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

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 vitiat-ing the ag

International Rules

The Design of Exceptions and Escape

Clauses in Trade Law

KRZYSZTOF J PELC

All treaties, from human rights to international trade, include formalexceptions that allow governments to legally break the rules that theyhave committed to, in order to deal with unexpected events Such insti-tutional “flexibility” is necessary, yet it raises a tricky theoretical ques-tion: how to allow for this necessary flexibility, while preventing itsabuse? Krzysztof Pelc examines how designers of rules in vastly differ-ent settings come upon similar solutions to render treaties resistant tounexpected events.

Essential for undergraduate students, graduate students, and scholars

in political science, economics, and law, the book provides a hensive account of the politics of treaty flexibility Drawing on a widerange of evidence, its multi-disciplinary approach addresses the para-doxes inherent in making and bending international rules

compre-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, Interna- tional Studies Quarterly, Journal of Conflict Resolution, European Jour- nal of International Relations, British Journal of Political Science, and Journal of International Economic Law, among others.

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

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 http://lccn.loc.gov/2016010173 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.

List of Tables page ix

2.1 The Debate over Flexibility in International Treaties 18

2.3 Resolving the Architectural Challenge 312.4 Assessing Theoretical Expectations 39

3 A Brief Intellectual History of Flexibility in Law 43

3.1 The Universality of Flexibility 44

4 The Twin GATT Exceptions: Fears and Solutions 93

4.1 Article XXI: The GATT Security Exception 934.2 Article XX: The Value of Constraint 1224.3 Conclusion: Article XXI vs Article XX 132

5 The Evolving Design of Flexibility 137

5.1 From Compensation to Contingency 1385.2 What Room for Efficient Breach? 1505.3 Trade Policymakers’ Remaining Flexibility Options 1615.4 Flexibility in Preferential Trade Agreements 1855.5 A Comparison Case: Flexibility in the Human Rights

6.1 Does Flexibility Fuel the Law of Constant Protection? 206

6.3 Flexibility and Unpredictability 220

7.1 Restraint in Allocation of Flexibility 2357.2 “Country Seeks Credibility”: How Governments Choose

8.1 Rules vs Behavior during the Great Recession 261

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 tothink of a treaty that does not address uncertainty through the insertion of

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 visions that allow participants to legally breach an agreement’s primary

pro-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

becom-ing 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: theyvalue 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 Itmirrors the dilemma which underlies both the national and the interna-

tional 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

interna-tional body, and bound by its rules – but not uncondiinterna-tionally Addressing

the design of flexibility in the specific context of one international regime

leads me to grapple with this foundational problem How to design

effec-tive constraints on power that can stand up to the events of the real world?

Wherever flexibility provisions allow participants to suspend the rulesduring 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

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 thetradeoff 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 agreethat 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

vitiat-ing 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 Thereason 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

Secu-rity Exceptions, p.600, available at www.wto.org/english/res_e/booksp_e/gatt_ai_e/

art21_e.pdf

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

behav-ior, 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

“seek-ing 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

vot-ers, investors, trade partnvot-ers, 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

provi-sions, the outcome is not widespread misuse; it is disuse: governments

pro-gressively 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 tutes one of the greatest demonstrations of global cooperation between

consti-states, and one that has been largely overlooked The success of

interna-tional cooperation is tradiinterna-tionally 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

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

prece-dent” 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, escapeclauses are designed to deal with hard times and exceptional circum-

stances 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

nonethe-less 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

extraordi-nary 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 visions required? How do the designers of rules guard against the abuse

of flexibility? And given the abundance of unconstrained flexibility

pro-visions, why do we see less abuse than we might expect? The book’s

argument addresses these questions, and in so doing puts forth three

para-doxes: 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

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 developingcountries 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 stanceagainst flexibility provisions in trade, having all but invented them: the

very first trade escape clause was included at the US’ behest in a

bilat-eral 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 questionthat 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

were so all-encompassing that a country could do whatever it wanted in the name of

achieving full employment” (Krueger, 2009).

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.

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

sci-entists 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

insti-tution 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

com-mon 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

uncon-strained 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

Mem-ber states are free to resort to these provisions at their whim

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

repre-sented as the most legalistic, binding, “hard law” regime in global

gov-ernance 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 ibility has not led to the system’s downfall The unconstrained flexibility

flex-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,”

coun-tries have actually relied on such “binding overhang” less than on trade

remedies, their costlier, more complex, more constrained alternative.8And

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

gov-ernments 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

sec-tion, 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

selection involved.

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

govern-ments to convey credible information to trade partners and domestic

audi-ences 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,

indi-vidual interests are not disciplined by a centralized authority and

coop-eration 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 thetrade 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

wit-nessed 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’ tinuous efforts to manage beliefs and expectations about one another

con-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

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

decision-maker 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

compen-sating aggrieved parties, but by conveying the nature of the circumstances

underlying escape

Domestic politics are one major reason for which countries join tradeagreements 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

Domes-tic poliDomes-tics 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

motivat-ing 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

flexi-bility 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

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

member-states.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

provi-sions 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

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,

com-pare 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’ preferencesover 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

insti-tutional 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

Costly though it may turn out to be to the trade regime, the nomenon of binding overhang is of great value to scholars It represents an

phe-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

acci-dent of history to better understand countries’ incentives over flexibility

An examination of country behavior in this respect leads one to clude that the true risks of unconstrained flexibility are not found where

con-they are often thought to be Given the considerable restraint observed

during the worst crisis since the Great Depression, warnings against

sud-den 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 tothe 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 tional cooperation It is a legal vacuum where the very protectionism

interna-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

flexibil-ity clauses within them, focusing on the trade regime I use the building

blocks outlined above to explain why states value the option of

suspend-ing the rules under some circumstances, but want to prevent its abuse

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

flexibil-ity: 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

ecclesiasti-cal 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

design-ers 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 twinexceptions 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

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

erod-ing 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

invoca-tion 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-makershave 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

evolu-tion 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

flex-ibility writ large Archival evidence of discussions by country

representa-tives 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

bind-ing 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

along-side the aforementioned mechanisms

I end Chapter 5 by comparing the flexibility provisions inserted intothe 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

pat-tern which will be familiar to students of inpat-ternational 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 sanguineabout: 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

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 somepredictions 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 ing the global financial crisis cannot be said to have been strictly the

dur-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

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

pro-visions In doing so, this chapter seeks to address the regime’s

architec-tural 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 thebenefits 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, andargue 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

flexibil-ity, 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

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

expecta-tions over flexibility hinge on underlying beliefs about why countries

commit to rules in the first place Modify that initial premise (by

con-sidering instead another type of economic agreement, such as investment

treaties), and expectations over the design of flexibility change

accord-ingly 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 wouldcontain 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

cli-mate 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 thatcountries 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

Yet a puzzle remained Although small economies wrought no benefitfrom 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 tics of trade, and the insight that leaders face competing incentives from

poli-domestic lobbies and the median voter Voters gain from trade

liberal-ization through lower prices, while domestic groups – especially

import-competing 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,

cam-paign 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 icy are futile: however much they proclaim to want to reduce economic

pol-distortion by lowering trade barriers, the political recompense from

offer-ing 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

barri-ers, but lobbies can provide funds that help leaders get (re)elected in a

way that can make up for decreased support from the electorate

Import-competing groups thus pull policy away from the median voter’s ideal

point, but compensate the government for the political costs of

distor-tions 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

Subsequent studies sought a number of ways to account for whygovernments might nonetheless want to forgo the rents from lobbies.

In one such account, since investors internalize expectations about the

likely import relief they can obtain from government, they over-invest

in uncompetitive sectors in anticipation of protection This results in

long-run distortion which cannot be fully compensated through political

donations (Maggi and Rodríguez-Clare, 1998) Leaders who extract

rents from trade protection thus end up worse off than if they had

not offered this protection Yet they face the same time-inconsistency

problem as earlier models assumed: no matter their initial promises,

they may later gain from giving in to the demands of special interests

clamoring for protection, if these are sufficiently politically organized In

this view, commitments through international institutions could increase

governments’ bargaining power vis-à-vis domestic industries, providing

political cover to reject powerful groups’ demands for protection

While the ability to reconcile economic trade theory with governments’

behavior in joining agreements marked an important turning point,

the hand-tying function of such agreements had long enjoyed support

from legal scholars (Hudec, 1987) and practitioners At the very start

of the WTO, in 1995, the institution’s founding Director-General, Peter

Sutherland, spoke of “disturbing signs of a rise in protectionist sentiments,

in various guises, in several developed countries,” and argued that “the

existence of such pressures in domestic politics are all the more reason

to reinforce the counterweight of the multilateral system.”1International

commitments were being portrayed as a solution to national leaders’

domestic temptations It is thus no exaggeration to say that the very

design of the WTO internalizes the political pressures that governments

may face at home, and attempts to arm countries against them

In the political science literature, a complementary explanationemerged around this time that focused on the assignation of blame by

voters during hard economic times (Mansfield, Milner, and Rosendorff,

2000, 2002) According to this view, governments make binding and

enforceable commitments to free trade to shield themselves from voters’

recriminations in the event of economic downturns In the midst of such a

downturn, voters are thought to be less likely to punish governments that

have tied their hands through an international trade agreement Voters

Address by Peter D Sutherland, Director-General, World Trade Organization to senior

media representatives, Davos, January 29th, 1995 WTO Doc TBT/W/191.

trust the institution to cry foul if the government breaks its commitments

by offering distortionary transfers to special interests If the alarm is not

sounded, voters conclude that the economic downturn was not caused by

such distortionary policies, and they are less likely to vote the government

out Empirical evidence of leadership turnover during hard times offers

support for such a view: leaders that sign trade agreements appear less

likely to be voted out in the wake of economic downturns The function

of international agreements, according to this view, is thus not only to

provide leaders with political cover to deny distortionary protection to

domestic groups, but also to allow policy-makers to credibly claim: “it

wasn’t me.”

In sum, we can account for sovereign countries’ otherwise bafflingdecision to voluntarily limit their policy options by joining international

agreements through two complementary theoretical explanations: treaties

serve (i) to credibly commit to free trade vis-à-vis trade partners, in

exchange for those countries doing the same and (ii) to tie the

gov-ernment’s hands vis-à-vis the domestic audience, in a way that reduces

the political costs of withholding distortionary trade protection from

powerful domestic industries These two explanation provide the building

blocks of our discussion of flexibility

Why Do Agreements Require Flexibility?

If treaties serve to tie leaders’ hands vis-à-vis interest groups by imposing

external costs on socially suboptimal behavior, then why would leaders

include mechanisms within those treaties to “loosen the ties that bind”?

The one reason behind the need for flexibility is uncertainty An inherent

feature of the social world, the existence of uncertainty means that no

treaty, or contract, can prescribe behavior for all possible future states of

the world Countries may have willfully committed to liberalized trade,

but what happens if a war breaks out with a trade partner? How must

signatories behave if the costs of adjustment in the domestic economy

suddenly grow unbearable? Of if a domestic recession renders otherwise

competitive industries too vulnerable? The designers of treaties must find

ways of allowing states to adjust their obligations to suit unexpected

cir-cumstances, while preserving the legitimacy of the agreement Flexibility

clauses are an answer to uncertainty The greater the uncertainty, the more

we might expect the inclusion of some form of flexibility

Until very recently, the thinking about how to deal with uncertainty

in treaties focussed on the longevity of treaties, rather than on clauses

that could temporarily suspend obligations The question has long been

whether a change of circumstances should lead to the abrogation of the

treaty as a whole It is with this question in mind that John Stuart Mill

spoke of how uncertainty should affect the design of treaties:

Nations cannot rightfully bind themselves or others beyond the period to which

human foresight can be presumed to extend; thus aggravating the danger which,

to some extent, always exists, that the fulfillment of the obligation may, by change

of circumstances, become either wrong or unwise.2

Mill had identified the essence of the problem: circumstances may arise

where remaining in compliance would have such costly consequences that

the parties to the treaty, could they have foreseen those circumstances,

would have agreed to allow a party to suspend its obligations Yet

“human foresight” cannot predict all such circumstances, and so the

contract remains necessarily incomplete Mill noted that most treaties

already addressed this issue through reservations of one form or another

Yet those treaties that did not, Mill bemoaned, could only be abrogated

through “some lawless act,” and no provision existed to consider the

offsetting harm that would result from maintaining commitments even

as circumstances had changed Since such changes were inevitable,

uncertainty invariably entailed the undermining of faith in all treaties

Mill went as far as making a direct link between the expectation of

longevity and likely compliance, proposing that observance of the treaty’s

terms is greater if there is a known end to the treaty’s obligations As he

put it, “there is a very much greater chance of [treaties] being faithfully

observed, if a legitimate and peaceful emancipation from them is looked

forward to at the end of a moderate length of time.”3Mill thus foresaw a

similar tradeoff to the one envisioned by contemporary political economy

scholars: some allowance for suspending obligations in the face of “undue

loss or injury” would reinforce the treaty in all cases short of such

circumstances, increasing overall compliance Inner limits render outer

limits more binding: the existence of reservations within the treaty implies

that compliance is expected in any case that does not fall under these

There remains a strand of the contemporary literature dealing directlywith Mill’s concerns over treaty longevity, in the study of sunset provi-

sions, most often in the context of the Nuclear Non-Proliferation Treaty.4

provi-sions, another means by which the law seeks to address uncertainty Yet the occurrence of

renegotiations is rare, since these often lead to forbidding transaction costs (Friedmann,

1989) This is so especially when any change to a country’s rights and obligations

For the most part, however, international treaties today deal with

uncer-tainty by allowing countries to temporarily suspend their obligations,

and subsequently return to compliance, keeping the overall treaty intact

through the breach

Exactly how do these flexibility mechanisms help treaty signatoriesdeal with uncertainty? One can distinguish two broad answers: the first

views flexibility as an institutional mechanism included as the behest

of domestic interest groups, while the second sees it as a mechanism

serving leaders In the first case, flexibility functions as insurance against

exogenous shocks Liberalization exposes the domestic economy to

greater market uncertainty by opening it to a broader set of foreign

competitors A sudden surge of imports, for instance, can cause

sig-nificant injury to a domestic import-competing industry Flexibility

instruments serve to address these unexpected costs The Canadian

paper industry can thus obtain some reassurance that, although

liber-alization renders it more vulnerable to competition from technologically

advanced Scandinavian paper producers, Canada has a means of reacting

to a sudden surge in cheap imports that threatens Canadian paper

mill jobs This insurance function affects behavior Just as homeowner

insurance makes it easier for individuals to buy a house, knowing

their investment is protected against catastrophic loss, the existence of

flexibility can temper the fears of risk-averse actors It can reduce domestic

opposition to liberalization from industries exposed to the uncertainty

represented by trade liberalization A corollary is that by reducing

opposition to liberalization, flexibility-as-insurance can also allow for

deeper commitments on the part of states than would otherwise be

possible.5

The second function of flexibility is to allow leaders a means of dealing

with domestic political pressure: the illustrative term envisions flexibility

as a “safety valve,” akin to those installed on boilers to keep them from

exploding if the pressure within grows unmanageable In this case, the

pressure is of a political nature: it is the pressure that domestic

indus-tries clamoring for import relief impose on their leaders The premise

is that international institutions themselves recognize that since leaders

automatically applies to all Members of a multilateral institution, as it does through the

Most Favored Nation (MFN) principle in international trade Renegotiations must then

be conducted with all affected parties.

depth corresponds to the difference between prescribed behavior under the agreement and

counterfactual behavior, had the state not joined the agreement.

must remain responsive to their domestic constituents, they may grow

unable to turn down demands for import relief when these reach some

peak level, no matter what the social welfare implications may be When

their political survival is at stake, national leaders will want a means of

responding The insurance and safety valve functions of flexibility

repre-sent two sides of the same coin: the first from the point of view of firms,

the second from that of decision-makers

Strictly speaking, it is (import-competing) firms that feel the sting oftrade liberalization on their skin Yet these firms can shift some of the costs

they incur from liberalization, or some of the benefits that would flow

from protection, onto governments, as by supporting the government, or

its opposition, through votes or campaign contributions As Downs and

Rocke (1995, 88) put it, the relevant uncertainty that renders the contract

incomplete in trade is “uncertainty about the future demands of interest

groups.” In the event of true political necessity, leaders want a means of

egress, and in turn they are willing to extend this option to their trade

partners The greater point is that while flexibility clauses in trade are,

on their face, concerned with the interests of firms, they are written and

agreed upon by governments Taken together, the insurance and safety

valve functions of flexibility provisions point to the same tradeoff By

allowing a suspension of the rules in one discrete case – for one industry

over a temporary period – some distortion results, but the overall

author-ity of the treaty is preserved

Similarly, both the insurance and safety-valve functions of flexibilityprovisions mean that agreements with flexibility clauses are easier to enter

into Few countries would ever join a trade agreement with no measure of

flexibility Engineers would refuse to build bridges with no timber spacers,

lest they fracture at the first abrupt temperature change, and governments

would refuse to join entirely rigid agreements, lest they need to violate

the rules at the first unexpected price shock Similarly, import-competing

industries that know they will have recourse to trade remedies in one form

or another in reaction to unexpected import surges will expend fewer

resources to lobby against the agreement to begin with The same goes for

democratic leaders, who will refuse to tie their hands in a way that keeps

them from responding to domestic demands for protection when their

political survival depends on it Flexibility provisions exist in recognition

of these competing considerations They make countries more likely to

sign onto trade agreements, and more likely to make more ambitious

commitments when they do sign (Kucik and Reinhardt, 2008; Baccini,

Dür, and Elsig, 2015)

The phenomenon is a generalizable one E.H Carr observed that theincentives of international political institutions over the barriers to mem-

bership mirror those of a very different institution: “The dilemma of

inter-national law is that of ecclesiastical dogma Elastic interpretation adapted

to diverse needs increases the number of the faithful Rigid

interpreta-tion, though theoretically desirable, provokes secession from the church.”

In other words, when the first-best solution of demanding compliance

under all circumstances cannot be attained, or results in a diminished

membership, the inclusion of flexibility provisions represents the

second-best outcome

Beyond insurance and safety valve functions and their combined effect

on the depth of commitments, there is also a more subtle side to flexibility

clauses While we usually look for the benefits of flexibility among its

users – be they industry, or the governments they can shift their costs

onto – another major reason for including some measure of wiggle-room

may be in the preservation of the rules themselves In the Discourses on

Livy, Machiavelli praised the Roman Senate for temporarily going against

custom and law in allowing people to take arms and defend themselves if

the city was in danger The Senate understood that given such necessity,

legislation had to be adapted to circumstance, since the reverse could not

be done It determined that “what they [the citizens] had to do, they

should do with its consent, in order that they should not, by

disobey-ing through necessity, get accustomed to disobeydisobey-ing through choice.”6

The Senate was insulating the legitimacy and long-term authority of its

law from the negative effects of a temporary crisis Similarly, if we think

that some degree of domestic pressure may make continued compliance

impossible for leaders, the agreement itself may gain from condoning such

escape, since non-compliance has a common delegitimizing effect In this

way, flexibility may work to improve what is commonly called the rule

of law within an agreement – to maintain the perceived legitimacy of

an institution, and the particular grip on the mind that rules exert The

institution’s ability to compel states to behave in a given way rests on this

legitimacy being preserved

Finally, beyond political factors, there also exists a strictly economicargument for the inclusion of flexibility provisions According to this view,

some measure of flexibility is welfare enhancing Yet the case here is less

clear: while we can imagine a scenario where an otherwise profitable firm

needs temporary relief to adjust to a sudden rise in foreign competition,

this explanation hinges on labor and capital market failures The stock

counter-argument to the infant-industry principle (Baldwin, 1969), which

is concerned with similarly temporary non-competitiveness of nascent

industries, can be applied here too: if, following a period of adjustment,

firms can be profitable in the face of foreign competition, why would

capi-tal markets not step in to fill the need? Since, moreover, knowing to whom

to offer temporary import relief requires knowing which firms are likely to

be profitable in the long term, one can also ask whether governments and

the international rules they put in place are best designed to make such

calculations Is any given definition of necessity well suited to identifying

those cases where temporary relief will bring overall efficiency?

Just as recent explanations for countries’s motives in joining economicagreements have turned to political, over economic, factors, so too this

book looks to political mechanisms as driving countries’ decision to

include some measure of flexibility in treaties as a means of responding

to uncertainty

Be it as a form of insurance or a safety valve, or as a means of serving the perceived authority of law, the existence of flexibility clauses

pre-betters three aspects of international agreements: overall compliance, as

per Mill’s early intuition; the size of the membership, as more countries

are able to join a given agreement; and the depth of such agreements, as

governments can afford to make more ambitious commitments, knowing

they have a means of recourse if the costs of adjustment grow unpalatable

2.2 the costs of flexibility

One premise of the architectural challenge is that flexibility is necessary:

sovereign states will not join agreements that do not offer some means of

escape under some circumstances The other is that flexibility has

down-sides: too much flexibility renders the agreement meaningless Just as the

motives for which countries join international agreements are usefully

separated into international and domestic categories, so too are the costs

of flexibility usefully approached through this dichotomy

Recall that the first reason why countries might join a trade ment is to alleviate a terms-of-trade prisoner’s dilemma among coun-

agree-tries Such dilemmas grow less acute when trade partners’ actions grow

more predictable While a trade agreement increases such

predictabil-ity, by allowing countries to make credible binding commitments and

providing information about their behavior, any measure of flexibility

within such agreements will necessarily decrease predictability anew It is

trivial to say that an infinitely flexible agreement, that is, one that would

allow countries to escape under any circumstances, would be equivalent

to there being no agreement at all More subtly, the costs of flexibility

flow from the unpredictability that accompanies any provision

allow-ing states to suspend their obligations, when set against the

counterfac-tual of an entirely rigid agreement While uncertainty about the state

of the world is what leads to the inclusion of flexibility clauses in the

first place, increased unpredictability in the behavior of states is what

results from the inclusion of these provisions The architectural challenge

may thus be restated as the search for a design of the rules that would

tame uncertainty, while minimizing the resulting unpredictability in state

behavior

In some cases, flexibility may mean that an exporter is less able topredict the treatment its exports will receive at trade partner’s border the

following day For reasons I explore in Chapter 6, the Israeli wine industry

has access to a great deal of flexibility, allowing it to ask for import

relief overnight from the Israeli government The result is that US wine

manufacturers complain that they face “confusion and unpredictability

about any pricing of product, a very effective deterrent to selling in the

Israel wine market.”7

Recent work has shown how one major benefit of trade agreements isthe way they decrease trade volatility (Mansfield and Reinhardt, 2008), a

concept intimately linked to the type of predictability at issue here Insofar

as trade agreements are in the business of reassuring trade partners about

the treatment they will receive at the border in the following period, the

existence of flexibility should decrease this benefit in an observable way

Firms cannot costlessly reallocate resources from one market to another,

or one product to another Investments in fixed assets cannot easily be

unwound Unexpected changes in market conditions abroad can thus lead

to operating losses (Francois and Martin, 2004) If a given exporter has

to choose between an export market that has more access to flexibility

provisions and one that has less, it is likely to choose the latter.8The cost

of flexibility provisions is thus directly measurable as a decrease in trade

flows The result is that access to flexibility may act as a tax on trade, even

as countries remain in full compliance Chapter 6 empirically describes

this tax on trade effect

as this book suggests may indeed be the case in some instances.

The cost of countries’ access to flexibility provisions also has asystemic component In fact, the archival record suggests that it is this

aspect that country representatives worry about most Their concern is

that countries’ invocation of flexibility provisions has a spillover effect,

making other countries’ use of flexibility more likely The trade regime

remains entrenched in diplomatic mores In this setting, state practice

matters a great deal: countries’ expectations over what constitutes

acceptable behavior is itself the product of prior country behavior, and

others’ reactions to that behavior When policy-makers speak of the

systemic cost flowing from the invocation of a flexibility provision, they

invariably worry over the precedent it might set.

The frequent invocation of this term merits unpacking, as the notion

of precedent is usually employed in a legal context, whereby a ruling

exerts a causal impact on subsequent rulings, binding judges to the

court’s past opinions (Schauer, 1987) This is evidently not the meaning

in which country representatives employ it here Even the WTO’s dispute

settlement mechanism, as with all of public international law, is not

formally bound by precedent Yet we know that governments not only

employ the language of precedent, but feel it to have a behavioral effect

even in non-legal settings (Johnstone, 2003) The recent discussion as to

whether the 2014 American sanctions in response to Russia’s adventurism

in the Ukraine would resist a WTO challenge is an apt illustration: while

there was wide agreement that the United States could invoke a successful

positive defense of its sanctions (otherwise a prima facie violation of its

WTO commitments under the most-favored nation principle) under

Article XXI’s security exception, Russian officials themselves hinted at

how the United States would be reluctant to invoke it, for fear of setting

a precedent that Russia could subsequently exploit.9

The fact that this discussion took place over the invocation of cle XXI, that least constrained of trade exceptions, is not happenstance

Arti-As I argue below, the fear of setting a “dangerous precedent” is greater,

the less formally constrained an exception or escape provision is

Policy-makers’ use of the legal concept of precedent is thus indicative of the

widespread belief that one country’s invocation of a flexibility provision,

especially one that is largely unconstrained, normalizes the use of that

provision, thus increasing the odds that others will rely on it in turn This

spillover effect can be thought of in terms of contagion, or diffusion of

state practice

The third type of international-level cost flowing from the existence offlexibility is more nuanced still, and has to do with the distributional con-

sequences of flexibility Any move away from rigid primary rules is likely

to represent a concession to power As I argue throughout the book, much

of the true restraints on the use of flexibility are of an informal nature

This is good news, insofar as there emerge norms to address otherwise

unconstrained policy options, but it also means that some countries may

fare better than others under such informality Stone (2011) explores the

ways in which the advantages that informality offers great powers are

actually woven into the contract by design, in exchange for the promise

by those powers to bind themselves Similarly, flexibility rules are a more

natural target of power than primary rules It is likely easier to redraw,

through state practice and legal argument, the boundaries of what

consti-tutes a legitimate use of antidumping, than to flout one’s obligations by

increasing tariffs outright above the level allowed It is also likely to be

an exercise at which countries with high legal capacity excel

Because flexibility covers a manifestly perplexing area of law, where ameasure is at once in breach (because it flouts a country’s primary com-

mitments) and in compliance (because it does so under the cover of formal

flexibility provisions), it is a natural target for the opportunistic exercise

of power and legal capacity Almost half of WTO disputes concern states’

allegedly invalid use of flexibility This is also why domestic pressure for

protection usually pushes for abusive recourse to flexibility provisions,

rather than outright violation of primary rules As per Kono (2006)’s

concept of “optimal obfuscation,” it is harder for the domestic political

opposition to capitalize on an incumbent’s misuse of a technical trade

remedy such as a countervailing duty than to denounce an explicit illegal

barrier that comes at a measurable cost to consumers In sum, although

it remains difficult to quantify, it is worth keeping in mind how part of

the cost of flexibility lies in the potential reinsertion of power into the

equation Insofar as legalization is often argued to decrease the returns to

power, the inclusion of flexibility provisions can increase the returns to

power once more

Finally, just as countries join trade agreements in part for domesticreasons, so too are the drawbacks of policy space partly of a domestic

nature Specifically, if governments sign onto trade agreements to decrease

the political costs of denying distortionary protection to powerful

inter-est groups, the existence of flexibility risks “untying” the hands of

lead-ers, and thus annulling the benefits derived from credible commitments

through the agreement If the government has full discretion over the

invocation of flexibility measures, then the means by which firms obtain

trade protection do not vary much under a trade agreement from what

they might be absent that agreement In both cases, industries lobby Given

how the objective of the agreement is to increase governments’ bargaining

power vis-à-vis industry, the inclusion of flexibility clauses may thus have

the opposite effect If all that industries have to do to create a state of

“political necessity” for policy-makers is put up sufficient pressure, then

an agreement may do little to increase governments’ bargaining power

The risk is that insofar as international treaties tie governments’ hands,

flexibility clauses – and especially, as I argue below, poorly designed

flexi-bility clauses – untie them The costs of flexiflexi-bility hinge on the constraints

put on its use

2.3 resolving the architectural challenge

The Cost Solution

How are these constraints designed? If flexibility left unchecked increases

unpredictability and is thought prone to abuse, then states that seek the

inclusion of some measure of flexibility have an incentive to restrict its

exercise The conventional wisdom has long been that the way to do so

is to render its exercise costly The seminal political science argument to

this effect is made by Rosendorff and Milner, who claim that “for escape

clauses to be useful and efficient they must impose some kind of a cost”

(Rosendorff and Milner, 2001, 831) In the economics literature, this has

become a commonplace statement As the most recent book-length

treat-ment of flexibility has it, a “real problem occurs when intra-contractual

remedies are such that they over- or undercompensate the victim

Over-compensatory escape clauses are ‘under’-enacted, whereas

undercompen-satory ones are used too often” (Schropp, 2009, 11) In this view, the

cost of escape represents a credible signal to affected trade partners of a

country’s intent to return to compliance in the following period It must be

set low enough to allow states to escape temporarily when forced to do so

(to avoid “under-enactment”), but high enough enough to prevent abuse

(“over-enactment”) States then resort to the escape clause whenever the

domestic benefit from relying on flexibility rises above the cost of doing so

The necessity of putting a cost on escape has been echoed by economistswho have sought to derive optimal levels of compensation to ensure the

possibility of recourse to escape while preventing its abuse (Herzing,

2005) Much of this research draws on Rosendorff and Milner’s two-stage

model, which separates the agreement bargaining phase from an infinitely

repeated trade cooperation game between countries This breach and pay

scheme is said to be self-enforcing Escaping members themselves have the

strongest incentive to offer compensation, since they are looking to make

their future return to compliance credible Delegation of authority is

minimal: the central institution need only record and publicize instances

of escape and compensation International legal regimes therefore need

not prevent all breaches, but merely enforce the compensatory transfers

that render such breaches efficient In this way, Rosendorff (2005)

claims a state could breach a WTO rule, “willingly” submit to dispute

settlement, and accept retaliation by its trade partners Or it could provide

compensatory concessions on some other products (Bello, 1996) Schropp

(2009), who is interested not only in the level of compensation, but in

the exact form it should take, goes so far as to rewrite his preferred

version of the WTO’s flexibility provisions The proposed Article XIX

language, which covers the GATT escape clause, calls for compensation

in an amount that would render all affected parties whole, in keeping

with the theory of efficient breach: “the Member shall provide means

of trade compensation which put the adversely affected Members in as

good a position as had the Member applying the measure performed as

promised.”10 Yet as I seek to demonstrate, the evolution of the escape

clause over the trade regime’s history has been in the opposite direction

As Rosendorff and Milner acknowledge, their solution also has much

to do with efficient breach theory The doctrine of efficient breach, drawn

from contract theory, claims that if countries find themselves in a

sit-uation where they value the option of violating their commitments so

strongly that they are willing to weather the punitive consequences of

noncompliance, the legal system enforcing the laws should facilitate the

violation, and offset its negative impact by having the violator

compen-sate the affected countries to leave them as well off as they would be

absent the breach If such a transfer can truly render the affected parties

“whole,” and still leave the violator better off, then the breach is said to be

“efficient,” since it leaves all parties at least as well off as if the violation

had not occurred (Bello, 1996; Rosendorff and Milner, 2001; Schwartz

and Sykes, 2002; Herzing, 2005).11

[1881], 272): “The only universal consequence of a legally binding promise is that the law makes the promisor pay damages if the promised event does not come to pass In every case it leaves him free from interference until the time for fulfillment has gone by, and therefore free to break his contract if he chooses.”

Rendering escape costly is a plausible, elegant, and logically consistentmeans of addressing the architectural challenge Next, I propose another,

and argue why we should expect this alternative scheme to be the one

espoused by the trade regime

The Contingent Flexibility Solution

Like many theoretical propositions, this one emerges from the clash

between existing theory and a confounding fact In this case, the

con-founding fact is that the trade regime, as described in Chapters 4 and

5, actually contains a great deal of policy space that, for all intents and

purpose, is free for countries to exploit These flexibility provisions do

not require compensation, and they do not impose a formal cost on

their users Yet despite this, we observe no abuse, as the basic premise of

existing theory would predict, even when demand for flexibility would

be expected to peak In fact, these unconstrained measures are not very

popular How can we square this outcome with the belief that only

an optimal level of cost can provide a solution to the architectural

challenge, striking the right balance between allowing countries to escape,

while preventing abuse? More generally, it does not seem that cost or

compensation has ever been a popular feature in the design of exceptions

and escape clauses through legal history Why would this most simple of

solutions be consistently shunned?

I argue that for all its elegance, the theory behind the cost solutionomits two key points The first is that countries join trade agreements

to tie their hands domestically, in order to liberalize their economy The

second is that escape does not take place in a vacuum: governments

avail-ing themselves of the option to temporarily suspend their commitments

must internalize the effects this has on the expectations of other

govern-ments, foreign exporters, and their domestic import-competing groups

In the diplomatic setting of the trade regime, founded as it is on

reci-procity, and where the concept of “precedent” looms large, state

prac-tice matters One country’s behavior affects everyone else’s view of what

constitutes acceptable behavior As for domestic lobbying pressure, it is

not fixed, but endogenously determined That is, domestic groups adjust

their lobbying efforts in view of their likely success These two points –

countries’ reasons for joining the agreement in the first place, and the

way each instance of escape affects the expectations of other actors –

lead to a different set of theoretical expectations from those advanced

thus far

The alternative to cost is a mechanism that hinges on context One

that allows the user of a flexibility provision to convey the nature of the

circumstances leading to escape – specifically, that these circumstances are

severe, and that they could not have been willfully manufactured by the

user of the escape clause In sum, the alternative to cost is a mechanism

that makes legitimate escape contingent on an exogenous shock I refer

to this as the contingent flexibility solution

Making escape contingent on observable necessity selects for thosecircumstances under which a state would have an incentive to breach the

contract even in view of the penalty spelled out in the contract; those

circumstances under which the costs of continued compliance are so high

that breaches are unavoidable No regime gains from classifying such

instances of breach under necessity as noncompliance These are the

set-tings where, had these circumstances been predicted in the initial contract,

compliance would not have been expected A contingent flexibility scheme

proxies for such circumstances through observable outcomes, such as, in

the case of the trade regime, an import surge that threatens to injure the

domestic industry.12 In sum, a contingent flexibility relies on the

occur-rence of an exogenous shock to classify all possible states of the world as

falling under either primary rules, or flexibility provisions

Thus far, the contingent flexibility scheme is equivalent to the costscheme, which also seeks to mimic a first-best complete contract (Schropp,

2009) There, the distinction is made on the basis of a government’s

willingness to pay some predetermined cost for the right to temporarily

breach the agreement If the cost is optimally set, then countries will be

allowed to escape when the necessity they face reaches a critical

agreed-upon level In efficient breach designs specifically, the cost is set at the

a third party adjudicatory body that is able to vet countries’ claims about their domestic circumstances As it happens, the trade regime has spent its sixty-year history building just such a body In fact, as the archival record shows in Chapter 5, GATT members saw how the development of a functioning dispute settlement mechanism that could adjudicate countries’ claims was the required factor that might allow an contingent flexibility scheme to take hold By comparison, a cost-based solution requires only a third party body to verify the payment of compensation This, in itself, is no simple task.

Given the flexibility user’s incentive to underestimate the compensation amount, and the affected parties’ incentives to overstate it, the cost scheme may require an arbitration body to adjudicate these competing claims Schropp (2009) describes such a body The WTO is actually well-equipped to handle this arbitration function, since through its GATT Article 22.6, it has successfully arbitrated over the amount of retaliation following findings of noncompliance Although Article 22.6 panels remain rare, the WTO has the proven ability to quantify the amount of harm occasioned by a breach.

compensatory amount that leaves all affected parties whole, that is, as

well off as they would have been absent the breach

But here is where the two schemes begin to differ Recall that countriesjoin trade agreements to tie their hands vis-à-vis their domestic interest

groups And recall that policy-makers do not experience uncertainty

directly, but through the demands made on them by these domestic

groups The contingent flexibility scheme makes escape contingent on

observable, exogenous hard times, such as an import surge We can

think of this as economic necessity: the type that import-competing

domestic groups feel on their skins By contrast, the cost scheme proxies

for necessity by looking to governments’ willingness to pay It proxies

for political necessity, circumstances where a government faces such

high domestic demands that its political survival depends on ceding to

them A cost-based flexibility design thus makes escape contingent on an

endogenous shock, that is, a surge in political demands that is affected

by the likely success of lobbying

Indeed, as opposed to import surges, which are the effect of nous market forces, the demands of domestic groups are endogenously

exoge-determined: domestic groups mobilize if their demands have a chance of

succeeding And one of the factors determining the odds of these demands

succeeding is the design of flexibility rules itself In a cost-based flexibility

scheme, domestic groups obtain protection if they are able to impose more

benefits or pain on the government than it costs the government to breach

the rule and compensate affected parties In sum, making escape

contin-gent on the payment of cost alone increases the payoffs from lobbying

for import relief, and against liberalization By contrast, in a contingent

flexibility scheme, willingness to pay plays no role The validity of escape

is determined by the presence of observable, exogenously produced hard

times In their absence, for example, if there is no observable import

surge, lobbying does not pay off, and industry mobilization is unlikely to

occur

The requirement for exogeneity achieves a key objective By definition,flexibility rules contingent on exogenously produced hard times select for

circumstances that carry no information about the likelihood of their

re-occurrence If a country’s escape from its commitments is the result

of exogenous events, such as an unforeseen import surge, it does not

affect the odds of that same country relying on flexibility in the following

period, and just as importantly, it does not make other countries any

more likely to do so either, since their own escape is also contingent on

an unforeseeable event that cannot be provoked The requirement for