Content uploaded by Thomas Rixen
Author content
All content in this area was uploaded by Thomas Rixen on May 25, 2016
Content may be subject to copyright.
1 23
The Review of International
Organizations
ISSN 1559-7431
Volume 8
Number 3
Rev Int Organ (2013) 8:363-387
DOI 10.1007/s11558-012-9158-5
The politicization of international
economic institutions in US public debates
Thomas Rixen & Bernhard Zangl
1 23
Your article is protected by copyright and all
rights are held exclusively by Springer Science
+Business Media New York. This e-offprint is
for personal use only and shall not be self-
archived in electronic repositories. If you wish
to self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at link.springer.com”.
The politicization of international economic institutions
in US public debates
Thomas Rixen &Bernhard Zangl
Published online: 22 December 2012
#Springer Science+Business Media New York 2012
Abstract Recent research has noted a trend of increased “politicization”of international
politics, i.e., decisions of international institutions are increasingly debated and contested
within civil society. What is lacking so far are explanations for this trend. In this paper we
derive four potential explanations and empirically test them. The first two, society-
centered, hypotheses focus on the process of socio-economic modernization on the one
hand and civil society structures on the other. The second pair of polity-centered hypoth-
eses focuses on the decision-making power of international institutions and on their
legitimacy. We measure politicization on the basis of a quantitative content analysis of
US quality newspaper articles about four decisions of different international institutions in
the issue area of international taxation. Our finding is that politicization is driven by the
increasing decision making authority of international institutions rather than by the lack of
legitimacy of their procedures or the factors emphasized by society-centered approaches.
Keywords Politicization .International institutions .Authority .Legitimacy .
Mobilization .International trade .International taxation
JEL classification F13 .F21 .F55 .F68 .H87
1 Introduction
Not all politics is politicized. Historically, many political decisions were made behind the
scenes by rulers and bureaucrats without being reported to or discussed by the public.
Only with 18th and 19th century enlightenment has politics become publicly contested,
Rev Int Organ (2013) 8:363–387
DOI 10.1007/s11558-012-9158-5
Electronic supplementary material The online version of this article (doi:10.1007/s11558-012-9158-5)
contains supplementary material, which is available to authorized users.
T. Rixen (*)
University of Bamberg, Feldkirchenstr. 21, 96052 Bamberg, Germany
e-mail: thomas.rixen@uni-bamberg.de
B. Zangl
Ludwig-Maximilians-Universität München, Oettingenstr. 67, 80538 Munich, Germany
e-mail: Bernhard.Zangl@gsi.uni-muenchen.de
Author's personal copy
i.e., politicized. Even today not all politics —the making of collectively binding
decisions with the claim of enhancing the common good —is politicized. Take, for
instance, the bulk of bureaucracy or court decisions, as well as a good number of
parliamentary or governmental decisions which never catch the eye of the general public.
In international politics, a lack of politicization remained very common until the
late 20th century. While some international decisions such as peace treaties after great
wars, e.g., the Treaty of Vienna in 1815 and the Versailles Treaty in 1918, did spark
public debate, the much greater part of decisions was made by diplomats in the
proverbial backrooms. The general public either did not take note of international
politics or, if it did, it rarely got involved in public debates, let alone heated ones. In
the meantime, the lack of politicization in international politics has evaporated.
Widely discussed recent manifestations of politicization are the protests; for instance,
at the World Trade Organization (WTO) ministerial conference in Seattle in 1999 as
well as the Group of Eight (G8) meetings in Geneva in 2001 and Heiligendamm in
2007 (see, e.g., Rucht 2005). But even beyond these widely publicized events,
decisions of international institutions are increasingly debated in the wider public.
Not only the European Union (Hooghe and Marks 2009), but also other international
institutions are becoming increasingly politicized (Tarrow 2005; Bédoyan et al. 2004;
Zürn et al. 2007).
Whereas scholars agree that there is a trend of increasing politicization of inter-
national politics, which began in the 1980s and 1990s (Keck and Sikkink 1998;
Reimann 2006), there is disagreement over how to best explain this trend. Broadly
speaking, two families of explanations are relevant: First, society-centered
approaches regard politicization to be triggered by the fundamental process of social
and economic development. As citizens’socioeconomic resources increase and their
interdependence via trade and division of labor is strengthened, they are better able to
participate in political decision making and have a greater interest in doing so (Lipset
1960; Inglehart and Welzel 2005). Just as these processes led to the politicization of
national politics, now, in the age of globalization, they are taken to the global level
and cause the politicization of international politics (see, e.g., Rosenau 2003; Furia
2005). Second, polity-centered approaches attribute the politicization of international
politics to the increasing scope and authority of international institutions. They stress
that international institutions gain in “bite,”because their decisions no longer rely on
intergovernmental consensus alone, but can increasingly marshal majoritarian and
supranational decision making competencies (Zürn et al. 2007: 150–152; Hooghe and
Marks 2009:6–8; Woods and Narlikar 2001: 572–579). The paper seeks to clarify
which of these approaches can better explain the politicization of international
politics. We aim to advance the debate by explaining the variance across cases, rather
than by analyzing general trends of politicization over time, as much of the previous
literature has done.
We compare the politicization of four international decisions on international
taxation: (1) the 1976 General Agreement on Tariffs and Trade (GATT) decision that
US tax preferences to so-called Domestic International Sales Corporations (DISCs)
violated international law; (2) the March 2000 WTO decision that the tax preferences
the US gave to so-called Foreign Sales Corporations (FSCs) were not compatible
with WTO law; (3) the 1998 Organization of Economic Cooperation and Develop-
ment (OECD) decision to curb so-called harmful tax competition; and finally (4) the
364 T. Rixen, B. Zangl
Author's personal copy
2001 decision of the OECD’s Global Forum on Taxation (GFT) that tax havens have
to exchange information on foreign taxable income with other countries. While it may
be thought that international taxation —as a very technical and publicly little noticed
issue area —would not be significantly politicized, we show below that in fact it is
just as politicized as other, better-known cases of international decisions, such as the
WTO-decisions on genetically modified food or hormone-treated beef. Thus, it seems
justified to generalize beyond the particular cases of taxation to the politicization of
international economic institutions in the OECD world of modern democracies.
We first lay out our concept of politicization and derive two related (and possibly
complementary) hypotheses for each of the two approaches. Second, we introduce
our comparative case study design. In the third section, we describe how we mea-
sured the level of politicization by means of a quantitative content analysis of US
quality newspaper articles on the four selected cases. We show that the level of
politicization was comparatively high in the OECD case, medium in the WTO case,
and low in the GFT and GATT cases. Fourth, we assess the explanatory power of the
hypotheses. We find that politicization is driven by the increasing decision making
authority of international institutions rather than by their lacking legitimacy, the level
of societal interdependencies or the structures provided by an increasingly global civil
society. The conclusion contains a brief discussion of the implications for future work
on the politicization of international politics.
2 Explanatory approaches to the politicization of international policies
Politicization is given if issues that did not catch the eye of the general public
previously are debated in the public sphere. Our definition requires three conditions
to be met. First, the issues themselves must be political, i.e., they must deal with
collectively binding decisions that come with the claim of promoting the common
good. Second, civil society must take note of them: that is, the issues must be debated
in the public sphere. Third, the issues must provoke a visible degree of contestation.
There must be conflicting positions about what the common good entails (cf.
Schmitter 1969: 165–6).
1
In short, politicization is present if collectively binding
decisions increasingly become the subject of controversial public debates.
The politicization of international institutions’decisions
2
and the related phenom-
enon of civil society activism have received significant attention in recent research.
On the one hand, IR scholars have overcome the state-centric view of international
politics and become more interested in how an increasingly transnational civil society
reacts to, becomes involved in, and influences decisions made by international
1
Notice that politicization is different from democratization. While the concept of democratization implies
institutional reforms that grant civil society a formal say in collective decision making, the concept of
politicization does not require this. We can have politicization without democratization, while we can also
have democracy without politicization: take the public protests against the current regime in Iran as an
illustration of the former and the bulk of decisions made in Western democratic countries as an example for
the latter.
2
Institutions are defined as persistent and connected sets of rules that prescribe behavioral roles, constrain
activity, and shape expectations (Keohane 1988: 386). The most important international institutions are
international organizations and international regimes.
The politicization of international economic institutions 365
Author's personal copy
institutions (Keck and Sikkink 1998; Reimann 2006; Steffek et al. 2007; Scholte
2002; Rosenau 2003). On the other hand, social mobilization scholars have overcome
their nation-centric view and address issues of transnational politicization of an
emerging global civil society (Tarrow 2005;Kriesietal.1995; Bédoyan et al.
2004). In both strands of research there is growing consensus that a process of
politicization in international politics is in fact occurring. Civil society actors do no
longer see international institutions as mere technical instruments to help states
cooperate in order to achieve common goals. Rather they are increasingly seen as
political institutions making binding decisions on “who gets what, when and how”
(Lasswell 1936). While truly transnational politicization —viz., transnational move-
ments or advocacy networks successfully mobilizing against the decisions of inter-
national institutions on a cross-national scale —is still considered to be rare,
domesticated transnational politicization —viz., national movements mobilizing
against international decisions —is considered to be on the rise (Bédoyan et al.
2004:39–41; Tarrow 1998: 192).
Yet, beyond the consensus that there is a trend toward increasing politicization,
there is no agreement on what its reasons are (see the discussion in Zürn et al. 2007:
149–158). It is neither clear why politicization began in the late 1980s, nor why it
appears to be more advanced vis-à-vis the decisions of some international institutions
but still rudimentary for others. Despite this lack of theory, one can roughly distin-
guish society-centered arguments, on the one hand, and polity-centered arguments,
on the other.
3
In the following we derive testable hypotheses from both approaches.
2.1 Society-centered approaches
From a society-centered perspective the process of socioeconomic development and
modernization triggers politicization, initially on the national, but now on the global
level. It is a bottom-up story; politicization has its roots in society and is carried
“upwards”to the political system. Two complementary hypotheses can be distin-
guished, one pointing to the importance of interdependence, the other focusing on the
structures provided by civil society organizations.
(1) Interdependence: Due to socioeconomic development and ensuing functional
differentiation and increasingly complex interdependencies, more individuals
are affected and become interested in the public policies that constitute and
regulate these interdependencies (Deutsch 1966; Lipset 1960). The demand for
information grows and is met by a growing mass media —newspapers, radio,
and television. Better education and more sophisticated means of communica-
tion enhance individuals’capabilities to act according to their autonomous and
specific interests. Individuals thus create, support and join interest groups which
represent their interests in the political realm. The result is a process of politi-
cization in which individual citizens and the interest groups representing them
get involved in public debates about public policies and try to influence them
(Inglehart and Welzel 2005:15–47; Tarrow 1998: 4).
3
A third approach is policy-centered and maintains that the level of politicization largely depends on the
issue area (Lowi 1972). While we plan to test policy-centered hypotheses in future work, we bracket them
here by controlling in our four-case-comparison for issue area characteristics.
366 T. Rixen, B. Zangl
Author's personal copy
The interdependence hypothesis states that this process of politicization,
which took place on the national level in many Western countries in the late
19th century, reached the international level in the 21st century —the age of
globalization (Ecker-Ehrhardt 2011). Functional differentiation creates complex
cross-border interdependencies among individuals and societies. This triggers
ordinary individuals’interests in international public policies as well as the
attention of the interest groups representing them in the political realm. Eco-
nomic globalization, i.e., transnational flows of goods, services and possibly
labor, raises peoples’material interests in the global economy and the rules
under which it operates. While economic globalization may increase overall
welfare, it also creates new inequalities and tensions between winners and losers
of this process. In developed countries, highly skilled and mobile labor should
benefit from globalization, while low or unskilled, immobile labor could suffer
wage loss or unemployment. Likewise, and closer to our cases, import-
competing industries will suffer from liberalization, whereas export-competing
industries will benefit. According to the interdependence hypothesis, these
distributive conflicts become politically salient and not only lead to new interest
groups or additional cleavages in domestic politics (Hooghe and Marks 2009;
Kriesi et al. 2012), but may also lead to the politicization of international
politics. In particular those international institutions that set and/or enforce the
rules of the global economy (such as the WTO, IMF, or OECD), or set the
agenda for liberalization or regulation (such as the G-8 and G-20 respectively)
may become the targets of societal politicization.
4
All in all, the interdependence hypothesis expects the politicization of inter-
national institutions to correlate positively with economic interdependence. To
the extent that interdependence has steadily increased since World War II,
politicization should be stronger in the 1990s and 2000s than it was in the
1970s and 1980s.
(2) Civil Society: The second society-based hypothesis complements the first. Its
proponents agree that socio-economic development and interdependence drive
the politicization of both national and international politics. Yet, they claim that
interdependence is only part of the story of societal politicization. It is a
necessary, but not a sufficient, condition for politicization. Due to collective
action problems, individuals are hard to mobilize even if well informed and
directly affected. Individuals may understand that a collective effort with like-
minded individuals against a certain policy furthers their interests, but they also
understand that they are better off, if others bear the costs of mobilization while
they profit from the efforts of others as free-riders. Whether politicization occurs
or not therefore depends crucially on the existence of civil society structures that
motivate and enable individuals to organize themselves and make their interests
heard (Tarrow 1998:16–23;Kriesietal.1995;Tilly1978). On this view,
historically speaking, the politicization of public policies in the 19th century
only became possible after the emergence of modern interest groups, clubs,
unions, associations, and parties, i.e., modern civil society organizations
4
One may also argue that domestic governments provoke citizens’criticism of international institutions by
shifting the blame for unpopular decisions to international or supranational organizations.
The politicization of international economic institutions 367
Author's personal copy
(Tarrow 1998:47–50).
This process of civil society building has now reached the international level,
facilitating the politicization of international politics (Rosenau 2003:57–62).
The emergence of numerous international non-governmental organizations
(NGOs) is seen to provide the backbone of the political mobilization of indi-
viduals on a transnational and global scale (Tarrow 2005:15–33; Risse-
Kappen 1995; Rosenau 1995). The civil society structure hypothesis expects
politicization of international institutions’policies to be stronger in the 1990s
and 2000s —the age of NGOs —than in the 1970s and 1980s, when fewer
NGOs existed (Reimann 2006). Also, politicization should be stronger in issue
areas with dense networks of interest groups than in issue areas with few NGOs
(Keck and Sikkink 1998).
2.2 Polity-centered approaches
Polity-centered approaches, developed by global governance theorists, argue that polit-
icization on the international and on the national level is mainly triggered by increasing
“bite”of the respective polities. Politicization is a top-down process; political institu-
tions gain decision making power and this provokes counter reactions “on the ground,”
whereby these institutions are subjected to increasing public scrutiny.
5
Two hypotheses
can be distinguished.
(1) Centralized Decision Making Power: According to the first polity-centered
hypothesis the accumulation of the power to make and enforce collectively
binding decisions by modern nation states was the spark that ignited the flame of
mass politicization in the 18th and 19th centuries (see, e.g., Poggi 1990:52–79;
Tilly 1990). The rise of state power engendered a counter reaction by individ-
uals who refused to subordinate themselves to the state. This became more
pressing as the state regulated ever more issues during the course of the 19th and
20th centuries. It no longer only dealt with ensuring security on its territory, but
also with taxation, public infrastructure, economic growth, education, healthcare
and culture etc. (Genschel and Zangl 2008). The centralization of decision
making power led to politicization, because individuals often had diverging
interests with respect to the policies pursued and wanted to influence the state.
A comparable process is now taking place internationally (Zürn et al. 2007:
149). Many of today’s international institutions are gaining decision making
power (Mathews 1997; Strange 1996). They make ever more decisions that not
only regulate issues of inter-state conduct —so-called at-the-border issues —
but also increasingly more intra-state matters —so-called behind-the-border
issues (Zürn 2004: 268–270). In consequence international institutions address
not only states, but also individuals. Moreover, decisions in international
5
We acknowledge that there are other top-down paths to politicization than just the reaction to decisions of
international institutions. Most important is the politicization of non-decisions or ineffectual decisions
(Rixen 2008b). For instance, with regard to climate change or the global financial crisis, politicization does
not occur because international decisions are seen as too intrusive; rather, politicization results from a lack
of international decision making that would effectively solve these problems. We do not investigate the
politicization of non-decisions in this paper.
368 T. Rixen, B. Zangl
Author's personal copy
institutions are no longer based solely on intergovernmental consensus. Supra-
national and majoritarian decision making is on the rise. For example, the
European Commission or the International Criminal Court can engage in su-
pranational decision making; others, such as the European Council of Ministers
or the United Nations Security Council can make majority decisions (Genschel
and Zangl 2008). In any case, centralized decision making power of interna-
tional institutions triggers politicization (Zürn et al. 2007:149–152). This
hypothesis states that the politicization of international institutions’decisions
is stronger, the more centralized their decision making is.
(2) Legitimate Procedure: The legitimate procedure hypothesis accepts and builds on
the conjecture that increasing decision making power engenders politicization. Yet,
proponents of this hypothesis maintain that the intensity of politicization depends on
the legitimacy of the decision making procedures (Franck 1990). Thus, the polit-
icization of the masses in the nation-state context during the 18th and 19th
centuries was largely due to the lack of legitimacy of the absolutist states of the
old regime. And the nation-state’s increasing ability to productively absorb societal
politicization in the 20th century is seen as a consequence of its constitutionaliza-
tion, i.e., the democratization of decision making with regard to rule-setting and the
judicialization of decisions in the realm of rule-application.
The same now applies to international institutions. As international institutions gain
in decision-making power, the (perceived) legitimacy of their procedures for rule-
making and rule-application becomes crucial for keeping politicization at bay (Zürn
et al. 2007: 152–156; Woods and Narlikar 2001: 572–579). International insti-
tutions are expected to face increasing politicization if their rule making is not
organized according to the democratic principle that those affected have a say
and if their rule application is not institutionalized according to the rule of law
principle that grants access to judicial review. Hence the legitimacy hypothesis
expects institutions with inclusive rule-making procedures to be less exposed to
politicization than institutions which exclude those affected. At the same time it
expects institutions with politically independent judicial rule-application proce-
dures to be less exposed to politicization than institutions in which rule-
application is organized as a political process controlled by the targets of the
rules themselves. To be sure, as they allow a wider variety of actors to partic-
ipate inclusive rule-making procedures and judicial rule-application procedures
can be a source of politicization, especially when compared to authoritarian
institutions that are willing and able to repress broad participation in and access
to rule-making and rule-application procedures.
6
Yet, as international institutions
do not have and never had the ability to repress the politicization of their decisions it
can be expected that increasing inclusiveness of rule-making procedures and in-
creasing judicialization of rule-application procedures leads to less politicization. As
these procedures are considered legitimate, those who are negatively affected are
more likely to accept binding decisions (Franck 1990). In addition, once they
6
Indeed, the consensus in the social movement literature is that the relation between democracy
and politicization is curvilinear. Politicization is highest in states with a mix of authoritarian and
democratic characteristics and decreases towards both ends of the democracy–autocracy spectrum
(see Tarrow 1998: 77).
The politicization of international economic institutions 369
Author's personal copy
participate in these procedures critics are less likely to take radical positions and
tend to become more moderate (see, e.g., Selznick 1948).
3 Case selection: International tax policies
To test the plausibility of the four hypotheses we compare US public debates over
four contested international tax policies. The debates relate to four decisions of
international institutions that were made to reduce international tax competition
among national authorities. To gain competitive advantages, governments adopted
national tax policies that, in the context of economic globalization, benefit the
individual country, but have negative externalities for other countries. As the resulting
competition was perceived to be harmful, international institutions made decisions to
abate the respective tax policies. We study public debates in US quality newspapers
that followed these decisions. US quality newspapers report about international and
domestic reactions to the decisions of IOs. As we know from existing research that
international policies rather lead to politicization in the domestic setting than to truly
transnational mobilization (Tarrow 1998: 192; Hooghe and Marks 2009) this focus
provides us with a sensitive instrument for measuring both transnational and also
domesticated transnational politicization.
7
Our cases are the following:
(1) GATT: We selected as our first case the 1976 GATT decision that US tax
preferences for DISCs violate international obligations. The decision was one
of the earliest by an international institution on tax issues. It came as a conse-
quence of a 1972 complaint filed by the European Community (EC) under the
GATT dispute settlement system, and it led to heated debates between the US
and the EC. In July 1973, after a series of fruitless consultations between the
disputants and upon a request by the EC, the GATT Council decided to establish
a panel to settle the dispute. The panel’s work was repeatedly hindered by the
disputing parties, but finally, in November 1976, it published a report stating
that US tax preferences for DISCs constituted export subsidies that were illegal
under the GATT. Despite facing heavy pressure from almost all of the other
GATT members, the US blocked the adoption of the panel report by the GATT
Council for more than five years. Even after it finally gave up its blockade in
1981, so that the GATT Council was able to adopt the report, the US still refused
to comply with it for an additional three years. Only in July 1984 did the US
finally repeal its tax law in order to come into conformity with, and meet its
obligations under, the GATT (Zangl 2006).
(2) WTO: We selected as our second case the March 2000 WTO decision that US
tax preferences for FSCs, which were created as a substitute for DISCs, previ-
ously under fire by the GATT, were not compatible with WTO law. This
decision reflected a complaint to the WTO by the European Union (EU) in
November 1997. In July 1998, after obligatory consultations between the
disputing parties failed, the EU requested that a panel be established to decide
7
To the extent that US newspapers report about transnational movements and other international reactions
to the decisions, this will be picked up by our indicator.
370 T. Rixen, B. Zangl
Author's personal copy
on the legality of the US tax provisions. In its October 1999 report, the panel
concurred with the EU that the preferential tax treatment for FSCs constituted
export subsidies that were illegal under WTO law. The US appealed, but to no
avail. In its February 2000 report, the WTO’s Appellate Body also requested the
US to bring its tax laws into conformity with WTO law. In November 2000,
although it had been highly critical of the report, the US replaced its FSC
scheme with the so-called Extra Territorial Income (ETI) scheme, which it
claimed was in compliance with WTO law (Zangl 2006).
(3) OECD: Our third case is the April 1998 OECD decision to curb harmful tax
competition (HTC). Mandated by the G-7, the OECD identified harmful tax
practices and wanted to make tax havens modify their national tax laws so that
they could no longer “poach”other countries’tax bases (OECD 1998).
8
Since
the OECD does not possess the legal competence to enforce decisions upon
states, a “naming and shaming”approach was adopted. In June 2000 the OECD
drew up a blacklist of uncooperative jurisdictions in order to exert pressure on
tax havens. Further, some OECD countries threatened to impose “defensive
measures,”such as terminating certain provisions for the avoidance of double
taxation, if tax havens refused to comply. But tax havens questioned the
legitimacy of such requests, dismissing them as an undue interference with
national tax sovereignty (see, e.g., Sharman 2006:83–86). Also, tax havens
argued that OECD governments were hypocritical: While the OECD insisted
that non-OECD members modify their tax systems, they did not expect the same
from the infamous OECD tax havens, Switzerland and Luxembourg, which had
abstained from the original report (Rixen 2008a,2011).
(4) GFT: Our fourth case is the July 2001 decision by the OECD’s GFT that tax havens
must comply with information exchange (IE) requirements concerning foreign
taxable income. In 2001, tax haven governments found an important supporter in
their conflict with the OECD. After fierce lobbying by business interests, the new
administration of George W. Bush declared that the United Stateswould not support
efforts to prescribe to other countries what their tax systems should look like. The
Bush administration suggested limiting the scope of the project on better adminis-
trative cooperation, transparency and IE between haven and non-haven countries.
Although this means that tax haven countries would have to compromise their
secrecy laws (bank secrecy and no IE with other countries) they could leave their
national tax systems (viz., low tax rates) unchanged. The OECD agreed to the
American proposal and began a process of evaluating tax havens under the umbrella
of the newly founded GFT, which represents a more inclusive and less confronta-
tional approach, as both OECD member countries and tax havens are members and
have a voice. By March 2004, essentially all tax havens, except for Andorra,
Liechtenstein, Liberia, Monaco and Marshall Islands, had made advance commit-
ments to cooperate with the standard (Rixen 2008a).
We selected these cases because we expected them to be normal cases with respect
to the level of politicization. As “regulatory policies with distributive consequences,”
8
In addition, OECD governments were asked to commit themselves to abolishing so-called preferential tax
regimes, i.e., those regimes trying to attract foreign capital by offering better tax treatment than was
available to domestic investors.
The politicization of international economic institutions 371
Author's personal copy
they are seizing the middle ground of politicization and are thus not extreme cases.
On the one hand, they are less politicized than decisions on “existential issues”as, for
instance, Security Council resolutions dealing with military interventions such as in
Libya or decisions of the International Monetary Fund with regard to the bail out of
states such as Greece, or the European Central Bank’s decision to buy government
bonds of states such as Italy or Spain.
9
Of course, tax issues can be existential issues
as the famous Boston Tea Party illustrates, but the tax issues we are dealing with here,
are not existential in this sense. Yet, on the other hand the selected cases can be
assumed to be more politicized than decisions which are concerned with predomi-
nantly “technical issues”such as the World Health Organization’s (WHO) health
recommendations, or the United Nations Education, Scientific and Cultural Organ-
ization’s (UNESCO) World Heritage List. Sure, international tax issues are often
highly technical, and some of the tax issues we selected here certainly are technical,
such as the determination of cross-border transfer prices, but in contrast to issues
which are predominantly technical, they have obvious distributive consequences,
creating identifiable winners and losers. Our cases thus are neither highly politicized
existential issues nor hardly politicized technical issues.
10
More important, the selected tax cases can also be seen as “normal”among
regulatory policies with distributive consequences. There are of course distributive
decisions made by international institutions that are better (less well) known than the
international tax policies we selected, but they are not necessarily more (less)
politicized. To check for this we assessed the politicization of the WTO Hormones
and the Genetically Modified Food cases. While the 163 articles we found in our
newspaper research for the Hormones case indicate that it is better known than our
cases, the 71 articles for the Genetically Modified Food case suggest that it is less
well know than some of our cases which come with 119 (HTC), 117 (FSC), 69
(DISC), and 48 (IE) articles respectively. Yet, measured by the number of evaluative
statements per article (one of our four main politicization indicators; see below), the
level of politicization of the Hormones and Genetically Modified Foods cases come
close to the average of politicization of our cases. With 0.58 and 0.52 they score
similar to the FSC case (0.64), higher than the DISC (0.25) and lower than the HTC
(1.92) and the IE (0.83) cases. This underscores that we did in fact select normal cases
and further suggests that the level of politicization in the selected cases can be seen to
be broadly representative of regulatory decisions with distributive consequences.
While the selection of normal cases allows for the (cautious) generalization of our
findings, the selected cases were also chosen in the spirit of a most similar design,
9
While such policies do almost always have distributional implications too, the high level of politicization
associated with them is due to the fact that they involve issues of “life-and-death”(be it in terms of physical
or economic survival).
10
Our distinction among these three categories of policies and the expectations with respect to the level of
politicization resonates with Lowi’s(1972) categorization of policies. “Regulatory policies with distributive
consequences”belong to the category of what Lowi calls regulatory policies. As he discusses, the level of
conflict in that category will vary depending on the degree of “redistributive”implication (Lowi uses the
term redistributive for what we call distributive). Our category of “existential”policies is very similar to his
“constituent”policies. Finally, what we call “technical”policies would in Lowi’s terminology be uncon-
troversial regulatory issues without distributive consequences. It is quite common in the literature to
distinguish among merely technical and expertise driven policies and more conflict-prone distributive
issues (Majone 1997).
372 T. Rixen, B. Zangl
Author's personal copy
which allows us to control for both policy and country characteristics. While the cases
display the variance needed to assess the plausibility of society-based and polity-
based explanations they are very similar in terms of their policies and the countries
that are affected. To begin with, assuming that “policy determines politics”(Lowi
1972) we selected tax policies that are similar, to the extent that not only do they stem
from the same issue area, but they also exhibit the same strategic structure. To gain
competitive advantages, national authorities had previously adopted tax policies that
benefited their own economy, but had negative externalities for others. In addition, all
four policies created diffuse benefits for the general public but concentrated costs for
multinational corporations.
11
This similarity allows us to rule out the possibility of
policy characteristics distorting the comparison between the levels of politicization in
our cases. Moreover, the selected cases are also similar to the extent that US interests
were directly affected by international institutions’tax policies. As the politicization
of the same decisions might have been different if France or Germany, let alone China
or Russia had been affected, making sure that the US is significantly affected in all
cases allows us to rule out that country characteristics distort the comparison between
the levels of politicization in our cases.
Thus, while these similarities allow controlling for policy and country character-
istics they do not limit our ability to generalize from our findings. As shown above
we assume that our tax cases are representative for regulatory decisions with distrib-
utive consequences —and thus for non-existential and non-technical decisions —by
international institutions. And we expect that the cases are representative not only for
decisions that affect the US, but modern OECD democracies in general. As discussed
the levels of politicization might be different in other OECD democracies, but we see
no reason why the causes of politicization should be different in the US than in
Germany or France for instance.
4 Assessing the politicization of international tax policies
To test the four hypotheses we have to assess the level of politicization provoked by
the DISC case under GATT, the FSC case under the WTO, the HTC case under the
OECD and the IE case of the GFT.
4.1 Operationalization
To measure politicization, we analyzed US quality newspaper articles. We draw on
newspapers because they are a good proxy for public debates about policies. While
we certainly have public debates also broadcasted on radio or television and increas-
ingly public debates are taking place on the internet, communication scholars agree
that the quality press can be considered a good representation of public communica-
tion because it acts as an agenda-setter for other media and provides a reasonable
synthesis of all these channels of communication (Kriesi et al. 2012,39–40;
Brüggemann et al. 2009, 396). Moreover, we draw on US newspapers not only,
11
This also shows in the data. Most interest groups commenting on the policies in the media are business
actors.
The politicization of international economic institutions 373
Author's personal copy
because they are easier to access in electronic newspaper archives but also because
the US is affected by the decisions taken in the cases we selected. By focusing on
quality newspapers we rely on a particularly sensitive measuring instrument; had we
opted for mass appeal publications, we would have found only very few articles on
these policies. Note that since these papers do not only report on the reactions to
international institutions’decisions within the US, we measure both truly transna-
tional politicization as well as domesticated transnational politicization. In US news-
papers we looked for reports, commentaries/editorials, and letters to the editor, that
were concerned with the four cases respectively. With a complex keyword search in
the electronic newspaper archives, LexisNexis and Factiva, we identified articles in
The Financial Times,
12
The New York Times,The Washington Post,Business Week,
and The Wall Street Journal that were published in the aftermath of the relevant
decisions by the four international institutions.
In the GATT case we looked at articles starting from November 1976, when
the panel declared the DISC regime a violation of trade law, up to July 1984,
when the US repealed its tax law in order to meet its international obligations
under the GATT. In the WTO case we selected articles from July 1998, when the
WTO established a panel to address the EU’s complaint, up to November 2000,
when the US repealed the illegal FSC provisions in its tax law. For the OECD
case we searched for articles stemming from a period between April 1998, when
the OECD published its HTC report, and June 2001, when it gave up the attempt
to force tax havens to change their tax systems and substituted it for the less
demanding IE requirements. For the GFT case we selected articles starting from
July 2001, when the information exchange program was inaugurated, up to
March 2004, when a progress report showed that almost all of the countries
had committed themselves to IE. Overall, we identified 353 articles that reported
on or referred to our cases in the time periods specified.
While the simple number of articles may tell us something about politicization in
the respective cases, we think that it is too broad an indicator. Therefore, in order to
get a more adequate picture of the level of politicization in each of the four cases, we
have developed four indicators that are based on a more complex coding scheme (see
Fig. 1). Our scheme takes evaluative statements about the four selected policies as the
unit of analysis rather than articles. We focus on opinions expressed about the
respective policies, because, as explained above, politicization is about the degree
of public contestation of political decisions. Based on the coding scheme we devel-
oped four indicators:
(1) Our first indicator focuses on the type of actors contributing evaluative state-
ments. Are they state or non-state, i.e., societal actors? The rationale for this
indicator is that the more societal actors —civil society and business —
participate, the more public are the respective debates. For example, consider
the following passage from an article in the Financial Times of 7 February 2000:
“‘Some countries offer privileges, some don’t,’said Patrice, a bank worker.
‘Taking that away takes away a whole lot of jobs, a whole lot of opportunities.’”
12
The Financial Times is from the UK, rather than the US. Yet, we used the US edition, which can be
assumed to be reflective of US public debates.
374 T. Rixen, B. Zangl
Author's personal copy
This was coded as the statement of a societal actor.
13
A higher (lower) share of
statements made by societal actors implies a higher (lower) level of
politicization.
(2) Our second indicator measures the average number of evaluative state-
ments per article about the respective policy. This shows us if an issue has
been taken note of in the public sphere, and it tells us something about the
level of contestation. Consider the following passage from the Financial
Times of 25 February 2000: “Charlene Barshefsky, US trade representative, said
Washington disagreed with the ruling ….”This was coded as an evaluative
statement (in this case by a state actor). A higher (lower) average number of
evaluative statements of this kind per article is regarded as indicative for a
higher (lower) level of politicization.
(3) Our third indicator is the ratio of negative to positive statements. The idea is that a
policy attracting more criticism is more contested, whereas, if media coverage
largely contains praise, politicization is lower.
14
In order to construct this indicator,
we coded negative and positive statements about the respective policy. In the
above examples, the statements were negative; the following passage from the
Financial Times of 25 February 2000, by contrast, was coded a positive
statement: “Charlene Barshefsky, US trade representative, said Washington …
respected its WTO obligations and did not want this issue to jeopardise ties with
the EU.”This statement was coded as positive, because it represents an attempt
to justify compliance with an adverse WTO decision. In any case, a higher
(lower) ratio of negative statements is thus seen as indicating a higher (lower)
level of politicization.
(4) Our fourth indicator focuses on the ratio of polity to policy statements.
Statements referring to the polities in which the policies are made can be
seen to be of a more principled kind than statements merely referring to
the desirability of a particular policy. We coded a statement as a polity
statement, if it referred not only to the contested policy, but also contained
an evaluation of the international institution that was responsible for the
respective policy.
15
Consider for example this statement: “For many, the
crackdown on tax havens is an issue of sovereignty. People ask why an
independent country should change to suit clubs of wealthy nations such as
the Organization for Economic Co-operation and Development”(Financial
Times of 7 February 2000). It was coded as a polity statement, because it not
only calls into question the OECD’s tax policy but also the OECD as the
appropriate polity for dealing with the tax regimes of states that are not OECD
members (viz., tax havens). Notice that one statement may at the same time be
coded as a positive and a negative evaluation, because it assesses the
policy and the polity differently. For example, the previous statement by
13
An evaluative statement ends, when a new speaker starts to speak.
14
Alternatively, one could argue that contestation is highest the more equal the shares of positive to
negative statements are. We contend, however, that contestation is highest if there is more criticism.
15
We only code polity statements when they are clearly part of a statement on the policy we are looking at.
Hence, a general criticism about the WTO is not coded even if it is made in an article that is also mentioning
the FSC case.
The politicization of international economic institutions 375
Author's personal copy
Charlene Barshefsky on the disagreement with the ruling but respect for
WTO obligations was coded as both a positive polity and a negative
policy statement.
16
We take a higher (lower) ratio of polity-to-policy statements
to indicate a higher (lower) level of politicization.
We use these indicators to compare politicization across the four cases.
17
On the
basis of the values for each indicator in the respective case, we construct a rank order
of politicization, from the most to the least politicized.
18
We also aggregate across the
four indicators by simply totaling the rank orders of the four simple indicators. We
acknowledge that we do not know whether any of the indicators has greater value
than others for capturing the underlying concept of politicization; therefore we have
assigned them equal weight in the aggregation.
4.2 Empirical findings
To ensure inter-coder reliability the coding procedure was as follows. Based
on the above coding scheme we first developed a set of coding rules and
carried out a pretest. This was done by both authors and a research assistant.
On the basis of the pretest we refined the coding rules. Then both authors
coded about 20 articles together, in order to develop a common understanding
of how to apply the rules. After that each author coded about half of the
remaining texts on his own. In the final round, the authors discussed all codes
to check for inter-coder reliability. Overall, we identified 360 evaluative state-
ments, 17 on the GATT’s DISC policy, 75 on the WTO’s FSC decision, 228
on the OECD’s HTC project and 40 on the GFT’s IE policy. Table 1shows the
structure of our sample.
16
This also means that it may be the case that we only coded one speaker (indicator 1) with two or more
statements. In consequence the overall number of evaluative statements is higher than the number of
identified speakers.
+–
Polity
Who Speaks?
State or Society
Policy Policy Polity
Fig. 1 Coding scheme for evaluative statements
17
Nota bene: As we use shares, ratios and percentages the mere number of articles on the respective
decision (and thus the duration of a case) does not impact on any of our four indicators. They are all geared
towards measuring the intensity of public contestation and not towards assessing the availability of public
information.
18
By simply using the rank orders of politicization, we disregard the differences between the absolute
values of the indicators. This may not be entirely unproblematic, but given the difficulties in defining a
plausible metric across the indicators, this seems to be a reasonable simplification.
376 T. Rixen, B. Zangl
Author's personal copy
The four indicators plus the combined indicator applied to our four cases of
international tax policies show that the levels of politicization differ quite remarkably
(see Table 2).
19
(1) GATT: The level of politicization regarding the GATT decision that US tax law
for DISCs violates international trade law is comparatively low. In our sample,
the media reports on the GATT decision are predominantly neutral: they contain
information on the dispute between the US and the EC, and they describe the
positions of the disputants and lay out the legal reasoning behind these posi-
tions. The share of evaluative statements per article at 0.25 is much lower than
our four-case average of 0.91.
The low level of politicization in the GATT case is also reflected in the fact
that the bulk of evaluative statements stems from state, rather than non-state
actors. Even so, most of the time the reports boil down to the US insisting that
the DISC scheme is GATT compatible, despite a panel report to the contrary.
Consider the following example: “The US again argued yesterday that DISCs do
not violate GATT because they merely compensate exporters for disadvantages
arising from the US company tax system”(Financial Times of 22 July 1984).
Thus, the share of societal actors’evaluative statements of 0.2 (3:15) in the
GATT case is somewhat lower than the average of 0.31.
20
Finally, the low level of politicization of the GATT case is made apparent by
the low ratio of negative-to-positive statements, 1.13 (9:8), as well as the low
ratio of polity-to-policy statements, 0.21 (3:14). Both ratios are clearly below
our four-case averages of 2.07 and 0.26, respectively. Therefore we find only
one statement in our sample of 17 DISC related evaluative statements that
extends the negative evaluation of the GATT policy to the GATT polity. Here
the US claims not only that the GATT tax policy is wrong but also argues that
the GATT is the wrong polity to deal with tax issues: “The US has called for a
new forum within the General Agreement on Tariffs and Trade to study taxation
systems, which may subsidize exports in violation of GATT rules”(Financial
19
That they lead to very similar rank orders is prima facie evidence for the reliability of our indicators. The
OECD or the WTO case almost always occupy the first or second rank, whereas the GFT and GATT cases
almost always end up in third or fourth place. Only indicator 2 which relies on the average number of
evaluative statements is an exception to this rule.
20
Remarkably, all of the non-state actor statements that we found were made by business actors, while civil
society actors were entirely absent from the debate.
Table 1 The sample Number of
articles
Articles
without
evaluative
statements
Articles with
evaluative
statements
Number of
evaluative
statements
GATT 69 55 14 17
WTO 117 86 31 75
OECD 119 28 91 228
GFT 48 24 24 40
Overall 353 193 160 360
The politicization of international economic institutions 377
Author's personal copy
Times 01 July 1982). Even in this statement the tone remains comparatively soft,
further attesting to the low level of politicization in the GATT case.
(2) WTO: The overall level of politicization for the WTO’s FSC decision is higher
than for the GATT case; but compared to the OECD’s HTC case, WTO
politicization remains moderate. As in the GATT case, many articles in our
sample were neutral, containing no evaluative statements. The share of evalu-
ative statements per article, 0.64 (75:117), is thus lower than the four-case
average of 0.91. But, when measured according to non-state actor involvement
in public debates, the WTO case displays a much higher level of politicization
than does the GATT case. In our FSC sample we found 33 evaluative statements
made by state actors. The general tenor of the reports is the US challenging the
WTO for criticizing US FSC legislation. Nevertheless an additional 34 state-
ments were made by societal (non-state) actors.
21
The share of societal actor
statements, 0.51, is not only above the average of 0.31, but in fact the highest of
all of the cases.
The high level of politicization for the WTO case is also reflected in the high
ratio of negative-to-positive statements, 2.26 (52:23), as well as the very high
ratio of polity-to-policy statements, 0.53 (26:49). Both ratios are clearly above
the averages, 2.07 and 0.26, respectively. Worthy of note, however, is that many
negative statements on the WTO policy come with a positive statement on its
polity. In some of these statements the positive evaluation of the WTO polity is
actually put forth as an argument for why the US should comply with the failed
FSC policy, as in the Charlene Barshefsky statement mentioned above.
(3) OECD: Compared to our other cases, politicization of the OECD’s HTC decision
is the highest. It has by far the greatest share of evaluative statements per article
(1.92 or 228:119). This is almost twice as much as the average. The majority of the
statements reported in the newspapers were made by state actors or representatives
of international organizations; the share of statements by civil society actors is
nonetheless the second highest, but at 0.35 (74:209) it is close to the average.
21
In the FSC case, not only business actors but also civil-society actors participated in public debates.
Fifteen evaluative statements stem from business actors, while 19 were made by civil society actors.
Table 2 Politicization in the four cases
GATT WTO OECD GFT Ø
(1) Share of societal speakers 0.20 (3rd rank) 0.51 (1st rank) 0.35 (2nd rank) 0.18 (4th rank) 0.31
(2) Evaluative statements
per article
0.25 (4th rank) 0.64 (3rd rank) 1.92 (1st rank) 0.83 (2nd rank) 0.91
(3) Ratio negative to
positive statements
1.13 (4th rank) 2.26 (2nd rank) 2.8 (1st rank) 2.08 (3rd rank) 2.07
(4) Ratio polity to
policy statements
0.21 (3rd rank) 0.53 (1st rank) 0.2 (2nd rank) 0.08 (4th rank) 0.26
Overall result
a
14 (4th rank) 7 (2nd rank) 6 (1st rank) 13 (3rd rank)
a
The overall result was determined by summing up the ranks for each case across the indicators. The lowest
sum indicates the highest level of politicization
378 T. Rixen, B. Zangl
Author's personal copy
The overwhelming majority of all statements came from business or state
representatives from tax havens, who usually paint a very negative picture of the
policy and the OECD, often referring to it as an imperialist organization
violating the sovereignty of their jurisdictions. However, this is not the only
reason for the high ratio of negative-to-positive statements. Even journalists and
other civil society speakers who generally express sympathy for the OECD
policy criticize it, in this case, for having excluded the tax havens from the
decision making process. Consider, for example the following passage from the
Financial Times of 22 April 2000: “The game became an ‘us and them’play by
the OECD against offshore financial centres. This undermined the sound prin-
ciples underlying the work by allowing the tax havens to accuse their pursuers
of hypocrisy.”Statements along these lines were very common and, conse-
quently, the OECD case exhibits not only the highest ratio of negative-to-
positive statements at 2.8 (168:60), compared to an average of 2.07, but also a
comparatively high ratio of polity-to-policy statements at 0.2 (38:190).
(4) GFT: Politicization in the GFT case is low. At 0.18 (7:38) it exhibits the lowest
share of societal speakers for all four cases. Many statements are explanatory,
indicating that the OECD has changed its approach to become more inclusive
and less strict on the kinds of policies permissible. As in the OECD’s HTC case,
many reported statements come from tax havens, but their tone has become less
critical. The following exemplifies both aspects: “‘A shadow was cast over the
sector by the unfair and unjust assault on us by the OECD, and this took its toll
on international business,’says Reginald Farley, the economic development
minister [from Barbados]. ‘We were pleased we were able to convince the
OECD that our laws, provisions and regulations meant that its concerns did
not apply to us. We did not have to sign any letter of intent or give any
undertakings.’”
Once the conflict between tax havens and normal tax states cooled
down, public interest decreased significantly, as the smaller number of
articles indicates (48 for the GFT case as compared to 119 in the OECD
case). Nevertheless, the number of evaluative statements per article
remained relatively high at 0.83 (40:48). Evaluative statements are about
twice as often negative than they are positive (2.08 or 13:27). However,
the reasons for critique vary significantly. Once the OECD had compro-
mised its initial goals, the critiquecamefromtwosides:ontheonehand,
the organization was criticized for letting tax havens off the hook too
easily; on the other, tax havens still questioned whether the OECD was
really serious about implementing a level playing field, blaming it for
being too soft on its own members, Switzerland and Luxembourg. Never-
theless, the ratio of polity-to-policy statements is lowest in the GFT case at
0.08 (3:37).
Taken together we can see that there are two cases exhibiting a high level of
politicization, the OECD and the WTO cases, and two cases with a low level,
the GATT and GFTcases. Although the level of politicization is high in both the
WTO and OECD cases, the form of politicization differs. Most importantly,
while the policy draws criticism in the WTO case, the polity is often viewed in
positive terms. In contrast, the OECD polity is viewed quite negatively.
The politicization of international economic institutions 379
Author's personal copy
5 Explaining the politicization of international tax policies
How do the four hypotheses fare in explaining the varying levels of politicization in
the four cases? We first consider the two society-centered hypotheses and then turn to
the polity-centered hypotheses.
5.1 Society-centered approaches
According to the interdependence hypothesis politicization should be higher the more
interdependent countries in the world are. For the cases at hand a good indicator for
interdependence are private capital flows, since it is the income from private cross-
border investments that make up the tax base that is at stake in all four cases. We use
the sum of foreign direct and portfolio investment of the world as a percentage of
world GDP as reported by the World Bank (http://data.worldbank.org).
22
We use the
average values for the respective years, in which our cases took place, to measure the
level of interdependence associated with each case. Interdependence is lowest in the
GATT case, with private capital being 1.26 % of GDP from 1976 to 1984, and
much higher in the three later cases. Of the later cases, we take interdependence to
be equal for the WTO and the OECD cases because they occurred at more or less
the same time. International private capital flows were at 2.55 % between 1998
and 2001. They were highest in the GFT case —2.71 % on average between 2001
and 2004. The interdependence hypothesis thus predicts the following rank order
of politicization:
GFT >OECD ¼WTO >GATT
The second society-centered hypothesis focuses on civil society structures.It
predicts that politicization is higher in those cases in which more organizations exist,
making it easier for individuals to overcome the collective action problem of mobi-
lization. In order to measure civil society structures we use as a proxy the number of
NGOs concerned with issues of taxation, active at the time of the respective cases. We
searched through our data set of newspaper articles and identified 41 NGOs that were
mentioned at least once in conjunction with at least one of our cases of international
tax policies. We then identified the dates when these NGOs started to operate (this
information was obtained from the respective NGO websites). The dates ranged from
the late 19th century when, for instance, the American Bar Association was founded
to the early 21st century when the Center for Freedom and Prosperity came into
existence. In 1976, when the GATT case emerged, only 23 of these NGOs existed. In
1998, by contrast, when the WTO and the OECD cases became virulent, the number
of NGOs that we found active with regard to issues of taxation had climbed to 38.
22
This is a good indicator for interdependence in our cases, because they are all about the taxation of
income from international investment. This is the case if tax havens do not tax the income from foreign
affiliates (FDI) or accounts (portfolio investment), or if the USA decide to not tax the income of FSCs or
DISCS (FDI). In both cases, export-competing industries with international affiliates are advantaged by
these tax breaks, whereas import-competing industries with no international affiliates are denied this
advantage. Thus, this indicator is very close to the substance of interdependence that is at stake in our
cases. In addition, if we were to use a broader indicator of interdependence such as the KOF Index of
Globalization (Dreher 2006), the rank order of interdependence would be the same.
380 T. Rixen, B. Zangl
Author's personal copy
This number increased to 41 by 2001, when the GFT was created. This means that,
for the cases at hand, the civil society and the interdependence hypothesis predict the
same rank order of politicization
23
:
GFT >OECD ¼WTO >GATT
Neither society-centered approach fares particularly well in accounting for the
respective levels of politicization in our cases. What these approaches predict differs
quite substantively from the level of politicization that we observed. This becomes
obvious if we compare the rankings predicted by the hypotheses to the respective
rankings according to our four indicators of politicization. In Tables 3and 4below,
we have evaluated each hypothesis by measuring the sum of the rank distances
24
between the theoretical predictions and the empirical evidence determined by our four
indicators. The lower the sum of the distances, the better the respective hypothesis
fares. This comparison reveals that neither hypothesis predicts any of the rankings
derived from our four indicators correctly. The rank distance is three for two of our
four indicators and five for the remaining two. The total sum of rank distances for all
four indicators amounts to 16. The politicization of international institutions does not
seem to be simply triggered by growing economic interdependence or the increasing
global civil society structures.
5.2 Polity-centered approaches
According to the centralized decision making power hypothesis, it is expected that
the politicization of international institutions is stronger, the more centralized their
decision making. In operational terms, three forms of decision making in international
institutions are to be distinguished: (1) intergovernmental decision making is the least
centralized of these types, insofar as decisions are always based on a consensus
among governments. Each and every government has the possibility to veto any
decision. (2) Majority decision making is more centralized, because decisions can be
made against the governments of states holding a minority position. (3) The most
centralized form of decision making in international institutions is supranational. For
this type of decision making, supranational bodies of an international institution —
not national governments —are vested with central decision making powers. These
bodies can make decisions that go beyond merely reflecting the interests of member
states’governments.
According to this operationalization, we find that GATT decision making is largely
intergovernmental, i.e., decentralized. This holds in particular for the so-called GATT
23
We would have arrived at the same rank order if we had focussed on all NGOs rather than just those
NGOs that were mentioned in our population of tax-related newspaper articles. According to the Yearbook
of International Organizations (Union of International Associations (UIA) 2010) the total number of NGOs
was 9521 in 1978, right after the GATT case emerged; it was 42100 at the beginning of the WTO and
OECD cases in 1998, and 49471 in 2002, just after the GFT case emerged. Had we based the analysis on the
overall funding available to NGOs, this would have also produced the same rank order (Reimann 2006).
24
The rank distances are calculated as follows: If a hypothesis predicts that the WTO lands on the first
rank, whereas it is actually on the third, then this corresponds to a rank distance of 2. Going through each of
the four ranks we determine this distance and then add them up. If a hypothesis predicts an equal level of
politicization for two cases, then they are both ranked equally, e.g., on rank two. In this event, the
subsequent case is lowered by one rank, e.g., placed on rank four.
The politicization of international economic institutions 381
Author's personal copy
rounds that determine the fundamental rules for the international trading regime, and
to a certain degree it also holds for GATT dispute settlement procedures in which the
decisions on the DISC regime were made. Although the reports on GATT compati-
bility of US tax law were drafted by a panel of independent experts, the establishment
of the panels and the adoption of their reports required a unanimous decision by the
GATT Council in which all member states are represented. If not entirely, the process
largely amounted to an intergovernmental one, because even disputants can block any
unwanted decision. Decision making in the WTO, by contrast, is much more cen-
tralized. Of course, many decisions in the WTO are made through intergovernmental
Table 3 Summary of predicted
rank orders Hypothesis Predicted rank order
1. 2. 3. 4.
Interdependence GFT >OECD =WTO >GATT
Civil society structure GFT >OECD =WTO >GATT
Centralized decisions WTO >OECD >GATT >GFT
Legitimate procedure OECD >GFT >GATT >WTO
Table 4 Evaluation of the four hypotheses
Indicator Rank order Distance to hypothesis
1. 2. 3. 4.
(1) Share of Societal speakers WTO >OECD >GATT >GFT Interdependence: 5
Civil society: 5
Centralized decisions: 0
Legitimate procedure: 6
(2) Evaluative statements per article OECD >GFT >WTO >GATT Interdependence: 3
Civil society: 3
Centralized decisions: 6
Legitimate procedure: 2
(3) Ratio of negative to
positive statements
OECD >WTO >GFT >GATT Interdependence: 3
Civil society: 3
Centralized decisions: 4
Legitimate procedure: 4
(4) Ratio of polity to
policy statements
WTO >OECD >GATT >GFT Interdependence: 5
Civil society: 5
Centralized decisions: 0
Legitimate procedure: 6
Overall Result —Sum of Rank Distances Interdependence: 16
Civil society: 16
Centralized decisions: 10
Legitimate procedure: 18
382 T. Rixen, B. Zangl
Author's personal copy
negotiations and thus require consensus among the governments of all member states.
But the WTO dispute settlement procedure which produced the decisions on the FSC
case constituted a supranational process. Like the GATT system, reports are drafted
by panels of independent experts or, in the case of an appeal, the independent judges
of the Appellate Body; but the disputants cannot block the establishment of a panel or
the involvement of the Appellate Body, nor can they obstruct the adoption of reports.
In fact, reports are automatically adopted unless the Dispute Settlement Body takes a
unanimous decision against doing so (Zangl 2008: 830–831).
Decision making in the OECD case is rather centralized. While in principle,
OECD decisions are made as a result of intergovernmental negotiations among all
members there is an important majoritarian twist to the HTC decision. In this case the
OECD defined harmful tax practices from which not only members, but also non-
members had to abstain, in order to avoid being black-listed. As most of the non-
members rejected this decision, they found themselves in a position similar to that of
a minority after a majority decision has been made. Therefore, we consider this case
to be of the majoritarian decision making type. The majoritarian twist was overcome
with the onset of the GFT process. After the US defaulted on the original policy, the
process became more inclusive. Tax havens were invited to participate in the newly
founded GFT. By practicing more decentralized modes of decision making, the GFT
served as an entirely intergovernmental arena in which members and non-members
could now block decisions that they were unwilling to comply with. The decision
making power hypothesis thus predicts the following rank order of politicization:
WTO >OECD >GATT >GFT
The second polity-centered hypothesis focuses on legitimate procedures. It pre-
dicts the level of politicization to be high in those cases where the legitimacy of the
institution’s rule-making and rule-application procedures, i.e., their constitutionaliza-
tion, is low. First, rule-setting decisions should follow the democratic principle of
inclusiveness: those affected by decisions should have a say in the decision making
process. This implies, first of all, that all targeted states are part of the rule-setting
process. Moreover, the more rule-making is centralized and the more targets are not
only states it also implies that civil society representatives should become involved.
Of course, the mere participation of civil society stakeholders in the decision making
processes of international institutions does not make them democratic in any de-
manding sense of the word. Nevertheless it approximates one important principle of
democracy, namely, that those affected by the rules should have an equal chance to
participate in the deliberations that determine those rules (Steffek et al. 2007; Scholte
2002). Second, decisions concerning the application of rules to specific circum-
stances must follow rule of law principles. Binding rules must be applied consistently.
This generally implies that decisions are made by politically independent judicial
bodies deciding exclusively on the basis of legal reasoning (Keohane et al. 2000).
Again, the mere existence of formally independent international bodies responsible
for judicial review is certainly not enough to ensure that the rule of law is actually in
place. Nevertheless, such bodies institutionalize rule of law principles, e.g., the
principle that like cases should be treated alike (Brownlie 1998).
According to these criteria, OECD decision making in the HTC case suffers from a
severe legitimacy deficit. After all, the OECD procedures did not follow the
The politicization of international economic institutions 383
Author's personal copy
democratic principles of inclusive rule making. Rather, OECD members tried to set
rules for states that could not take part in the rule making process, because they were
not even members of the organization. As rule making was hegemonic if not imperial
it severely violated the democratic principle of granting voice to all affected parties.
Moreover, the rule making process did not include civil society actors or other
stakeholders who could have lent it some legitimacy. In addition, OECD rule
application does not conform to rule of law principles. As there is no institutionalized
dispute settlement procedure, let alone a judicial one, rule application in the OECD
remains largely political. By contrast, the GFT which took the relevant decisions in
the IE case can be considered much more legitimate. While civil society actors were
still excluded, all affected states have now become part of the rule making process.
The GFT was able to overcome the OECD’s imperialist touch, because any state
could block decisions at any time.
25
The same applies to rule-making under GATT
and the WTO. Yet, as opposed to the GFT both can draw on dispute settlement
procedures that give rule application in the DISC and FSC cases respectively
additional legitimacy. However, in this respect the WTO is to be considered more
legitimate than GATT. The relevant GATT dispute settlement procedures are con-
trolled by the disputants themselves. They may block not only the involvement of any
GATT dispute settlement panel at their own discretion but also any decision the
GATT panel may propose. Thus, rule application remains a political process that
institutionalizes rule of law principles only to a limited degree. By contrast, the WTO
dispute settlement process is dominated by politically independent bodies —panels
and the Appellate Body —which consist of politically independent actors, viz., legal
experts and judges, who are committed to resolving disputes on the basis of existing
WTO law and engage in legal reasoning. In short, the WTO dispute settlement
procedure is comparatively more legitimate (Zangl 2008: 830–831). Overall the lack
of legitimacy hypothesis predicts the following rank order:
OECD >GFT >GATT >WTO
Among the two policy-centered hypotheses the centralized decision making
hypothesis proves to be more successful in explaining the varying degrees of
politicization in our four cases than the legitimacy hypothesis. Table 4shows
that the centralized decision making procedure hypothesis provides a much better fit
with respect to most indicators than the legitimate procedure hypotheses. While in the
case of the legitimate procedure hypothesis rank distances amount to four for all the
four indicators, in case of the centralized decision-making hypothesis rank distances
are zero for two indicators —thus providing a perfect fit —and amount to four and
six for the remaining two indicators. Thus the combined rank distances for the
centralized decision-making hypothesis equals 8 while amounting to 18 for the
legitimate procedure hypothesis. The politicization of international institutions seems
25
Note also that, contrary to the OECD case, there was no credible threat of sanctions in the GFT case.
Although OECD governments did not officially withdraw the threat of sanctions after policy had been
redirected towards information exchange, it was nevertheless agreed that no sanctions would be imposed
unless all OECD countries —including the tax havens, Luxembourg and Switzerland —removed the
harmful features from their own tax systems. This made the threat of sanctions ineffectual.
384 T. Rixen, B. Zangl
Author's personal copy
to be driven by the centralization of their decision making —a trend which cannot be
curbed by attempts to make the relevant decision making procedures more legitimate.
The overall comparison thus lends mild support for the polity-based and against
the society-based approach, because combined rank distances are 28 compared to 32.
The more specific comparison of hypotheses however reveals a clear-cut winner: the
centralized decision making hypothesis’prediction fits much better the level of
politicization in the four cases than the predictions of any of the other three alternative
hypotheses. With a sum of rank distances of 10 it fares much better than the
interdependence hypothesis, the civil society structure hypothesis and the legitimate
procedure hypothesis (Tables 3and 4). The politicization of the decisions of interna-
tional institutions thus largely reflects their decision making authority. As the com-
paratively high sum of rank distances for the legitimate procedure hypothesis shows,
this is so irrespective of whether they adopted legitimate procedures. It seems as if
authoritative institutions do not succeed in absorbing politicization by becoming
more legitimate.
6 Conclusion
While the politicization of international politics and institutions is a topic of interest
to many scholars in international relations, the empirical analysis of its causes has so
far received too little attention. Our paper engages in such empirical work. We focus
on four cases of tax-related decisions of international institutions, which are repre-
sentative of the large group of “regulatory policies with distributive consequences”—
the GATT decision on DISCs, the WTO decision on FSCs, the OECD decision on
HTC and the GFT’s decisions on IE. Operationalizing politicization in terms of public
contestation as reported in US quality newspapers, we show that politicization was
high in the OECD case, somewhat lower but still comparatively high in the WTO
case, but much lower in both the GFT and the GATT cases.
26
Our analysis shows that
the varying levels of politicization cannot be explained by society-centered
approaches focusing on cross-border societal interdependencies and the structures
provided by an increasingly global civil society. Rather, our findings lend mild
support to the polity-based approach. More specifically, the analysis supports the
centralized decision-making hypothesis which states that politicization is driven by
the increasing authority of international institutions.
There is one important caveat to this finding that we want to highlight and that
should be addressed in future research. The polity- and society-centered approaches
(and the respective hypotheses) are not necessarily incompatible. It might be fruitful
for future research to consider combinations of society- and polity-based premises
that accompany varying levels of politicization. For instance, it seems plausible that
some degree of interdependence and civil society organization is needed before
centralized decision making and legitimate procedures can have an impact on the
26
In addition, our data indicates that the politicization of the four decisions is not only the result of civil-
society activism, but was largely driven by business actors. A comparison of civil society driven and
business actor driven processes of politicization appears to be a promising avenue for future politicization
research.
The politicization of international economic institutions 385
Author's personal copy
level of politicization. In fact, this threshold may not have been reached in the GATT
case. Society-based approaches appear to offer a good explanation for why the GATT
case was much less politicized than the OECD and WTO cases, and also the GFT
case; this is a result that polity-based approaches cannot explain. While this is an
issue to be addressed in future research, it is important to stress that our paper is a first
step towards the goal of empirically investigating the interactions of different factors
contributing to the politicization of international institutions. Clearly distinguishing
the individual factors and drawing out their observed implications is a necessary first
step to investigate their interaction.
Acknowledgments We presented earlier drafts of the paper at the 2009 General Conference of the
German Political Science Association (DVPW) in Kiel, the “Transnational Conflicts and International
Institutions”Colloquium at the Social Science Research Center Berlin (WZB), the International Relations
Colloquium at the Geschwister-Scholl-Institute (GSI) at LMU Munich, the 2010 Conference on “Politics
beyond the Nation State”in Bremen and the 2010 SGIR Pan-European International Relations Conference
in Stockholm. We received helpful comments from participants at these events and from Martin Binder,
Klaus Dingwerth, Matthias Ecker-Erhardt, Monika Heupel, Tine Hanrieder, Martin Höpner, Andreas
Kruck, Peter Mayer, Fritz Scharpf, Duncan Snidal, Lora Viola and Michael Zürn. In addition, three
anonymous reviewers provided exceptionally helpful comments. Manuel Domes, Johannes Jüde,
Simon Primus, Anne Siemons and Mary Kelley-Bibra provided research and technical assistance.
We thank all of them.
References
Bédoyan, I., Aelst, P. V., & Walgrave, S. (2004). Limitations and possibilities of transnational mobilization.
The case of EU Summit Protesters in Brussels 2001. Mobilization, 9(1), 39–54.
Brownlie, I. (1998). The rule of law in international affairs. International Law at the Fiftiest Anniversary of
the United Nations. The Hague: Martinus Nijhoff.
Brüggemann, M., Hepp, A., Königslöw, K. K., & Wessler, H. (2009). Transnationale öffentlichkeit in
Europa: forschungsstand und perspektiven. Publizistik, 54, 391–414.
Deutsch, K. W. (1966). Nationalism and social communication. An inquiry into the Foundations of
Nationality (2nd ed.). Cambridge: MIT Press.
Dreher, A. (2006). Does globalization affect growth? Empirical evidence from a new index. Applied
Economics, 38(10), 1091–1110.
Ecker-Ehrhardt, M. (2011). Cosmopolitan politicization: how perceptions of interdependence foster citi-
zens' expectations in international institutions. European Journal of International Relations.
doi:10.1177/1354066110391823.
Franck, T. (1990). The power of legitimacy among nations. Oxford: Oxford University Press.
Furia, P. A. (2005). Global citizenship, anyone? Cosmopolitanism, privilege and public opinion. Global
Society, 19(4), 331–359.
Genschel, P., & Zangl, B. (2008). Transformations of the state —From monopolist to manager of political
authority. TranState Working Paper No. 76. Bremen. http://www.sfb597.uni-bremen.de/download/en/
publikationen/080918_AP_notPublished_Genschel_Zangl_TransformationsState.pdf (last accessed 17
March 2010).
Hooghe, L., & Marks, G. (2009). A Postfunctionalist Theory of European Integration: from permissive
consensus to constraining dissensus. British Journal of Political Science, 39(1), 1–23.
Inglehart, R., & Welzel, C. (2005). Modernization, cultural change and democracy. The human develop-
ment sequence. Cambridge: Cambridge University Press.
Keck, M. E., & Sikkink, K. (1998). Activists beyond borders: Advocacy networks in international politics.
Ithaca: Cornell University Press.
Keohane, R. (1988). International institutions: two approaches. International Studies Quarterly, 32(4),
379–396.
386 T. Rixen, B. Zangl
Author's personal copy
Keohane, R. O., Moravcsik, A., & Slaughter, A. M. (2000). Legalized dispute resolution: interstate and
transnational. International Organization, 54(3), 457–488.
Kriesi, H., Koopmans, R., Dyvendak, J. W., & Giugni, M. G. (1995). New social movements in Western
Europe. A comparative analysis. London: UCL Press.
Kriesi, H., Grande, E., Dolezal, M., Helbling, M., Höglinger, D., Hutter, S., et al. (2012). Political conflict
in Western Europe. Cambridge: Cambridge University Press.
Lasswell, H. (1936). Politics: Who gets what, when, how. Cleveland: McGraw-Hill.
Lipset, S. M. (1960). Political man: The social bases of politics. Garden City: Doubleday.
Lowi, T. (1972). Four systems of policy, politics, and choice. Public Administration Review, 33(4), 298–310.
Majone, G. (1997). From the positive to the regulatory state: causes and consequences of changes in the
mode of governance. Journal of Public Policy, 17(2), 139–167.
Mathews, J. T. (1997). Power shift. Foreign Affairs, 76(1), 50–66.
OECD. (1998). Harmful tax competition. An emerging global issue. Paris: OECD.
Poggi, G. (1990). The state. Its nature, development and prospects. Stanford: Stanford University Press.
Reimann, K. D. (2006). A view from the top: international politics, norms and the worldwide growth of
NGOs. International Studies Quarterly, 50(1), 45–67.
Risse-Kappen, T. (Ed.). (1995). Bringing transnational relations back in. Non-state actors, domestic
structures and international institutions. Cambridge: Cambridge University Press.
Rixen, T. (2008a). The political economy of international tax governance. Basingstoke: Palgrave/
Macmillan.
Rixen, T. (2008b). Politicization and institutional (non-)change in international taxation. WZB Discussion
Paper, SP IV 2008–306.http://bibliothek.wzb.eu/pdf/2008/iv08-306.pdf (last accessed on 26 January
2010).
Rixen, T. (2011). From double tax avoidance to tax competition: explaining the institutional trajectory of
international tax governance. Review of International Political Economy, 18(2), 197–227.
Rosenau, J. N. (1995). Governance in the twenty-first century. Global Governance: A Review of
Multilateralism and International Organizations, 1(1), 13–43.
Rosenau, J. N. (2003). Distant Proximities: Dynamics beyond Globalization. Princeton: Princeton
University Press.
Rucht, D. (2005). Transnational social movements in the era of globalization. In M. Herkenrath, C. König,
H. Scholtz, & T. Volken (Eds.), The future of world society (pp. 183–197). Zürich: Intelligent.
Schmitter, P. C. (1969). Three neofunctional hypotheses about international integration. International
Organization, 23(1), 161–166.
Scholte, J. A. (2002). Civil society and democracy in global governance. Global Governance: A Review of
Multilateralism and International Organizations, 8(3), 281–304.
Selznick, P. (1948). Foundations of the theory of organizations. American Sociological Review, 13(1), 25–35.
Sharman, J. C. (2006). Havens in a storm: The struggle for global tax regulation. Ithaca: Cornell University
Press.
Steffek, J., Kissling, C., & Nanz, P. (Eds.). (2007). Civil society participation in European and global
governance. Houndmills: Palgrave Macmillan.
Strange, S. (1996). The retreat of the state: The diffusion of power in the world economy. Cambridge:
Cambridge University Press.
Tarrow, S. (1998). Power in movement. Social movements and contentious politics. Cambridge: Cambridge
University Press.
Tarrow, S. (2005). The new transnational activism. Cambridge: Cambridge University Press.
Tilly, C. (1978). From mobilization to revolution. Reading: Addison-Wesley.
Tilly, C. (1990). Coercion, capital and European states. Cambridge: Basil Blackwell.
Union of International Associations (UIA) (2010). Yearbook of international organizations online:http://
uia.be/node/52 (last accessed 31 March 2010).
Woods, N., & Narlikar, A. (2001). Governance and the limits of accountability: the WTO, the IMF, and the
World Bank. International Social Science Journal, 53(170), 569–583.
Zangl, B. (2006). Die Internationalisierung der Rechtsstaatlichkeit. Streitbeilegung in GATT und WTO.
Frankfurt: Campus.
Zangl, B. (2008). Judicialization matters! A comparison of dispute settlement under GATT and the WTO.
International Studies Quarterly, 52(4), 825–854.
Zürn, M. (2004). Global governance and legitimacy problems. Government and Opposition, 39(2), 260–
287.
Zürn, M., Binder, M., Ecker-Erhardt, M., & Radtke, K. (2007). Politische ordnungsbildung wider willen.
Zeitschrift für Internationale Beziehungen, 14(1), 129–164.
The politicization of international economic institutions 387
Author's personal copy
