Making and bending international rules the design of exceptions and escape clauses in trade law

Making and Bending
International Rules
The Design of Exceptions and Escape
Clauses in Trade Law

K R Z YS Z TO F J. P E L C


Making and Bending International Rules
All treaties, from human rights to international trade, include formal
exceptions that allow governments to legally break the rules that they
have committed to, in order to deal with unexpected events. Such institutional “flexibility” is necessary, yet it raises a tricky theoretical question: how to allow for this necessary flexibility, while preventing its
abuse? Krzysztof Pelc examines how designers of rules in vastly different settings come upon similar solutions to render treaties resistant to
unexpected events.
Essential for undergraduate students, graduate students, and scholars
in political science, economics, and law, the book provides a comprehensive account of the politics of treaty flexibility. Drawing on a wide
range of evidence, its multi-disciplinary approach addresses the paradoxes inherent in making and bending international rules.
Krzysztof J. Pelc is William Dawson Scholar and Associate Professor
in the Department of Political Science at McGill University, Montréal.
His research focuses on the politics of international economic rules and
his work has been published in International Organization, American
Political Science Review, Journal of Politics, World Politics, International Studies Quarterly, Journal of Conflict Resolution, European Journal of International Relations, British Journal of Political Science, and
Journal of International Economic Law, among others.


One Liberty Plaza, New York NY 10006, USA
Cambridge University Press is part of the University of Cambridge.
It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781107140868

© Krzysztof J. Pelc 2016
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2016
A catalog record for this publication is available from the British Library.
Library of Congress Cataloging in Publication Data
Names: Pelc, Krzysztof J., author.
Title: Making and bending international rules : the design of exceptions and
escape clauses in trade law / Krzysztof J. Pelc.
Description: Cambridge, United Kingdom : Cambridge University Press, 2016. |
Includes bibliographical references and index.
Identifiers: LCCN 2016010173 | isbn 9781107140868 (Hardback : alk. paper)
Subjects: LCSH: Foreign trade regulation. | Foreign trade
regulation–Language. | World Trade Organization.
Classification: LCC K3943 .P45 2016 | DDC 343.08/7–dc23 LC record
available at />isbn 978-1-107-14086-8 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy
of URLs for external or third-party Internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.


Contents

page ix
xi

List of Tables

Acknowledgments
1

The “Architectural Challenge” of International Rules
1.1 Introduction
1.2 The Trade Regime’s Architectural Challenge
1.3 The Dirty Secret of the Trade Regime
1.4 The Design of Escape Provisions
1.5 An Accident of History
1.6 Overview of Chapters

1
1
5
7
10
12
13

2

A Theory of the Design of Flexibility
2.1 The Debate over Flexibility in International Treaties
2.2 The Costs of Flexibility
2.3 Resolving the Architectural Challenge
2.4 Assessing Theoretical Expectations

18
18
27

31
39

3

A Brief Intellectual History of Flexibility in Law
3.1 The Universality of Flexibility
3.2 “Necessity Knows No Law”
3.3 Changed Circumstances

43
44
45
75

4

The Twin GATT Exceptions: Fears and Solutions
4.1 Article XXI: The GATT Security Exception
4.2 Article XX: The Value of Constraint
4.3 Conclusion: Article XXI vs. Article XX

93
93
122
132

vii
/B
9D

B5898 :DB
C ,
B: G 9 5 5 56 9 5

75 6D 8 9 BD 7BD9 270 4
C ,
75 6D 8 9 BD 7BD9 9D

9D

1 6D5D 9 B
C , 8B BD

25

. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9


Contents

viii

5

The Evolving Design of Flexibility
5.1 From Compensation to Contingency
5.2 What Room for Efficient Breach?
5.3 Trade Policymakers’ Remaining Flexibility Options
5.4 Flexibility in Preferential Trade Agreements
5.5 A Comparison Case: Flexibility in the Human Rights
Regime

137
138
150
161
185

6

The Bad News
6.1 Does Flexibility Fuel the Law of Constant Protection?
6.2 The Empirical Data
6.3 Flexibility and Unpredictability

206
206
209
220

7

The Good News
7.1 Restraint in Allocation of Flexibility
7.2 “Country Seeks Credibility”: How Governments Choose
among Flexibility Options

234
235

The Great Recession and Beyond
8.1 Rules vs. Behavior during the Great Recession
8.2 What Does the Future Hold?

259
261
264

8

/B
9D

189

243

Bibliography

267

Index

279

B5898 :DB
C ,
B: G 9 5 5 56 9 5

75 6D 8 9 BD 7BD9 270 4
C ,
75 6D 8 9 BD 7BD9 9D

9D

1 6D5D 9 B
C , 8B BD

25

. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9


1
The “Architectural Challenge” of International Rules

“Boundless intemperance
In nature is a tyranny.”
— Macbeth, Act IV

1.1 introduction
Rules are undone by unexpected events. In the realm of international
politics, droughts, floods, coups, wars, epidemics, price shocks, financial
crises, and surges of imports are as many events that can upset the laws
governing the behavior of states. There is broad agreement that in the
midst of unexpected circumstances, the same rules that normally bind
countries may need to be temporarily suspended, to allow governments
to deal with exigency.
In fact, one of the constants running through all types of agreements
is the inclusion of formal clauses that specify just how signatories will be
allowed to break the very rules they have agreed on. Such escape clauses
are prevalent in international trade, the regime this book examines most
closely. But they are also found in the investment regime, the human
rights regime, ancient Roman law, early canon law, religious rules of
every stripe, and in the precepts of just war theory. Even absolute laws
and moral rules recognize the need for their own suspension in some
circumstances. These different sets of rules are a testament to the first
paradox I examine in this book: rules become more effective by being
imperfect. Entirely rigid agreements break apart at the first hurdle.

In the international realm, in particular, one would be hard pressed to
think of a treaty that does not address uncertainty through the insertion of
1
/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


The “Architectural Challenge” of International Rules

2

formal escape provisions of one form or another. In fact, the international
treaty governing international treaties, the Vienna Convention on the Law
of Treaties, includes a notorious flexibility clause addressing changes of
circumstances.
In the Vienna Convention, as in other agreements, the inclusion of provisions that allow participants to legally breach an agreement’s primary
rules leads to a tricky theoretical question. We know that some measure
of wiggle-room can be highly beneficial to treaties, to the point of becoming an essential condition for their existence. The ability to temporarily
escape an agreement’s obligations in hard times renders it less vulnerable
to unforeseeable events. Flexibility allows for deeper commitments by the
treaty’s signatories, by providing a form of insurance that comes into
effect if the costs of adjustment suddenly run too high. It also lowers
barriers to entry, enlarging the membership, and with it, the gains from
cooperation. Yet build in too much flexibility, and the agreement can be
rendered ineffective, like a boiler with too many pressure-release valves.
States thus face conflicting incentives over flexibility provisions: they
value the option of relying on them in unexpected hard times, yet they
also have a constant incentive to abuse this option, and they fear that
other states will do the same. The ways in which international rules seek
to allow for some flexibility, while limiting its abuse, is the subject of this
book.

The debate over the design of flexibility is the very stuff of politics. It
mirrors the dilemma which underlies both the national and the international political process: there are gains to be had from delegating power;
yet delegate too much power, and the risk is tyranny. This fundamental
compromise animates political thought from classical philosophy to the
Federalist papers. In each case, the designers of rules seek to negotiate a
similar compact, one where power is delegated to a national or international body, and bound by its rules – but not unconditionally. Addressing
the design of flexibility in the specific context of one international regime
leads me to grapple with this foundational problem. How to design effective constraints on power that can stand up to the events of the real world?
Wherever flexibility provisions allow participants to suspend the rules
during unexpected hard times, they lead to similar fears. Negotiators of
the Vienna Convention in the 1960s thus warned against abuse of its
flexibility clause, contained in Article 62, claiming the provision was too
vague, and insufficiently constrained. So did the negotiators of the General
Agreement on Tariffs and Trade (GATT), in July 1947, as they agreed to
insert a national security exception into what was then the world’s most

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


1.1 Introduction

3

ambitious trade agreement. As the representative of the United States,
which had written the first draft of the provision, declared to the assembly:
We have got to have some exceptions. We cannot make it too tight, because we
cannot prohibit measures which are needed purely for security reasons. On the
other hand, we cannot make it so broad that, under the guise of security, countries
will put on measures which really have a commercial purpose.1

Countless negotiators and designers of rules have contemplated the
tradeoff at the center of this book. If the agreement is too tight, it will be
undone by events. If it is too flexible, it will be undone by abuse. In the
case of the GATT security exception, despite being so clearly conscious of
the challenge before them, by all accounts the negotiators failed at their
task. The national security exception, which is applicable to this day and
allows countries to be the sole judges of whether there exists a threat to
their security, is considered far too loose and insufficiently constrained.
One of the foremost theorists of the GATT, John Jackson, has denounced
it as a “catch-all clause” that is “so broad, self-judging, and ambiguous
that it obviously can be abused.”2
Jackson is in good company. Political scientists and economists agree
that when flexibility rules are too loose, they inevitably lead to abuse.
The standard account has long been that unless reliance on a flexibility
provision is made difficult, states will exploit it. As the seminal account
of escape clauses in international politics has it, unless there are formal
constraints on flexibility, states “will invoke it all the time, thus vitiating the agreement.”3 The associated assumption is that given the choice,
countries will always opt for the least constrained and cheapest available
option for escaping their obligations. As a recent book length treatment of
flexibility provisions concludes, “it is thus evident that an injuring country
will always go for the escape instrument which promises ‘most mileage’,
i.e. the fewest enactment costs, the lowest compensation, and the largest
scope of application.”4
In this book, I argue that even this “evident” premise is wrong. The
reason is that governments’ choices over escape do not take place in a
vacuum, and governments know it. Policymakers often speak of wanting
to avoid a “dangerous precedent.” They do not mean this in the strict legal
sense. What they mean is that by exercising an ill-defined, unconstrained
1 E/PC/T/A/PV/33, p. 20-21, in: “Analytical Index of the GATT,” Article XXI Secu-

rity Exceptions, p.600, available at www.wto.org/english/res_e/booksp_e/gatt_ai_e/
art21_e.pdf
2 Jackson (1997a, 230).
3 Rosendorff and Milner (2001).
4 Schropp (2009).

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


The “Architectural Challenge” of International Rules

4

exception, countries risk normalizing its exercise, making it more likely
that others will exercise it in turn. Governments are perpetually trying to
manage one another’s expectations of what constitutes acceptable behavior, and the formal rules are but one part of this. Practice gains prominence
wherever the rules are ambiguous. This leads me to the second paradox
of flexibility: countries turn to flexibility provisions not in spite of their
constraints, but because of them. I describe this as governments “seeking paperwork”; we observe it in the human rights regime as much as
in international trade. States seek to credibly convey to their audiences
that the current instance of escape does not increase the odds of escape
recurring. They do this by demonstrating that the event that precipitated
escape, the source of necessity – the drought, the country-wide strikes,
the surge of imports – is not only genuine, but that it could not have
been willfully manufactured. The function of escape clauses is to allow
escapees to demonstrate this one key point: escape today does not make
escape tomorrow more likely. Otherwise, the audience – made up of voters, investors, trade partners, or foreign governments – will update its
expectations, to the detriment of the escaping country, about the odds of
seeing further violations justified by similar events. When this happens,
risk premia rise, investment drops, trade flows decrease, and governments

get ousted.
Accordingly, in the absence of constraints on the use of flexibility provisions, the outcome is not widespread misuse; it is disuse: governments progressively abandon policies that do not benefit from credible constraints,
and that do not allow them to manage their audiences’ expectations. Such
has been the fate of the Vienna Convention’s escape clause, and of the
GATT’s security exception. In fact, I show that countries have at times
preferred to be found in formal violation, rather than to have to rely on
the security exception, even when the circumstances would have justified
doing so. More striking still, given the choice between a less constrained
and a more constrained flexibility clause, countries frequently turn to the
latter. In the book’s empirical analysis, I show that we can reliably account
for this choice by considering states’ incentives.
Countries’ behavior with respect to unconstrained flexibility constitutes one of the greatest demonstrations of global cooperation between
states, and one that has been largely overlooked. The success of international cooperation is traditionally assessed by asking whether countries
comply with, or break, the rules they have imposed on one another.
Hence the oft-repeated phrase according to which most countries obey
most rules most of the time. The argument in this book implies that an

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9

C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


1.2 The Trade Regime’s Architectural Challenge

5

equally important, and potentially more telling measure of international
cooperation lies in the legally allowed actions that countries don’t take,
when those actions can precipitate socially undesirable outcomes. Such is
the case when states choose not to exercise an ill-defined exception even as
they are legally entitled to do so, out of fear of setting a “dangerous precedent” and making its use by everyone else more likely. In this, countries
are driven by a concern over reciprocity that is more fundamental than

the constraints of formal rules. Invoking an unconstrained flexibility
provision may be the best option in the short term, but governments
realize that it may carry negative long term effects.
This leads me to the third paradox of flexibility. On their face, escape
clauses are designed to deal with hard times and exceptional circumstances. Yet their true concern is with normalcy. Treaty negotiators know
they can do little to affect behavior during emergencies. They internalize
the old legal maxim according to which “necessity knows no law.” Escape
clauses are invoked in those instances where, by construction, the law
would hold little sway over behavior. But escape clauses are nonetheless required to carve out and distinguish these instances from normal
circumstances, and thus to preserve the rules’ authority over the greater
part, by far, of the circumstances states find themselves in. Without an
explicit clause suspending the rules in hard times, necessary violations risk
rendering similar violations during less-than-hard times more acceptable.
In short, flexibility provisions exist to prevent behavior under extraordinary circumstances from spilling over onto normal times. They are not
concerned with hard times per se, but with what comes after.
Three questions are at the heart of this book. Why are flexibility provisions required? How do the designers of rules guard against the abuse
of flexibility? And given the abundance of unconstrained flexibility provisions, why do we see less abuse than we might expect? The book’s
argument addresses these questions, and in so doing puts forth three paradoxes: Rules gain from imperfection. States turn to flexibility provisions
not in spite, but because of their constraints. And flexibility clauses are
concerned not with necessity per se, over which they hold little sway, but
with what comes after. Next, I briefly rehearse this argument in the setting
of the international trade regime.

1.2 the trade regime’s architectural challenge
Treaties stand or fall by their flexibility provisions, and nowhere more so
than in the international trade regime. When the Doha Round trade talks

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:

The “Architectural Challenge” of International Rules

6

collapsed in Geneva in July 2008, the disagreement at fault turned out to
have been over the precise extent to which states could break the treaty if
they faced hard times. Developing countries had asked for the creation of
a clause that would have enabled them to suspend all their obligations in
times of need, and developed countries objected to the terms of this clause.
As a result, negotiators from 153 nations went home empty-handed.5
Insisting on such “license to breach” is not a peculiarity of developing
countries. The 2008 talks were far from an isolated case. The failure in
1947 of what was to be the world’s first multilateral trade agreement
and the third pillar of the Bretton Woods institutions, the International
Trade Organization (ITO), can be chalked up to another wrangle over
flexibility. In that instance, the US Congress could not stomach what it saw
as the overly broad balance-of-payments and full employment exceptions
pushed for by Europe, and never ratified the treaty as a result (Diebold,
1952; Ruggie, 1982).6
This is not to say that the United States ever held any principled stance
against flexibility provisions in trade, having all but invented them: the
very first trade escape clause was included at the US’ behest in a bilateral trade agreement with Argentina in 1941. By 1947, President Harry
Truman had signed an executive order requiring that an escape clause be
included in all future trade agreements to which the United States was a
signatory. As long as there have been formal rules binding sovereign states,
there have been additional rules put in place allowing states temporary
breaches of their commitments.
How to allow flexibility, but prevent its abuse? This is the question
that Pascal Lamy, the World Trade Organization (WTO) Director General
until 2013, called the institution’s “architectural challenge.”7 The term

is apt. It conveys how international rules do not emerge fully formed,
but are deliberately designed, much like buildings and bridges. Whereas
bridges are devised to weather gusts of wind and the pull of gravity,
international rules are designed to withstand members’ often conflicting
incentives, and the limited enforcement capabilities proper to an anarchic
5 See Wolfe (2009).
6 “It was rightly objected by many that the ‘full employment exceptions’ in the second part

were so all-encompassing that a country could do whatever it wanted in the name of
achieving full employment” (Krueger, 2009).
7 “The architectural challenge is to shape trade agreements that strike the right balance
between flexibility and commitments. If contingency measures are too easy to use, the
agreement will lack credibility. If they are too hard to use, the agreement may prove
unstable as governments soften their resolve to abide by commitments.” Foreword by the
Director General. WTO World Trade Report 2009, xi.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


1.3 The Dirty Secret of the Trade Regime

7

global system. Poorly designed bridges will collapse. Similarly, rules that
are not structurally sound will lead to the fracturing of the agreement. An
added complication arises from the fact that international rules are the
outcome of bargaining among states, rather than the product of a single
designer, and the design of flexibility has a way of favoring some countries
over others. Little wonder that flexibility is among the greatest points of
contention in international treaties.

There already exists an answer to the architectural challenge. Building
on the sensible premise that unless reliance on flexibility is made difficult,
states will invoke it all the time, the solution envisioned by political scientists and economists alike is to render escape costly (Rosendorff and
Milner, 2001; Rosendorff, 2005; Schropp, 2009). If countries that need
to temporarily exit their commitments under an agreement were made to
pay some “optimal cost,” then the benefits of flexibility can be attained,
all the while reassuring trade partners that the exercise of flexibility is
temporary, and that escaping states will re-enter compliance as soon as
it becomes feasible. This solution has been for some time a foregone
conclusion. And the effort of the corresponding research program, which
has grown rapidly in recent years, has turned to exactly how an institution would arrive at the “optimal cost” that would satisfy the double
requirement of the architectural challenge: low enough to allow flexibility
when needed, high enough to prevent abuse. This research program has
led to parallel beliefs over country behavior. Scholars have assumed that
given the choice, countries will always opt for the least constrained and
cheapest available option for escaping their obligations.

1.3 the dirty secret of the trade regime
The observation of state behavior should lead us to re-examine these common assumptions. The solutions to the architectural challenge proposed
by theorists, such as making escape costly, are not the ones pursued by
governments. Similarly, predictions that governments will invoke unconstrained flexibility provisions “all the time” have not come to pass. In fact,
these common beliefs cannot contend with what I call the dirty secret of
the trade regime.
The truth is that there is sufficient flexibility inserted into countries’
commitments to sink the global trade system without breaking a single
country obligation. Countries actually have at their disposal an arsenal of
flexibility measures which, it turns out, are largely unconstrained. Member states are free to resort to these provisions at their whim.

/B
B5898 :DB

G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:

The “Architectural Challenge” of International Rules

8

These are not limited to the aforementioned security exception, which
is found in GATT Article XXI. It is a small concern in comparison to
a mostly overlooked fact about countries’ tariff schedules, which looms
large in this book’s empirical analysis. As it turns out, there exists a large
gap between countries’ bound duties (the maximum tariffs they can levy),
and their applied duties (the tariffs actually levied at the border). As a
result, the average WTO member today can raise its average tariff by 18
percent overnight, without falling foul of any of its obligations. This is
a striking fact in itself, given how the trade regime is traditionally represented as the most legalistic, binding, “hard law” regime in global governance. On the highway of international trade, the average car could be
going at twice its current speed without actually breaking the speed limit.
Despite the absence of checks on their use, the existence of such flexibility has not led to the system’s downfall. The unconstrained flexibility
provisions of the trade regime have not been invoked abusively, and their
respective agreements have not been vitiated. The Article XXI security
exception has been invoked exactly once in the WTO era, and then, not
formally. Meanwhile, its sister provision, the GATT General Exceptions
(Article XX), did not see any use until it grew significantly constrained
through rounds of litigation during the GATT era, and then again during
the WTO period: the more restricted it became, the more governments
turned to it. As for the gap between bound and applied tariffs that would
allow members to raise the average tariff by 18 percent for “free,” countries have actually relied on such “binding overhang” less than on trade
remedies, their costlier, more complex, more constrained alternative.8 And
this, even during the worst economic crisis since the Great Depression.
The “catch-all” exceptions through history have fared similarly, rarely
leading to the abuse we might expect. Time and again, governments have
confuted warnings of spirals of defection, and refrained from exercising

loose exceptions. Norms have emerged against their invocation, until governments all but abandoned them.
In fact, states in the trade regime exercise restraint at every turn. They
do not attempt to maximize their access to flexibility, and appear instead
to act in accordance with findings I present in the book’s analysis section, where I demonstrate that simply having access to unconstrained
flexibility acts as a tax on trade. Even governments’ domestic allocation
of flexibility reflects similarly strategic behavior: governments minimize
8 As I show in the analysis in Chapter 7, this holds even once we account for the country

selection involved.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


1.3 The Dirty Secret of the Trade Regime

9

access to unconstrained sources flexibility precisely for those industries
most likely to push for its use. States also rely on flexibility measures in a
consistent fashion. When they do turn to unconstrained measures, it tends
to be under observable hard times, where necessity is self-evident. In the
absence of such observable necessity, countries seek not easy loopholes,
but institutional checks and domestic investigations. These allow governments to convey credible information to trade partners and domestic audiences about the circumstances driving their invocation of an escape clause.
It is difficult to reconcile countries’ observable self-restraint with what
we know about international relations. Under international anarchy, individual interests are not disciplined by a centralized authority and cooperation is deemed unlikely. In such a state of nature,9 it is the function
of institutions to credibly tie leaders’ hands through hard, enforceable
rules. In the absence of such hard rules, we expect every country to follow
its individual incentives, and together to produce a socially suboptimal
outcome.
Yet given the menu of unused flexibility provisions scattered across the

trade regime, it is no exaggeration to say that the ties that bind states can
be broken at any moment. The regime nonetheless achieves its objectives:
in international trade, we observe none of the rampant protectionism witnessed in a world devoid of multilateral rules, such as in 1930, when the
Smoot Hawley Tariff led to a protectionist wave that aggravated the Great
Depression. If the high level of contemporary cooperation is not reducible
to hard rules enforced by credible enforcement, nor to the reluctance to
pay for escape made costly, how do we account for it?
What underlies the set of trade rules and exceptions is countries’ continuous efforts to manage beliefs and expectations about one another.
Abuse of exceptions is ultimately not held back by legal constraints alone,
but by countries’ continual willingness to seek such constraints, even as
unconstrained mechanisms remain available. Straying from expectations,
for instance by relying on loosely defined exceptions in the absence of
true necessity, comes at a measurable cost to trade, even as such actions
may remain entirely legal. The study of flexibility thus holds an important
lesson for global governance as a whole. The country behavior we observe
has far more to do with reciprocity and informal cooperation than the
past decade’s focus on legalization and binding rules would lead us to
believe.10
9 Milner (1991, 71)
10 Abbott et al. (2000); Goldstein et al. (2000). Cf. Finnemore and Toope (2001).

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270

75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


The “Architectural Challenge” of International Rules

10

1.4 the design of escape provisions
What leads countries to opt for constrained flexibility provisions, even as

unrestricted alternatives are available, also serves to explain the specific
design of these provisions. The conventional solution of rendering escape
costly ignores a unique feature of the international trade regime, which is
that the temptation to cheat on agreements comes not from the decisionmaker per se, but rather from domestic groups that exert pressure on the
decision-maker. The designers of trade agreements take this feature into
account when deciding on the shape of flexibility rules. This leads them
to opt for contingent flexibility over cost-based flexibility. Existing rules
prompt countries to convey the validity of their escape not by compensating aggrieved parties, but by conveying the nature of the circumstances
underlying escape.
Domestic politics are one major reason for which countries join trade
agreements to begin with. Commitments at the international level increase
governments’ bargaining position vis-à-vis powerful import-competing
domestic groups asking for trade protection. The domestic level is also,
conversely, the main reason why countries include flexibility clauses in
these agreements, to act as an insurance policy against unexpected events,
when the political costs of compliance grow insurmountable. Domestic politics also account for the specific design of flexibility clauses: as
I demonstrate, rendering escape costly rewards lobbying for protection.
This is why governments opt instead for rules that make escape contingent
on the presence of observable hard times.
Specifically, the rules of the trade regime, as in a host of other legal
systems, have evolved to make escape contingent on the exogeneity of
underlying circumstances. That is, on whether the circumstances motivating escape were unforeseeable, and whether they were, or could have been,
willfully produced. Did the import surge in steel arise from unforeseen
developments? Was the price shock the result of uncontrollable factors?
If not, the invocation of the escape clause may be formally challenged
as a violation. Such requirements, far from constituting an impediment
to the use of the escape provision, are the very reason governments can
turn to it. Whereas states formally commit to an institution once, at the
moment of signing, they then continually recommit to it by shying away
from unconstrained exceptions, and opting instead for contingent flexibility mechanisms, the better to reassure their trade partners and domestic

audience. Ulysses is perpetually refastening his own ties.
The virtue of the contingent flexibility design that has emerged
in the WTO is reducible to a simple logic: since exogenous events

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


1.4 The Design of Escape Provisions

11

are independent, they carry no information about the likelihood of
their re-occurrence, and cannot, by definition, be willfully generated
by governments or domestic groups seeking protection. Making the
legitimate invocation of flexibility contingent on the occurrence of such
events achieves two objectives. First, it prevents the risk of contagion,
whereby one instance of escape makes such behavior more acceptable,
and more likely to be espoused by others – a constant concern for a
diplomatic institution so deeply entrenched in the notion of precedent and
reciprocity. Second, a design that makes escape contingent on exogenous
events prevents opportunistic behavior, whereby domestic actors would
try and exploit the option of escape by pushing governments to exercise it
short of true necessity, simply to gain a competitive advantage. Institutions
such as the WTO thus attempt to forestall the possibility of abuse of
flexibility, and the likelihood of spirals of defection, by allowing only
those instances of escape that arise from “unforeseen” events that threaten
to cause injury.
As I show, this simple logic underlies much of the WTO treatment of
flexibility. As with all rules, the design of flexibility does not emerge ex
nihilo. It reflects existing incentives and underlying concerns – in this case,

concerns about managing others’ expectations. One of the main functions
of institutions is to reduce unpredictability in the behavior of memberstates.11 Flexibility can amount to a step backwards in this regard: a
given country reacting to hard times by raising barriers to protect its
steel industry will make its trade partners wary that more protection will
follow, unless there is some means of clearly circumscribing the exercise
of flexibility to this single instance.
It is possible to assess state incentives in this regard by examining state
behavior in the absence of constraints on flexibility. When I do so, a
striking fact emerges. The cost of sowing unpredictability in one’s trade
regime appears high enough to lead countries to be discriminating in their
reliance on flexibility, according to the very logic underlying formal rules,
even in cases where those formal rules fall short. The logic according to
which observable exogenous shocks validate the use of flexibility provisions applies more widely than the rules that embody it. Ultimately, it is
the desire to manage the expectations of trade partners, investors, and
exporters that drives both the design of rules covering flexibility, and the
behavior of states in the absence of such rules.

11 Mansfield and Reinhardt (2008).

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


The “Architectural Challenge” of International Rules

12

1.5 an accident of history
The modern trade regime affords us a rare opportunity to glimpse into
countries’ incentives over flexibility. This book leads me to examine a
number of different systems of rules across different time periods, but

I return to the trade regime to empirically test my expectations. One
reason for this is the sheer length of GATT/WTO history, and the wide
availability of data this entails. Today, scholars have access to millions of
observations covering both trade policy and trade flows for every member
in the institution since its inception: we can examine Cameroon’s trade
policy on unbleached cross twill woven cotton fabrics across time, compare it to that of its neighbors, and see how it impacts every trade partner’s
exports of the same product over time. The wealth of these data is also,
conversely, this book’s great methodological challenge. I also rely heavily
on the rich archival records of trade negotiations between member-states
from the 1940s onwards. These show country representatives explicitly
debating some of the very questions I examine here.
The other factor that allows us to discern countries’ preferences
over flexibility is largely the result of an accident of history. During
the Uruguay Round, which began in 1986 and concluded with the
inception of the WTO, new member-states, and especially developing
countries, were allowed to bind their tariffs at very high rates in exchange
of getting rid of import quotas and other non-tariff barriers. Such
“tariffication” is an established process by which countries convert
all forms of trade protection into tariffs. This harmonization renders
subsequent comparisons between states’ policies and further rounds
of tariff abatement considerably easier. Yet in the case of the Uruguay
Round, tariffication created a new source of wiggle-room, in the form of a
large gap between maximum bound duties, and the applied duties actually
levied by governments at the border. This gap, called tariff “water,” or
“binding overhang,” means that today, the average member could raise its
duties by 18 percent overnight without falling foul of its commitments.
This led some observers to refer to the process as “dirty tariffication”
(Ingco, 1996). A number of WTO members, as well as the WTO’s
Secretariat, have since bemoaned the existence of this wide gap between
obligations and behavior, and the unconstrained tariff flexibility it has

entailed. As I demonstrate in the book’s empirical analysis, there is every
reason to think that allowing such levels of binding overhang was an institutional mistake, an accident with considerable unintended consequences.
There has been much backtracking in this respect by member-states since
the WTO’s inception, in an effort to seal off the cracks in the bulwark.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


1.6 Overview of Chapters

13

Costly though it may turn out to be to the trade regime, the phenomenon of binding overhang is of great value to scholars. It represents an
unprecedented opportunity to observe governments’ preferences. What
happens when states within an otherwise highly legalized forum with
sophisticated monitoring mechanisms have access to what is effectively
“free” flexibility? The book’s empirical analysis capitalizes on this accident of history to better understand countries’ incentives over flexibility.
An examination of country behavior in this respect leads one to conclude that the true risks of unconstrained flexibility are not found where
they are often thought to be. Given the considerable restraint observed
during the worst crisis since the Great Depression, warnings against sudden increases of tariffs across the board are likely to prove unfounded, just
as predictions that countries would turn wantonly to abuse the GATT’s
national security exception have not been borne out. Instead, the true cost
comes from the considerable uncertainty that unconstrained flexibility
generates, exerting a daily cost in the form of a tax on trade. This cost
is weathered disproportionally by agricultural sectors in developing and
middle-income countries.
In sum, the occurrence of the accident of history which has led to
the existence of binding overhang is what has made a great part of the
analysis in this book possible. It is what allows me to measure the cost
of uncertainty flowing from unconstrained flexibility (Chapter 6); the

way in which countries exercise restraint in negotiating for additional
overhang if they have flexibility from other sources; and the way in which
states choose when to use “free” flexibility vs. constrained flexibility
(Chapter 7).
The existence of binding overhang represents a hard test for international cooperation. It is a legal vacuum where the very protectionism
usually targeted by the trade regime is legally allowed. That countries
do not avail themselves of this policy space nearly as much as one would
expect holds considerable implications for our understanding of global
governance.

1.6 overview of chapters
The remainder of this book proceeds as follows. In Chapter 2, I outline
a theory of the design of international agreements and of the flexibility clauses within them, focusing on the trade regime. I use the building
blocks outlined above to explain why states value the option of suspending the rules under some circumstances, but want to prevent its abuse.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


The “Architectural Challenge” of International Rules

14

The question is, how do they manage this balancing act? I also tackle the
concept of efficient breach, which looms large in the debate over flexibility: should we expect trade agreements to allow countries to pay for the
right to temporarily breach their obligations, if everyone else is left as well
off as they would have been absent a breach? The common answer is yes;
I argue that taking into account the domestic political underpinnings of
trade agreements suggests the opposite. I also discuss how the notion of
precedent holds the key to governments’ puzzling restraint with regards to
vague, unconstrained, easy to invoke flexibility provisions. The outcome

of Chapter 2 is a series of empirical expectations over both the design
and the invocation of escape clauses which are then tested in subsequent
chapters.
In Chapter 3, I draw a brief intellectual history of flexibility in law,
tracing two of its central tenets through time: the notion of necessity, and
the notion of changed circumstances. As I show, these two concepts turn
up, again and again, in unexpected settings, from medieval ecclesiastical canon law to Machiavelli’s writings. My first objective in tracing the
intellectual history of flexibility is to demonstrate that for nearly as long
as there have been rules to constrain behavior, there have been additional
rules put in place to sanction transgressions in specified circumstances.
And these have systematically led to discussions about the “architectural
challenge” underlying such exceptions. There is universal concern over
the abuse of loosely defined exceptions. One solution that emerges with
striking frequency is to make the validity of escape contingent on some
exogenous necessity, that is, a state of overwhelming need that could not
have been willfully created. Another lesson concerning the true aim of
flexibility comes out of examining the intellectual history of flexibility in
law. While we are used to thinking of exceptions and escape clauses as
created for the benefit of their eventual users – the way tax loopholes are
offered to the wealthy – an examination of rules of a normative character,
like religious law, suggests a different reading. What comes across is that
flexibility is included not to protect its users, but rather to protect the
sanctity, or the normative pull, of the rules themselves from what designers realize is inevitable noncompliance under some circumstances. What
Chapter 3 draws out are the limits of law – those cases where the rules
must adapt to behavior, since the opposite is known to be unfeasible.
Chapter 4 jumps forward to the twentieth century, to consider the twin
exceptions of the GATT: the national security exception of Article XXI
and the general exceptions of Article XX. The argument with regards
to these two provisions, which are similar in many respects but have

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


1.6 Overview of Chapters

15

radically different histories, is one I make throughout the book. Countries
seek constraints. And against all expectations, governments are loath to
rely on unconstrained flexibility. The reason governments cite for this
restraint? The fear of setting a dangerous precedent. Countries fear eroding the contours of the exception in a way that would normalize its usage,
and might lead others to do the same. This is what explains the paucity of
invocations of Article XXI in the face of alarmist warnings that its abuse
would spell the end of the trade regime. Article XXI remains a failure;
not because it has been abused, but because it has fallen into disuse. The
contrast is made with Article XX, the general exceptions, which was also
criticized as overly loose and prone to abuse at its creation, and which
was also left largely unused by member-states, until the jurisprudence
from a series of legal rulings began adding constraints on its invocation. Remarkably, as the general exceptions became progressively more
constrained, governments became more likely to invoke them. What the
stricter requirements on the use of Article XX accomplished was to reduce
the risk that one invocation would engender another. Jurisprudence saved
Article XX from desuetude.
Chapter 5 then fills out the menu of flexibility options policy-makers
have at their disposal in the trade regime today. I first focus on the regime’s
quintessential escape clause, the safeguard. There, I show how the evolution in the design of the safeguard from 1947 to the late 1990s, away from
compensation and towards an examination of the circumstances leading
up to escape, serves as an apt illustration for the regime’s treatment of flexibility writ large. Archival evidence of discussions by country representatives provides valuable evidence of the awareness with which negotiators
undertook the reform of the safeguard. Today, countries cannot invoke
the safeguard merely by promising to compensate affected countries: they

must show that the safeguard is the result of an exogenous shock that was
“unforeseen.”
Beyond safeguards, Chapter 5 considers the two other trade remedies,
antidumping and countervailing duties, which together form the most
used flexibility provisions today. I then describe the emergence of binding overhang, and how WTO members have reacted to its availability.
Rounding out the flexibility policy menu, I consider the way in which
mechanisms outside of the trade regime, such as currency devaluations,
can achieve the same results as flexibility regimes, which sets up parts of
the empirical analysis. I also briefly review the option of renegotiations,
and discuss whether it should be regarded as a flexibility provision alongside the aforementioned mechanisms.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


The “Architectural Challenge” of International Rules

16

I end Chapter 5 by comparing the flexibility provisions inserted into
the WTO to those in preferential trade agreements (PTAs). This is an
opportunity to ask what factors drive variation in the design of flexibility
across the trade regime’s 600 PTAs. Finally, I spend the last part of the
chapter comparing the design of flexibility provisions in trade to that of
derogations in the human rights regime. There, I argue that the absence
of reciprocity in human rights carries considerable implications for the
design and use of flexibility. In this case, observers’ pessimism may be
warranted.
Chapters 6 and 7 contain the bulk of the book’s quantitative analyses,
and they serve as counterweights to one another: Chapter 6 delivers the
bad news, Chapter 7 the good. In Chapter 6, I demonstrate that the trade

regime includes more flexibility than is usually thought. In fact, countries
have access to sufficient policy-space to sink the trade system without
ever breaking a rule. And the mere availability of this high amount of
flexibility, and especially of unconstrained flexibility, acts as a tax on trade,
the magnitude of which has long been underestimated. This is of special
concern given how the countries that have most access to unconstrained
flexibility are developing countries. This lack of constraints on flexibility
is usually seen as a concession granted to poor countries, yet in a pattern which will be familiar to students of international trade, developing
countries may emerge as the net losers of such “special and differential
treatment.” I also show that there is considerable evidence for flexibility
provisions fueling the Law of Constant Protection, a phrase coined in
Bhagwati (1989) that suggested that if one source of trade protection were
eliminated, another would simply pop up elsewhere. Considering the case
of India and then Ecuador, I demonstrate that this appears to be the case
even within countries, at the industry level. I then exploit variation in
tariff lines’ implementation to demonstrate that the same seems to hold
across all WTO members. The demonstration that flexibility provisions
impose a tax on trade, and that they allow countries to backtrack on their
most ambitious commitments, is bad news for the trade regime.
Chapter 7 delivers the good news. Indeed, there is much to be sanguine
about: the wide availability of unconstrained flexibility has not led to the
regime’s collapse, even in the midst of the Great Recession. Even those
countries that could significantly raise their tariff rates overnight without
falling foul of their obligations have in most cases turned to contingent
flexibility mechanisms instead. Why? The explanation offered in the
book is that countries value escape clauses not in spite of, but because of,
their constraints and requirements. Chapter 7 provides evidence for this
belief. When countries have access to an alternate form of flexibility, as

/B

B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:

1.6 Overview of Chapters

17

with those states that have a freely floating currency that allows them to
devalue in cases of need, they are shown to be systematically less likely
to set aside large amounts of binding overhang, even after controlling for
a battery of country characteristics. Moreover, countries’ allocation of
wiggle-room across industries follows a similar story: countries are seen
offering unconstrained flexibility precisely to the industries least able to
abuse it. By contrast, the same flexibility is withheld from industries most
likely to abuse it, even as, or precisely because, these industries tend to
hold the most domestic political clout. Finally, patterns of use follow the
same story: countries are loath to turn to free, unconstrained sources of
flexibility, except when the circumstances they find themselves in show
self-evident necessity.
Chapter 8 takes stock of these findings, and uses them to make some
predictions about the likely evolution of flexibility in global governance
in coming years. I pay special attention to the lessons of the 2008 financial
crisis. A book about flexibility is necessarily also a book about hard times,
since it is in view of such hard times that escape provisions are included
in treaties in the first place. Here I ask, have the levees held? The answer
appears to be yes. Looking at the entire WTO era, we have observed far
less reliance on escape mechanisms than we might have expected during
the Great Recession. More interesting still is that this is not an artifact
of this most severe of crises, but a generalized phenomenon: the same
domestic crisis leads to less reliance on flexibility mechanisms of all sorts
if trade partners are also in the midst of similar hard times.

Following on the book’s main findings, the restraint witnessed during the global financial crisis cannot be said to have been strictly the
result of binding rules. This is because multilateral trade rules contain
far more policy space than is usually assumed. Taking full advantage
of the flexibility legally allowed by the regime would have led to dire
consequences. Formal legal rules have not done the heavy lifting; rather,
informal cooperation appears to have driven restraint. During the crisis,
the trade regime relied less on the hard law that stands as the hallmark
of institutional strength, and performed instead more as the diplomatic
institution that it is, providing information about, and a focal point for,
state behavior. As one negotiator declared during the GATT negotiations
in the late 1940s in a heated discussion about the risks of the vaguely
worded national security exception, the “spirit” in which countries would
invoke these provisions was the only true bulwark against abuse. And so
it was during the Great Recession. Time and again, states presented with
the option to escape their legal obligations “for free” turn away from it.
The task of this book is to help explain why.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9

C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


2
A Theory of the Design of Flexibility

This chapter develops a theory of institutional flexibility that seeks to
account for two phenomena. The first is the design of rules covering
flexibility; the second is the behavior of states invoking flexibility provisions. In doing so, this chapter seeks to address the regime’s architectural challenge: how can flexibility clauses be designed to allow countries
temporary reprieve when needed, while precluding abuse?
The architectural challenge emerges from the attempt to balance the
benefits and the costs of flexibility. The benefits are well established.
The costs are less often discussed, and include the unpredictability that

results from governments having the ability to suspend their obligations.
Governments’ seek to manage such unpredictability while retaining the
option to escape under hard times.
I discuss the main existing explanation for the design of flexibility, and
argue that it is at odds with the domestic factors that lead countries to
sign agreements in the first place. The remainder of the theory formulates
expectations over state behavior in the absence of constraints on flexibility, which I argue obeys the same logic that drives the design of rules in
the first place.

2.1 the debate over flexibility in
international treaties
Country negotiators are not alone in disagreeing about the design of
flexibility in trade. It is also the source of much debate among scholars.
In important respects, the disagreements over the design of rules can be
18
/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:


2.1 The Debate over Flexibility in International Treaties

19

brought down to a debate about a prior question: why do countries tie
their hands through international trade agreements in the first place?
Why Do Countries Sign Trade Agreements?
This question goes beyond the scope of this book, and much of the early
institutional literature has been dedicated to answering it. Yet expectations over flexibility hinge on underlying beliefs about why countries
commit to rules in the first place. Modify that initial premise (by considering instead another type of economic agreement, such as investment
treaties), and expectations over the design of flexibility change accordingly. The argument I put forward is premised on the importance of the
domestic political drivers of countries’ willingness to make international

commitments.
As Paul Krugman once put it, if there were a economist’s creed, it would
contain the lines “I understand the Principle of Comparative Advantage”
and “I advocate Free Trade” (Krugman, 1987). The liberalization of trade
leads to lower prices and greater consumer choice, it hedges against climate shocks or other interruptions of production, and leads to an efficient
allocation of capital. In view of these considerable benefits, economists
have long faced the hard task of explaining why free trade has never
emerged in any lasting fashion in practice. For as long as nation-states
have traded, they have also put up barriers to trade against one another.
Why would countries spurn the first-best policy, and why is there a need
for an institution, such as the GATT-WTO, to push countries to do what
is already in their best interest?
Since the nineteenth century, economic theory has recognized that
countries can benefit from exploiting their market power, that is, their
ability to affect world prices, by setting an “optimal tariff.” Such a tariff
would improve a country’s terms-of-trade, or the amount of imports it
could obtain in exchange of its exports, to the detriment of its trade
partners. The possibility of such a “beggar-thy-neighbor” policy leads
to a terms-of-trade prisoner’s dilemma, whereby every country with
sufficient market power has as its dominant strategy to impose an optimal
tariff, regardless of what other countries do, leaving everyone worse
off. Countries want to capture the gains from trade that accompany
trade liberalization, yet their individually rational behavior leads to a
socially irrational outcome. Accordingly, the original explanation for why
countries join trade agreements portrayed international commitments as
a means of alleviating the terms-of-trade prisoner’s dilemma.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B:

A Theory of the Design of Flexibility

20

Yet a puzzle remained. Although small economies wrought no benefit
from setting a unilateral tariff, given their lack of market power, these
economies often put up just as high trade barriers as large countries,
and only abated their tariffs upon making binding commitments through
binding treaties. Something else than the pursuit of a terms-of-trade
advantage was pushing them to protectionism.
The answer came from a closer examination of the domestic politics of trade, and the insight that leaders face competing incentives from
domestic lobbies and the median voter. Voters gain from trade liberalization through lower prices, while domestic groups – especially importcompeting industries and labor unions – gain from the protection from
foreign competition that trade barriers can supply. Import relief shields
domestic producers, but does so at the cost of distortions to the economy,
in the form of higher prices and misallocated capital investment. Leaders
want to get re-elected, and they must thus trade-off one form of political
support against the other. They can further their odds of election either by
appealing to the median voter’s preferences, or by appealing to domestic
groups that can provide blocks of votes and, most importantly, campaign contributions. These contributions, in turn, can be used to inform
uninformed voters, deter political competition, and act as a signal of a
candidate’s viability.
The implication is that leaders’ promises to voters about trade policy are futile: however much they proclaim to want to reduce economic
distortion by lowering trade barriers, the political recompense from offering protection to domestic interest groups may be such, that leaders will
gain from going back on their promises. The resulting “time-inconsistent”
nature of leaders’ promises provided the beginning of a solution to the
question of why governments sign international agreements.
In what remains to this day the reference model of the politics of trade,
called the “Protection for Sale” model, Grossman and Helpman (1994)
theorized which industries receive trade protection in a small competitive
economy. In their model, the general electorate loses from trade barriers, but lobbies can provide funds that help leaders get (re)elected in a

way that can make up for decreased support from the electorate. Importcompeting groups thus pull policy away from the median voter’s ideal
point, but compensate the government for the political costs of distortions to the domestic economy. In some circumstances, governments may
thus do better under a protection for sale scheme than under a free-trade
equilibrium, if they can extract higher rents that allow them to stay in
power.

/B
B5898 :DB
G 9 5 5 56 9 5

C,

C,

75 6D 8 9 BD 7BD9 270
75 6D 8 9 BD 7BD9 9D

4 9D
1 6D5D 9
C, 8 8B BD

B

/97
. 3

5

,

,

G6 97 B

9 .5 6D 8 9 .BD9 9D

B: