Democracy From The Outside-In? International Organizations and Democratization

Democracy from the Outside-In?
International Organizations and

Democratization
Jon C. Pevehouse
Democratic transitions have become a widely studied phenomenon in comparative

politics. The “third wave” of democratization spurred a considerable body of
research examining the origins and consequences of these transitions.1 One topic
that has received little attention within this literature, however, is international
factors that inßuence domestic regime transition. Philippe Schmitter adequately
summarizes the weight given to international factors in the democratization process
in much of the literature:
[O]ne of the Þrmest conclusions that emerged . . . was that transitions from
authoritarian rule and immediate prospects for political democracy were
largely to be explained in terms of national forces and calculations. External
actors tended to play an indirect and usually marginal role . . .2
Democracy is seen as the outcome of a domestic political process that is not

inßuenced by actors outside the nation-state. Recent literature in comparative
politics, however, questions this sweeping conclusion.3 Unfortunately, this new
literature has not yet developed core theories or examined cross-national empirical
data exploring the association of international factors with democratic transitions or
democratic consolidation.4
Despite this lack of analytical attention, democracy promotion has become a
foreign policy goal for many existing democracies, with international organizations
I would like to thank Edward MansÞeld, Donald Sylvan, Tim Frye, Michael Barnett, Leigh Payne,
William Howell, Beth Simmons, and Dan Reiter for their helpful comments. This paper also beneÞted
greatly from seminars at the University of California-San Diego and the Program on International
Politics, Economics and Security of the University of Chicago. In addition, the editors and three
anonymous reviewers at IO provided very insightful feedback.
1. See Huntington 1991; Geddes 1999.
2. Schmitter 1986, 5.
3. See Whitehead 1996a; Pridham, Herring, and Sanford 1994.
4. Pridham 1991a, 21.
International Organization 56, 3, Summer 2002, pp. 515–549

© 2002 by The IO Foundation and the Massachusetts Institute of Technology
516 International Organization
(IOs) as an important vehicle for achieving these ends.5 For example, the IO-
democracy link has been an important justiÞcation for the enlargement of organi-
zations such as the North Atlantic Treaty Organization (NATO), the European
Union, the Council of Europe, and even the expansion of the North American Free
Trade Agreement (NAFTA) into a Western Hemispheric Free Trade Association.
Both in academic and policy circles, this link is an oft-asserted—yet rarely
explained—association.
In the Þeld of international relations, traditional theories of international institu-
tions (for example, neoliberal institutionalism) do not offer much insight into how
or under what circumstances these organizations may promote democracy. Most
theories of international organizations concentrate on international outcomes (such
as interstate conßict and cooperation), and do not provide a basis from which to
draw strong causal linkages between international agents and domestic actors. One
body of international relations scholarship that does provide a basis from which to
consider these “outside-in” linkages, however, is the “second image reversed”
framework, which provides a general rubric for theories that discuss external
inßuences on state formation, structure, and institutions. This tradition, along with
theories of international institutions that open up the “black box” of the state,
provides a strong basis from which to build a theory that explicates how IOs may
inßuence the democratization process.
In addition to the lack of theoretical attention given to the IO-democratization
link, little empirical work investigates the relationship between IOs and democra-
tization. While existing literature is rich in detailed case study, there are no
cross-national empirical studies suggesting the conditions under which this rela-
tionship might hold. This lack of systematic analysis of IOs and democratization is
puzzling given the importance of the question to comparative politics, international
relations, and policymakers.
The article proceeds in four sections. First, I examine the general relationship
between IOs and democratization. Second, I explore three possible causal mecha-
nisms that elucidate the linkages between IOs and democratization. Third, I discuss
the “supply-side” issue concerning which organizations should be expected to be
associated with democratic transitions. Finally, I present a statistical test of the
argument linking membership in regional IOs and democratization.6
5. Christopher 1995.
6. I refer to intergovernmental organizations whose members are mainly geographically proximate as
regional IOs. Although the theoretical side of the argument should apply to IOs generally, I limit my
empirical test of the argument to regional organizations, excluding larger IOs such as the International
Monetary Fund (IMF), the World Bank, and the United Nations (UN). I discuss the justiÞcation for this
in the next section.
International Organizations and Democratization 517
Democracy from the Outside-In?
Most democratization studies present regime transitions as the outcome of a

domestic political process that is not inßuenced by actors outside the nation-state.
Internal forces play a dominant role in determining the prospects of transition,
whether those forces are structural (for example, the state of the economy) or
contextual (for example, virtu or fortuna).7 Even literature sympathetic to interna-
tional factors relies on domestic processes to explain regime change.8 External
actors are an intervening or facilitating variable rather than the central causal force
that brings a democratic transition.
The second image reversed literature provides an excellent starting point for
examining the relationship between IOs and democratization. This framework
encompasses theories contending that international factors inßuence domestic po-
litical outcomes. The international factors and the domestic political outcomes that
fall under the second image reversed rubric span a broad number of variables. Peter
Gourevitch’s initial survey of this literature considered factors such as military
intervention, international economic trends, and the (anarchic) nature of the inter-
national system.9 A variety of domestic political outcomes were also discussed
within the framework, including electoral outcomes, trade policies, domestic coa-
litions, and regime type/change. Although a review of the corpus of second image
reversed literature developed after Gourevitch’s effort is beyond the scope of this
article, I brießy discuss a piece of this literature on regime type and change to give
the overall ßavor of the argument.
Gourevitch’s two central discussions of regime change revolve around the
inßuence of international economics and the nature of the international state system.
In the latter realm, a litany of hypotheses concerning global economic processes
have played a key role in thinking about regime type and domestic political
institutions. Ranging from Alexander Gerschenkron’s work on the timing of
industrialization to James Kurth’s study of the product cycle and political authority,
many scholars have used international economics to explain the structure and
change of domestic political institutions.10 Recent strands of this literature include
work in comparative politics on economic crises and regime change.11 In these
works, political regimes are structured or altered to achieve the best possible
economic outcomes given the constraints and the dynamic nature of the interna-
tional economic system.
The anarchic nature of the international system—and the resulting drive for state
security—also provide a link from the international to the domestic sphere. Dating
from the late nineteenth century, the Seeley-Hintze Law holds that the greater the
7. On the importance of contextual factors in domestic regime transitions, see Linz and Stepan 1996.
8. Whitehead 1996a.
9. Gourevitch 1978. See also Almond 1989.
10. See Gerschenkron 1962; Kurth 1979.
11. Gasiorowski 1995.
insulation of a nation-state from outside inßuence, the less political power would be
centralized within the state.12 More recently, William Thompson argued that the
presence of external security threats to states can inhibit and erode moves toward
democracy. Democracy can suffer setbacks during security crises because leaders
will often consolidate their own power to mobilize resources to meet (or make)
external threats.13
Gourevitch emphasizes that, despite these potentially powerful external factors
affecting regime type, “[external pressures] are unlikely to be fully determining
. . . Some leeway of response to pressure is always possible, at least conceptu-
ally.”14 Thus any theory that purports to explain how international factors inßuence
fundamentally domestic decisions must contain some link to the domestic political
process. International forces create constraints and opportunities for democratiza-
tion through both economic and military-security processes. Yet this is only part of
the picture. One must also deÞne how the actors within the state cope with the
presence of these outside inßuences.
Unfortunately, the most well-developed literature on international institutions—
neoliberal institutionalism—largely ignores domestic politics.15 Most neoliberal
institutionalist research has focused on international outcomes, so it is unclear
whether the same causal mechanisms (decreased transaction costs, ameliorating
information asymmetries, etc.) link these institutions with the domestic political
process. Institutional theorists have recently called for more empirical research to
outline “well-delineated causal mechanisms” to explain the impact of international
institutions, especially with reference to domestic politics.16
Other theories of international institutions, however, do discuss domestic politics.
Literature from both new institutionalism and sociological institutionalism provide
starting points for specifying the inßuence of IOs on processes within the nation-
state. Theories within the new institutionalist framework explicitly link IOs (or
international institutions more broadly) to domestic politics through the causal
mechanism of commitment enhancement, due either to problems of time-inconsis-
tent preferences or to limit pressures arising from a pluralist society. Three works
within this framework stand out.
Robert Putnam’s two-level games framework suggests how a domestic actor’s
“win sets” could be constrained by IOs and institutions.17 Similarly, Judith Gold-
stein’s studies linking multilateral trade institutions to hand-tying in domestic
politics are especially important in bridging the international-domestic institution
gap.18 Goldstein’s argument asserts that executives can use these trade institutions
(for example, NAFTA or the World Trade Organization) as a tool to pursue broad
12. Almond 1989, 242– 44.

13. Thompson 1996; for a contrary position, see Reiter 2001.
14. Gourevitch 1978, 911.
15. Milner 1998.
16. See Martin and Simmons 1998, 749, 757; Keohane and Martin 1995.
17. Putnam 1988.
18. Goldstein 1996 and 1998.
general-welfare goals by constraining Congress’s ability to act. Finally, Etel

Solingen’s study of regional security arrangements holds that domestic coalitions
(which push for accession to these organizations) can be “bank-rolled” by interna-
tional institutions to pursue their own goals within domestic politics (even when the
domestic society may not be an example of a “pluralist” model).19 These studies
show how institutions external to the state may impinge upon choices made by
domestic actors.
The second body of literature that provides plausible causal mechanisms linking
IOs with domestic outcomes is sociological institutionalism. Martha Finnemore
shows how domestic preferences can actually emanate out of interactions within
international institutions and organizations.20 Some of this research, however,
centers on international norms rather than formal institutions.21 Still, many exam-
ples in the empirical literature focus on formal IOs such as the United Nations
Education, ScientiÞc and Cultural Organization; the International Labour Organi-
zation; or the World Bank.22 In these works, IOs effectively serve as a conduit of
norm diffusion and promotion focused on domestic actors.23 Thus domestic policies
may change, not out of a constraint imposed by the institution, but rather from a
fundamental change in a domestic actor’s preferences.
Although each of these bodies of theory provides an excellent starting point for
developing a theory connecting IOs to democratization, no work from either
tradition approaches the particular question of linkages between IOs and democratic
transitions. In the next section, I outline speciÞc causal mechanisms linking regional
organizations to regime change, paying close attention to both domestic politics and
extant theories of international institutions.
Regional Organizations and Regime Change:

Causal Mechanisms
Three potential causal mechanisms can explain the inßuence of IOs on regime
change. First, pressures (both diplomatic and economic) generated from these
organizations can, in combination with internal forces, compel autocratic regimes to
liberalize. Second, IO membership can lead to the acceptance of liberalization by
certain elite groups, as it can lower the risks that these groups face during the
democratization process. This acceptance of liberalization—labeled acquiescence—
can occur in two distinct ways: through a hand-tying process or through the
19. Solingen 1998.

20. Finnemore 1996a and 1996b.
21. Strang and Chang 1993, 237.
22. See Finnemore 1996b; Risse, Ropp and Sikkink 1999.
23. See McNeely 1995; Keohane and Hoffmann 1993. Also, Ikenberry and Kupchan 1990 contend
that hegemons can socialize elite groups in other states through the manipulation of material incentives.
This is a similar causal process, but the incentive manipulation occurs through the regional organization,
which encourages and even conditions beneÞts on certain behavior.
socialization of domestic elites. Before turning to the causal mechanisms, I brießy

discuss the scope of organizations that I include in this analysis.
Scope of Theory and Analysis

I limit the focus of my analysis to regional IOs for both practical and theoretical
reasons. Regional organizations are those in which membership consists of geo-
graphically proximate states. These organizations are the most common type of IO
in the world system.24 Examining the most ubiquitous type of organization provides
obvious advantages in terms of inferences drawn from any statistical Þndings. In
addition, the vast majority of theoretical literature on international inßuences on
democratization concentrates on regional organizations as opposed to larger, global
organizations.25 This is largely a result of the common focus of comparative politics
scholars on discrete regions when studying democratization. Given these practical
considerations, I concur with Laurence Whitehead that “the importance of such
international dimensions of democratization seems much clearer at this regional
level than at the world-wide level of analysis.”26
From a theoretical perspective, one also would expect the causal processes
discussed below to function more readily in regional organizations. Because
regional IOs tend to operate with small numbers and higher levels of interaction than
global organizations, causal processes such as socialization, binding, monitoring,
and enforcement are more likely in regional organizations.27 In addition, regional
IOs are more likely to possess leverage to pressure member states to democratize as
the vast majority of economic and military agreements are made under the auspices
of, or to create, regional organizations. These beneÞts are key to the functioning of
several of the causal mechanisms. For these reasons, as a Þrst empirical cut at this
question, I exclude non-regional organizations from my discussion and analyses.28
I also do not include international and regional Þnancial institutions (IFIs) in this
analysis. Although there is tremendous speculation and debate over the impact of
these organizations (such as the IMF, regional development banks, etc.) on democ-
ratization, there is little systematic research on this topic.29 I exclude these types of
institutions for two reasons. First, IFIs are likely to inßuence democracy through
fundamentally different causal processes. For example, although the assistance from
IFIs is often conditioned, membership is rarely conditional on anything other than
24. Shanks, Jacobson, and Kaplan 1996.

25. See Pridham 1991a; Whitehead 1996a. The one exception to this is the UN. A handful of authors
have examined UN efforts to promote democratization. See Russett 1998 and Russett and Oneal 2001.
The vast majority of academic and policy discussion, however, centers on regional organizations such as
the Council of Europe, the Organization of American States (OAS), and NATO.
26. Whitehead 1996b, 395.
27. See Nye 1987; McCormick 1980, 79.
28. A handful of the organizations included in this study have members that are “extra-regional.” Most
of these exceptions are cases where the United States or Canada is a member of an otherwise regional
military institution (for example, NATO). Nye deÞnes these organizations as “quasi-regional.” Nye 1987, 8.
29. For exceptions, see Kahler 1992; Haggard and Kaufman 1992.
paying the costs associated with membership.30 This assistance-based conditionality

is sometimes associated with democratization, yet the conditions of adjustment
loans are almost always economic in nature. While these economic conditions may
have clear political implications, political transitions to democracy (as measured
here) would be a second-order effect. This is important because an empirical test of
this IFI-democratization proposition would be different than the one undertaken
here and would need to reßect the multistage processes linking economic and
political liberalization.
Another difÞculty of testing the IFI-democratization proposition is that IFI
activity could merely be an intervening variable for economic crises. That is, IFI
activity may correlate with democratization because this activity is often spurred by
economic crises—which can have an independent effect on the prospects for
democracy. One would need to disentangle the inßuence of economic crises from
that of IFI policies, which again would entail a different empirical test than that
undertaken here.31 With these limitations on the theoretical and empirical scope of
the argument in mind, I now discuss some possible causal mechanisms that can link
regional organizations to democratization.
Pressure and Liberalization in Autocratic Systems

One common conclusion of the transitions literature is that elite schisms are an
impetus for political liberalization.32 Liberalization is not meant to unseat ruling
elites from power, but rather is meant to be a closed-ended process, a “controlled
opening of political space.”33 Liberalization can lead directly to democratization,
proceed slowly for many years, or end with more repression on the part of the
regime.34 Ideally, authoritarian leaders hope to expand their power through limited
change—increasing legitimacy and forestalling calls for further reform.
Although much of the literature is mixed as to what causes the split within the
ruling bloc, many argue that exogenous shocks (for example, economic disruptions
or political uprisings) force elites to take some action to restore the legitimacy of
30. Regional organizations condition membership: rarely do regional IOs condition a speciÞc beneÞt
while allowing full membership. Regional IOs tend to create various levels of membership (observer
status, associate membership, and full membership), each of which carries different obligations and
beneÞts. I include associate memberships in my analysis and more fully discuss this issue later in this
article.
31. These practical and theoretical issues concerning non-regional IOs and IFIs aside, I do re-run my
statistical analyses after including membership in several non-regional organizations (for example, the
UN) and IFIs (for example, the World Bank, the IMF, and regional development banks). As I discuss in
the following section, however, exclusion of these organizations does not bias the results in any
noticeable fashion. Despite this empirical robustness, further theoretical and empirical investigation is
needed to determine whether and how non-regional organizations and IFIs may inßuence the democra-
tization process.
32. See Kaufman 1986; Przeworski 1986; O’Donnell and Schmitter 1986.
33. Przeworski 1991, 57.
34. Mainwaring 1992.
their regime.35 Disagreements as to the prudent course of action then arise within the
authoritarian bloc. Some regimes may be able to weather the crisis (depending on
factors ranging from the nature of the current autocratic regime, economic condi-
tions, and/or the past performance of the regime), while other regimes may decide
to liberalize in an attempt to restore their legitimacy.
External pressure from regional IOs exerted on authoritarian regimes that are
undertaking liberalization can be a force for democratization. Once a regime has
begun to open political space, pressure and coercion by an IO of which the regime
is a member can push authoritarians to liberalize more than they otherwise would.
This pressure can undermine authoritarian rule in two ways. First, it can create
economic difÞculties for the regime if part of the pressure by the organization is the
suspension of trade and Þnancial beneÞts. This can further exacerbate economic
crises that can undermine the regime. Second, diplomatic pressures and the resulting
international isolation can help to further de-legitimize a regime at home since,
during these times of crisis, a regime’s international posture may be especially
important.36 If allies and institutional partners treat the regime as a pariah state, this
can inßuence public and elite perceptions of the regime within the state. Either
pressure can weaken an authoritarian regime’s grip on power, ultimately pushing the
regime to democratize.
Regional IOs can apply pressure in a variety of ways ranging from overt
de-legitimization of the regime by IO members through diplomatic pressure to direct
economic sanctions against the regime or even expulsion from the organization. To
understand how the pressure mechanisms work, two questions must be addressed:
(1) Why do member states pressure other members to undertake democratization?
and (2) Why is the regional organization the mechanism by which the pressure
occurs?
Why would democratic states pressure non-democratic states to become democ-
racies, especially in regional organizations? First, as a way to boost their own
international status and distance themselves from allies or neighbors, young democ-
racies may pressure former authoritarian partners to make a similar transition. As
Geoffrey Pridham argues, the act of foreign policy reorientation can lend internal
legitimacy to new democracies.37 Thus new democracies will have incentives to
treat autocracies (especially former political allies) as pariah states to establish their
own legitimacy.
Even established democracies may make the promotion of democracy a major
foreign policy priority for reasons of domestic legitimacy.38 In the words of
Whitehead, “[S]uccess in supporting democracy abroad has served to reinforce the
legitimation of the democratic order at home, and to boost national pride and
35. See Mainwaring 1992; O’Donnell and Schmitter 1986.

36. Whitehead 1996d, 273.
37. Pridham 1995.
38. BloomÞeld 1994.
self-conÞdence.”39 Certainly the efforts of the United States to promote democracy

can be viewed largely in this light.40
Second, if scholarly research concerning the economic and political advantages of
democracy is correct, then one would expect democracies to rationally desire to
have more democracies in the world. Research shows that democracies prefer to
trade, cooperate, and ally with one another.41 In addition, democracies better
promote economic growth and stability.42 Thus, expanding the number of democ-
racies expands interaction opportunities for existing democracies. Given the oppor-
tunity, a democracy may well attempt to push a non-democratic neighbor or trade
partner to liberalize.
Why will regional IOs be a potent source of pressure against authoritarian
regimes? Two factors make these institutions more likely to be chosen and to be
successful. First, regional institutions provide a highly visible forum to air com-
plaints against member states. This forum provides states of all sizes with a low-cost
“voice” opportunity.43 The beneÞts of international institutions in terms of lowering
transaction costs have been elucidated in other works on international institutions.44
Because these institutions provide an accessible forum for public condemnation or
economic sanctions, they provide a ready conduit for pressure from one regime on
another.45
Second, multilateral efforts will often be a favored mechanism of democracies
because they minimize the perception of overt “meddling” on the part of one
particular actor. For example, the United States has been criticized for its unilateral
efforts at democracy promotion in Latin America.46 If efforts to promote democracy
are widely perceived as illegitimate and a violation of sovereignty in the target state,
such intervention often backÞres, creating support for the authoritarian regime. With
the support of regional institutions, however, similar efforts can gain legitimacy
because of their multilateral nature.47
One scenario for IO-led pressure is the case of re-democratization after a member
suffers a breakdown of democracy. A prime example would be the OAS pressure on
Guatemala after the self-coup of Jorge Serrano. In May of 1993, Serrano dissolved
39. Whitehead 1996c, 248.

40. Smith 1994. There are, of course, cases where democracies have attempted to subvert young
democracies if other (often geopolitical) objectives seem more pressing (for example, Allende’s Chile or
Arbenz’s Guatemala). Yet there are also cases where other objectives are subverted to push for political
liberalization (for example, Somoza’s Nicaragua or the Shah’s Iran).
41. On trade, see Morrow, Siverson, and Tabares 1998; Polachek 1996. On cooperation, see Russett
1993; Leeds 1999. On alliances, see Simon and Gartzke 1996; Gaubatz 1996.
42. See Keefer and Knack 1995; Barro 1997.
43. Grieco 1996, 286 – 89.
44. See Keohane 1984; Martin 1992.
45. This skirts the classic issue of collective action problems in coordinating economic sanctions.
Because institutions are likely to help identify cheaters (for example, through the construction of focal
points), however, concerns over free riding will be lessened within an institution. Martin 1992. I revisit
this assumption later in this article.
46. Drake 1998, 79 – 81.
47. Pastor 1989.
Guatemala’s legislature and courts, and announced that he would rule by decree.48
Led by the democratic members of the organization, the OAS lodged high-proÞle
protests and moved to levy sanctions against the regime.49 After Þve days, Serrano
was forced from ofÞce by the military, which reinstalled a civilian president. Many
observers credit the OAS response as an important part of Serrano’s calculations to
step down.50
Another case of an IO effectively de-legitimizing and pressuring an autocratic
state is that of the European Union and Greece. After the overthrow of Greek
democracy in 1967, the European Union suspended Greece’s Association Agree-
ment. Many scholars of the Greek transition to democracy (which occurred in 1973)
claim that this suspension was important in eventually undermining the colonels’
regime: “Greece suffered immediate Þnancial consequences from the freezing of the
Greece-EC Financial Protocol and the agricultural harmonisation talks . . .The
political implications were even more ominous . . . exclusion from the rapidly
integrating Community was a singularly dangerous prospect.”51 Some scholars have
gone so far as to credit this action as a key pressure against the colonel’s regime
throughout the period of dictatorship.52
Regional IOs are an enabling mechanism increasing the probability that democ-
racies can and will push non-democracies to liberalize during times of domestic
upheaval. These institutions help to de-legitimize autocratic regimes through vari-
ous means, including diplomatic pressure, economic sanctions, or even expulsion
from the organization. Although this may not be the most important determinant in
convincing autocrats to loosen their grip on power, at critical junctions within the
liberalization process, it provides a powerful impetus for political liberalization.
Societal Elites and Acquiescence to Liberalization

During decisions to begin political liberalization— or during the process itself—
certain powerful elite groups may attempt to stop this course of action. Membership
in regional organizations can decrease the likelihood of this veto in two ways. First,
regional IOs can create credible guarantees to key groups that can assuage their fears
of democracy. Second, through a socialization process, regional IOs can help to
persuade elites to become less inimical to the process of liberalization by altering
their belief systems. In this section I discuss these two processes as they relate to
business elites and the military, respectively.53
Authoritarians (whether in single-party systems or military dictatorships) depend
upon the support of other groups in society for their power. One theory explaining
48. Halperin and Lomasney 1998, 137.

49. Cameron 1994, 169.
50. See Farer 1996; Cameron 1998.
51. Verney and Couloumbis 1991, 109.
52. Coufoudakis 1977, 130 –31.
53. These processes are not, however, necessarily conÞned to these speciÞc groups.
the rise of autocracies is that these regimes best protect the interests of these
important groups. For example, business elites may fear that democracy will bring
populists or leftists to power, endangering their property rights or Þnancial interests.
The military may fear democratic transitions because of potential threats— espe-
cially subordination to civilian supremacy—to its institutional interests. If these
groups perceive that their interests are threatened by political liberalization, they are
likely to stand in the way of these efforts.54 Membership in IOs can help to calm the
fears of the groups either by serving as an external guarantor of rights and
preferences, or by altering preferences through a socialization process.
Business Elites. For many business elites in authoritarian systems, democracy

conjures up images of populism and radicalism. Research on bureaucratic-authori-
tarianism in Latin America, for example, argues that business elites supported coups
against democracies in the 1960s and 1970s because they felt the military would
protect their interests from “the masses.”55 Concerns over policies such as nation-
alization and (land and/or income) redistribution caused middle-class business
interests and internationalist economic coalitions to support authoritarian takeovers,
often by the military. Much of the major democratic transitions literature of the
1980s assumes that business interests will naturally ally themselves with authori-
tarian regimes, which are better suited to protect their interests.56 In a recent
adaptation of this proposition, Leigh Payne argues that business leaders have no
strong preferences for any particular type of government, but will support arrange-
ments that defend their economic interests.57
Thus when confronting a situation where liberalization is an option for an
authoritarian regime, economic elites will calculate their interests in deciding
whether to support liberalization or remain backers of the hard-line authoritarian
guard:
In some cases, the worst that an elite can expect under a strategy of toleration
is an unpleasant loss of status and political power that leaves its economic base
and religiocultural values secure. In other cases, the call for toleration of
political opposition fuels deep-seated fears within the ruling elite about its
economic viability, the continued existence of hallowed institutions, or even
personal survival. . . . A political elite will have some estimation of its pro-
spective capacity to protect its basic interests both by building institutional
safeguards into the emerging democratic process and by actively competing in
it . . .58
54. Kaufman 1986, 86. On the general issue of minority interests gaining commitments during
regime/institution formation, see Fearon and Laitin 1996.
55. See O’Donnell 1979; Whitehead 1989, 85.
56. See O’Donnell and Schmitter 1986, 27; Payne 1994, 2.
57. See Payne 1994; Malloy 1987, 252–53.
58. Marks 1992, 51.
If elites can Þnd some way of guaranteeing their economic and/or political well-
being, they are more likely to acquiesce to liberalization. Of course, elites face the
difÞculty of ensuring that reformers respect these “institutional safeguards.” If they
attach a particularly low probability to the survival of these safeguards, they may
rationally refuse to liberalize at the outset. One observer of Latin American politics
offers an extreme example: “If open elections seriously threaten complete loss of
private property, capitalists will all become authoritarians.”59
International agreements can serve as a credible external guarantee of safeguards
for elites, especially economic elites, since they raise the costs of domestic policy
change. The external guarantee provided by membership in regional (in this case
economic) organizations lessens the perception on the part of business elites that
democracy will be dangerous to their interests. These guarantees can prove essen-
tial: “A vital element in the process of democratic consolidation is therefore to
induce [business interests and propertied classes] to conÞne their lobbying within
legitimate bounds and to relinquish their ties with the undemocratic right. External
reassurance (and if possible guarantees) may provide a critical inducement at the
beginning of a consolidation process, although the need for this should diminish as
democratization advances.”60
The literature on regional economic agreements has long argued that these
institutions function as a domestic commitment device. Membership in a regional
organization helps to lock in economic policies and rights enacted by domestic
elites.61 For example, free trade agreements codify commitments to free trade and
set up a system of veriÞcation to monitor the implementation of such reform. These
mechanisms help to lock in commitments among states that free trade will continue
even in the face of domestic opposition and new administrations.62
A critical example of these guarantees arises in the area of property rights, where
regional economic agreements also help to provide commitments that governments
will not engage in opportunistic behavior. Since a common goal of many regional
economic agreements is to lure foreign investment into the region, these institutions
provide explicit guarantees about property and investment. To lure multinational
Þrms to invest in a region, these arrangements must provide guarantees against
opportunistic behavior on the part of host governments— guarantees that also would
apply to domestic Þrms. As such, these organizations can provide important
reassurances concerning property rights and investments.
Laurence Whitehead has argued that this process was essential to democratization
efforts in Southern Europe. Because the EC “offered critical external guarantees to
the business and propertied classes of southern Europe . . . democracy would lose
59. Sheahan 1986, 163.

60. Whitehead 1989, 84.
61. Goldstein 1998, 143– 44; MansÞeld, Milner, and Rosendorff 2002.
62. Fernández and Portes 1998; MansÞeld 1998; Milner 1998, 29.
much of its sting for the rich.”63 The EC insisted on adequate compensation for any
property taken by the state, and insured the free movement of capital and goods.
This externally monitored and enforced guarantee provided credible protection for
economic elite interests, which induced their acquiescence in the democratization
process in Southern Europe. This was especially true in Spain and Portugal, where
economic elites had traditionally been hostile to democracy.64 For Spanish elites
who were a potential roadblock to democracy, the stipulations of the EC “provided
guarantees and reassurances to those who faced the post-authoritarian future with
apprehension.”65
These commitments to trade and property may reassure economic elites that, even
in the worse case scenario of a populist-oriented, democratic government, their
interests will be protected. Of course, any government (democratic or authoritarian)
can withdraw from these international agreements, but they would pay a high cost
for doing so. Thus membership in IOs reduces the probability for opportunistic
behavior, increasing the odds that business elites will acquiesce to liberalization and
democracy.
The Military and Socialization. The other group that can be inßuenced by
membership in a regional organization is the military. Similar to business elites, the
military is a powerful group concerned with protecting its interests and institu-
tions.66 Often, the military stands in the way of political liberalization out of fears
for its autonomy (for example, subjugation to civilian rule) and for protection of its
institutions (for example, against reprisals for its role in past authoritarian govern-
ments).67 Regional IOs, especially security-oriented organizations, can help per-
suade the military to acquiesce to democratization by not only providing externally
supported guarantees, but by helping to reorient military ofÞcers away from their
interest in domestic politics.
Regional security organizations can assure the military of continued support
either through the domestic regime or alliance partners. To maintain a credible
military force as a part of an alliance, a state must provide adequate resources to its
military and is often required by its allies to do so. These requirements help provide
military ofÞcers with institutional protection. The military may also receive direct
Þnancial or technical beneÞts from its alliance partners, as was the case during
Hungary’s democratic transition. Through the Partnership for Peace (PFP) program,
Hungary received technical military assistance, while through NATO it has received
assistance in the modernization process. In addition, NATO required that military
63. Whitehead 1996d, 271. Although Southern European states were not full members of the EC,
Greece, Spain, and Portugal each had association agreements prior to their transitions.
64. In the Spanish case, see Whitehead 1986. In the Portuguese case, see Manuel 1996, 75.
65. Powell 1996, 297.
66. Dassel and Reinhardt 1999.
67. Przeworski 1991, 31; Whitehead 1989, 81– 84.
spending in Hungary be stabilized and even increased.68 This reversed a downward

spiral in the military budget, which had created dissatisfaction among military
ofÞcers. Thus PFP/NATO membership was important for the completion of the
transition in Hungary.69
An even stronger impact of regional military organizations on the military and its
attitudes toward democracy can come in the form of socialization. Regional
alliances and military organizations, especially those that conduct joint training
operations or maintain permanent institutions (such as NATO), can help to socialize
military leaders in member states as to the role of the military in domestic society.70
While this may or may not be the goal of membership, this socialization process
may occur through interactions in the institution.
Socialization amounts to persuading military leaders that the role of the military
is not to act as an internal police force, but rather to protect the state from outside
forces.71 As Pridham argues, “A more stable way for these [Southern European]
governments to internationalize the military role was through integration in a
European organization such as NATO.”72 Moreover, the idea of civilian supremacy
over military missions and institutions is often an issue of contention in transitional
states. By interacting with military leaders of other states who subscribe to these
types of doctrines, military elites in autocratic or recently autocratic states are more
likely to internalize these doctrines, making them more likely to accept full
democracy. This issue may be important not only in the military’s acceptance of an
initial move towards liberalization, but in completing the transition to democracy.
Perhaps nowhere has this dynamic been more important than the transition to
democracy in Spain. Long excluded from NATO membership, after Franco’s death
and the beginning of the transition to democracy, NATO accession became a foreign
policy goal of the new Spanish regime. Although Spain’s transition had taken place
more than six years earlier, a cadre of military ofÞcers attempted a coup against the
young democracy in 1981. This incident highlighted the need to control the Spanish
military and keep it away from the domestic political process.
NATO became the vehicle to achieve this goal, as “the belief surfaced in
government circles that entry to NATO would help secure the new democracy as it
would ‘modernize’ the Army through growing international contacts and direct its
attention away from domestic politics.”73 Such reasoning proved correct as Spain’s
68. Robert Wright, A New Security Blanket, Financial Times, 7 December 1998, 3.
69. Vetschera 1997, 19; Braun 1999, 19.
70. On the importance of the exchange of information and ideas within institutions in inßuencing
behavior, see McNeely 1995, 35–36.
71. A similar process can occur in the context of democratic consolidation. For example, the threat of
punishment from the IO in the case of a military coup can convince the military to stay in the barracks.
Pevehouse 2002. Here, the argument is that during the transition process itself, the socialization process
convinces the military that coups are “off-limits.”
72. Pridham 1994, 196.
73. Pridham 1991b, 228. See also Treverton 1986, 32–33. Interestingly, Spain’s main opposition
Socialist Party (PSOE), which had initially opposed NATO membership, acquiesced to accession after
the 1981 coup attempt largely because of its potential impact on the Army ofÞcers. Tovias 1984, 167.
army did undergo signiÞcant modernization and reorientation after NATO mem-
bership. Through joint maneuvers, modernization, and improvements in military
technology, the Spanish military became reoriented away from domestic politics.74
In this section I have discussed some causal mechanisms by which regional
organizations may be associated with democratic transitions. However, not all IOs
are likely to be associated with these transitions. IOs may lack the resources and/or
political will to serve as an external promoter or supporter of democracy. For
example, we should not expect IOs such as the Warsaw Pact to condone democratic
transitions. With this in mind, I now consider the “supply-side” part of my
argument.
The Supply-Side of Democratization
Only certain regional IOs are likely to be associated with democratic transitions and
consolidation. While it would be simple to test whether membership in a regional
organization was associated with the transition to democracy, since almost every
state in the world is a member of multiple regional organizations, this is not a useful
strategy.75 Intuition tells us that only some organizations are likely to engender in
any or all of the aforementioned causal mechanisms.
I argue that organizations with a higher democratic density are more likely to be
associated with democratic transitions. Democratic “density” refers to the percent-
age of permanent members in the organization that are already considered demo-
cratic.76 In effect, the more homogeneously democratic the organization, the more
likely it will “supply” these causal mechanisms to assist in the democratization
effort.
First, homogeneously democratic regional organizations are more likely to place
conditions on membership that will be associated with the application of external
pressure. States will be more likely to pressure other states if conditions require
democratic governance. If organization membership is mixed in terms of regime
type, it will be unlikely to have enacted these conditions, since approving conditions
on membership would be more difÞcult in these “mixed” organizations.
Of course, regional IOs may choose not to pressure a state that is undergoing the
liberalization process or that has suffered a breakdown of democracy even if
74. Tovias 1984, 167; Hurrell 1996, 161; Pridham 1994, 199 –200.
75. Also, coding IOs for their membership requirements (for example, democratic conditionality) is
not fully satisfying because IOs may not enforce the conditions they create. For example, Portugal (under
Salazar) was a founding member of NATO although the organization’s Preamble describes it as an
alliance of democratic nations. Thus meaningful analyses would require a proxy to measure both the
presence of conditions and the likelihood of enforcement.
76. Note that this measure does not refer to the level of democracy within the organizational structure
or procedures. Although discussions of the “democratic deÞcit” of IOs may be important for this
argument, I leave this topic for later analysis. From this point forward, any reference to democratic IOs
should be understood as the aggregate level of democracy among the members rather than a trait of the
organizational structure.
democratic conditionality exists. Thus enforcement of these conditions is also

critical. Empirically, IOs do inßict punishment on those members that break
democratic conditions of agreements. For example, the EC suspended the Greek
association agreement in 1967 after the military coup.77 In addition, the Council of
Europe suspended Turkey’s involvement in that organization after the September
1980 coup.78 Although the supply of enforcement may be problematic, my conten-
tion is that the more homogeneously democratic the regional IO, the more likely it
will pressure offending states.
I previously discussed one incentive of democracies to overcome this potential
collective action problem (expectations about trade, peace, or cooperation with other
democracies). Another factor that could increase the likelihood of enforcement is the
transparency of democracies.79 Democratic states are less likely to openly shirk on
enforcing conditions of an IO. If a state is deciding whether to help enforce a
conditionality clause (for example, suspending free trade or imposing trade sanc-
tions), it is easier to witness the behavior of fellow democratic members than
autocratic members, ceteris paribus. If one member state attempts to circumvent the
external pressure applied by the organization by working with the offending state,
other members may be less likely to push for punishment. Because cheating is easier
to detect in more homogeneously democratic organizations, members will have
fewer fears of cheating, increasing the odds of enforcement.
The other causal mechanisms (reassurance and socialization) are also more likely
to function in highly democratic IOs. SpeciÞcally, elites contemplating liberaliza-
tion will be especially reassured by regional IOs composed of democracies because
democracies are more likely to fulÞll their international commitments. Because
democratic leaders face potential audience costs for not following through on their
international obligations or for reneging on these commitments, they are more likely
to fulÞll such commitments.80 For elite groups within states facing liberalization
pressures, assurances offered by IOs composed of mainly democracies will be
preferred over assurances by “mixed” organizations, since there will be an ex ante
perception that democracies are more likely to uphold their commitments.
Finally, if socialization is the causal process linking these institutions with
democratization, these processes will be more common the more democracies exist
in the organization. If there are more interactions with democratic actors, then the
transmission of values and norms about the democratic process is more likely.
Of course, regional organizations may choose not to enforce these conditions. If
the members of the organization deem the costs too high or other “strategic
considerations” mitigate the likelihood of enforcement, IOs may do little to promote
77. Whitehead 1993, 154.

78. Karaosmanoglu 1991, 162.
79. See Smith 1998; Schultz 1998 and 1999.
80. See Fearon 1994; Gaubatz 1996. For empirical support of this general proposition, see Leeds
1999. For a speciÞc empirical application, work on regional trade agreements has found that democracies
are more likely to join these arrangements with one another for reasons of commitment. MansÞeld,
Milner, and Rosendorff 2002.
democracy. The “supply-side” inßuence on the IO-democracy link is especially

important for policymaking. If the external guarantees and threats of punishment are
not credible, regional IOs may no longer be efÞcacious in their policies toward
democratization. Nonetheless, my argument is that “democratic” regional organi-
zations will be more likely to enact and enforce policies favorable to the develop-
ment of democracy.
Testing the Argument
If the preceding argument is correct, an association should exist between democratic

regional organizations and democratization. To test this hypothesis, I use a data set
measuring both democracy and involvement in regional organizations. The data set
includes every member of the Correlates of War system from 1950 to 1992. I code
transitions to democracy based on both Polity98 data and Gasiorowski’s regime
transitions data set.81 I use both data sets as a robustness check, since both
emphasize different substantive aspects of democracy. Polity98 emphasizes insti-
tutional structures, while Gasiorowski’s focus includes institutions, free elections,
and civil liberties. In addition, the data sets lack a complete spatial overlap—
Gasiorowski codes only the developing world, excluding Europe and North Amer-
ica.82 To test my argument, I estimate the following model:
DemTransit ⫽ ␣0 ⫹ ␤1IOScoreit-1 ⫹ ␤2⌬IOScoreit-1 ⫹ ␤3pcGDPit
⫹ ␤4⌬pcGDPit ⫹ ␤5Contagionit ⫹ ␤6PastDemit
⫹ ␤7RegConßictit ⫹ ␤8IntViolenceit-1
⫹ ␤9Milregit ⫹ ␤10Indepit ⫹ ␮it
The dependent variable, DemTransit, measures the probability of a democratic
transition. Gasiorowski’s data include exact dates of democratic transitions. For the
Polity98 data, I code transitions as movements above a cut point (⫹6) on the
composite democracy score included in the Polity98 data. The observed value of the
dependent variable is dichotomous. It equals 1 if state i completes a transition to
democracy in year t and zero otherwise.
The Þrst independent variable, IOScoreit-1, represents the score of the most
democratic IO of which state i is a member in the year t-1. This variable is
computed by Þrst taking the average democracy score (from Polity98) of all
members of a regional organization, except state i. Each state will then have a
number of IO-democracy averages equal to the number of organizations of which it
is a member.83 IOScoreit-1 represents the highest average of those organizations of
81. See Marshall 1999; Gasiorowski 1996.

82. This will serve as an excellent robustness check against the charge that the IO-democracy
association is a Europe-only phenomenon. If the Þndings are consistent across both data sets, one may
infer that similar processes are occurring in both the developed and developing world.
83. Data on IO membership are taken from Banks (various years) and Banks and Mueller 1998.
which state i is a member. The expectation is that membership in democratic IOs

will make the probability of a democratic transition more likely. Thus the coefÞcient
estimate for this variable should be positive and statistically signiÞcant.
The second independent variable, ⌬IOScoreit-1, is computed to isolate the effects
of joining an IO. This variable is included because the process of joining a highly
democratic IO may be an important impetus to make the transition to democracy.
This variable is the one-year difference of IOScoreit-1. If a state joins a new IO that
is more democratic than any previous IO, the value of ⌬IOScoreit-1 will be positive.
Including this change variable as well as the level variable allows a comparison of
the effects of joining a democratic organization with the overall democratic density
of the IO.84 The expectation is that the sign for this coefÞcient should be positive.85
The next two independent variables, pcGDPit, and ⌬pcGDPit, represent economic
factors that may be relevant to democratization. Beginning with Seymour Martin
Lipset’s work, the theory that economic well-being creates demands on the part of
citizens for responsive government has been described by some as “almost beyond
challenge.”86 Some have challenged this conclusion, however, including Adam
Przeworski and Fernando Limongi, who show that economic development is not
related to democratic transitions, but rather to sustaining democracy.87 For these
reasons, I include pcGDPit, which should be positively associated with transitions to
democracy. The expectation of ⌬pcGDPit is less clear, since although growth rates
and modernization are hypothesized to bring pressures for democratization, low
economic growth is especially likely to hit authoritarians very hard.88
The variable Contagionit controls for possible diffusion effects from other
established democracies. This phenomenon has various names— contagion, diffu-
sion, demonstration effects, and snowballing—and holds that the presence of
democracies is likely to encourage democracy in nearby countries.89 Contagionit is
a continuous variable based on the number of democracies in state i’s geographic
84. Change in the level of democracy could represent a change in the level of democracy of the same
IO. That is, ⌬IOScoreit-1 may have a positive value because of democratization in other member states
rather than a new membership for the state in question. While this is a concern, because the Contagionit
variable controls for surrounding states’ democratization, this should not be a problem.
85. The operationalization of these independent variables does not allow one to discriminate which
causal process better accounts for any correlation found in the data. As no organizations data exists on
this question, the decision was made to test for broad relationships before more narrow statistical tests
were conducted. In addition, the task of investigating speciÞc causal mechanisms should include
case-oriented research on various democratic transitions. Pevehouse 2000 outlines several cases.
86. See Burkhart and Lewis-Beck 1994, 903; Lipset 1959.
87. Przeworski and Limongi 1997. Economic factors also may be related to the breakdown of
autocracy. Londregan and Poole 1990 and 1996 Þnd this to be the case. See also Haggard and Kaufman
1995.
88. Both pcGDPit and ⌬pcGDPit are taken from Summers et al. 1995, and supplemented with data
from the World Bank 1998; Mitchell 1995 and 1998.
89. Huntington 1991. An IO’s effect on democracy could take this form as well, but by controlling for
this phenomenon independently, I account for any contagion effects independent of IO membership.
region in year t. 90 Its yearly value is the same for every country in a given region.91
This variable should be positively associated with democratic transitions, since
higher numbers of extant democracies should lead to a hospitable environment for
the movement to democracy, or even pressures from neighboring states to democ-
ratize.92
Many who study transitions to democracy argue that past experience with
democracy makes a future transition more likely. Since the country has previously
developed democratic institutions, these traditions will often survive a period of
autocracy and can re-emerge as a foundation for democracy. Przeworski and
colleagues, however, point out that a history of democracy in an authoritarian
regime also means a history of democratic breakdown; it is not clear whether this
factor works for or against the likelihood of a democratic transition.93 To control for
these dynamics, I include PastDemit in the model. The variable equals 1 if state i has
previously been a democracy, otherwise, the variable is coded zero.94
The next two independent variables tap the level of external and internal conßict
experienced by a nation-state. RegDispit measures the number of disputes in state i’s
region in year t. 95 As previously noted, some scholars argue that regional instability
and international conßict will not augur well for the development of democracy—
states require a peaceful international environment to democratize.96 Given the
extant theory, I expect this variable to have a negative effect on the probability of
democracy. That is, the more disputes within a region (which may or may not
involve state i), the less likely a democratic transition.
To measure the effect of internal violence, I include the variable IntViolenceit-1.
This variable, taken from Banks, is coded 1 if state i suffers from anti-government
protests, strikes, riots, guerilla insurgencies, or assassination attempts on major
government ofÞcials in year t-1. 97 Although violence is often required to unseat an
90. I follow the Correlates of War coding of regions, which includes: North America, South America,
Africa, Europe, Asia-PaciÞc, Asia, and the Middle East. See Small and Singer 1994. Although Small and
Singer consider the Oceania area as an independent region (Australia, New Zealand, Fiji, etc.), I combine
these few states with the Asia region.
91. Depending on which data set is used to measure the dependent variable, however, the value of
Contagionit may vary since each data set differs on which states are considered democracies.
92. This variable is only moderately correlated with IOScore (r ⫽ .29 in Gasiorowski; r ⫽ .11 in
Polity98). This lack of correlation arises because smaller subsets of states in a region may have common
IO memberships, while other states in the region may be less democratic, driving down that region’s
overall level of democracy.
93. Przeworski et al. 1996, 43– 44.
94. The state must have been a democracy between 1950 and year t. States that were democracies that
suffered breakdowns before 1950 are coded zero.
95. Regions are deÞned in fn. 90. MIDS data are taken from Jones, Bremer, and Singer 1996.
96. Thompson 1996. Of course, there could be another, more complex causal story: If peace means
prosperity, then the economic development and democracy link again becomes important. I do not
attempt to evaluate this theory, per se, and leave the question of causal mechanisms aside.
97. Banks 1994 provides a description of his coding criteria for each type of event.
autocrat, it will be rare for democracy to arise given the civil and economic upheaval
accompanying large-scale internal violence.98
Some scholars argue that different types of autocratic regimes are more or less
susceptible to democratization. For example, Barbara Geddes contends that military
regimes “carry with them the seeds of their own disintegration.”99 She argues that
returning to the barracks is always a fairly attractive option for some military
leaders, so when faced with a crisis, military regimes are more likely to collapse
than other forms of autocracy. To control for this dynamic, I include MilRegit, which
equals 1 if state i is controlled by the military in year t. 100
The Þnal independent variable, Indepit, measures the length of time state i has
been an independent nation-state. As an extension of the idea that past experience
with democracy may be correlated with democratization, longer periods of political
independence may be positively related to the probability of a transition to
democracy. Longer periods of political independence will allow time to foster civil
society and independent political parties.101 Finally, ␮it is a stochastic error term.
This model includes many of the factors hypothesized by both the case study and
quantitative literature to inßuence regime transition. All control variables (or a close
variant) appear in previous empirical work on regime transitions—the vast majority
of which tap the domestic context of regime transition.102 The inclusion of these
variables is important because membership in democratic regional organizations
will likely not have the strongest inßuence on the transition process. Even those
theorists who point to the importance of external factors in transitions rarely argue
these are the determining factors in these events. Thus my model attempts to build
in a signiÞcant number of domestic factors to guard against omitted variable bias.
This creates added conÞdence that any relationship between regional IOs and
democratization is not an artifact of model speciÞcation.
The sample for these model estimations of necessity excludes some countries,
since states that are already democratic cannot experience a transition to democracy.
The sample includes only states that can undergo a democratic transition (autocra-
cies and anocracies),103 and the N of each model differs depending on which data
98. Linz and Stepan 1996, 107–108. Because violence may result from a transition, the lag is used
to avoid endogeneity.
99. Geddes 1999, 122.
100. These data are taken from Banks 1994, but because Banks’ data are coded only through 1989,
I have updated the data through 1992, relying on Banks and Mueller 1998 as well as the CIA Factbook
1999.
101. Small and Singer 1994 trace the origin of the international state system to 1815 and thus, for all
nations who were independent before this date, their date of independence is set to 1815.
102. See Burkhart and Lewis-Beck 1994; Gasiorowski 1995; Londregan and Poole 1996; Przeworski
and Limongi 1997.
103. Technically, those countries that are autocracies or anocracies at t-1 are included in the sample.
That is, if a state is a democracy in year t because of a transition, it is included, since in the prior year
it was either an autocracy or anocracy. Transition years—those years where the outcome of the transition
is uncertain—are also included.
TABLE 1. Descriptive statistics and associations
Polity98 Gasiorowski
Number of states in analysis 106 86

Transitions
North America 3 3
South America 18 22
Europe 9 —
Asiaa 9 15
Africa 6 13
Middle East 5 7
TOTAL 50 60
Average IOScore
All states in sample 8.21 8.10
States democratizing in t ⫹ 1 13.05 11.20
Mean values in analysis
IOScore 8.21 8.10
⌬IOScore 0.22 0.21
pcGDP 2,440.75 2,182.58
Contagion 46.59 33.00
PastDem (0/1) 4.8 1.8
RegConßict 0.73 0.75
IntViolence (0/1) 0.53 0.57
Milreg (0/1) 0.25 0.31
Independence 52.3 48.9
a
Includes Oceania states.
set is used (since their deÞnition of democracy differs). Table 1 provides some
descriptive statistics of the data.
The Þrst section of Table 1 shows the distribution of democratic transitions across
each region, and the second section of the table shows the simple bivariate
relationship between IOScore and democratization. Clearly, states that will undergo
a transition to democracy in the subsequent year are members of more homoge-
neously democratic regional organizations. The full model of democratic transitions,
moreover, shows a similar relationship.
Statistical Results
Because the observed value of the dependent variable is dichotomous, I use logistic
regression to estimate the model.104 Table 2 presents the estimates of the model.
Note that the Þrst column presents estimates using the Polity98 data, and the second
104. In each speciÞcation of the model, I checked for the inclusion of a correction for temporal
dependence. Beck, Katz, and Tucker 1998. In each case, a cubic spline function and one to three knots
are never statistically signiÞcant.
TABLE 2. Estimates of the determinants of the transition to democracy,

1950 –1992
Polity98 Gasiorowski Polity98 Gasiorowski
IOScoreit⫺1 0.089*** 0.065** 0.099***a 0.112***a

(3.74) (2.25) (2.75) (2.67)
⌬IOScoreit⫺1 0.087 0.073 ⫺0.325a ⫺0.228a
(1.41) (1.38) (⫺1.48) (⫺1.18)
pcGDPit ⫺2.79 ⫻ 10⫺5 ⫺3.34 ⫻ 10⫺5 ⫺2.49 ⫻ 10⫺5 ⫺3.76 ⫻ 10⫺5
(⫺0.45) (⫺0.65) (⫺0.42) (⫺0.70)
⌬pcGDPit ⫺0.0005* 3.44 ⫻ 10⫺5 ⫺0.0004 6.88 ⫻ 10⫺5
(⫺1.81) (0.45) (⫺1.32) (1.08)
Contagionit 0.057** 0.085 0.069*** 0.098**
(2.21) (1.60) (2.63) (1.98)
PastDemit 1.634*** 0.551** 1.757*** 0.555**
(5.18) (2.00) (5.49) (1.98)
RegConßictit ⫺0.456*** ⫺0.094 ⫺0.459*** ⫺0.100
(⫺2.74) (⫺1.02) (⫺2.80) (⫺1.10)
IntViolenceit⫺1 2.359*** 1.073*** 2.285*** 1.049***
(4.10) (2.75) (4.10) (2.70)
Milregit ⫺1.856*** ⫺1.709*** ⫺1.872*** ⫺1.655***
(⫺2.75) (⫺3.26) (⫺2.76) (⫺3.17)
Indepit 0.005 0.003 0.006* 0.004
(1.37) (0.77) (1.71) (1.11)
Constant ⫺7.022*** ⫺5.135*** ⫺7.487*** ⫺6.004***
(⫺12.64) (⫺12.21) (⫺12.27) (⫺9.74)
N ⫽ 2776 2299 2777 2302
Log likelihood ⫺195.44 ⫺244.50 ⫺197.25 ⫺243.90
␹2 104.94*** 63.31*** 103.99*** 69.49***
Note: Figures in parentheses are asymptotic z-statistics computed using Huber standard errors.
a
In columns 3 and 4, IOScoreit⫺1 and ⌬IOScoreit⫺1 include non-regional organizations and IFIs.
*** p ⱕ .01
** p ⱕ .05
* p ⱕ .10; two-tailed tests.
column uses the Gasiorowski data set. For both data sets, membership in democratic
IOs is positively related to the likelihood of a transition to democracy. In addition,
both coefÞcient estimates are highly statistically signiÞcant. Thus membership in
democratic IOs, while controlling for domestic economic trends, internal and
external violence, and past experience with democracy, is signiÞcantly correlated
with the transition to democracy.
To illustrate how important the IO variable is to the prospect of experiencing a
transition to democracy, I show the predicted probabilities of a transition in Tables
3 and 4, based on the Þrst two columns of Table 2. The Þrst lines in each table
present the “baseline” likelihood of a democratic transition for a state in the data set
in which the authoritarian system was not a military regime, the state has no past
TABLE 3. Predicted probabilities of transitions based on the estimate, in column

1 of Table 2, using the Polity98 data
Probability of transition
All variables at mean (baseline)a 0.0021

Increase in IOScore by one standard deviation from mean 0.0034
No IO membership 0.0010
Past experience as democracy 0.0107
Past experience ⫹ military regime 0.0017
Military regime ⫹ internal violence 0.0035
Military regime 0.0003
a
Milregit, IntViolit, and PastDemit, all equal 0 in the baseline model. All other variables held at
mean level.
experience with democracy, and did not experience internal violence.105 Note the
change in probability of a transition in the second line of each table, where IOScore
is increased by one standard deviation from its mean value. For the Polity98 data,
the likelihood of a transition increases by more than sixty percent, while for the
Gasiorowski data the result is a more modest forty percent.
The third line of each table presents the predicted probability of a transition in the
case where the state is not a member of a regional organization included in the IO
sample.106 These results are quite striking. For the Polity98 data, membership in an
IO with the mean level of IOScore increases the odds of a transition over a state with
no IO memberships by more than 100 percent. For the Gasiorowski data, the
comparable change is nearly seventy percent. IOs appear to have a very strong
impact on the likelihood of a democratic transition, and this impact grows as the
organization’s membership becomes more democratic, even while accounting for
important domestic dynamics.
Most control variables are of the expected sign, but not all of the estimates
achieve statistical signiÞcance. One variable that consistently does not take on the
expected sign, however, is per capita gross domestic product (GDP). In all cases this
term is negative, but in no case is the estimate statistically signiÞcant. Thus it
appears that higher economic development is negatively related to democratic
transitions. The lack of statistical signiÞcance should instill caution when interpret-
ing these results, leaving one to conclude that the overall level of development in a
country has little direct impact on the likelihood of transitions. This Þnding is
consistent with the Przeworski and Limongi argument that higher levels of devel-
105. These are the modal values for the dummy variables in the data set. All other continuous
variables in the model are set to their mean value (shown in section 3 of Table 1).
106. Examples of states that are a party to no IOs over some periods of the data are Israel, North
Korea, Cuba, North Vietnam, Taiwan, and China.
TABLE 4. Predicted probabilities of transitions based on the estimates in column

2 of Table 2, using the Gasiorowski data
Probability of transition
All variables at mean (baseline)a 0.0117

Increase in IOScore by one standard deviation from mean 0.0160
No IO membership 0.0069
Past experience as democracy 0.0201
Past experience ⫹ military regime 0.0037
Military regime ⫹ internal violence 0.0062
Military regime 0.0021
a
Milregit, IntViolit, and PastDemit, all equal 0 in the baseline model. All other variables held at
mean level.
opment are associated with the continuation of, rather than the transition to,
democracy.107
Similarly, change in per capita GDP also shows a weak inßuence on the
probability of a democratic transition. For the Polity98 data, this variable is negative
and statistically signiÞcant. Because authoritarian governments often come to power
in the midst of economic crises, economic success can function to legitimate their
continued existence, lessening the probability of a democratic transition.108 Like-
wise, poor economic performance (falling growth rates) increases the likelihood of
a democratic transition, a Þnding consistent with past statistical research.109 For the
Gasiorowski data, however, this parameter estimate is not statistically signiÞcant.
One possible explanation of this divergence of results across data sets lies in their
differing geographic coverage. The lagging growth rates of Eastern and Southern
European states immediately prior to their transitions bolster these results in the
Polity98 data. Since these cases are absent in the Gasiorowski data, the statistical
association is much weaker.
For both data sets, past experience with democracy and a previous military regime
have substantial impacts on the probability of a transition. This Þnding concerning
Milregit is in accord with past empirical literature, which Þnds that these regimes are
more difÞcult to overthrow than single-party or personalistic regimes. However, the
Þnding contradicts Geddes’ argument that military regimes may be more susceptible
to transitions to democracy.110 In addition, past experience with democracy (Past-
Demit) is positively related to the propensity for a transition. This Þnding is in line
with theoretical expectations concerning this variable, namely that a history of
107. Przeworski and Limongi 1997.

108. O’Donnell 1979.
109. Gasiorowski 1995.
110. See Gasiorowski 1995; Geddes 1999.
democracy bodes well for the probability of future transitions to democracy. In fact,
past experience with democracy has an enormous effect on the predicted probability
of a transition. Line 4 of Tables 3 and 4 shows that this variable increases the
probability of a transition dramatically for both data sets.
The existence of domestic violence (IntViolenceit-1) has a strong impact on the
likelihood of a democratic transition. This variable is consistently positive and
highly statistically signiÞcant, indicating that greater levels of domestic violence
increase the probability of a transition. This is not in concordance with previous
theoretical expectations—it was hypothesized that domestic violence would make a
transition less likely. In many of the cases included here, some level of domestic
hostilities likely preceded the transition to democracy. One conclusion that can be
drawn is that, given the large substantive impact of this variable, it is possible that
many transitions to democracy in the sample were preceded by civil violence.
One additional control variable warrants discussion. The estimate of the effect of
regional conßicts (RegConßictit) on the prospects for democracy is statistically
signiÞcant in the Polity98 data, but not in the Gasiorowski data.111 The estimates
consistently have the predicted sign across both data sets—increasing regional
hostilities lower the probability of democracy. Again, the difference in spatial
coverage may account for the fact that only in the Polity98 data is this variable
signiÞcant. Europe experiences the highest number of militarized disputes during
the vast majority of the period under analysis, largely because of incidents arising
out of the Cold War. Thus the lack of democratization for years in Eastern Europe
coincided with a high rate of disputes.112 Given that these cases are absent in the
Gasiorowski data, this relationship is much weaker in the models estimated using
that data.
Overall, these models provide Þrm evidence that involvement in democratic
regional organizations can encourage the process of democratization. This is strong
initial support for the theory, especially given that the model controls for other
factors that could be associated with transitions (especially those factors that are also
correlated with IO membership). The argument is not, however, that IOs play the
strongest role in the prospects for regime transition. Indeed, the results of these
analyses indicate (as Tables 3 and 4 highlight) that internal factors such as past
experience with democracy, previous type of regime, and changes in economic
growth rates play a far more substantial role. The contention here is that the regional
IO dimension cannot be excluded from analyses of transitions since I have
demonstrated that these second image reversed processes are theoretically impor-
tant.
111. Using interstate war data, rather than MIDS data, to measure regional conßict yields estimates
that are not statistically signiÞcant.
112. This causal association is tenuous, yet interstate disputes here are essentially serving as a proxy
for Cold War activity, which no doubt inßuenced democratization in Europe during this period.
Additional Tests
To assess the robustness of these results, I estimate several additional models. The
Þrst model adds region-speciÞc Þxed effects to this previous variant to test whether
there are systematic differences in the probabilities for transitions across each
region. By introducing a dummy variable for each region (deÞned above), one can
account for factors that may be important only in certain geographic regions that are
not included in the model, yet may be correlated with membership in regional
IOs.113 Controlling for these unspeciÞed factors may eliminate any association
between IOs and democratization. In these models, however, the key independent
variable of interest, IOScoreit-1, remains statistically signiÞcant for both data sets.
To assess an alternative hypothesis, I estimate one variant of the model that
substitutes NIOit-1 for ⌬IOScoreit-1. This variable measures the number of IOs of
which state i is a member in year t-1. If any regional organization can fulÞll the
functions outlined in my theory, regardless of the organization’s democratic density,
this should test for this possibility. In no case does the presence of this variable
mitigate the inßuence of IOScoreit-1, which provides strong support for the notion
that identity of the IO matters in terms of promoting transitions. Not all regional IOs
will be able (or willing) to fulÞll the functions outlined above.114
It is also important to ensure that these results are not undermined by simultaneity
bias. If countries Þrst undergo transitions to democracy, then join highly democratic
IOs, a statistical association may exist even though the causal process is reversed.
To guard against this possibility, I lag the value of IOScore. This is the best
insurance against a simultaneity problem, since during the year of transition the
IOScore from the previous year is used. Thus the causal process must begin with a
high IOScore, then democratization. Second, I estimate a model identical to the
original model, but I reverse the dependent and major independent variable
(IOScore).115 To show no reverse causation, I measure IOScore in year t, and the
democratic transition in year t-1. In this model, the estimate of the democratic
transition variable is not statistically signiÞcant, indicating that democratic transi-
tions have little inßuence on IOScore.
Another possible complicating factor in exploring the regional IO-democracy
relationship is that states desiring membership in these organizations may enact
113. On the use of region-speciÞc dummies, see Feng and Zak 1999. I use the region as the unit of
analysis for substantive and methodological reasons. The substantive reason is that for the democrati-
zation literature, the region is often the relevant unit of analysis. See O’Donnell and Schmitter 1986;
Huntington 1991. Methodologically, if one moves down a level of analysis (to unit Þxed effects), much
of the data will be lost because any country that does not experience a transition will be excluded from
the data set. While this is not a problem econometrically, it does introduce potential bias in the results
since one is summarily excluding “the dog that doesn’t bark.”
114. In addition to NIOit-1, the model was estimated substituting a measure for the average IOScore
of each state in year t-1. This measure allows one to take into account all IOs of which state i is a
member. Although this variable is always of the predicted sign, it is never statistically signiÞcant. This
is further corroboration that only one democratic IO is needed to supply the functions linking IOs with
democratic transitions.
115. In addition, ⌬IOScoreit-1 is excluded. Its inclusion makes no difference in the results.
“anticipatory reforms” to gain entrance into a regional IO. In a similar vein, one
could argue that many regional IOs require democratization before accession to the
organization. While these processes would support my general argument, the causal
process would be different: rather than encouraging transitions after joining an
organization, the prospective beneÞts held out by the organization would induce
political change. While this process could be at work in some cases, I have
attempted to control for this possibility. First, the preceding endogeneity check
suggests that this anticipatory process is not occurring since one would expect to
Þnd newly democratic states with higher IOScores. Second, the introduction of the
lagged IOScore and the change in IOScore help control for this causality issue since
it is the previous year’s IOScore that “predicts” transitions in the present year.116
Third, and most importantly, I include in each state’s list of IO memberships their
associate or observer memberships in regional IOs. Many states may enter regional
IOs as associate or observers. For example, before Spain, Portugal, or Greece were
allowed full membership into the European Community, they were granted partial
privileges in the organization through a trade agreement.
In all of these cases, the state is coded as if it was a member of the organization.
The causal process identiÞed above can work even if full membership has not been
granted. The same pressure from members and acquiescence effects can take place
even for associate members. The preceding example of Greece supports this
assumption: the EC could not suspend Greece’s membership (as it was not a full
member), but it could threaten to withhold beneÞts granted to its associate members.
The anticipatory reforms engendered by the regional IO may anticipate full mem-
bership, but not associate membership. If one re-estimates the model, excluding
associate and observer memberships in organizations, the results remain consis-
tent—full membership only in democratic regional IOs is still signiÞcantly related
to democratic transitions.117
Finally, as pointed out in the discussion of selection of IOs for this analysis, a set
of models were estimated in which IOScoreit-1 and ⌬IOScoreit-1 included both
non-regional organizations and IFIs. Columns 3 and 4 of Table 2 show these
estimates. For both the Polity98 and Gasiorowski data, the central variable of
interest, IOScoreit-1, remains statistically signiÞcant. The estimate of ⌬IOScoreit-1
does change sign, but remains statistically insigniÞcant, indicating that joining IOs
has little inßuence on the prospects for transition whether or not non-regional IOs
and IFIs are included in the sample. The estimates of the control variables remain
nearly identical to those in columns 1 and 2 of Table 2.
116. The model is also relatively robust with respect to increasing these lag lengths. The results are
robust for the Polity98 data using up to Þve-year lags (the longest period tested) and for the Gasiorowski
data using up to three-year lags.
117. From an econometrics standpoint, this anticipatory reform argument is not responsible for the
statistical association between IO membership and democracy. Not only are the key independent
variables lagged, but also, once a state Þnishes a transition to democracy, it leaves the sample. Thus if
a state gains IO membership years after a transition to democracy, it would not be captured in this
particular sample. The results excluding association membership are available from the author on request.
One should proceed with caution in reading too much into these auxiliary
estimations. As suggested earlier, non-regional (global) organizations and IFIs may
inßuence democratization in fundamentally different ways than regional organiza-
tions. To truly learn how these organizations inßuence domestic regime choices,
further theorizing and additional empirical work must be conducted. I present these
results here only to indicate that the exclusion of these organizations does not bias
the central Þnding of my analyses.
Conclusion
The interaction of international and domestic politics has garnered increased

attention in the Þeld of international relations and comparative politics over the past
Þfteen years. One important area of focus within this broad topic has been the
relationship between IOs and regime change. From both policy and theory perspec-
tives, however, the absence of cross-national research has weakened the debate
about the efÞcacy of international actors in promoting democracy. In this article I
have attempted to alleviate this dearth of empirical research.
Drawing from theories of rational institutionalism and sociological institutional-
ism, I argue that regional organizations can inßuence the domestic political process
even in realms of elite behavior. I outline a series of causal mechanisms—pressure
from the international institution, assuaging fears of national elites, or socializing a
group of national elites— by which regional IOs can inßuence the dynamics of
political liberalization. Short, illustrative examples demonstrate the plausibility of
such causal processes, while elucidating some ways in which regional institutions
interact with domestic politics to inßuence outcomes.
Results of this study provide insight into two related areas of international
relations theory: institutionalist theory and research on the democratic peace. By
assessing how differences in the membership of institutions create varied outcomes
with respect to democratization, I show how variations in institutions (on at least one
dimension) can inßuence state behavior. Institutional theorists have lamented a lack
of empirical investigation on whether differences among institutions along a variety
of dimensions may lead to diverse outcomes.118 By delineating a well-deÞned set of
casual mechanisms linking the international and domestic spheres, I have attempted
to push one part of institutionalist theory in the direction of the second-image
reversed framework.
Recent research on the democratic peace—speciÞcally the “Kantian Tripod”
concept—assumes that IOs play a key role in the maintenance of peace between
democracies.119 However, this research ignores many links between IO membership
118. Martin and Simmons 1998.

119. Russett, Oneal and Davis 1998.
and democratic development.120 Moreover, in all of this work, organizations are

equally weighted in their importance. My analysis contributes to this literature by
establishing the existence of feedback in the Kantian Tripod and by delineating
along what dimension (level of democracy within the membership) variation in IOs
matters for a speciÞc outcome (democratization).
Although not speciÞcally addressed here, these causal processes linking the
international and domestic spheres (pressure, credible commitments, and socializa-
tion) may play a role in other issue areas. These Þndings from the issue of regime
change, while important theoretically, may have limitations due to the unique nature
of the domestic political situation during times of transition. Further research should
examine these possible “outside-in” linkages across a variety of policy issues.
Finally, if policymakers continue to push the expansion of regional and interna-
tional organizations as a tool to expand democracy around the globe, three issues are
particularly important. First, policymakers must attend to the issue of forum choice.
The theory and results presented here suggest that smaller, more homogeneously
democratic regional organizations will have a stronger inßuence in promoting
democratic transitions. Policymakers must be cognizant that not all IOs are equal:
some will be more effective than others in promoting democracy.
Second, the policy community must consider the issue of causal mechanisms.
Some advocates of the IO-democracy link have suggested that, by increasing
external security, IOs increase the likelihood of democracy and of consolidation.121
I suggest that other causal mechanisms are at work and thus, policymakers must be
attuned to the particular nature of linkages between IOs and democracy. Stressing
the security beneÞts of IOs, for example, while ignoring their importance in making
credible commitments or socializing elites, risks using IOs in ways that may not
increase the likelihood of democratization.
Finally, policymakers cannot ignore the “supply-side” issues in this process. To
be effective promoters of democracy, these organizations and their member states
must make commitments to condition membership and enforce those conditions. If
the supply-side of the equation is attended to in a vigilant matter, the expansion of
regional organizations may bode well for democracy around the globe.
References
Almond, Gabriel A. 1989. The International-National Connection. British Journal of Political Science 19
(2):237–59.
Asmus, Ronald D., Richard L. Kugler, and F. Stephen Larrabee. 1993. Building a New NATO. Foreign
Affairs 72 (4):28 – 40.
Banks, Arthur S. 1994. Cross-polity Data Codebook. Unpublished manuscript, Binghamton University,
Binghamton, N.Y.
120. For an exception, see Russett and Oneal 2001, 200 –206.
121. Asmus, Kugler, and Larrabee 1993.
———. Various years. Political Handbook of the World. Binghamton, N.Y.: CSA Publications.
Banks, Arthur S., and Thomas C. Mueller. 1998. Political Handbook of the World. Binghamton, N.Y.:
CSA Publications.
Barro, Robert J. 1997. Determinants of Economic Growth: A Cross-Country Empirical Study. Cam-
bridge: MIT Press.
Beck, Nathaniel, Jonathan N. Katz, and Richard Tucker. 1998. Taking Time Seriously: Time-Series
Cross-Section Analysis with a Binary Dependent Variable. American Journal of Political Science 42
(4):1260 – 88.
BloomÞeld, Richard J. 1994. Making the Western Hemisphere Safe for Democracy? The OAS
Defense-of-Democracy Regime. Washington Quarterly 17 (2):157– 69.
Braun, Aurel. 1999. Introduction: The Continuing Dilemmas of Transition. In Dilemmas of Transition:
The Hungarian Experience, edited by Aurel Braun and Zoltan Barany, 1–28. Lanham, Md.: Rowman
& LittleÞeld.
Burkhart, Ross, and Michael Lewis-Beck. 1994. Comparative Democracy: The Economic Development
Thesis. American Political Science Review 88 (4):903–10.
Cameron, Maxwell A. 1994. Democracy and Authoritarianism in Peru: Political Coalitions and Social
Change. New York: St. Martin’s Press.
———. 1998. Self-Coups: Peru, Guatemala, and Russia. Journal of Democracy 9 (1):125–39.
Christopher, Warren. 1995. America’s Leadership, America’s Opportunity. Foreign Policy 98:6 –27.
CIA World Factbook. 1999. Available from 具www.odci.gov典. Accessed 30 October 1999.
Coufoudakis, Van. 1977. The European Economic Community and the ‘Freezing’ of the Greek
Association, 1967–1974. Journal of Common Market Studies 16 (2):114 –31.
Dassel, Kurt, and Eric Reinhardt. 1999. Domestic Strife and the Initiation of Violence at Home and
Abroad. American Journal of Political Science 43 (1):56 – 85.
Drake, Paul W. 1998. The International Causes of Democratization, 1974 –1990. In The Origins of
Liberty: Political and Economic Liberalization in the Modern World, edited by Paul W. Drake and
Mathew D. McCubbins, 70 –91. Princeton, N.J.: Princeton University Press.
Farer, Tom. 1996. Collectively Defending Democracy in the Western Hemisphere: Introduction and
Overview. In Beyond Sovereignty: Collectively Defending Democracy in the Americas, edited by Tom
Farer, 1–25. Baltimore, Md.: Johns Hopkins University Press.
Fearon, James D. 1994. Domestic Political Audiences and the Escalation of Interstate Disputes. American
Political Science Review 88 (3):577–92.
Fearon, James D., and David D. Laitin. 1996. Explaining Interethnic Cooperation. American Political
Science Review 90 (4):715–35.
Feng, Yi, and Paul J. Zak. 1999. The Determinants of Democratic Transitions. Journal of Conßict
Resolution 43 (2):162–77.
Fernández, Raquel, and Jonathan Portes. 1998. Returns to Regionalism: An Evaluation of Nontraditional
Gains from Regional Trade Agreements. The World Bank Economic Review 12 (2):197–220.
Finnemore, Martha. 1996a. National Interests in International Society. Ithaca, N.Y.: Cornell University
Press.
———. 1996b. Norms, Culture, and World Politics: Insights from Sociology’s Institutionalism. Inter-
national Organization 50 (2):325– 47.
Gasiorowski, Mark J. 1995. Economic Crisis and Political Regime Change: An Event History Analysis.
American Political Science Review 89 (4):882–97.
———. 1996. An Overview of the Political Regime Change Dataset. Comparative Political Studies 29
(4):469 – 83.
Gaubatz, Kurt T. 1996. Democratic States and Commitment in International Relations. International
Organization 50 (1):109 –39.
Geddes, Barbara. 1999. What Do We Know About Democratization after Twenty Years? Annual Review
of Political Science 2:115– 44.
Gerschenkron, Alexander. 1962. Economic Backwardness in Historical Perspective. Cambridge, Mass.:
Belknap Press.
Goldstein, Judith. 1996. International Law and Domestic Institutions: Reconciling North American
“Unfair” Trade Laws. International Organization 50 (4):541– 64.
———. 1998. International Institutions and Domestic Politics: GATT, WTO, and the Liberalization of
International Trade. In The WTO as an International Organization, edited by Anne O. Krueger,
133– 60. Chicago: University of Chicago Press.
Gourevitch, Peter A. 1978. The Second Image Reversed: The International Sources of Domestic Politics.
International Organization 32 (4):881–912.
Grieco, Joseph. 1996. State Interests and Institutional Rule Trajectories: A Neorealist Interpretation of the
Maastricht Treaty and European Economic and Monetary Union. Security Studies 5 (3):261–306.
Haggard, Stephen, and Robert Kaufman. 1992. Economic Adjustment and the Prospects for Democracy.
In The Politics of Economic Adjustment, edited by Stephen Haggard and Robert Kaufman, 319 –350.
Princeton, N.J.: Princeton University Press.
———. 1995. The Political Economy of Democratic Transitions. Princeton, N.J.: Princeton University
Press.
Halperin, Morton, and Kristen Lomasney. 1998. Guaranteeing Democracy: A Review of the Record.
Journal of Democracy 9 (2):134 – 47.
Huntington, Samuel. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman,
Okla.: Oklahoma University Press.
Hurrell, Andrew. 1996. The International Dimensions of Democratization in Latin America: The Case of
Brazil. In The International Dimensions of Democratization: Europe and the Americas, edited by
Laurence Whitehead, 146 –74. Oxford: Oxford University Press.
Ikenberry, G. John, and Charles Kupchan. 1990. Socialization and Hegemonic Power. International
Organization 44 (3):283–315.
Jones, Daniel, Stuart Bremer, and J. David Singer. 1996. Militarized Interstate Disputes, 1816 –1992:
Rationale, Coding Rules, and Empirical Patterns. Conßict Management and Peace Science 15
(2):163–213.
Kahler, Miles. 1992. External Inßuence, Conditionality, and the Politics of Adjustment. In The Politics
of Economic Adjustment: International Constraints, Distributive Politics, and the State, edited by
Stephen Haggard and Robert Kaufman, 89 –136. Princeton, N.J.: Princeton University Press.
Karaosmanoglu, Ali L. 1991. The International Context of Democratic Transition in Turkey. In
Encouraging Democracy: The International Context of Regime Transition in Southern Europe, edited
by Geoffrey Pridham, 159 –74. New York: St. Martin’s Press.
Kaufman, Robert. 1986. Liberalization and Democratization in South America: Perspectives from the
1970s. In Transitions from Authoritarian Rule: Comparative Perspectives, edited by Guillermo
O’Donnell, Philippe Schmitter, and Laurence Whitehead, 85–107. Baltimore, Md.: Johns Hopkins
University Press.
Keefer, Philip, and Stephen Knack. 1995. Institutions and Economic Performance: Cross-Country Tests
Using Alternative Institutional Measures. Economics and Politics 7 (3):207–27.
Keohane, Robert O. 1984. After Hegemony: Cooperation and Discord in the World Political Economy.
Keohane, Robert O., and Stanley Hoffman. 1993. Conclusion: Structure, Strategy, and International
Roles. In After the Cold War: State Strategies and International Institutions in Europe 1989 –1991,
edited by Robert Keohane, Joseph Nye, and Stanley Hoffman, 381– 406. Cambridge, Mass.: Harvard
Press.
Keohane, Robert O., and Lisa L. Martin. 1995. The Promise of Institutionalist Theory. International
Security 20 (1):39 –51.
Kurth, James. 1979. The Political Consequences of the Product Cycle: Industrial History and Political
Outcomes. International Organization 33 (1):1–34.
Leeds, Brett Ashley. 1999. Domestic Political Institutions, Credible Commitments, and International
Cooperation. American Journal of Political Science 43 (4):979 –1002.
Linz, Juan J., and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern
Europe, South America, and Post-Communist Europe. Baltimore, Md.: Johns Hopkins University
Press.
Lipset, Seymour M. 1959. Some Social Requisites for Democracy: Economic Development and Political
Legitimacy. American Political Science Review 53 (1):69 –105.
Londregan, John B., and Keith T. Poole. 1990. Poverty, the Coup Trap, and the Seizure of Executive
Power. World Politics 42 (2):151– 83.
———. 1996. Does High Income Promote Democracy? World Politics 49 (1):1–30.
Mainwaring, Scott. 1992. Transitions to Democracy and Democratic Consolidation: Theoretical and
Comparative Issues. In Issues in Democratic Consolidation: The New South American Democracies
in Comparative Perspective, edited by Scott Mainwaring, Guillermo O’Donnell, and Arturo Valen-
zuela, 294 –341. South Bend, Ind.: University of Notre Dame Press.
Malloy, James M. 1987. The Politics of Transition in Latin America. In Authoritarians and Democrats:
Regime Transition in Latin America, edited by James Malloy and Mitchell Seligson, 235–58.
Pittsburgh, Pa.: University of Pittsburgh Press.
MansÞeld, Edward D. 1998. The Proliferation of Preferential Trading Arrangements. Journal of Conßict
Resolution 42 (5):523– 43.
MansÞeld, Edward D., Helen V. Milner, and B. Peter Rosendorff. 2002. Why Democracies Cooperate
More: Electoral Control and International Trade Agreements. International Organization 56 (3):
477–514.
Manuel, Paul C. 1996. The Challenges of Democratic Consolidation in Portugal. Westport, Conn.:
Praeger.
Marks, Gary. 1992. Rational Sources of Chaos in Democratic Transition. In Reexamining Democracy:
Essays in Honor of Seymour Martin Lipset, edited by Gary Marks and Larry Diamond, 47– 69.
Newbury Park, Calif.: Sage Publications.
Marshall, Monty. 1999. Polity 98 Dataset. Available from 具www.bsos.umd.edu/cidcm/polity典. Accessed
1 July 1999.
Martin, Lisa L. 1992. Coercive Cooperation: Explaining Multilateral Economic Sanctions. Princeton,
N.J.: Princeton University Press.
Martin Lisa L., and Beth A. Simmons. 1998. Theories and Empirical Studies of International Institutions.
International Organization 52 (4):729 –57.
McCormick, James. 1980. Intergovernmental Organizations and Cooperation Among Nations. Interna-
tional Studies Quarterly 24 (1):75–98.
McNeely, Connie. 1995. Constructing the Nation-State: International Organization and Prescriptive
Action. Westport, Conn.: Greenwood Press.
Milner, Helen V. 1998. Regional Economic Co-operation, Global Markets, and Domestic Politics. In
Regionalism and Global Economic Integration, edited by William Coleman and Geoffrey Underhill,
19 – 41. London: Routledge.
Mitchell, Brian R. 1995. International Historical Statistics: Africa, Asia and Oceania 1750 –1988. New
York: Stockton.
———. 1998. International Historical Statistics: The Americas 1750 –1993. London: Macmillan
Reference.
Morrow, James D., Randolph M. Siverson, and Tressa E. Taberes. 1998. The Political Determinants of
International Trade: The Major Powers, 1907–1990. American Political Science Review 92 (3):649 –
61.
Nye, Joseph S., Jr. 1987. Peace in Parts: Integration and Conßict in Regional Organization. Reprint.
Latham, Md.: University Press of America.
O’Donnell, Guillermo. 1979. Modernization and Bureaucratic-Authoritarianism: Studies in South
American Politics. Berkeley, Calif.: Institute of International Studies.
O’Donnell, Guillermo, and Philippe C. Schmitter. 1986. Transitions from Authoritarian Rule: Tentative
Conclusions About Uncertain Democracies. Baltimore, Md.: Johns Hopkins University Press.
Pastor, Robert, ed. 1989. Democracy in the Americas: Stopping the Pendulum. New York: Holmes and
Meier.
Payne, Leigh. 1994. Brazilian Industrialists and Democratic Change. Baltimore, Md.: Johns Hopkins
University Press.
Pevehouse, Jon C. 2000. Democracy from Above? Regional Organizations and Democratization. Ph. D.
diss., Ohio State University.
———. 2002. With a Little Help from My Friends? Regional Organizations and the Consolidation of
Democracy. American Journal of Political Science 46 (3):611–26.
Polachek, Solomon W. 1996. Why Democracies Cooperate More and Fight Less: The Relationship
between International Trade and Cooperation. Review of International Economics 5 (3):295–309.
Powell, Charles. 1996. International Aspects of Democratization: The Case of Spain. In The International
Dimensions of Democratization, edited by Laurence Whitehead, 285–314. Oxford: Oxford University
Press.
Pridham, Geoffrey. 1991a. Encouraging Democracy: The International Context of Regime Transition in
Southern Europe. New York: St. Martin’s Press.
———. 1991b. The Politics of the European Community, Transnational Networks and Democratic
Transition in Southern Europe. In Encouraging Democracy: The International Context of Regime
Transition in Southern Europe, edited by Geoffrey Pridham, 212– 45. New York: St. Martin’s Press.
———. 1994. The International Dimension of Democratization: Theory, Practice, and Inter-regional
Comparisons. In Building Democracy? The International Dimension of Democratization in Eastern
Europe, edited by Geoffrey Pridham, Eric Herring, and George Sanford, 7–31. New York: St. Martin’s
Press.
———. 1995. The International Context of Democratic Consolidation: Southern Europe in Comparative
Perspective. In The Politics of Democratic Consolidation: Southern Europe in Comparative Perspec-
tive, edited by Richard Gunther, P. Diamandouros, and Hans-Jurgen Puhle, 166 –203. Baltimore, Md.:
Johns Hopkins University Press.
Pridham, Geoffrey, Eric Herring, and George Sanford, eds. 1994. Building Democracy? The Interna-
tional Dimension of Democratization in Eastern Europe. New York: St. Martin’s Press.
Przeworski, Adam. 1986. Some Problems in the Study of the Transition to Democracy. In Transitions
from Authoritarian Rule: Comparative Perspectives, edited by Guillermo O’Donnell, Philippe Schmit-
ter, and Laurence Whitehead, 47– 63. Baltimore, Md.: Johns Hopkins University Press.
———. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin
America. Cambridge: Cambridge University Press.
Przeworski, Adam, Michael Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 1996. What Makes
Democracies Endure? Journal of Democracy 7 (1):39 –55.
Przeworski, Adam, and Fernando Limongi. 1997. Modernization: Theories and Facts. World Politics 49
(2):155– 83.
Putnam, Robert. 1988. Diplomacy and Domestic Politics: The Logic of Two-Level Games. International
Organization 42 (3):427– 60.
Reiter, Dan. 2001. Does Peace Nature Democracy? Journal of Politics 63 (3):935– 48.
Risse, Thomas, Stephen Ropp, and Kathyrn Sikkink. 1999. The Power of Human Rights: International
Norms and Domestic Change. Cambridge: Cambridge University Press.
Russett, Bruce. 1993. Grasping the Democratic Peace : Principles for a Post-Cold War World.
———. 1998. A Neo-Kantian Perspective: Democracy, Interdependence, and International Organiza-
tions in Building Security Communities. In Security Communities, edited by Emmanuel Adler and
Michael Barnett, 368 –94. Cambridge: Cambridge University Press.
Russett, Bruce, John Oneal, and David R. Davis. 1998. The Third Leg of the Kantian Tripod:
International Organizations and Militarized Disputes, 1950 –1985. International Organization 52
(3):441– 67.
Russett, Bruce, and John R. Oneal. 2001. Triangulating Peace: Democracy, Interdependence, and
International Organizations. New York: Norton.
Schmitter, Phillipe. 1986. An Introduction to Southern European Transitions from Authoritarian Rule:
Italy, Greece, Portugal, Spain, and Turkey. In Transitions from Authoritarian Rule: Southern Europe,
edited by Guillermo O’Donnell, Philippe Schmitter, and Laurence Whitehead, 3–10. Baltimore, Md.:
Johns Hopkins University Press.
Schultz, Kenneth A. 1998. Domestic Opposition and Signaling in International Crises. American
Political Science Review 92 (4):829 – 44.
———. 1999. Do Democratic Institutions Constrain or Inform? Contrasting Two Institutional Perspec-
tives on Democracy and War. International Organization 53 (2):233– 66.
Shanks, Cheryl, Harold K. Jacobson, and Jeffrey H. Kaplan. 1996. Inertia and Change in the Constel-
lation of International Governmental Organizations, 1981–1992. International Organization 50 (4):
593– 627.
Sheahan, John. 1986. Economic Policies and the Prospects for Successful Transition from Authoritarian
Rule in Latin America. In Transitions from Authoritarian Rule: Comparative Perspectives, edited by
Guillermo O’Donnell, Philippe Schmitter, and Laurence Whitehead, 154 – 64. Baltimore, Md.: Johns
Hopkins University Press.
Simon, Michael W., and Erik Gartzke. 1996. Political System Similarity and the Choice of Allies: Do
Democracies Flock Together, or Do Opposites Attract? Journal of Conßict Resolution 40 (4):617–35.
Small, Melvin, and J. David Singer. 1994. Correlates of War Project: International and Civil War Data,
1816 –1992. Study no. 9905. Ann Arbor, Mich.: Inter-University Consortium for Political and Social
Research.
Smith, Alastair. 1998. International Crises and Domestic Politics. American Political Science Review 92
(3):623–38.
Smith, Tony. 1994. America’s Mission: The United States and the Worldwide Struggle for Democracy
in the Twentieth Century. Princeton, N.J.: Princeton University Press.
Solingen, Etel. 1998. Regional Order at Century’s Dawn : Global and Domestic Inßuences on Grand
Strategy. Princeton, N.J.: Princeton University Press.
Strang, David, and Patricia Mei Yin Chang. 1993. The International Labor Organization and the Welfare
State: Institutional Effects on National Welfare Spending, 1960 –1980. International Organization 47
(2):235– 62.
Summers, Robert, Alan Heston, Daniel A. Nuxoll, and Bettina Aten. 1995. The Penn World Table (Mark
5.6a). Available from pwt.econ.upenn.edu. Accessed 1 December 2000.
Thompson, William R. 1996. Democracy and Peace: Putting the Cart Before the Horse? International
Organization 50 (1):141–74.
Tovias, Alfred. 1984. The International Context of Democratic Transition. In The New Mediterranean
Democracies: Regime Transition in Spain, Greece, and Portugal, edited by Geoffrey Pridham,
158 –71. London: Frank Cass.
Treverton, Gregory. 1986. Spain: Domestic Politics and Security Policy. London: International Institute
for Strategic Studies.
Vetschera, Heinz. 1997. Security Policy and Democratic Control. In Civil-Military Relations in Post-
Communist States, edited by Anton Bebler, 15–21. Westport, Conn.: Praeger.
Verney, Susannah, and Theodore Couloumbis. 1991. State-International System Interaction and the
Greek Transition to Democracy in the Mid-1970s. In Encouraging Democracy: The International
Context of Regime Transition in Southern Europe, edited by Geoffrey Pridham, 103–24. New York:
St. Martin’s Press.
Whitehead, Laurence. 1986. International Aspects of Democratization. In Transitions from Authoritarian
Rule: Comparative Perspectives, edited by Guillermo O’Donnell, Philippe Schmitter, and Laurence
Whitehead, 3– 46. Baltimore, Md.: Johns Hopkins University Press.
———. 1989. The Consolidation of Fragile Democracies: A Discussion with Illustrations. In Democracy
in the Americas: Stopping the Pendulum, edited by Robert Pastor, 76 –95. New York: Holmes and
Meier.
———. 1993. Requisites for Admission. In The Challenge of Integration: Europe and the Americas,
edited by Peter Smith, 149 – 82. New Brunswick, N.J.: Transaction Publishers.
———. 1996a. The International Dimensions of Democratization: Europe and the Americas. Oxford:
Oxford University Press.
———. 1996b. Democratic Regions, Ostracism, and Pariahs. In The International Dimensions of
Democratization: Europe and the Americas, edited by Laurence Whitehead, 395– 412. Oxford: Oxford
University Press.
———. 1996c. Concerning International Support for Democracy in the South. In Democratization in the
South: The Jagged Wave, edited by Robin Luckham and Gordon White, 243–73. Manchester:
Manchester University Press.
———. 1996d. Democracy by Convergence and Southern Europe: A Comparative Politics Perspective.
In The International Aspects of Democratization: Europe and the Americas, edited by Laurence
Whitehead, 265– 84. Oxford: Oxford University Press.
World Bank. 1998. The World Development Indicators. CD-ROM. Washington, D.C.: World Bank.

Democracy From The Outside-In? International Organizations and Democratization

Uploaded by

Copyright:

Available Formats

Democracy From The Outside-In? International Organizations and Democratization

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Democracy From The Outside-In? International Organizations and Democratization

Uploaded by

Copyright:

Available Formats

Democracy from the Outside-In?

International Organizations and

Democratic transitions have become a widely studied phenomenon in comparative

Democracy is seen as the outcome of a domestic political process that is not

International Organization 56, 3, Summer 2002, pp. 515–549

Democracy from the Outside-In?

Most democratization studies present regime transitions as the outcome of a

12. Almond 1989, 242– 44.

general-welfare goals by constraining Congress’s ability to act. Finally, Etel

Regional Organizations and Regime Change:

19. Solingen 1998.

socialization of domestic elites. Before turning to the causal mechanisms, I brießy

Scope of Theory and Analysis

24. Shanks, Jacobson, and Kaplan 1996.

paying the costs associated with membership.30 This assistance-based conditionality

Pressure and Liberalization in Autocratic Systems

35. See Mainwaring 1992; O’Donnell and Schmitter 1986.

self-conÞdence.”39 Certainly the efforts of the United States to promote democracy

39. Whitehead 1996c, 248.

Societal Elites and Acquiescence to Liberalization

48. Halperin and Lomasney 1998, 137.

Business Elites. For many business elites in authoritarian systems, democracy

59. Sheahan 1986, 163.

spending in Hungary be stabilized and even increased.68 This reversed a downward

The Supply-Side of Democratization

democratic conditionality exists. Thus enforcement of these conditions is also

77. Whitehead 1993, 154.

democracy. The “supply-side” inßuence on the IO-democracy link is especially

Testing the Argument

If the preceding argument is correct, an association should exist between democratic

81. See Marshall 1999; Gasiorowski 1996.

which state i is a member. The expectation is that membership in democratic IOs

TABLE 1. Descriptive statistics and associations

Number of states in analysis 106 86

TABLE 2. Estimates of the determinants of the transition to democracy,

Polity98 Gasiorowski Polity98 Gasiorowski

IOScoreit⫺1 0.089*** 0.065** 0.099***a 0.112***a

TABLE 3. Predicted probabilities of transitions based on the estimate, in column

All variables at mean (baseline)a 0.0021

TABLE 4. Predicted probabilities of transitions based on the estimates in column

All variables at mean (baseline)a 0.0117

107. Przeworski and Limongi 1997.

The interaction of international and domestic politics has garnered increased

118. Martin and Simmons 1998.

and democratic development.120 Moreover, in all of this work, organizations are

You might also like

IOScoreit⫺1 0.089* 0.065 0.099a 0.112a