A, which specifies the amount of effort to be
provided, in exchange for a fixed payment. Unfortunately, it is too costly for P to
directly measure A’s effort (think of an employee’s initiative and creativity, or an
offshore supplier’s investment in responsible labor practices). Thus, the only way in
which P can contractually induce A to provide effort is through an “incentive
contract”that links A’s compensation to some outcome that is sensitive to his effort.
The natural candidate outcome for such an incentive contract is the output y, that
is, A’s individual contribution to P. While it is plausible that P observes such
contribution, however, it is less plausible that she can describe it and record it in a
contract in such a way that a third party with little knowledge of the firm (a court) can
assess compliance with the contract’s terms. For instance, while a manufacturer
might realize through her repeated interaction with various suppliers that supplier
A’s components are better customized and engineered than those of supplier B, it
might be difficult for the manufacturer to objectively demonstrate that B’s compo-
nents have contributed more than A’s components to the market success of the final
product. We capture this basic idea by assuming that A’s output yis observable but
non-verifiable: the parties can informally agree on an incentive contract that rewards
A based on output but cannot expect a court to enforce such a contract, which must
therefore be relational.
While output-based incentive contracts cannot be enforced by courts, there may
be variables other than individual output that are sensitive to A’s effort and that can
be objectively and easily measured (real-world examples are an agent’s sales, a unit’s
or firm’s accounting profits, or a supplier’s timely delivery). For simplicity, we
assume there is one such variable, which we label p(for “measured performance”).
Importantly, we assume that while pis sensitive to A’s effort, it is a noisy proxy for
such effort, in the sense that A may also be able to affect it through an unproductive
action m(again, a positive real number), which we call “manipulation.”For instance,
agents on a sales commission may make inaccurate statements to induce more
customers to buy the product, and a supplier incentivized for timeliness may
economize on quality control. Formally, we assume p¼eþgm, where the
parameter g(for “gaming”) measures the extent of misalignment between individual
output and measured performance. Like effort, manipulation also requires some time
and dedication, and hence it is costly to A. We assume for expositional simplicity
that manipulation has the same cost as productive effort: cmðÞ
1
2m2.
Given the environment described above, the contractual relationship between P
and A proceeds as follows. At the beginning of each period, P offers an incentive
contract to A, which specifies A’s compensation as a function of output and mea-
sured performance: w(y,p)sþbp þBy. The term sin A’s compensation is a
contractual fixed payment (a salary if A is a subordinate employee, an input price if
A is a supplier). The term bis a contractual “piece rate”paid for each realized unit of
the performance measure p. Lastly, the term Bis a relational bonus, which is paid
based on P’s observation of A’s realized individual output, y. Because output cannot
be verified by a court (as discussed above), the relational bonus Bmust be self-
enforcing—that is, A must trust P to pay it as specified by the compensation formula.
626 R. Gil and G. Zanarone
Thus, the model allows for arm’s-length incentive contracts that solely rely on court
enforcement (b>0 and B¼0), purely relational contracts (B>0 and b¼0) that
solely rely on self-enforcement, and a combination of the two (b>0 and B>0).
Before analyzing the interaction between contractual and relational incentives, it
is useful to briefly discuss, as a benchmark, the ideal allocation of effort for the
principal-agent relationship. If there were no contracting frictions (i.e., if effort were
directly contractible), P and A would select the “first-best”actions, denoted by e
⁎
and
m
⁎
, that maximize the sum of their individual payoffs, or “joint surplus”:
Ve,mðÞyceðÞcmðÞ¼e1
2e2þm2:ð25:1Þ
Clearly, the first-best level of manipulation is zero given that manipulation is
costly but does not produce any value: m
⁎
¼0. Regarding productive effort, it is
efficient for A to exert it up to the point where its marginal contribution to the joint
surplus (derivative of ywith respect to e) equals its marginal cost (derivative of c(e)),
that is, e
⁎
¼1. The joint surplus resulting from these first-best actions is therefore
V⁎1
2.
25.2.1.2 Arm’s-Length Contracting
To build some intuition on how contracting frictions (the fact that effort and output
are non-contractible and that the performance measure pis manipulable) affect
incentive provision, we begin with analysis of the case where P can only use
contractual incentives: b>0 and B¼0. In this case, if A accepts the compensation
offered by P, she chooses effort and manipulation to maximize his own utility, that is,
the difference between compensation and the actions’costs:
u¼sþbp ceðÞcmðÞ¼sþbeþgmðÞ
1
2e2þm2:ð25:2Þ
The solution of A’s problem is given by the effort and manipulation response
functions e
F
(b) and m
F
(b), which equate A’s marginal benefit under a given contrac-
tual piece rate b(i.e., the derivatives of the variable compensation bp with respect to
eand m, respectively) to the marginal cost (defined as above):
eFbðÞb,mFbðÞbg:ð25:3Þ
Note that, because contractual incentives must be based on the distortionary
performance measure p, an increase in the piece rate braises not only A’s effort
but also her manipulation, and more so the greater the misalignment between
measured performance and A’s individual output, as measured by parameter g.
Anticipating A’s response to contractual incentives, P chooses the payment terms
sand bto maximize her own profit (the difference between output and
compensation),
25 Governing Relationships: Theory, Evidence, and Opportunities 627
π¼ysbp,ð25:4Þ
subject to the constraint that A’s total compensation is high enough to cover his costs
(or else A would reject P’s contractual offer). This condition, which can be stated
mathematically as u0, is known in the economics of contracts as “participation
constraint.”Clearly, it is optimal for P to offer the minimum fixed payment that
satisfies A’s participation constraint, which given (25.2) is:
sFce
FbðÞ þcm
FbðÞ bp:
After substituting s
F
into (25.4), P’s problem simplifies to choosing the piece rate
to maximize the joint surplus, given A’s anticipated response to the piece rate, that is,
P maximizes:
Ve
FbðÞ,mFbðÞ ¼eFbðÞce
FbðÞ cm
FbðÞ:
After substituting the response functions from (25.3), P’s problem can be further
rewritten as that of maximizing:
b1
2b21þgðÞ ð25:5Þ
The solution is given by the piece rate that equates P’s marginal benefit (deriv-
ative of output, the first term in (25.5)) to her marginal cost (derivative of A’s cost of
responding to incentives, the second term in (25.5)), that is:
bF1
1þg2,
which generates joint surplus VF1
21þg2
ðÞ
.
When output and measured performance are perfectly aligned, such that A cannot
increase the latter through manipulation (g¼0), the optimal arm’s-length contract
features high-powered incentives (b
F
¼1) that elicit first-best actions from A
(substituting b¼1 into (25.3), we obtain e
F
¼1, m
F
¼0). As the misalignment
parameter ggrows, the contractual piece rate decreases (b
F
<1), and the (now
second-best) optimal contract sacrifices some productive effort to contain
unproductive manipulation.
25.2.1.3 Optimal Combination of Contractual and Relational Incentives
Let us now consider the more general case in which P can use both contractual piece
rates (b) and relational bonuses (B) to incentivize A—that is, P offers compensation
w(y,p)sþbp þBy. If A accepts, she chooses the two actions to maximize his
payoff given the contract, that is:
u¼sþbp þBy ceðÞcmðÞ¼sþbeþgmðÞþBe 1
2e2þm2:ð25:6Þ
628 R. Gil and G. Zanarone
As usual, the solution is given by the actions that equate A’s marginal benefit and
cost. For manipulation, these are the same as under arm’s-length contracting
(because the relational bonus is conditional on output, which does not depend on
manipulation). Thus, A’s manipulation continues to be m
F
(b). For productive effort,
A’s marginal cost is the same as before, but the marginal benefit is now larger
because if A exerts one more unit of productive effort, she now increases his
compensation through both the piece rate and the relational bonus:
eRb,BðÞbþB:
Anticipating A’s response (and taking into account, as before, that the optimal
salary is the minimum value of sthat induces A to accept the agreement), for a given
relational bonus B, P chooses the piece rate bto maximize the joint surplus:
eRb,BðÞce
Rb,BðÞcm
FbðÞ:ð25:7Þ
After substituting A’s response functions into the surplus function above and
equating P’s marginal benefit and marginal cost of b, the optimal piece rate for a
given relational bonus is:
bBðÞ
1B
1þg2:ð25:8Þ
It is also useful to define the joint surplus that P appropriates as a function of the
piece rate in (25.8) and the relational bonus B:
VRBðÞeRbBðÞ,BðÞce
RbBðÞ,BðÞcm
FbBðÞðÞ:
While our analysis is not complete yet, the piece rate formula in (25.8) already
provides some insights. If output and measured performance are completely mis-
aligned (gtends to infinity), or if B¼1, so that the relational bonus alone can induce
A to exert firs-best effort without manipulation, P does not benefit from adding a
distortionary contractual piece rate to A’s compensation: b(B¼1) ¼0. Aside from
these limit cases, adding a contractual piece rate to the relational bonus is typically
beneficial: b(B)>0, decreasing in the misalignment parameter gand in the size of
the relational bonus B.
To complete our analysis of the interaction between contractual and relational
incentives and draw empirical implications from the model, we now study P’s
choice of the relational bonus B. Recall that, because P’spromisetopayBcannot
be enforced by a court, the relational bonus must be self-enforcing. In relational
contracting models (which are special cases of repeated game models), self-
enforcement arises from the threat of terminating the relationship: both parties
agree at the outset (explicitly or implicitly) that if P ever fails to pay the promised
bonus, A will never believe any of P’s promises in the future, and thus P will revert
to the arm’s-length incentive contracts analyzed in Sect. 25.2.1.2 (which do not
25 Governing Relationships: Theory, Evidence, and Opportunities 629
require trust and credibility to be enforced), earning the joint surplus V
F
forever
after. This threat to terminate the relational contract ensures self-enforcement if in
any given period, P’s present discounted gains from the relational contract are
larger than the relational bonus she is tempted to default on in the current period.
Taking into account that P’s per period gain from honoring the promised bonus is
the difference between surplus under relational contracting, V
R
(B), and surplus
under arm’s-length contracting, V
F
, and that P always discounts next-period gains
by δ, the condition for a relational bonus Bto be self-enforcing can be formulated
mathematically as:
δVRBðÞVFþδ2VRBðÞVFþδ3VRBðÞVFþ... B,
which can be rewritten as
δ
1δVRBðÞVFB:ð25:9Þ
Thus, P’s problem is to choose the relational bonus Bto maximize the per period
joint surplus V(B), subject to self-enforcement condition (25.9).
The problem above highlights the trade-off P faces when choosing how much to
rely on the relational bonus to incentivize A. On the one hand, increasing B(and thus
decreasing the contractual piece rate b) is attractive for P, as it allows her to elicit
more productive effort without having to also compensate A’s unproductive manip-
ulation that the piece rate contract incentivizes. On the other hand, promising a larger
relational bonus Bincreases P’s present temptation to renege and, hence, makes it
harder to keep the relational contract self-enforcing (the self-enforcement condition
(25.9) becomes harder to satisfy as Bgrows).
If the parties value the future of their relationship enough (because the exogenous
probability of continuation, δ, or the relationship’s per period value, V
R
(B), is high,
or alternatively, because the “fallback option”of reverting to arm’s-length
contracting, V
F
, is not too attractive), then a relational bonus high enough to induce
A to exert the first-best effort level is self-enforcing. Thus, there is no need for
distortionary piece rates—the optimal compensation agreement is purely relational:
B
R
¼1, and b
R
¼0. For lower value of the parties’future relationship, the self-
enforcement constraint (25.9) becomes binding—that is, a bonus that elicit first-best
effort is not credible for A, and the highest relational bonus consistent with (25.9)is
B
R
<1, the largest positive number that satisfies (25.9) with equality. In such a
scenario, adding a contractual piece rate b
R
>0 to the relational bonus becomes
valuable (despite the wasteful manipulation activity it induces) because given the
limited credibility of P’s promises, the piece rate allows P to elicit a higher effort
from A than the relational bonus alone.
It is interesting to contrast these findings with the common view of collaborative
relationships as “hand-shake agreements”(Macaulay, 1963). While economic
modeling does support Macaulay’s insight that collaborating parties do not “play
by the book,”relying on contracts to dictate their behavior, it also suggests that
630 R. Gil and G. Zanarone
court-enforceable contracts play an important role in making relational collaboration
credible (Klein, 2000).
25.2.1.4 Empirical Implications
The simple model developed above suggests specific predictions on the comple-
mentarity between contractual and relational incentives that one can test empirically.
The first prediction is that among similar relationships governed by the same contract
terms, those with a longer shadow of the future (higher levels of δin self-enforce-
ment condition (25.9)) will be more cooperative (higher effort e, lower manipulation
m) and more productive (higher output and surplus). If, instead, these collaborations
did not rely on relational incentives to complement contractual ones (i.e., if they
were pure spot-market, arm’s-length relationships), the shadow of the future should
not affect their productivity holding the contractual incentives constant. One can test
this prediction by regressing effort, cooperation, or surplus in collaborative relation-
ships governed by similar contract terms on measures of the shadow of the future.
We discuss examples of how to develop such an empirical strategy in the next
section.
A second prediction of the model is that parties respond to contractual incentives
differently when these are used in combination with relational ones. One can see that
from our formal analysis above by noticing that under a given contractual piece rate
b, A chooses effort e
F
(b)¼bif no relational bonus is present, whereas A chooses the
higher effort e
R
(b,B)¼bþBif a relational bonus B>0 is also present. One could
test this prediction on the complementarity between contractual and relational
incentives by checking whether the observed cooperation in a collaboration is
systematically profitable or unprofitable under the court-enforceable contract terms
used by the parties. Finding that observed cooperation is unprofitable under the
contract would be consistent with relational incentives playing a complementary
role. We elaborate further on this empirical strategy in the next section.
A third prediction of the model is that relational incentives (B) affect the parties’
choice of contractual incentives (b). To see this point analytically, recall from
Eq. (25.8) that the optimal contractual piece rate is bBðÞ
1B
1þg2, which varies
(in our setting, decreases) in the relational bonus B. Thus, collaborations character-
ized by different levels of the optimal relational bonus (for instance, due to different
shadows of the future, as measured by δ) will select different contractual piece rates.
The insight that the design of court-enforceable contracts is affected by the presence
of an overarching relational contract is quite general and has been analyzed in
several refinements or applications of the canonical model.
3
3
See, for instance, Battigalli and Maggi (2008) and Kvaløy and Olsen (2009) for general models of
the interaction between contractual and relational provisions in incomplete agreements. See also
Zanarone (2013) for a model of the interaction between relational contracts and the contractual
allocation of decision rights in franchising. Finally, Argyres et al. (2020) provide for a model of the
interaction between relational contracts and outsourcing scope in buyer-supplier relationships.
25 Governing Relationships: Theory, Evidence, and Opportunities 631
A fourth and last prediction of the canonical model pertains to the relationship
between contractual technology (monitoring, legal systems, courts) and relational
contracting. Specifically, the model predicts that when relational incentives are used,
improvements in the contractual technology may increase or decrease surplus, that
is, contractual technology and relational contracts may be substitutes or comple-
ments (Baker et al., 1994,2002). To see this point in our formal analysis, consider a
reduction in the misalignment parameter g. If P were solely relying on the contrac-
tual piece rate bto incentivize A, a reduction in misalignment would reduce
unproductive manipulation (m
F
(b)¼gb decreases in g). This reduction in manipu-
lation would encourage P to raise the piece rate, which in turn would increase
productive effort and surplus (joint surplus under the optimal arm’s-length piece
rate contract, VF¼1
21þg2
ðÞ
, decreases in g). When relational incentives are also used,
however, arm’s-length contracting is P’s fallback option after reneging on the
relational bonus. Thus, an increase in V
F
(due to the lower misalignment g) reduces
P’s net long-term gain from the relational contract (the left-hand side of condition
(25.9)), making it harder to keep a given relational bonus Bself-enforcing.
Therefore, the overall effect of a reduction in misalignment on surplus is ambig-
uous and non-monotonic. While initial improvements in the contractual technology
(starting from very low levels) may increase productive effort and surplus (i.e., the
manipulation-reduction effect dominates the fallback-option-enhancement effect),
further improvements may backfire: the possibility for P to revert to arm’s-length
contracting may become too attractive, and as a result, relational bonuses that were
self-enforcing under poorer contractual technologies may become unsustainable
(i.e., may no longer satisfy self-enforcement condition (25.9) above). To put it
more colorfully, a decent-but-not-perfect contractual technology may crowd out
superior relational contracting. This result provides a possible microeconomic foun-
dation for legal arguments in favor of a pure private ordering in business trans-
actions, whereby parties willingly “opt out of the legal system”and commit
themselves to solely rely on relational enforcement mechanisms (Bernstein, 1992).
Despite its important influence on relational contracting theory, this implication of
the canonical model has so far escaped serious empirical scrutiny, as discussed
below.
25.2.1.5 Reinterpreting the Canonical Model: Adaptation
and Organization Design
The insights of the canonical model of relational incentive contracts, as discussed
above, extend to different but related contracting problems. For instance, several
models analyze the role of relational contracts in settings where court-enforceable
contracts are incomplete both ex ante (parties cannot specify and commit to desired
actions under all possible future contingencies) and ex post (once it becomes clear
what actions should be taken, it is not possible to fill contractual gaps via spot
negotiation, for instance, because there is insufficient time). Examples of such
settings are franchising networks, supply chains, strategic alliances, and other
organizational arrangements that are “hybrids”between spot market contracting
632 R. Gil and G. Zanarone
and intrafirm hierarchies (for a comprehensive review and critical discussion of the
literature on hybrids, see Ménard, 2013).
For instance, Baker et al. (2011) develop a general model of relational contracts in
which multiple parties agree to adapt non-contractible decisions to realized contin-
gencies and explore which contractual allocation of the authority to make those
decisions among the parties supports the best relational contract. Zanarone (2013)
develops a related model of distribution networks in which decisions are contract-
ible, but pro-distributor laws constrain the set of feasible contracts, and studies the
interaction between relational contracts and the contractual allocation of decision
rights. Argyres et al. (2020) model adaptation in buyer-supplier relationships and
show that broad-scope contracts, whereby a given buyer concentrates its purchases
into one supplier, enable the parties to sustain more efficient relational adaptation,
compared to narrow-scope contracts. Like the canonical model discussed above,
these models of relational adaptation find that contractual and relational provisions
should coexist and that the design and performance of court-enforceable contracts is
affected by the presence of an overarching relational contract.
A second area toward which the canonical model of relational contracts has been
extended is organization design and particularly the choice between centralization
and delegation of decision authority in organizations. Firms sometimes incentivize
their employees (or suppliers) through an appropriate organization design rather than
through incentive contracts. Aghion and Tirole (1997) made this point clear by
showing that if a “boss”delegates to a “subordinate”the authority to implement a
decision, she can incentivize the subordinate to exert creative efforts that may
ultimately benefit the boss.
4
While Aghion and Tirole (1997) make the simplifying
assumption that a boss can contractually commit to delegate authority, in most
organizations, delegation (and, more broadly, any configuration of the organizational
chart) is the boss’s discretionary decision and, thus, can always be retracted. In other
words, delegation of authority is a relational promise, pretty much like incentive
bonuses in the canonical model. Baker et al. (1999) developed the first model of
relational delegation, studying how the expected duration and value of the relation-
ship between a boss and a subordinate affects the credibility of the boss’s promise to
delegate. This model shares one of the insights we derived from the canonical model
above: as the expected duration and value of the boss-subordinate relationship
increases, the promise to delegate authority to the subordinate becomes more
credible and, thus, the boss can elicit greater levels of initiative and effort from the
subordinate.
Drawing upon Baker et al. (1999), Li et al. (2017) develop a model showing that
organizations may fail to efficiently recentralize previously delegated authority for
fear that employees will interpret centralization as a breach of trust and quit the firm.
4
While Aghion and Tirole (1997), like us in this section, study the incentive-provision function of
delegation, organizational economists have explored other important functions such as coordination
(Garicano, 2000), adaptation (Dessein, 2002), and the interaction between the two (Alonso et al.,
2008; Rantakari, 2008). See Bolton and Dewatripont (2013); Gibbons et al. (2013); Garicano and
Van Zandt (2013); and Mookherjee (2013), for extensive reviews of this literature.
25 Governing Relationships: Theory, Evidence, and Opportunities 633
Rantakari (2023) develops a “radical”model of relational organization design in
which even the boss’s authority over her/his subordinates (which is assumed to be
contractual in both Aghion & Tirole, 1997, and Baker et al., 1999) must be part of a
self-enforcing relational agreement. In this agreement, centralized authority arises in
equilibrium when the employee makes a self-enforcing promise to follow the boss’s
instructions (without attempting to investigate the matter and second-guess
the boss).
25.2.2 The Clarity Problem and the Role of Formality
Models of relational contracts (including the one presented here) make the simpli-
fying assumption that parties agree at the outset on what constitutes cooperation or
defection (under all contingencies). More recent papers (Gibbons & Henderson,
2013; Helper & Henderson, 2014) have questioned this assumption, pointing out
that like court-enforceable contracts, relational contracts are also “incomplete.”This
would mean that parties cannot foresee at the outset all the contingencies their
relational contract will have to cover as time goes by. Additionally, language
ambiguities and imperfections may further limit the parties’ability to reach a
common understanding and interpretation of what actions their agreement
prescribes.
This problem of conflicting interpretations, also known as the “clarity problem”
(Gibbons & Henderson, 2012), appeared to be present at First Boston in the 1990s:
while the company’s new owner (Credit Suisse) believed bankers’bonuses should
be reduced in a year of poor financial results, First Boston bankers believed that,
because they were as talented and hardworking as their peers at other banks, they
should receive similar bonuses regardless of profitability differences. After receiving
bonuses below their expectations for two consecutive years, most top bankers at First
Boston quit the firm (Stewart, 1993). Another example of the clarity problem, in the
context of organizational design rather than contracting (see our discussion above),
is the “Spaghetti organization”experimented by Oticon, a Danish producer of
hearing aid equipment, in the 1990s (Foss, 2003). Oticon decentralized decision-
making at the worker level to foster innovation by allowing workers to form teams,
compete for human capital within the organization, and pursue the projects they
desired. After a promising start, duplication of effort drove the organization to
overrule several decisions made by workers. Importantly, while Oticon’s top man-
agement viewed such overruling as perfectly consistent with the decentralized
organizational structure they created, employees interpreted it as breach of a prom-
ise. Foss (2003) reports that, following the overruling, several employees lost morale
and motivation and some of them decided to quit the firm.
As the First Boston and Oticon examples illustrate, the clarity problem is espe-
cially consequential in relational contracts because it may lead not only to occasion-
ally inefficient decisions but also to the complete breakdown of valuable
collaborations. To understand this point, recall that in a relational contract, promises
of cooperation are credible only if they are backed by a threat to terminate the
634 R. Gil and G. Zanarone
relationship in the event of defection. For instance, in our relational incentive model,
both parties believe that if P fails to pay the bonus when A has produced the agreed-
upon output, A will no longer trust P’s promises in the future, and relational
incentives will never again be sustainable. While this mutually understood threat
of termination is useful to keep relational contracts self-enforcing, it may lead to
unnecessary and wasteful termination when a clarity problem is present. Imagine
that P and A in our model disagree on the definition and/or interpretation of output.
Then, it could happen that after A chooses his effort, P “sees”low output, y¼y, and
hence pays a low bonus, By, whereas A “sees”high output, y¼
y, and thus expects a
high bonus, B
y. When that is the case, A may interpret P’s payment of By(instead of
B
y) as defection and terminate the relational contract in response. Then, despite the
fact that nobody has willingly defected, P and A will be forced to revert to arm’s-
length contracting, and earn low surplus, in all subsequent periods.
Given the relationship-disruption risks posed by the clarity problem, a frontier
research question for economists and social scientists interested in this phenomenon
is how contractual language and design together with management practices could
help parties to reach a shared framing of the relationship and a shared interpretation
of mutual obligations. For instance, recent papers study corporate culture as a set of
shared understandings among collaborating parties and the role of organizational
leaders in moving from one culture to another (Gibbons et al., 2021a,b).
An interesting case study of clarity-enhancing management practices is Netflix,
which revolutionized the entertainment industry and attracted hundreds of millions
of subscribers in over 190 countries. In a recent book (Hastings & Meyer, 2020),
cofounder Reed Hastings explains that because the company promotes the autonomy
of workers and managers—from deciding how to travel to a business event to
deciding whether to hire an expensive consultant—it has developed a principle
that limits unnecessary or opportunistic decisions: “When spending company
money, always act in Netflix’s best interest.”Employees who are found to be in
breach of this principle can be fired. Hastings soon realized that the definition of
“Netflix’s best interest”is highly incomplete: how can employees know in each
unforeseen situation what course of action the company’s executive consider to be
consistent with the principle? Hastings explains that Netflix has solved this clarity
problem by asking its managers to provide extensive history-based examples to new
hires, which clarify what the company considered to be its “best interest”in past
circumstances, and hopefully offer guidance on what such best interest could be in
not-yet-foreseen future circumstances that share some similarities with past ones. As
examples accumulate, clarity should build up. Hastings also notices how managers
typically do not fire subordinates after behaviors that appear to be against the
company’s interest—they instead give those subordinates the benefit of the doubt,
explaining why their behavior was inappropriate, and resort to the threat of termi-
nation only after the subordinate has been long enough at Netflix, such that clarity
concerning the “best interest”principle can be more safely presumed.
A few recent papers at the frontier between law, economics, and sociology have
also explored how contractual formalism may help parties in a collaborative
25 Governing Relationships: Theory, Evidence, and Opportunities 635
relationship to achieve clarity. For instance, Hadfield and Bozovic (2016) suggest
that because formal and legalistic language is explicit and clear, it may enable the
parties to agree on a common classification of contingencies. This may explain why
formal agreements, often drafted by lawyers, are commonly used even in relation-
ships where the threat of litigation is not credible, and hence formalism should not
serve the purpose of facilitating courts’verification and enforcement. Bernstein and
Peterson (2023) similarly argue that long-term buyer-supplier contracts include a
variety of “managerial provisions”that appear to clarify to the parties themselves
(rather than to courts) how to define good performance, and what supplier’s conducts
deemed unsatisfactory by the buyer may lead to termination or forgiveness. A clarity
role of contractual formalism is also invoked by Frydlinger and Hart (2024), who
argue that by including formal guiding principles (such as integrity and loyalty) in
their contract, parties frame their relationship as one of mutual collaboration, thus
activating some innate principles of mutual altruism and cooperation that would
remain inert in the absence of such provisions.
Besides contracts, the clarity-building role of formalism can also be seen in the
context of organization design. Clearly framed organizational charts and decision-
making processes may ensure that managers and employees in a firm develop a
shared understanding of who makes what decisions in each contingency. This, in
turn, may prevent employees from interpreting a manager’s decision to correct or
reject some of their proposals as the breach of a promise of delegation, as in the
Oticon case discussed above. In fact, Oticon itself is a good illustration of how a
more formalized organizational structure may help firms to sustain a relationally
agreed decision-making process. After employees lost motivation due to the frequent
overruling of their proposals by the upper management, Oticon formalized and
restructured the scope of employees’delegated authority. On the one hand, under
the new structure, employees’projects would be routinely approved, and hence there
was no ambiguity left on the existence of a promise to delegate. On the other hand,
projects would have to be chosen within formal company guidelines and overseen by
centrally appointed team leaders. Thus, employees would be clearly informed about
the fact that there are limits to delegation under the agreed-upon organizational
structure and what those limits are.
25.3 From Theory to the Data
A classic study by Poppo and Zenger (2002) provides the first evidence that court-
enforceable contracts and relational contracts complement each other. Using survey
instruments to measure the extent and complexity of contracts and the strength of
relational governance in buyer-supplier dyads, they find that more detailed and
complex formal contracts are used in relationships that the parties perceive as
more collaborative, and trust based. While this is an important pioneering study, it
leaves an important question open: which economic mechanism drives the
co-presence of contractual and relational provisions?
636 R. Gil and G. Zanarone
The canonical model developed in Sect. 25.2 focuses on facilitation of self-
enforcement as the key mechanism underlying the complementarity that studies
like Poppo and Zenger (2002) uncover. As discussed above, the model also provides
predictions that allow one to test for this mechanism. In this section, we discuss (1) a
selection of empirical papers that provide insight on how to test these predictions and
(2) research opportunities arising from untested implications or extensions of the
canonical model as well as the clarity problem discussed in Sect. 25.2.1.4 above.
Before we proceed, it is important to acknowledge that the list of empirical papers
discussed here is not (and does not intend to be) an exhaustive review of the literature
(such type of reviews, as discussed in the introduction, can be found in Gil &
Zanarone, 2017,2018, and Macchiavello, 2022). Instead, we selected papers that
in our view address most clearly and directly the theoretical insights we aim to shed
light on. Moreover, while we cite some early papers (see right below!), our discus-
sion primarily focuses on recent work that is not extensively covered in the existing
reviews.
We organize our discussion of the empirical evidence below around each of the
theoretical predictions highlighted in Sect. 25.2.1.4.
25.3.1 Surplus Increases in the Shadow of the Future (for Given
Contract Provisions)
The first prediction of the canonical model is that even when collaborations are
governed by court-enforceable contracts (and holding those contracts fixed), their
performance and surplus increase in the present discounted value (hereafter, PDV) of
the future relationship between the parties. Finding evidence in support of these
predictions would be a way to empirically test for the interaction and complemen-
tarity between court-enforceable and relational contracts.
The first paper providing a rigorous test of this prediction (i.e., one based on
plausibly exogenous variation in the PDV of contractual relationships) is Gil and
Marion (2013).
5
They study how contractors’relationships with subcontractors
affect the competitiveness of their bids for procurement contracts (a proxy for the
generated surplus). The importance of contractor-subcontractor relationships in this
industry makes it an ideal setting to study informal agreements (Eccles, 1981). Using
data for 10 years of highway procurement contracts from Caltrans (California
Department of Transportation) auctions, they use the number and value of local
projects (at the construction district level) auctioned by Caltrans over the next
calendar year as an exogenous proxy for changes in the PDV of contractor-
subcontractor relationships. This is an exogenous proxy because projects are offered
5
Earlier papers, some of which we discuss below (Corts & Singh, 2004; Kalnins & Mayer, 2004),
have used past interactions as a proxy for future interactions. A problem with this approach is that
past interactions may also proxy for mutual learning. Other papers (e.g., Poppo et al., 2008) have
used endogenous measures such as the parties’expectation on the duration of their business
relationship.
25 Governing Relationships: Theory, Evidence, and Opportunities 637
when highway and road repairs are needed, and not based on the locally available
pairs of contractors and subcontractors. Gil and Marion (2013)first notice that
contractors and subcontractors do use court-enforceable contracts, which are
disclosed as part of the contractor’s auction bids. Then, they show that despite the
presence of these contracts, the shadow of future interactions between contractors
and subcontractors (proxy for the PDV of their relationship) significantly improves
performance. Then, they show that, holding past interactions (and hence accumu-
lated mutual learning) constant, contractors post lower bids in highway procurement
auctions when more and more valuable forthcoming projects are scheduled (shadow
of the future). Because a contractor interacts with several subcontractors during the
temporal span of the data, and because there are several contractors in any given
construction district, the regression analysis in Gil and Marion (2013) can include
contractor and construction district fixed effects—that is, they hold constant all
possible time-invariant contractor and location characteristics, including the boiler-
plate court-enforceable contract used by a contractor with its subcontractors.
An important methodological contribution to the empirical literature on relational
contracting is Macchiavello and Morjaria (2015). This paper does not study the
interaction between contractual and relational provisions and is covered extensively
in a companion chapter of this handbook. Nevertheless, we briefly discuss it here
because its methodology has been used by subsequent papers studying the interac-
tion problem, as discussed below. Macchiavello and Morjaria (2015) analyze a rich
dataset of contracts between Kenyan rose exporters and buyers. Their setting allows
them to compute the exporters’temptation to renege on their informal agreements
with buyers as the stipulated quantity times the spot market auction price of roses.
Because self-enforcement implies that the PDV of a given buyer-exporter dyad must
be at least as large as the reneging temptation (see, for instance, self-enforcement
condition (25.9) in our model above), Macchiavello and Morjaria (2015) use their
estimates of the reneging temptation as lower bounds for the PDV. To validate their
interpretation of the reneging temptations as lower bounds for the PDV, they provide
evidence that following an unforeseen increase in the spot market price, informal
contracts between exporters and buyers proportionally reduce the stipulated quantity,
such that the self-enforcement condition is restored. This “lower-bound approach”to
measuring the PDV of contractual relationships is potentially replicable in other
settings because present reneging temptations, unlike future interactions, can often
be measured based on market and contractual data (such as the rose spot market price
and quantity data used by Macchiavello & Morjaria, 2015).
Gil et al. (2022) use this methodology to study the collaborations between US
major airlines and their outsourced regional partners around the 2008 financial crisis.
They show that although majors and regionals sign long and detailed formal
contracts, the extent of their collaboration around the crisis is importantly affected
by the PDVof their future relationship. In particular, Gil et al. (2022) observe that the
2008 financial crisis reduced profitability in the airline industry (as in many other
industries), calling for a downsizing of the majors’route portfolios. Then, they show
empirically that majors selectively cut regional routes from the low-PDV relation-
ships while preserving those in the high-PDV relationships. This result is robust to
638 R. Gil and G. Zanarone
the inclusion in their empirical specifications of route (market) and major airline
fixed effects—that is, Gil et al. (2022) look at the effect of PDV on the resilience of
route portfolios while holding constant all possible characteristics of the focal route
and, most importantly, all possible characteristics of the major airline, including the
boilerplate court-enforceable contract it uses with its regional partners.
To measure the PDV of major-regional relationships, Gil et al. (2022) use the
lower-bound approach developed by Macchiavello and Morjaria (2015) and
discussed above. We briefly describe their implementation as that can be insightful
for empirical researchers interested in applying the lower-bound approach to other
relational contracting settings. As noted by Forbes and Lederman (2009), a key
non-contractible cooperative decision for major and regional airlines is the exchange
of landing slots during the slot rationing that follows inclement weather (70% of
winter days at several major airports in the USA). Gil et al. (2022) assume that
because slot exchanges between majors and outsourced regionals are
non-contractible, they must be governed by relational agreements. They proxy the
regionals’maximum temptation to renege on such agreements (i.e., the regional’s
present savings from denying slots to the major in the contingency in which
inclement weather strikes all airports served by the partnership) by aggregating
bad weather conditions (rainfall and snow) across all routes in the major-regional
partnership. Similar to Macchiavello and Morjaria (2015), Gil et al. (2022) use their
proxy for the regional’s reneging temptation as a lower bound for the PDV of the
major-regional relationships because in a self-enforcing relational contract, the PDV
must be at least as large as the reneging temptation.
25.3.2 Under Relational Contracting, Responses to and the Design
of Court-Enforceable Contracts Differ from Arm’s-Length
Contracting
A few early papers provide evidence consistent with the prediction that the court-
enforceable provisions one observes under relational contracting differ from those
that one should observe under arm’s-length contracting. For example, Kalnins and
Mayer (2004)—in the context of IT service procurement—show that downstream
firms rely less on formal fixed-price terms to motivate their suppliers to cut costs
when they are engaged in repeated interactions with the suppliers. Similarly, in a
study of offshore oil drilling contracts, Corts and Singh (2004)find that projects are
less likely to be governed by a turnkey contract (akin to fixed-price) than by a
day-rate contract (akin to cost-plus) when the oil company and the driller have
worked together in the past. More recently, Ryall and Sampson (2009) study joint
technology development contracts in the telecommunications and microelectronics
industries and find that court-enforceable provisions, including penalties, are more
frequent and detailed when the firm and the partner interact more frequently. A
common feature of these studies is that the extent of relational contracting is proxied
by past interactions between the parties rather than measures of the shadow of the
future. This is somewhat problematic given that the self-enforcement mechanism
25 Governing Relationships: Theory, Evidence, and Opportunities 639
underlying relational contracting relies on the latter. A clear opportunity for future
researchers is therefore to leverage on the methodologies discussed in Sect. 25.3.1
above to develop tests for how the shadow of the future affects contractual design.
A recent paper by Kosova and Sertsios (2018) uses a different approach to study
how relational contracting affects the design of court-enforceable contracts in the
context of hotel franchising. Rather than measuring the extent of relational
contracting via repeated interactions, they measure it through the franchisor’s diffi-
culty of monitoring franchisees, and hence of enforcing their desired performance
contractually. They find that hotel franchisors grant upper-tier franchise contracts
(and higher future quasi-rent rents that can be lost upon termination) to locations that
are farther from the headquarters and hence harder to monitor. Their interpretation of
the results is that as monitoring costs increase, a relational contract requiring
franchisees to exert effort may cease to be self-enforcing, and this, in turn, induces
franchisors to raise the franchisees’quasi-rents to keep the relational contract self-
enforcing.
Barron et al. (2020) provide rare evidence on a related implication of our model—
namely, that parties in a relational contract cooperate beyond the letter of their court-
enforceable agreement. They study contracts between movie distributors and a large
Spanish exhibitor, which are governed by formal revenue sharing provisions spec-
ifying the share of each movie reel’s revenues that is kept by the exhibitor in any
given week the movie is shown. Barron et al. (2020)find that when the exhibitor
shows movies whose continuation would be unprofitable given their formally
contracted revenue share and box office turnouts, the distributor rewards the exhib-
itor by unilaterally and retroactively renegotiating the exhibitor’s revenue share
upward after the movie has ended its run at its theaters. This result indicates that
the exhibitor responds to relational incentives (the discounts) rather than contractual
ones (the revenue share) in making its movie-showing decisions. It also indicates
that to sustain relational contracts, distributors initially select contract terms that
would not be optimal in the absence of relational contracting: once a movie ends its
run, a retroactive increase in the revenue share clearly cannot affect the exhibitor’s
contractual incentives to show that movie.
6
25.3.3 A Better Contractual Technology May Crowd Out Relational
Contracts
A key challenge in testing the “crowding out”prediction of the canonical model is
how to measure the quality of arm’s-length contracting (the fallback option two
6
Moreover, Barron et al. (2020) also proxy the lower bound of the PDV of distributor-exhibitor
relationships through the maximum discount granted by the exhibitor in the first half of the sample
period. They find that the exhibitor is more likely to keep showing high-risk movies when the
relationship with the distributor has higher PDV. This result further supports the hypothesis that
distributors and exhibitors govern their relationships through a combination of formal and relational
contracts.
640 R. Gil and G. Zanarone
parties would revert to if their relational agreement broke down). A natural measure
is the quality of third-party enforcement, that is, the speed and reliability with which
contracts are enforced in a given country or industry at a certain point in time. All
else equal, higher-quality third-party enforcement should increase the surplus two
parties can achieve by relying on arm’s-length contracts that solely rely on court
enforcement. Several studies have attempted to measure variations in court quality
both across countries (Djankov et al., 2003; Johnson et al., 2002) and within
countries (Berkowitz & Clay, 2006; Choi et al., 2008). In addition to developing
contemporary measures of court quality based on the flexibility and low formalism
of court proceedings, Djankov et al. (2003) have also argued that the common-law
legal tradition exhibits lower formalism than the civil-law tradition due to exogenous
historical reasons. Accordingly, they have proposed to use common-law legal origin
as an instrumental variable for contemporary court quality that may allow to identify
the causal effect of court quality on outcomes of interest. The historical analysis
underlying this approach rests on strong assumptions and has been criticized
(Arruñada, 2007; La Porta et al., 2007).
Despite the possibility to proxy for variations in the value of arm’s-length
contracting, we are not aware of papers that have developed an empirical strategy
to test the crowding out hypothesis. Two studies that provide interesting insights
(albeit without developing a full test) are Antràs and Foley (2015) and Michler and
Wu (2020). Antràs and Foley (2015) study a large US frozen food exporter’s
provision of trade credit to its international clients. They find that repeated interac-
tions between the exporter and a given client (a proxy for relational contracting)
increase trade credit but only if court quality in the client’s country is low enough.
Their evidence is consistent with crowding out: when arm’s-length contracting
becomes less dysfunctional, a too-attractive fallback option for the client may
prevent the buyer from offering superior relational trade credit agreements, and as
a result, marginal improvements in relational enforcement no longer affect trade
credit provision.
Using survey data on groundwater irrigation contracts in Bangladesh, Michler
and Wu (2020)find that households rely less on fixed-price contracts and more on
output sharing contracts, in villages where the punishment of contract breach is more
severe (social ostracism instead of negotiated symbolic penalties or fines). Since the
main problem in groundwater irrigation is to ensure consistent and reliable water
delivery, Michler and Wu (2020) interpret fixed-price agreements—which do not
provide incentives for water delivery—as relational incentive contracts (in the sense
that water delivery under these agreements must be incentivized through a self-
enforcing arrangement), and crop sharing agreements, which do provide such
incentives, as court-enforceable incentive contracts. Under this interpretation, their
results are consistent with crowding out: improvements in the value of arm’s-length
contracting (more severe punishments in case of breach of the output sharing
obligation) reduce the parties’reliance on relational incentives.
While the results in Antràs and Foley (2015) and in Michler and Wu (2020) are
consistent with the crowding out hypothesis, they are also consistent with a much
simpler hypothesis: if third-party enforcement becomes good enough (due to higher
25 Governing Relationships: Theory, Evidence, and Opportunities 641
court quality, as in Antras and Foley, or due to stronger punishments, as in Michler
and Wu), parties may be able to achieve the first-best level of cooperation (high trade
credit in Antras and Foley’s setting, reliable water delivery in Michler and Wu’s
setting), and hence they may simply not need relational contracting. In terms of our
model from Sect. 25.2,ifgdrops down to zero, arm’s-length contracting achieves
the first best, and there are no gains from adding a relational bonus to the contractual
piece rate. It would therefore be desirable to replicate the kind of test that Antràs and
Foley (2015) and Michler and Wu (2020) perform using continuous measures of
both relational contracting and third-party enforcement. Observing that marginal
improvements in third-party enforcement first increase and then reduce the use of
relational contracts would be consistent with crowding out but not with first-best
formal contracting.
25.3.4 Relational Contracts and Organizational Structure
As discussed in Sect. 25.2.1.5, the canonical model of relational contracting can be
reinterpreted to analyze a corporate boss’s promise to delegate authority to a
subordinate (Baker et al., 1999). Several empirical papers have documented how
firms allocate decision authority efficiently, delegating it to subordinates when that is
likely to add value. Some examples are McElheran (2014) on delegation of IT
investments; Lo et al. (2016) on pricing authority in the industrial equipment
industry; Guadalupe et al. (2014)onfinance and human resource decisions in
Fortune 500 companies; Kala (2022) on strategic decisions in Indian state-owned
enterprises; and Yang et al. (2023) on invention development at a global ICT
company. However, no paper (to the best of our knowledge) has directly and
explicitly investigated “relational delegation,”that is, whether delegation requires
self-enforcing relational contracts to be sustainable and, particularly, whether the
shadow of the future (i.e., the PDV of boss-subordinate relationships) facilitates
delegation.
The empirical study that perhaps gets closest to documenting relational delega-
tion is Bloom et al. (2012). Using cross-country plant-level data on about 4000
firms, they find that higher levels of bilateral trust between a multinational’shome
country and a subsidiary’s country increase decentralization of authority from the
headquarters to the subsidiary’s managers. These results are consistent with the
hypothesis that relational contracts are necessary to sustain delegation. However,
while their explanatory variable, bilateral trust between two countries (as measured
by the World Values Survey), probably correlates with the PDVof the relationship
between a boss in country 1 and a subordinate in country 2, alternative interpre-
tations of this variable, and of the findings in Bloom et al. (2012), are possible. For
instance, as discussed by the authors, lack of bilateral trust may also reflect a
CEO’s belief that the subsidiary’s manager has misaligned interests, which would
lead the CEO to reduce delegation even if the decision to delegate were contract-
ible. Future studies should therefore aim to develop direct measures of the PDVor
642 R. Gil and G. Zanarone
discount factor in a parent-subsidiary relationship (perhaps drawing upon the
measurement methodologies discussed in Sect. 25.3.1 above) and integrate these
measures into the cross-country delegation data of Bloom et al. (2012), which are
publicly available.
Overall, the paucity of empirical evidence on how relational contracts sustain
delegation and corporate organizational structures represents in our view an impor-
tant opportunity for future research.
25.3.5 Future Challenges: Empirical Analyses of the Clarity Problem
In the field of relational contracting, empirical research typically comes after a wave
of theoretical models. Given the difficulty of collecting data on this phenomenon,
creative and sometimes indirect tests need to be designed, and theoretical models
provide guidance in the development of such tests. Theoretical research on the
clarity problem is still in its infancy, and we do not have yet a canonical model of
this problem and its solutions, like we do for the credibility/self-enforcement
problem (see Sect. 25.2 of this chapter). It is therefore not surprising that we have
virtually no empirical investigations of clarity. This is in a sense exciting because we
foresee that clarity will be one of the main areas of research on contracting in the
upcoming years.
Perhaps the only published empirical paper on clarity to date is Gibbons et al.
(2023). In their lab experiment, a buyer and a seller play a trading game in several
periods, unable to predict how their environment will change over time. Gibbons
et al. (2023) begin by asking players what pricing and quality decisions they would
consider to be consistent with the relational contracts under different hypothetical
scenarios. Then, they find that after one of these scenarios is actually realized in the
experiment, buyer-seller pairs that lacked a shared understanding of the relational
contract (i.e., pairs plagued by a clarity problem) are less likely to make decisions
that match the other player’s expectation (i.e., they are less likely to reach an
equilibrium), resulting in lower joint surplus.
In the second part of the experiment, Gibbons et al. (2023) explore how formal
agreements the parties may reach at the outset on how to play the game may
attenuate the clarity problem. They find that buyer-seller pairs who agree on a
broad principle (i.e., a general criterion on how to adapt price and quality to future
contingencies) rather than a narrow rule (specific price and quality levels) are more
likely to reach a shared understanding of the relational contract and efficient deci-
sions as the game proceeds.
One potential consequence of the clarity problem, as documented by Gibbons
et al. (2023), is failure to reach an equilibrium and to efficiently adapt decisions to
realized contingencies. An even worse consequence, as discussed in Sect. 25.2
above, might be the misinterpretation of cooperation as defection (due to the lack
of a common understanding of the relational contract) and the consequent break-
down of valuable relationships. Because of the severity of such a consequence,
25 Governing Relationships: Theory, Evidence, and Opportunities 643
parties may attempt to preempt the clarity problem by choosing actions that the other
party is less likely to interpret as defection, even if these actions are inefficient.
A recent working paper by Espinosa, Groshjean, Yuan, and Zanarone (2024)
provides evidence consistent with such preemptive behavior. They use US data on
individual meetings between lobbyists and Congress legislators to investigate col-
laborative relationships between the two sets of parties. In these collaborations, the
politician provides political support for the lobbyist’s clients, and the lobbyist pro-
vides useful information and advice to the politician. Because a typical lobbyist is
time-constrained and has many clients, the efficient relational contract is one where
the lobbyist meets the politician only in the “right”periods—that is, those where the
politician sits in a committee that is relevant to the lobbyist’s clients. However,
politicians switch committees (and lobbyists switch clients) with some frequency,
implying that ex ante, the two parties may lack a shared understanding of what “right
period”might mean. Espinosa et al. (2024) hypothesize that to prevent relationships
from breaking down due to this clarity problem, lobbyists may engage in hyper-
conservative behavior, meeting a given politician even when the period is not right.
Consistent with this hypothesis, they find that lobbyist-politician pairs that have
interacted frequently in the past (and hence have had a chance to develop a common
understanding) efficiently interrupt meetings when the politician rotates to a com-
mittee that does not matter to the lobbyist’s client (and resume meetings when the
opposite kind of rotation occurs). In contrast, pairs that have had few past interac-
tions (and hence are less likely to share a common understanding of the relational
contract) do not interrupt meetings after the politician rotates to a less relevant
committee.
Future empirical investigations of the clarity problem (and the role of contractual
formality in solving this problem) could collect high-frequency data on cooperative
actions that are contingent on hard-to-predict future states and investigate whether
clarity-enhancing formal provisions increase the frequency of efficient outcomes
(i.e., adaptation of decisions to the realized state) and/or decrease the frequency of
wasteful disagreement and relationship termination. Implementing this kind of empir-
ical test poses at least three challenges. First, it is difficult to obtain high-frequency
data on state-contingent cooperation as that information is typically proprietary and
confidential. Second, it is difficult to find sources of exogenous variation in contractual
formalism that allow the researcher to identify causal effects (as opposed to correla-
tions). Third, given that contractual provision also play an enforcement-enhancing role
in relational agreements (as discussed in Sect. 25.2), it is hard to assess whether any
observed effect of formal contracts on adaptation and termination is due to the clarity-
enhancing or the enforcement-enhancing role of such contracts.
Regarding the first challenge (collecting high-frequency data on state-contingent
cooperation), a promising empirical setting to tackle it would be the US airline
industry where, as discussed above, a key cooperation problem is the coordinated
exchange of landing slots between major airlines and regional partners under
continuously changing weather conditions. Real-time data on slot exchanges at US
airports are available and are being used to study governance (Argyres et al., 2024).
644 R. Gil and G. Zanarone
Regarding the second challenge (finding exogenous variation in contractual formal-
ism), researchers could exploit natural experiments such as regulation-induced
changes in contractual design (Zanarone, 2009) or conduct field experiments to
generate random variations in formalism. Several field experiments on the effects
of organizational forms and management practices have been successfully
conducted in the development literature (Macchiavello, 2022). The methodologies
developed in this literature could serve as a guidance for empirical researchers
interested in exploring the clarity-enhancing role of formalism. Regarding the third
challenge (disentangling the clarity-enhancing and enforcement-enhancing roles of
formalism), a possible strategy would be to identify settings in which the threat
of litigation is not feasible nor credible, and thus any enforcement-enhancing rule of
formalism can be excluded by design. For instance, it could be argued that certain
cross-border contractual agreements are not backed by a credible threat of litigation.
Moreover, Ben-Shahar and Bernstein (2000) argue that companies for whom trade
secrets and proprietary technology and know-how are especially valuable do not
have a credible threat to sue their suppliers for non-performance because the risk that
confidential information will become public through the discovery process is pro-
hibitive for them.
While not an easy task, these empirical strategies are potentially feasible and may
enable future researchers to make progress (beyond laboratory experiments) toward
understanding the role of contractual formality in generating shared understandings
and interpretations of relational contracts, thereby solving the clarity problem.
25.4 Concluding Remarks
In this paper, we have presented a simple version of the canonical model of relational
contracting (Baker et al., 1994,2002; Levin, 2003), which provides predictions on
the complementarity between contractual and relational incentive provisions, and we
have discussed selected empirical studies through the lenses of this model. While we
have found support for the hypothesis that contractual and relational provisions
interact with and complement each other, we have also noted several opportunities
for future empirical research on governance. In particular, future researchers may
investigate the possibility that imperfect court enforcement institutions crowd out
relational contracting, the role of relational contracting in sustaining delegation
promises in organization, and the role of contractual formalism in solving clarity
problems (as opposed to enforceability problems) in relational agreements. We hope
that our discussion of theoretical and empirical approaches to relational contracting
will encourage future cohorts of researchers within and outside economics to
continue to expand our understanding of this important and fascinating
phenomenon.
Acknowledgment of Funding Open access made possible through a generous donation in honor
of the Ronald Coase Institute.
25 Governing Relationships: Theory, Evidence, and Opportunities 645
References
Aghion, P., & Tirole, J. (1997). Formal and real authority in organizations. Journal of Political
Economy, 105(1), 1–29.
Alonso, R., Dessein, W., & Matouschek, N. (2008). When does coordination require centralization?
American Economic Review, 98(1), 145–179.
Antràs, P., & Foley, C. F. (2015). Poultry in motion: A study of international trade finance practices.
Journal of Political Economy, 123(4), 809–852.
Argyres, N., Bercovitz, J., & Zanarone, G. (2020). The role of relationship scope in sustaining
relational contracts in Interfirm networks. Strategic Management Journal, 41, 222–245.
Argyres, N., Gil, R., & Zanarone, G. (2024, Forthcoming). Does supply concentration encourage
cooperation? Evidence from airlines. Management Science.https://doi.org/10.1287/mnsc.2022.
03455.
Arruñada, B. (2007). Pitfalls to avoid when measuring institutions: Is ‘doing business’damaging
business? Journal of Comparative Economics, 35(4), 729–747.
Baker, G., Gibbons, R., & Murphy, K. J. (1994). Subjective performance measures in optimal
incentive contracts. Quarterly Journal of Economics, 109, 1125–1156.
Baker, G., Gibbons, R., & Murphy, K. J. (1999). Informal authority in organizations. The Journal of
Law, Economics, and Organization, 15(1), 56–73.
Baker, G., Gibbons, R., & Murphy, K. J. (2002). Relational contracts and the theory of the firm.
Quarterly Journal of Economics, 117,39–84.
Baker, G., Gibbons, R., & Murphy, K. J. (2011). Relational adaptation. SSRN: https://ssrn.com/
abstract¼4256008
Barron, D., Gibbons, R., Gil, R., & Murphy, K. J. (2020). Reel authority: Relational renegotiation in
the movie industry. Management Science, 66(5), 1868–1889.
Bartlett, A. (1998). Lincoln electric: Venturing abroad. Harvard Business School Case #398–095.
Battigalli, P., & Maggi, G. (2008). Costly contracting in a long-term relationship. RAND Journal of
Economics, 39, 352–377.
Ben-Shahar, O., & Bernstein, L. (2000). The secrecy interest in contract law. Yale Law Journal, 109,
1885–1925.
Berkowitz, D., & Clay, K. (2006). The effect of judicial Independence on courts: Evidence from the
American states. The Journal of Legal Studies, 35, 399–440.
Bernstein, L. (1992). Opting out of the legal system: Extralegal contractual relations in the diamond
industry. The Journal of Legal Studies, 21(1), 115–157.
Bernstein, L., & Peterson, B. (2023). Managerial contracting: A preliminary study. Journal of Legal
Analysis, 14, 176.
Bloom, N., Sadun, R., & Van Reenen, J. (2012). Americans do IT better: US multinationals and the
productivity miracle. American Economic Review, 102(1), 167–201.
Bolton, P., & Dewatripont, M. (2013). Authority in organizations. In R. Gibbons & J. Roberts
(Eds.), Handbook of organizational economics. Princeton University Press.
Bull, C. (1987). The existence of self-enforcing implicit contracts. Quarterly. Journal of Econom-
ics, 102, 147–159.
Choi, S., Gulati, G., & Posner, E. (2008). Professionals or politicians: The uncertain empirical case
for an elected rather than an appointed judiciary. Journal of Law, Economics and Organization,
26, 290–336.
Corts, K., & Singh, J. (2004). The effect of repeated interaction on contract choice: Evidence from
offshore drilling. Journal of Law, Economics, and Organization, 20, 230–260.
Dessein, W. (2002). Authority and communication in organizations. The Review of Economic
Studies, 69(4), 811–838.
Djankov, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2003). Courts. Quarterly Journal of
Economics, 118(2), 453–517.
Eccles, R. (1981). The Quasifirm in the construction industry. Journal of Economic Behavior and
Organization., 2(4), 335–357.
646 R. Gil and G. Zanarone
Espinosa, M., Groshjean, T., Yuan, Z., & Zanarone, G. (2024). How the clarity problem affects
relational contracts: Evidence from Lobbying. mimeograph.
Fast, N., & Berg, N. (1975). The Lincoln electric company. Harvard Business School Case
#376–028.
Forbes, S., & Lederman, M. (2009). Adaptation and vertical integration in the airline industry.
American Economic Review, 99, 1831–1849.
Foss, N. (2003). Selective intervention and internal hybrids: Interpreting and learning from the rise
and decline of the Oticon spaghetti organization. Organization Science, 14, 331–349.
Frydlinger, D., & Hart, O. (2024). Overcoming contractual incompleteness: The role of guiding
principles. Journal of Law, Economics and Organization, 40(3), 625–647.
Garicano, L. (2000). Hierarchies and the Organization of Knowledge in production. Journal of
Political Economy, 108(5), 874–904.
Garicano, L., & Van Zandt, T. (2013). Hierarchies and the division of labor. In R. Gibbons &
J. Roberts (Eds.), Handbook of organizational economics. Princeton University Press.
Gibbons, R. (2005). Incentives between firms (and within). Management Science, 51,2–17.
Gibbons, R., & Henderson, R. (2012). Relational contracts and organizational capabilities. Orga-
nization Science, 23, 1350–1364.
Gibbons, R., & Henderson, R. (2013). What do managers do? In R. Gibbons & J. Roberts (Eds.),
Handbook of organizational economics. Princeton University Press.
Gibbons, R., Matouschek, N., & Roberts, J. (2013). Decisions in organizations. In R. Gibbons &
J. Roberts (Eds.), Handbook of organizational economics. Princeton University Press.
Gibbons, R., LiCalzi, M., & Warglien, M. (2021a). What situation is this? Shared frames and
collective performance. Strategy Science, 6(2), 124–140.
Gibbons, R., Siegel, J., & Weber, R. (2021b). Strategy meets culture (for breakfast): Understanding
the relationship and highlighting its potential. Strategy Science, 6(2), 111–118.
Gibbons, R., Grieder, M., Herz, H., & Zehnder, C. (2023). Building an equilibrium: Rules
vs. principles in relational contracts. Organization Science, 34(6), 2231–2249.
Gil, R., & Marion, J. (2013). Self-enforcing agreements and relational contracts: Evidence from
California highway procurement. Journal of Law, Economics and Organization, 29, 239–277.
Gil, R., & Zanarone, G. (2017). Formal and informal contracting: Theory and evidence. Annual
Review of Law and Social Science, 13, 141–159.
Gil, R., & Zanarone, G. (2018). On the determinants and consequences of informal contracting.
Journal of Economics & Management Strategy, 27(4), 726–741.
Gil, R., Kim, M., & Zanarone, G. (2022). Relationships under stress: Relational outsourcing in the
U.S. airline industry after the 2008 financial crisis. Management Science, 68(2), 1256–1277.
Granovetter, M. (1985). Economic action and social structure: The problem of embeddedness.
American Journal of Sociology, 91(3), 481–510.
Guadalupe, M., Li, H., & Wulf, J. (2014). Who lives in the C-suite? Organizational structure and the
division of labor in top management. Management Science, 6, 824–844.
Hadfield, G. K., & Bozovic, I. (2016). Scaffolding: Using informal contracts to support informal
relations in support of innovation. Wisconsin Law Review, 2016(5), 981–1032.
Hastings, R., & Meyer, E. (2020). No Rules Rules. Penguin Press.
Helper, S., & Henderson, R. (2014). Management practices, relational contracts, and the decline of
general motors. Journal of Economic Perspectives, 28(1), 49–72.
Johnson, S., McMillan, J., & Woodruff, C. (2002). Courts and relational contracts. Journal of Law,
Economics, and Organization, 18, 221–277.
Kala, N. (2022, forthcoming). The impacts of managerial autonomy on firm outcomes.
Econometrica.
Kalnins, A., & Mayer, K. (2004). Relationships and hybrid contracts: An analysis of contract choice
in information technology. Journal of Law, Economics, and Organization, 20, 207–229.
Klein, B. (1980). Transaction cost determinants of "unfair" contractual arrangements. American
Economic Review Papers and Proceedings, 70(2), 356–362.
25 Governing Relationships: Theory, Evidence, and Opportunities 647
Klein, B. (2000). The role of incomplete contracts in self-enforcing relationships. Revue
d’Économie Industrielle, 92,67–80.
Kosova, R., & Sertsios, G. (2018). An empirical analysis of self-enforcement mechanisms: Evi-
dence from hotel franchising. Management. Science, 64,43–63.
Kvaløy, O., & Olsen, T. (2009). Endogenous verifiability and relational contracting. American
Economic Review, 99, 2193–2208.
La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2007). The economic consequences of legal
origins (NBER working paper 13608).
Lafontaine, F., & Slade, M. (2013). Inter-firm contracts: The evidence. In R. Gibbons & J. Roberts
(Eds.), The handbook of organizational economics. Princeton.
Lazear, Edward P., & Oyer, Paul. (2013) “Personnel Economics”. The Handbook of Organizational
Economics, edited by Robert Gibbons and John Roberts, Princeton: Princeton University Press,
pp. 479–519.
Levin, J. (2003). Relational incentive contracts. American Economic Review, 93(3), 835–857.
Li, J., Matouschek, N., & Powell, M. (2017). Power dynamics in organizations. American Eco-
nomic Journal: Microeconomics, 9(1), 217–241.
Lo, D., Dessein, W., Ghosh, M., & Lafontaine, F. (2016). Price delegation and performance pay:
Evidence from industrial sales forces. The Journal of Law, Economics, and Organization, 32(3),
508–544.
Macaulay, S. (1963). Non-contractual relations in business: A preliminary study. American Socio-
logical Review, 28,1–19.
Macchiavello, R. (2022). Relational contracts and development. Annual Review of Economics, 14,
337–362.
Macchiavello, R., & Morjaria, A. (2015). The value of relationships: Evidence from a supply shock
to Kenyan rose exports. American Economic Review, 105, 2911–2945.
MacLeod, B., & Malcomson, J. (1989). Implicit contracts, incentive compatibility, and involuntary
unemployment. Econometrica, 57(2), 447–480.
MacNeil, I. (1978). Contracts: Adjustments of long-term economic relations under classical,
neoclassical, and relational contract law. Northwestern University Law Review, 72, 854–906.
Malcomson, J. (2013). Relational incentive contracts. In R. Gibbons & J. Roberts (Eds.), Handbook
of organizational economics. Princeton University Press.
Masten, S. (2021). Adaptation, adjudication, and private ordering: Contractual relations through the
Williamson lens. Journal of Institutional and Theoretical Economics, 2, 283–296.
McElheran, K. (2014). Delegation in multi-establishment firms: The organizational structure of IT
purchasing authority. Journal of Economics and Management Strategy, 23(2), 225–257.
Ménard, C. (2013). Hybrid modes of organization. In R. Gibbons & J. Roberts (Eds.), Handbook of
organizational economics. Princeton University Press.
Michler, J. D., & Wu, S. Y. (2020). Governance and contract choice: Theory and evidence from
groundwater irrigation markets. Journal of Economic Behavior and Organization, 180,
129–147.
Mookherjee, D. (2013). Incentives in hierarchies. In R. Gibbons & J. Roberts (Eds.), Handbook of
organizational economics. Princeton University Press.
Poppo, L., & Zenger, T. (2002). Do formal contracts and relational governance function as sub-
stitutes or complements? Strategic Management Journal, 23, 707–725.
Poppo, L., Zhou, K., & Ryu, S. (2008). Alternative origins to Interorganizational trust: An
interdependence perspective on the shadow of the past and the shadow of the future. Organi-
zation Science, 19,39–55.
Rantakari, H. (2008). Governing adaptation. The Review of Economic Studies, 75(4), 1257–1285.
Rantakari, H. (2023, forthcoming). Simon says? Equilibrium obedience and the limits of authority.
Journal of Law, Economics, and Organization. https://doi.org/10.1093/jleo/ewad026.
Ryall, M., & Sampson, R. (2009). Formal contracts in the presence of relational enforcement
mechanisms: Evidence from technology development projects. Management Science, 55,
906–925.
648 R. Gil and G. Zanarone
Schwartz, A. (1992). Relational contracts in the courts: An analysis of incomplete agreements and
judicial strategies. Journal of Legal Studies 21(2), 271–318.
Stewart, J. (1993). Double dare at First Boston. Financial Review, September 17. https://www.afr.
com/politics/double-dare-at-first-boston-19930917-jl05w
Telser, L. (1981). Margins and futures contracts. Journal of Future Markets, 1(2), 225–253.
Williamson, O. (1979). Transaction-cost economics: The governance of contractural relations.
Journal of Law and Economics, 22(2), 233–61
Williamson, O. (1991). Comparative economic organization: The analysis of discrete structural
alternatives. Administrative Science Quarterly, 36, 269–296.
Yang, J., Criscuolo, P., & Silverman, B. (2023). The Organizational Structure of Development and
Firm Innovation. mimeograph.
Zanarone, G. (2009). Vertical restraints and the law: Evidence from automobile franchising. Journal
of Law and Economics, 52, 691–700.
Zanarone, G. (2013). Contract adaptation under legal constraints. Journal of Law, Economics and
Organization, 29, 799–834.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
25 Governing Relationships: Theory, Evidence, and Opportunities 649
