Following the Leader: International Order, Alliance Strategies, and Emulation

By Raymond C. Kuo

Nations have powerful reasons to get their military alliances right. When security pacts go well, they underpin regional and global order; when they fail, they spread wars across continents as states are dragged into conflict. We would, therefore, expect states to carefully tailor their military partnerships to specific conditions. This expectation, Raymond C. Kuo argues, is wrong.

Following the Leader argues that most countries ignore their individual security interests in military pacts, instead converging on a single, dominant alliance strategy. The book introduces a new social theory of strategic diffusion and emulation, using case studies and advanced statistical analysis of alliances from 1815 to 2003. In the wake of each major war that shatters the international system, a new hegemon creates a core military partnership to target its greatest enemy. Secondary and peripheral countries rush to emulate this alliance, illustrating their credibility and prestige by mimicking the dominant form.

Be it the NATO model that seems so commonsense today, or the realpolitik that reigned in Europe of the late nineteenth century, a lone alliance strategy has defined broad swaths of diplomatic history. It is not states' own security interests driving this phenomenon, Kuo shows, but their jockeying for status in a world periodically remade by great powers.

• Language: English
• Publisher: Stanford University Press
• Release date: Aug 3, 2021
• ISBN: 9781503628571

Book preview

FOLLOWING THE LEADER

International Order, Alliance Strategies, and Emulation

Raymond C. Kuo

STANFORD UNIVERSITY PRESS

Stanford, California

Stanford University Press

Stanford, California

©2021 by the Board of Trustees of the Leland Stanford Junior University.

All rights reserved.

No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or in any information storage or retrieval system without the prior written permission of Stanford University Press.

Printed in the United States of America on acid-free, archival-quality paper

Library of Congress Cataloging-in-Publication Data

Names: Kuo, Raymond C. (Raymond Cheng), author.

Title: Following the leader : international order, alliance strategies, and emulation / Raymond C. Kuo.

Description: Stanford, California : Stanford University Press, 2021. | Includes bibliographical references and index.

Identifiers: LCCN 2020048021 (print) | LCCN 2020048022 (ebook) | ISBN 9781503628434 (cloth) | ISBN 9781503628571 (epub)

Subjects: LCSH: Alliances. | International relations. | Alliances—Case studies. | World politics.

Classification: LCC JZ1314 .k865 2021 (print) | LCC JZ1314 (ebook) | DDC 355/.031—dc23

LC record available at https://lccn.loc.gov/2020048021

LC ebook record available at https://lccn.loc.gov/2020048022

Cover design: Rob Ehle

Typeset by Newgen in 10/13.5 Adobe Garamond Pro

For my dad

Contents

Acknowledgments

1. Transhistorical Patterns in Alliance Strategy

2. The Theory of Strategic Alliance Diffusion

3. The Diffusion of Alliance Strategy: Systemic Patterns and Evidence

4. Great Powers and Strategic Constraints: The Bismarckian Era, 1873–1890

5. Cold War Credibility: NATO, SEATO, and CENTO, 1949–1965

6. Diffusion to the Periphery: Security Cooperation in Southern Africa, 1992–2004

7. The Dominant Strategy and Alliance Failure

8. The Dominant Alliance Strategy: Policy Implications and Theoretical Extensions

Appendix

Notes

Bibliography

Index

Acknowledgments

GEORGE LAWSON, ONE OF MY professors at the London School of Economics, once said that good research has a high ratio of thought to writing. In that same vein, I hope that these brief words of thanks convey the depth of my gratitude to everyone who supported and helped me bring this book to fruition.

To John Ikenberry: Thank you for being a tireless supporter of this project. I could always rely on your encouragement and enthusiasm, even when I harbored deep doubts myself.

To my scholarly mentors: I am truly lucky to have benefited from your guidance. Andrew Moravcsik and Jacob Shapiro, thank you for pushing, for not letting complacency or superficial satisfaction set in when deeper thought and greater effort could be accomplished instead. Michaela Mattes and Paul Poast, your professional and personal advice made this book immeasurably stronger and myself a better political scientist. To Mark Crescenzi, Keren Yarhi-Milo, Ashley Leeds, and Jeremy Pressman: I could only have written this book by following the paths you set. And a special thanks to Robert Keohane: With one question, he unlocked this book’s puzzle.

To my friends, inside and outside academia: Thank you for your guidance on this book and, even more importantly, for keeping me sane while writing it. I would particularly like to recognize the support of Brian Blankenship, Jaquilyn Waddell Boie, Luke Brown, Simon Collard-Wexler, Michael Hunzeker, Chris Kendall, Alexander Lanoszka, Danielle Lupton, Alexandre Su, and Meredith Wilf.

To Stanford University Press: Thank you for shepherding this book so effectively and speedily. Alan Harvey, Caroline McKusick, and Jessica Ling, I have never had a better publishing experience than with you. Thank you to my incredible faculty and administrative colleagues at Fordham University, the University at Albany–SUNY, and Princeton University. Liu Bingwan, Zhu Lei, Mihalis Alisandratos, and Robert Gray provided outstanding research assistance.

I am undeservedly lucky to be surrounded by strong, brilliant women. Thank you to my mom, for teaching me assertiveness; to my sister, who inspires me every day with her perseverance and patience; to Layla and Ariya, my two daughters, who are the joy of my life. And to Shanta, my incomparable wife. My life is unimaginable without you.

Finally, this book is dedicated to my dad, Dr. Kuonan Kuo. While a doctorate in chemistry was a path to a better life, his first interests were always politics, history, and law. Had he had the freedom to pursue these intellectual passions, he would have made a great social scientist.

1

Transhistorical Patterns in Alliance Strategy

1.1 Introduction

On June 17, 1950, the seven members of the Arab League signed the modern Middle East’s first collective security pact: the Treaty of Joint Defense and Economic Cooperation. The alliance was a failure. It never brought the Arab militaries together into an effective, combined force. It struggled to deter, let alone combat, threats. Israel decisively defeated its members in 1967, then again during the War of Attrition and the October War. Even worse, Egypt and then Jordan signed separate peace treaties with Jerusalem despite threats of alliance expulsion. The pact did little to temper the Arab Cold War, and, indeed, members regularly coerced or even attacked one another, as with the 1990 Iraqi conquest of Kuwait and opposing Saudi and Egyptian interventions during the North Yemen Civil War.

In many ways, this treaty was designed to fail. Its strategy was wholly unsuited to its members. Composed of small and weak countries, none had advanced military capabilities, modern and diversified economies, nor even significant diplomatic influence. The league could not consolidate its diverse forces and national commands, and, even before signing, members recognized that they lacked the political cohesion required for combined military planning and operations.

Yet they directly emulated NATO’s design. Despite internecine conflict, the league founded an integrated command structure—the United Arab Command—modeled on NATO’s Supreme Headquarters Allied Powers Europe. Despite limited trade and economic ties, the treaty established an Economic Council to coordinate member industry and agriculture, much like Article 2 of the North Atlantic Treaty encourages. Most importantly, the pact copied NATO’s Article 5 defense guarantee, stating that signatories consider any [act of] armed aggression made against any one or more of them or their armed forces, to be directed against them all. Despite wide differences in member characteristics, geostrategic threats, and economic conditions, both Western Europe and the Arab League pursued identical alliance strategies.

We see this same pattern of strategic convergence on infertile ground across history. In 1833, two small Italian city-states signed a defense pact modeled after the Quadruple Alliance (1815), the preeminent pact of its time. Similarly, in 1864, Colombia and Ecuador concluded a defense partnership. Like the Quadruple and Italian alliances, it stipulated the exact number of naval forces each party would provide in a conflict. But neither had a functioning navy at signing. Indeed, Colombia had decommissioned its naval school and abolished the Ministry of the Navy almost forty years before.¹ Prior to World War 1, nearly all alliances were secret. Since then, secret pacts form a negligible, at times nonexistent, part of the global alliance network.²

Interstate military alliances are central to international security. They deter conflict, aggregate power to challenge adversaries, underpin regional and global order, and foster security communities that create durable, peaceful relations among countries. These pacts have also spread wars across continents and dragged states into conflicts they would have rather avoided, generating new bloodshed. Given these stakes, nations have powerful incentives to get their alliances right. Overcommitting risks entrapment, but institutionalized coordination can offer more robust and durable cooperation. Limited treaties offer diplomatic flexibility, but raise the risk of abandonment by partners at critical and inopportune moments.

To make effective pacts, countries adopt specific alliance strategies: Means-ends beliefs about the guarantees, political promises, and organizational features that best allow their partnerships to overcome cooperation problems and generate military security. States deterring threats publicly announce defensive obligations. Allies hoping to conquer neighbors spell out the division of spoils, cost-sharing formulas, and troop contributions. States facing domestic or international resistance to their security cooperation keep their pacts secret. Nations mix and match these features to address the specific threats, domestic constraints, and military capabilities they face. Given the range of these factors, states should carefully tailor their military partnerships. We expect wide variation in alliance strategy, organizational design, and military guarantees.

This expectation is wrong. The examples above are not isolated coincidences. In any given year, 75 percent of states pursue the same alliance security strategy. They provide similar guarantees, create similar internal coordinating mechanisms, and use alliances to achieve similar foreign policy goals. Countries converge on a single, legitimate approach to create security through military pacts: a dominant alliance strategy or form. Across periods divided by major wars, this form varies sharply. In some, it features a realpolitik approach to military cooperation emphasizing fluid realignment and alliance abandonment. In others, it fosters integrative, cohesive, and institutionalized security communities. Within those periods, however, the dominant strategy’s prevalence locks in a single conception of legitimate and credible alliance cooperation. It thus constitutes an integral component of international order, defining broad swaths of diplomatic history, as well as structuring and constraining military policy, strategic alignment, and the likelihood of war for decades.

This book asks: What causes the dominant alliance strategy? What spreads it? And what effects does its variation have on international security and order? Its answer is grounded in major war. As Ikenberry (2001) notes, a single great power emerges victorious from these global conflicts. Its decisive, comprehensive victory grants a windfall of power and authority. This new hegemon then creates a core alliance gathering other great but lesser powers to meet its most important security challenge in the postwar strategic environment. This central pact forms the nucleus of the dominant strategy.

The leading state does not impose this form on secondary and peripheral states. Indeed, it actively avoids doing so. But such countries draw their own lessons from the new global leader. They elevate the core alliance as the standard of credible and legitimate military cooperation. Emulation—the diffusion of a single alliance strategy—proceeds according to two paths. Secondary countries—those allied to a core alliance member—follow a credibility diffusion process. Because their partners have a guarantee from the hegemon, these states worry about relative reliability: the rank or prioritization of their pact within their partner’s broader alliance portfolio. In essence, they ask: What are the limits of my ally’s commitment, and how can I ensure my pact stands above or at least equal to its competing obligations? They push partners to emulate the core alliance’s strategy and features in their own separate pacts.

Peripheral countries—allied nations not connected to either core or secondary states—emulate to gain status benefits. Mimicking the central alliance strategy reinforces the sovereign status of these states, demonstrating that they abide by the outward markers of effective alliances and pursue foreign policy goals validated by the hegemon. In so doing, peripheral nations hope to bolster their internal and external legitimacy, raising their standing within the international system. In total, the dominant form subverts theoretical expectations. Instead of tailored pacts designed according to means-ends utility maximization, states ignore their individuated conditions and coalesce on a single alliance strategy.

This chapter establishes the phenomenon of interest. Using the Alliance Treaty Obligations and Provisions (ATOP) dataset, section 1.3 demonstrates that most alliances converge on a single design. Few states—and none of the great powers—defy this dominant alliance strategy in their formal military partnerships. Existing scholarship provides little analytical traction over the dominant form’s existence and variation, although section 1.4 outlines two possible explanations. If enough states face the same security challenges, these common threats could produce institutional isomorphism. Alternatively, hegemons could impose their preferred designs. Substantial empirical and theoretical problems confront both approaches. Consequently, the following chapter will introduce a new theory drawing upon network analysis and social theories of rank and status to explain the book’s puzzle, the dominant alliance strategy. Finally, section 1.5 will outline how the book is organized.

Far more structure emerges from major wars than we suspect. Security regimes are not driven solely by hegemonic or great power actions but also by decisions from multiple strata in the interstate system. These decisions, moreover, follow their own, separate logics, with states responding to a mix of power, credibility, and status incentives depending on their network position. A complex, interactive process generates more uniform security behavior and international order than previously understood.

1.2 Alliance Security Strategies

This project defines alliances as: Written agreements, signed by official representatives of at least two independent states, that include promises to aid a partner in the event of military conflict, to remain neutral in the event of conflict, to refrain from military conflict with one another, or to consult/cooperate in the event of international crises that create a potential for military conflict.³ Within this definition, states can choose from a wide variety of alliance design features and functional roles to facilitate cooperation, account for individual member vulnerabilities or liabilities, settle intra-allied disagreements, and more effectively deter threats. Together, these design choices create an alliance security strategy: A means-ends mapping or belief about how particular military guarantees, organizational features, and political commitments (i.e., means) can be applied in alliances (i.e., ways) to allow cooperation, bolster cohesion, and, ultimately, generate security (i.e., ends).

The literature on alliances has grappled with many of these design elements, grouped into three broad categories: alliances’ functional roles or purposes, their management, and their cohesion. Prior studies generally adopt a functionalistrationalist approach: States select particular design features because they believe they will effectively resolve specific cooperation problems. For example, with respect to the functional purposes of alliances, defensive realists expect most countries to form defensive military agreements targeting specific threats.⁴ Weitsman (2004)’s seminal alliance typology argues that high external threats prompt states to have clearly defined military obligations and conditions for activation. By contrast, Benson (2012) and Kim (2011) argue that entrapment fears spur partners to adopt conditional or ambiguous security guarantees, the better for a state to constrain adventurous or risk-taking behavior by an ally. Pacts should therefore include clear limits on the casus foederis (i.e., events triggering the military commitment); be very little thinly institutionalized, if at all; and primarily commit states to mutual consultation and possibly defense. Abandonment should also be quite common.⁵

Although sharing many predictions with their defensive counterparts, offensive realists might expect somewhat more offensive and neutrality pacts, particularly as rising powers attempt to establish regional or even global hegemony. To seize strategic surprise, these pacts may be secret. To address distributive conflicts among partners, these alliances may agree to specific divisions of spoils and/or conquered territory in advance. In either case, both forms of realism anticipate a realpolitik approach to alliance strategy. As threats and capabilities increase, states create sharply delimited alliances to either defend or advance specific national interests, as well as fluidly rebalance against emerging adversaries (including former allies).

Other studies follow Koremenos et al. (2001) in exploring how states use certain provisions to overcome management problems. Wallander and Keohane (1999) and Wallander (2000) claim that nonmilitary coordinating institutions, larger membership size, and longer expected duration are responses to high issue density and policy spillover. States establish coordinating mechanisms with broad delegated portfolios to develop comprehensive policies and avoid duplication of effort. Bearce et al. (2006) show how states use information-sharing provisions to mitigate crisis escalation, while Ikenberry (2001) and Schroeder (1976) highlight institutional binding as a way to manage imbalances of power and policy coordination among allies. Finally in this area, Bensahel (2007) and Moeller (2016) explore alliance warfighting performance, how elements like unified command structures, equipment interoperability, common doctrine, and joint training can act as force multipliers and contribute to battlefield success and military victory or defeat. Insofar as security partners face, say, high interdependence causing spillover or threats requiring significant multilateral coordination, they have a wide range of institutional strategies and options to address these concerns.

The third category includes alliance strategies to handle cohesion and credibility problems. States can leverage reputations of resolve to facilitate alliance-making, while institutionalization can help countries that have previously reneged on their commitments find new partners.⁶ Gibler (1996) explores alliances that settle territorial disputes, which typically contain detailed provisions demarcating land to be exchanged and national borders. Lake (2009) and the small states literature examine asymmetric obligations in partnerships between powerful and peripheral countries.⁷ The latter can offer few security benefits to the former, and so instead exchange political acquiescence for military assistance. Even outside of patron-client relations, distributional conflicts over burden-sharing can prompt member-specific contributions.⁸

Under certain conditions, states may want to reduce their credibility and alliance cohesion. Domestic political costs and the strategic benefits of flexibility may lead states to limit institutionalization and other markers of commitment.⁹ Gibbs (1995) finds that those same incentives lead states to include secret provisions or enact completely hidden pacts. Democracies can rely on costly ratification procedures and securing the support of veto actors to demonstrate commitment, obviating institutionalization.¹⁰

The choice of alliance strategy has critical effects on interstate security and global stability. Ikenberry (2001) places institutionalized alliances at the heart of constitutional international orders. By contrast, fluid, realist pacts may spur and spread warfare,¹¹ although others highlight the role of carefully worded guarantees to prevent this from happening.¹² Military coordination should increase the likelihood of victory, while information-sharing provisions reduce intra-allied conflict and stabilize security relations.¹³ At the limit, certain alliance strategies can foster security communities where war is unthinkable between members.

All these theories expect states to carefully tailor their pacts to specific threats, constraints, and capabilities. Partners should mix and match features to optimize the security their alliances produce. For example, a country facing a powerful threat but suffering reputational problems might ask for defensive commitments, coupled with explicit conditions on invocation, consultative fora, and clear decision-making procedures to address concerns about reliability. Those working with potentially untrustworthy counterparts to acquire territory may create offensive or neutrality pacts that commit partners to specific spheres of influence or areas of conquest in advance. Certain design choices might be more popular than others, but such popularity typically reflects a higher prevalence of underlying causes. For instance, advances in transportation and communications technology make it easier for alliances to establish management bodies. In addition, some features may correlate with others. Offensive realism might expect states to adopt a common set of designs in their alliance strategy: neutrality and offensive obligations, minimal institutionalization, perhaps also secrecy, provisions clearly dividing gains, and/or delineated contributions. But this would only apply to a select set of states embarking on aggressive expansion or regional hegemony. Their targets would likely adopt opposing features (e.g., defensive guarantees with robust institutionalization), preserving variation in partnership design. Given the range of individuated state challenges, threats, vulnerabilities, and capabilities, the literature would expect diversity in alliance security strategies.

1.3 The Dominant Alliance Strategy

This expectation is wrong. Alliance characteristics move and disappear from the international security system as blocks. Design choices are strongly correlated, such that analyzing features individually misses broad, transhistorical patterns. Most importantly, rather than diversity, the majority of states and pacts ignore their individuated circumstances and converge on a unique, prevailing alliance strategy with a single set of cooperative functions, normative expectations, and organizational traits. This dominant alliance strategy or form is not a mixed bundle of what features happen to be widespread at a given time. These features conceptually—and as discussed in the next chapter—theoretically connected through normative beliefs about the optimal alliance strategy; that is, how states can best configure military cooperation to create security through these partnerships. The dominant strategy anchors and constrains alliance design choices, forming a standard of security cooperation that, this book argues, all states recognize and most states adopt.

This dominant strategy alternates between two broad approaches. The first is a realpolitik approach marked by conditional commitments, limited institutionalization, narrow scope, and occasionally predatory features like the division of spoils. Alliances serve limited, short-term functions, with realignment and abandonment not merely expected but even valued to prevent concentrations of power and maintain stability.

By contrast, the second is an integrative approach to security cooperation. Under this strategy, alliances define long-term, even indefinite, patterns of military relations among members. They feature coordinating mechanisms with wide policy scope, using tighter coordination—rather than realignment—to adapt the pact to emerging challenges. This coordination can be extensive, even intrusive, touching upon the states’ core security prerogatives, such as integrating military commands; delegating decision-making on troop training and deployment; harmonizing intra-allied production and financing of equipment; and encompassing broader economic, political, and social exchange. Under this approach, alliances constitute the institutional foundation for robust policy relationships and security communities.

The dominant alliance form’s exact features vary by time period, but the Utrecht (1715–55), Bismarckian (1873–94), and Interwar (1919–38) eras followed the realpolitik approach. An integrative dominant strategy prevailed in the Concert (1815–66), Cold War (1946–91), and post–Cold War (1992+) periods. We can use the Alliance Treaty Observations and Provisions (ATOP) dataset to visualize and formalize these patterns from 1815–2003.¹⁴ The graphs below present the proportion of alliances featuring a particular characteristic.¹⁵ The alternation between realpolitik and integrative strategies is most pronounced with general coordinating bodies: alliance bodies with the delegated, decision-making authority to manage and coordinate intra-allied policy and military planning and operations, often through a permanently seconded bureaucracy. As seen in figure 1.1, 80.66 percent of alliances had a formal organization in the Concert period. After the Franco-Prussian War of 1870–71, no pacts possessed such bodies until 1907. After World War II, this feature’s prevalence suddenly spikes from 8.99 percent in 1946 to 90.64 percent in 1948 and is even higher today. Figure 1.1 displays proportions, so even as the number of states and alliances increase by orders of magnitude, few go against the dominant alliance strategy, whether realpolitik or integrative.

This pattern applies to more than a dozen other institutional features. Integrative alliance strategies aim to create robust, overlapping intra-allied cooperation. Periods with an integrative dominant form should therefore see disproportionately higher levels of mediation/arbitration mechanisms to settle internal disputes, as well as cooperation on nonmilitary policy issues. Because of this, following Wallander and Keohane (1999), these pacts should often have more members and survive longer. Moreover, we should expect sharp drops in these features’ prevalence during realpolitik eras, which manage policy spillover and externalities differently.

And that is what we see in figure 1.2. With alliance size, for example, most partnerships had a dozen members or more during the early 1800s and again after 1945. During the realpolitik eras, they had three on average. Cooperation on nonmilitary issues was similarly prevalent during eras with an integrative dominant form, but only 11.16 percent of alliances had them outside of those years. Provisions for internal mediation were far more common outside the Bismarckian and Interwar periods than within them. Similarly, collective security pacts—those alliances not directed at specific targets (at least nominally) but instead concerned with any disturbance of the peace—are much more frequent during the Concert period and following World War II. In addition, since 1947, a majority of alliances have not specified an end date. Members expect their pacts to structure their security relations indefinitely. Prior to that year, fewer than one