Introduction
AFTER the Second World War disrupted global and regional security orders, new US-led regional security arrangements emerged with varying institutional forms. The multilateral security institutions that took shape in the North Atlantic Treaty Organization contrasted with the bilateral pacts that connected the United States and its Asian allies.Footnote 1 Multilateral security and political institutions also emerged in the Western Hemisphere after the war—particularly the Inter-American Treaty of Reciprocal Assistance (the Rio Treaty) and the Organization of American States (oas). Nevertheless, debates about the institutional forms of postwar regional security orders have overlooked this contemporaneous case. This is more than an empirical oversight.
Theoretical frameworks that emphasize external threat, burden-sharing, and shared identity as jointly necessary for the emergence of multilateral regional security institutions do not explain the inter-American case.
The post-World War II inter-American system after 1947 centered on a collective security treaty and after 1948 included a strengthened, multipurpose multilateral organization. These regional institutions were shaped by power asymmetries, but they also ensconced norms of sovereign equality and nonintervention. At the time of the creation of these institutions, most Latin American leaders saw them as way to gain greater voice through multilateral decision-making and to create institutional and legal constraints on US unilateralism. Historical and international relations scholarship largely treats postwar inter-American institutions as a continuation of hegemonic or imperial leadership. The empirical examination in this article contests that interpretation and uncovers two important aspects of the period. First, the continuation (indeed, expansion) of inter-American institutions was far from certain given disagreements among US policymakers who variously favored new global arrangements, a renewed inter-American system, or the preservation of unilateral prerogative. Second, the new inter-American system was not a US imposition. Instead, crucial support and impetus for multilateral inter-American security institutions originated with Latin American policymakers who wanted to expand, build upon, and repurpose prewar regional institutions in the context of a nascent global order. Although US power and interventionism figured among the concerns of the Latin Americans, they also sought to institutionalize engagement, voice opportunities, restraint, and rules-based order with the United States and with one another. The postwar multilateral security architecture emerged as a grand bargain that institutionalized and extended US influence while recognizing Latin American demands.
This case does not conform to prominent explanations for the emergence of multilateral regional security institutions. Drawing on multinational archival evidence and a historical institutionalist framework, I show how shared historical antecedents of regionalism shaped the emergence of multilateral security institutions. In the Americas, these included a legacy of Pan-American cooperation, designs, and practices that were situated in juridical, normative, and diplomatic traditions partially shared among the United States and Latin American republics. In the North Atlantic too, historical antecedents—wartime collaboration, interwar negotiations, histories of arbitration, and earlier proposals for North Atlantic cooperation—prefigured nato’s multilateral regional security institutions.Footnote 2 The paucity of such antecedents spanning the Pacific corresponds with the absence of security multilateralism in the region, namely the failure of the Pacific Pact, and the subsequent reliance on bilateralism.Footnote 3 My argument contributes to broader debates about international institutional creation and design by showing how, during critical junctures, antecedents shape actors’ choices and facilitate processes of layering and conversion. These conceptual tools, adapted from historical institutionalism, improve ir accounts of how history matters for the development of regional institutions by distinguishing antecedents from identity and by illuminating processes of institutional creation and change. By employing underutilized Brazilian, Chilean, Colombian, and Mexican foreign ministry archives along with records from the US Department of State and the Harry S. Truman Presidential Library, this article makes an empirical contribution on postwar inter-American relations and extends into the security realm emerging research on Latin American contributions to the norms, practices, and institutions of global governance.Footnote 4
The article continues as follows. The next section outlines the debate over the emergence of divergent forms of postwar security orders. After showing that leading explanations do not fit the inter-American case, I advance an alternative explanation based in historical institutionalism (hi) by examining the formation of postwar inter-American security institutions in depth. The article concludes with a brief comparison of the post-WWII cases to demonstrate how antecedents help to explain institutional variation in US-led security regionalism.
Orders and Regions
Although the US-led, post–World War II international order is often characterized as multilateral, substantial variation in regional security arrangements existed from the start. ir scholars offer numerous explanations for variations in institutional emergence and form. Supplyside explanations typically treat hegemons or “pivotal states” as security providersFootnote 5 that act for systemic reasons or due to the internationalist interests of governing coalitions.Footnote 6 Derrick Frazier and Robert Stewart-Ingersoll point to the leading power’s capability, role, and orientation as determinants of regional order.Footnote 7 Thomas Pedersen labels West German postwar strategies as “cooperative hegemony,” emphasizing the leading power’s role in fostering regional order.Footnote 8 Conversely, explanations of the regional demand side focus on common external and domestic threats and negative security externalities.Footnote 9 In comparative regionalism, Amitav Acharya and Iain Johnston emphasize power distributions, cooperation problems, identity, and regime-type heterogeneity to explain distinctive regional institutional designs. Miles Kahler and Andrew MacIntyre emphasize spillover and feedback effects in generating support from economic interests for regional integration.Footnote 10
Within the literature on security regionalism, one prominent debate compares the emergence of multilateral security institutions in nato with their absence in Asia.Footnote 11 But the “Why is there no nato in Asia?” debate overlooks the inter-American experience during the same juncture.Footnote 12 From 1944 to 1948, Western Hemisphere states conceived, negotiated, and implemented a collective security pact and multifaceted, multilateral organizations. This contemporaneous inter-American case—slightly predating nato—allows for a strong test of existing explanations for the emergence of multilateral security institutions.
Christopher Hemmer and Peter Katzenstein effectively treat the emergence of multilateral security institutions as dichotomous: the North Atlantic developed multilateral security governance while Asia did not. Although North Atlantic and inter-American security arrangements differed in important respects, both regional systems fit prominent definitions of multilateralism, including the conceptualization these authors employ. For Robert Keohane, multilateralism is essentially coordination among three or more states.Footnote 13 John Ruggie invokes a “qualitative dimension” in which coordination occurs “on the basis of certain principles of ordering relations among those states.”Footnote 14
For security multilateralism specifically, Ruggie emphasizes “some expression or other of collective security or collective self-defense.”Footnote 15 For these latter two scholars, multilateral institutions can be global or regional in scope; Ruggie further stresses that principles of conduct apply generally among member states.
nato and the inter-American system had important similarities starting with their founding texts, the Rio Treaty (1947) and the North Atlantic Treaty (1949). Both created regional, collective defense guarantees based on the principle that an attack on one is an attack on all, with mandatory mechanisms for multilateral consultation. Both drew juridical legitimacy from United Nations Charter Article 51 on regional collective defense, have mechanisms for referral to the UN Security Council, and specify the creation of an organization for executing these functions.Footnote 16 These similarities contrast with the failed Pacific Pact and, later with the Southeast Asian Treaty Organization (seato), which lacked a collective defense clause and retained unilateral and bilateral decision-making.Footnote 17
Hemmer and Katzenstein (building on Ruggie) emphasize that the emergence of regional security institutions, namely nato, was part of the construction of a North Atlantic region that included the United States and Canada. Although the United States is sometimes considered an extra-regional power in Europe and a natural part of the Americas, in fact, Pan-American and inter-American constructs naturalized the United States’ place in an American region in a manner similar to the construction of the North Atlantic region. Pan-American regionalism competed with constructs that placed the US outside of “Latin” or “Hispanic” regions; proponents of these alternatives even proposed security arrangements directed at excluding the United States and European powers from their region. In this sense, the United States was no more of an extra-regional power in the North Atlantic than it was in the inter-American region. Materially, the distinction is similarly ambiguous, particularly regarding the US role in South America. Although the US exercised long-standing predominance around the Caribbean, it is doubtful that its influence or interests were deeper in South America in the late 1940s than in Europe or Asia, where it had fought wars, held massive debts, and garrisoned hundreds of thousands of troops.
Still, the inter-American system and nato emerged with important differences that developed over time. In contrast to the North Atlantic Treaty, which emphasizes collective defense against external aggression, the Rio Treaty emphasizes peaceful resolution of disputes among signatories and explicitly considers the possibility of intraregional conflicts. It is more specific about decision-making procedures and juridical precedents, with implicit reference to nonintervention and sovereign equality.Footnote 18 But more important than initial textual differences is how the arrangements evolved in their first years. Following hemispherewide consultation in 1945, the American states approved a collective security compact in 1947 that was complemented by a reinvigorated regional body, the oas, the following year. Although the oas effectively functioned as the decision-making forum for the Rio Treaty, the oas was created to oversee a range of functions beyond defense—quite different from nato’s decision-making body, the North Atlantic Council. The oas functioned through contentious voting instead of the consensus and unanimity that characterized nato.Footnote 19 Inter-American military coordination—centered on the Inter-American Defense Board, a wartime holdover—was far less centralized than its counterpart in Europe.Footnote 20 Proposals for an integrated command structure faltered. In contrast, nato evolved greater military command integration and centralization than was first envisioned, spurred by the Soviet nuclear test, the Berlin blockade, and the Korean War. Functioning on a permanent basis after mid-1950, the North Atlantic Council coordinated large resource and troop commitments while postwar military aid in the Americas was minimal and bilateral.Footnote 21nato developed into an operationalized military alliance with a centralized headquarters; the Rio Treaty did not.
What explains the variation in institutional forms emerging from the postwar juncture? In comparing nato and Asia, Hemmer and Katzenstein discount several US-centric “universal and indeterminate explanations.” Ruggie argues that multilateralism is how US leaders enacted a particularly American hegemony,Footnote 22 but an emphasis on consistent US practices cannot account for multilateralism’s failure to emerge in Asia. Although congruent with multilateralism in the Americas, this explanation fails against the details of the case.Footnote 23 Explanations concerning US beliefs, Eisenhower’s New Look strategy, which reduced conventional overseas military commitments, and, alternatively, neighbors’ fears of resurgent enemies, are underdetermined. Nor do they explain the emergence of multilateralism in the Americas. Instead, Hemmer and Katzenstein advance an “eclectic explanation” focused on “power, threats, and identity.” Their framework remains especially useful for examining the broader literature on the emergence of regional security orderFootnote 24 because it captures liberal arguments about institutional efficiency, realist arguments about responses to threats, and constructivist arguments about identity.
Hemmer and Katzenstein’s first condition concerns the presence of cooperative great powers that help to shoulder security burdens: the United States shares authority if it expects lower costs. Similarly, Galia Press-Barnathan hypothesizes that multilateralism is more likely if power disparities between the US and its regional partners are moderate and “power disparities among the regional partners are low.”Footnote 25 Although neither Europe nor Asia could offset costs immediately after the war, US policymakers viewed several European states as once-and-future great powers. In Asia, this applied only to Japan. As such, great power status helps to account for US multilateralism in Europe versus bilateralism in Asia.Footnote 26 But US policymakers doubted any Latin American country would be a great power. This pessimism about Latin America’s great power potential and burden-sharing capacity contrasts with the creation of inter-American multilateral security institutions. Press-Barnathan’s emphasis on regional commitment to cooperation, as opposed to material capacity, fares better against the inter-American case—but it is ambiguous in Asia given divided preferences over regional security arrangements between Japan and its neighbors.Footnote 27
Hemmer and Katzenstein’s second condition highlights the response to external threats. Realist accounts often argue “a regional power assumes the burden of defending the area from external security threats.”Footnote 28 According to Hemmer and Katzenstein, higher and qualitatively different threat perceptions in Europe help to explain why multilateralism emerged there and not in Asia. (Counterintuitively, a war in Asia [Korea] spurred deeper multilateralism within nato in its first years while failing to do so in Asia itself.Footnote 29) But even in Europe, there is no “direct line from a certain type of threat (cross-border Soviet attack) to a particular institutional form (multilateralism).”Footnote 30 Kai He and Huiyun Feng add that US policymakers respond to higher levels of threat—“loss” in prospect-theory terms—by accepting greater risk via multilateralism.Footnote 31 Again, the inter-American case complicates these explanations. US policymakers perceived little threat from, within, or to Latin America in the immediate postwar period. Although threats in the Americas in 1945 to 1948 rank as the lowest of the three regions (the North Atlantic, Asia, and the Americas), multilateral security institutions were constructed, defying He and Feng’s explanation. US leaders desired Latin American cooperation—particularly access to raw materials and strategic territory—in any future conflict, but this goal was in response to recent wartime experiences, not to perceptions of an imminent threat to the region.
Hemmer and Katzenstein emphasize a third condition, “perceptions of collective identity.” US and European policymakers understood themselves as composing a common Western, Christian civilization, which meant “the Europeans could be trusted with the additional power a multilateral institution would give them.”Footnote 32 US prejudice against Asians sank attempts to replicate nato there.Footnote 33 Southeast Asia was seen first through a colonial, and then, through a Cold War lens. Strategic rationales were offered for the failed Pacific Pact and the 1954 creation of seaTO, but civilizational affinities were conspicuously absent.
Identity plays little role in explaining postwar inter-American multilateral security institutions. Racial biases characterized US views of Latin Americans, as they did for US views of Asians. Although Latin American statesmen, often descendants of Europeans, used the language of “Christian civilization,” many US policymakers viewed Latin Americans as racially inferior and childlike—a belief infamously encapsulated in George Kennan’s 1950 report on his Latin America tour.Footnote 34 These views were not universal in the State Department; Kennan was not a Latin Americanist and many in the department had greater appreciation of the countries to which they were assigned. Racial and cultural prejudice contrasted with a “Western Hemisphere idea” that identified the Americas as a “new world” of republics that interacted according to different principles.Footnote 35 But condescending paternalism dominated US policymaking beginning with Theodore Roosevelt’s corollary to the Monroe Doctrine, and through President Woodrow Wilson’s emphasis on “teaching” Latin Americans proper political culture and justified numerous US interventions in the circum-Caribbean in the decades before Franklin Roosevelt’s election as president.Footnote 36 Shedding light on the effects of this racial prejudice is a core facet of two decades of “revisionist synthesis” in historical scholarship on US-Latin American relations.Footnote 37 It surfaces in archival records in the postwar period, too. Latin Americans were rarely treated as full members of the Western civilization that, for US policymakers, united the North Atlantic.
The causes highlighted in the “no nato” debate were strikingly absent in the Americas (see Table 1), making it puzzling that multilateral security institutions emerged in the region. I argue that instead, shared historical antecedents of regionalism played a key role in the creation of these new institutions. Alone, such antecedents are not a sufficient cause, but they are one productive cause in the context of a permissive critical juncture, as discussed below. Despite those antecedents, the eventual outcome was not predetermined: US commitment to inter-American institutions was in doubt from the 1944 Washington Conversations on International Peace and Security Organization (Dumbarton Oaks Conference) to the 1945 United Nations Conference on International Organization (San Francisco Conference) and beyond. But from this critical juncture, a collective security treaty and multilateral organization emerged. Burden-sharing, external threat, and collective identity were not clear drivers of US policy, nor were they the crucial rationale offered by Latin Americans, illustrating how antecedents favor multilateral institutions even in the absence of other commonly emphasized conditions. Diplomats and policymakers turned to the importance of prewar antecedents to build multilateral institutions.
Table 1 Explanations of Postwar Security Arrangements
Historical Trajectories and Regional Orders
One study of institutional gestation in the Americas concludes that “legacies matter.”Footnote 38 Indeed, many accounts of regional institutional formation acknowledge the importance of history but incorporate it into their explanations only in idiosyncratic ways that sometimes overlap with identity. For example, comparative regionalists Acharya and Johnston mention two types of historical influence on institutional design: historical memory and path dependence.Footnote 39 Both are forces for continuity. Historical memory is closely related to identity, evinced in suggestions of an “asean way” that favors certain institutional forms.Footnote 40 To these, one could add gradualism and accretion, in which institutional structures inexorably build over time. Kahler and MacIntyre compare such treatments of history to “crude diagrams of the ascent of man.”Footnote 41
Building on recent developments in historical institutionalism, my explanation advances existing treatments of history’s role in the formation of regional institutions by showing how certain antecedents gain causal weight in critical junctures. Hi’s approach to explanation bridges Ir’s salient rationalist-constructivist paradigmatic divide,Footnote 42 which is also reflected in Hemmer and Katzenstein’s eclectic explanation. The more precise treatment of history in this article allows historical antecedents to be distinguished from regional identities, as emphasized by Hemmer and Katzenstein and ir constructivists generally. In the inter-American case, antecedents do not covary with identity and can be analyzed separately.
hi was long known for an analytical model in which relative stasis is punctuated by brief periods of rapid institutional change.Footnote 43 Scholarship initially emphasized mechanisms, especially path dependence,Footnote 44 which favor institutional continuity after such critical junctures. In an hi account of European security institutions, Arnand Menon notes this division between exogenous shocks in creation moments followed by long periods of stability.Footnote 45 Although hi has developed explanations for gradual and incremental change,Footnote 46 its accounts of rapid change continue to emphasize exogenous shocks and critical junctures.Footnote 47 Most famously, John Ikenberry argues that in the critical junctures that follow major wars, the combination of state capacities and interests shapes lasting postwar orders.Footnote 48 Giovanni Capoccia and Daniel Kelemen define critical junctures as “relatively short periods of time during which there is a substantially heightened probability that agents’ choices will affect the outcome of interest.”Footnote 49 Although there is debate in hi about the nature of critical junctures and how to identify them, I adopt the contingency-focused approach because it allows scholars to examine “near misses” of institutional creation or reform, counterfactuals, and choices not taken.Footnote 50 The post-WWII period is one such critical juncture, allowing for the comparison of three divergent regional outcomes. Divergent institutional forms are possible because in such critical junctures, loosened structural constraints “increase the causal power of agency.”Footnote 51
Building on this view of critical junctions, I integrate Dan Slater and Erica Simmons’ concept of “critical antecedents” with hi mechanisms for incremental change that have largely been examined outside critical junctures.Footnote 52 Critical antecedents are “factors or conditions preceding a critical juncture that combine with causal forces during a critical juncture to produce long-term divergence in outcomes.”Footnote 53 Although Slater and Simmons argue these antecedents may affect causal mechanisms, they do not specify the types of mechanisms one expect. But elsewhere, hi describes several mechanisms for incremental change—drift, conversion, layering, and exhaustion.Footnote 54 I argue that similar mechanisms, especially layering and conversion, can function within critical junctures, causally connecting antecedents to later institutional development. Layering describes a process in which “new elements are attached to existing institutions and so gradually change their status and structure”; conversion refers to a transformation of existing institutions for new purposes.Footnote 55 Slater and Simmons suggest antecedents can shape actors’ range of choices during critical junctures;Footnote 56 I agree and add that antecedents can also shape mechanisms within critical junctures. The creation of regional security institutions during the postwar critical juncture illustrates how these mechanisms operate. As Ikenberry demonstrates, systemic wars produce rare opportunities for major changes in international order.Footnote 57 Even then, new order is not created ex nihilo. Bits may be novel, but diplomats and policymakers construct new institutional structures atop shared antecedents through layering and conversion.
That said, antecedents do not all matter equally or produce multilateral outcomes in every critical juncture. As with hi generally, “individual causal factors normally must be analyzed as part of larger combinations” with attention to contextual conditions and to interactions of causal factors.Footnote 58 One can consider these factors in terms of “permissive” and “productive” causes.Footnote 59 The presence of permissive causes produces the contingency of a critical juncture, while productive causes lead to changes within that juncture. The emergence of any given multilateral institution will have multiple and complex causes. The postwar Asian, inter-American, and North Atlantic cases shared permissive conditions, but productive causes varied. Burden-sharing, threat, and common identity shaped nato’s creation but were not necessary conditions for security multilateralism everywhere. Building on Ikenberry, I emphasize the interaction of antecedents with preference compatibility among secondary states and the leading state. Going beyond interest-based accounts that stress the importance of positive feedback for expanding institutions while treating their origins as exogenous,Footnote 60 I offer a stronger account of institutional emergence by specifying permissive conditions and highlighting the interaction of antecedents with other productive causes.
Orfeo Fioretos notes “external developments, including major crises, may cause groups in some countries to experience preference transformations over national designs if they do not value historic institutions highly, while in other countries the same events may serve to strengthen preferences for extant designs if groups value historic institutions.”Footnote 61 When actors value historic institutions during critical junctures, antecedents of regionalism make multilateral outcomes more likely; “antecedent conditions define the range of institutional alternatives available to decision makers [in a critical juncture] but do not determine the alternative chosen.”Footnote 62 As Manuela Moschella and Eleni Tsingou illustrate regarding the 2008-2009 financial crisis, antecedents may favor less radical change in moments of crisis.Footnote 63 In other cases, the permissive conditions of the critical juncture facilitate the consolidation of inchoate antecedent practices and designs. For that to happen, agents must value antecedents. The nature of the antecedents matters for institutional form. I refer to antecedents that are shared and relate to the construction of a region. Shared antecedents include prior events, designs, and practices that involve a critical mass of the actors relevant to the critical juncture. They include incipient cooperation among these actors, such as efforts to ameliorate regional conflicts through treaties or mediation; creation of mechanisms for peaceful resolution or arbitration of disputes; joint actions to address transnational problems; and the creation of venues, organizations, or networks between and beyond state actors.
My explanation can be tested against outcomes and processes. Strong “diagnostic evidence”Footnote 64 would connect the presence of antecedents throughout a causal chain, from actors’ preferences through mechanisms (layering and conversion), to the emergence of multilateral institutions. Evidence should show that actors base claims and expectations on antecedents that may be valued because they represent previous investments that constrain actors’ choices and because they shape perceptions of future costs and benefits; layering and converting antecedents should be justified as a more efficient and secure solution than starting anew. As is often true in process tracing, the role of antecedents must be untangled from competing explanations against the case narrative.Footnote 65 If actors respond to new threats and to a new distribution of capabilities with new designs—unconstrained by antecedents—it would support rationalist explanations linked to threat and institutional efficiency. If designs are shaped more by shared identities and the trust they foster than by antecedents of regionalism, it would support constructivist explanations. I turn to the inter-American case to explore these explanations in more detail.
The Reformation of the Inter-American System
Although nato is considered the multilateral security institution par excellence, the 1947 Rio Treaty preceded it; in fact, the State Department cited the Rio Treaty as a model during nato’s creation.Footnote 66 The initial agreements impressed similar obligations upon their signatories, although eventually nato became more prominent and centralized and received greater resources. Comparatively, ir scholarship gives little attention to inter-American multilateral security institutions. The prevailing historical treatment—drawn on US sources and focused on US decisions—emphasizes the solidification of US dominance instead of institutional emergence and development.Footnote 67 Regarding the 1947 conference in Rio and the 1948 Ninth International Conference of American States (Bogota Conference), where the oas was created, Lester Langley writes, “The United States generally got what it wanted; Latin Americans did not.”Footnote 68 Brazilian historian Gerson Moura concludes that the new regional order “was no more than the juridical and political framework for irreversible United States hegemony over the continent.”Footnote 69 Drawing on research in Brazilian, Chilean, Colombian, Mexican, and US archives, I argue that these interpretations overstate US dominance and read later Cold War trends backward into the system’s creation. Despite clear US material primacy and some Cold War concerns, the multilateral security arrangements of the late 1940s imperfectly reflected the goals of Latin American leaders who sought to expand and convert prewar antecedents into multilateral institutions for the postwar world.
Antecedents of Inter-American Multilateralism
Inter-American postwar collective security was ensconced in a multilateral system centered on the Rio Treaty (also known as the Inter-American Treaty of Reciprocal Assistance, or tiar) and the oas. This system converted and was layered over prewar institutions and practices, including the Pan-American Union (pau) and Pan-American conferences. It drew on earlier norms and practices of Latin American international jurisprudence, sovereign equality and nonintervention, and inter-American consultation, as well as on the region’s global engagement at the 1907 Hague conference and in the League of Nations.Footnote 70 US support for greater Latin American participation at The Hague emerged in part from connections between US Secretary of State Elihu Root and his Latin American counterparts, who emphasized affinities as fellow republics with a shared dedication to advancing international law.Footnote 71 These antecedents emerged in the context of US expansionism and Latin American ambivalence about US power. In 1889, the United States established a precedent of semi-regular meetings of American foreign ministers and a commercial office in Washington, but interest was muted. As US power became clear following the 1898 Spanish-American War, inter-American relations gained salience. In the early twentieth century, US occupations of Nicaragua (1912-1933), Haiti (1915-1934), and the Dominican Republic (1916-1924), as well as shorter interventions, particularly in Mexico (1914, 1916-1917) sparked local resistance movements that enjoyed substantial regional sympathy.Footnote 72 Latin American diplomats sought prohibitions on intervention, an end to gunboat diplomacy, and recognition of sovereign equality.Footnote 73 Intellectuals, including José Martí of Cuba and Rubén Darío of Nicaragua, rejected US pretensions to leadership.Footnote 74 But opposition was far from uniform, and many Latin Americans envisaged the benefits of cooperation with a more restrained northern colossus. This tension shaped early regional practices, including eight Pan-American conferences between 1889 and 1938; the pau; joint mediation of conflicts, including the Argentine, Brazilian, and Chilean mediation between the US and Mexico in 1914; the 1933 Anti-War Treaty of Nonaggression and Conciliation, signed in South America; and influential inter-American legal networks.
Through incremental layering and conversion, the Washington D.C. based commercial office of the 1890s referenced above, became the pau— a multifaceted international organization—and in 1923 the American states agreed to appoint permanent representatives to its governing board.Footnote 75 Despite strong US influence, the pau and Pan-American conferences created customs of regional consultation. Through early twentieth-century Pan-Americanism, “Latin American states sought to bind the United States while maintaining their own freedom of action.”Footnote 76 Although always somewhat lax, over time these institutions became more formal and multilateral. A 1933 study concluded that the pau acted as “an independent international administrative organ of a more or less permanent character based on the principle of equality of American states.”Footnote 77 At Pan-American summits, Latin American diplomats stridently questioned US unilateralism without renouncing the benefits of engagement. The summits drew US presidential visits in 1928 and 1933 and spurred “binding mechanisms”Footnote 78 in the form of a nonintervention declaration, the crown jewel of FDR’s Good Neighbor Policy. Although disparate and fragmentary, the Pan-American system allowed Latin America to consult, push, pressure, and sometimes oppose the United States during the 1920s and 1930s.
Although US influence grew during the first decades of the twentieth century, pre-WWII US predominance is often overstated, geographically and temporally. In South America, it was contested by European powers and often resisted by larger South American republics. Europe, including the Axis powers, provided South America with markets and military partners during the interwar period. World War II’s geostrategic and economic realities—curtailed transatlantic shipping; US overseas basing, including major deployments to northeast Brazil; and US wartime expansion—increased US influence at the expense of Europeans far beyond what World War I had done. Inter-American diplomacy reflected this. In 1940, diplomats concurred that an attack by a “non-American state” against the Americas “shall be considered as an act of aggression.”Footnote 79 At the Third Meeting of the Foreign Ministers of the American Republics, held in Rio de Janeiro six weeks after the Pearl Harbor attack, the ministers declared, “solidarity must be translated into facts.”Footnote 80 The foreign ministers’ meeting veered into internal security with restrictions on Axis nationals and firms.Footnote 81 Latin American states were valuable supporters of Allied efforts. By late January 1942, most—with the important exceptions of Argentina and Chile—had broken relations with the Axis powers. The region’s most significant roles were economic and geostrategic. Latin America supplied critical wartime commodities (food, oil, copper, tin, bauxite, rubber) at controlled prices. US bases on Latin American territory were crucial for moving troops and matériel, patrolling sea lanes, and hunting German U-boats. Inter-American coordination occurred through meetings of foreign ministers, regular diplomatic consultation, and via the pau. This coordination included postwar planning beginning in mid-1943.Footnote 82 After the war, these antecedents favored multilateral security institutions, although they did not guarantee them.
World War II also shaped domestic politics in many Latin American countries. Democratization flowered, or at least repression diminished, from roughly 1944 to 1947. As Latin American leaders shaped inter-American arrangements, they also responded to internal pressures, including through a regional debate over squaring support for democratization with principles of sovereignty and nonintervention.Footnote 83 By late 1947, however, traditional conservative and military elites were ascendant while the left faced renewed repression. Democracy receded in several countries.Footnote 84 Nevertheless the pre-1947 center-left and the ensuing conservative wave both favored cooperation with the United States. Roosevelt’s New Deal and Good Neighbor policies, alongside proclaimed Allied goals, resounded with Latin American democrats during the war and in its immediate aftermath. Anticommunist conservatives—especially influential militaries—also favored the United States. Latin American militaries had grown more invested in security regionalism through training and Lend-Lease transfers, creating expectations of further benefits.Footnote 85 Governments of different domestic stripes held similar inter-American security preferences, especially while global Cold War tensions remained a secondary concern.
From War to Postwar: A Critical Juncture
World War II upended global order, creating a critical juncture in which new arrangements quickly took shape. Postwar planning in the United States began seriously in 1943 and some Latin American governments began to formulate plans for international order by the end of that year.Footnote 86 At the Bretton Woods Conference in July 1944, Latin American states pushed for developmentalist content that drew on New Deal and inter-American antecedents. Eric Helleiner stresses how US-Latin American cooperation in response to the Great Depression informed US policymakers’ views and plans for global institutions. Earlier cooperation, including a frustrated initiative for a regional development bank, created space for Latin Americans to pursue their economic priorities.Footnote 87 In security and diplomacy, Latin Americans almost universally favored a strong postwar regional system embedded in international law. The US government was divided about how to balance regional and global levels in the postwar order. Under Secretary of State Sumner Welles and Assistant Secretary of State Nelson Rockefeller favored regional multilateralism, though their position weakened when Welles resigned in late 1943, wounded by administration infighting and the threat of personal scandal. Special Assistant to the Secretary of State Leo Pasvolsky and Secretary of State Edward Stettinius advocated predominant global security arrangements. The US debate pitted advocates of the Monroe Doctrine (in traditional or multilateralized form) against those who feared that an autonomous inter-American system would legitimate European imperial preferences and Soviet claims in Eastern Europe.
Under Pasvolsky’s guidance, the Dumbarton Oaks plan, largely drafted between February and October 1944 with minimal consultation beyond Great Britain, the Soviet Union, and the United States, threatened to obviate the inter-American system in favor of a dominant global body. The lack of consultation and the plan itself frustrated Latin Americans and demonstrated the value of a privileged, regional forum with the United States. Brazilian Ambassador Carlos Martins, supported by Latin American counterparts, protested the violation of inter-American norms of consultation in the preparation of the Dumbarton Oaks plans. Their demands spurred a series of briefings between Latin American diplomats and US officials in Washington. Shortly after the Dumbarton Oaks plan was circulated in October 1944, Mexican Secretary of Foreign Affairs Ezequiel Padilla called for an international summit to consider how the new world organization envisioned at Dumbarton Oaks would affect “the economic unity of the [American] Continent and its contributions for a permanent peace in the world.”Footnote 88 Padilla’s call resounded with established practices for hemispheric diplomatic conferences, converting that antecedent into a voice opportunity for Latin Americans. The United States agreed to a consultative meeting, but sought to delay “actual changes in the structure of the inter-American system … [until] after the world security organization is more fully developed.”Footnote 89 Still, Latin Americans succeeded in scheduling an inter-American meeting in Mexico City before the San Francisco conference to avoid a fait accompli.
Although seen as a pro-US figure, Padilla emerged as a leading regionalist in the Mexico City and San Francisco conclaves. The Mexican response to the Dumbarton Oaks draft, sent on October 31, 1944, highlighted declarations from the seventh and eighth International Conferences of American States and inter-American jurisprudence as a foundation for any future world order. No antecedent was more important than inter-American nonintervention, in which Latin America had invested much and from which it expected continuing benefits: “This principle, the cornerstone of the Inter-American System, deserves to be in the foreground among those which the New World can offer as a contribution of its own to the formation of the International Organization that is created.” Mexico sought to convert nonintervention from a negative guarantee to a positive principle of international organization. Mexico embedded its claims in the “same ideas” Roosevelt had praised earlier that month, seeking to entrap the United States in the president’s proregionalist rhetoric.Footnote 90 Both strategies would eventually bear fruit.
Brazil’s aspirations to a central, global role fostered some ambivalence in its policy on the conjunction of regional and global orders. In mid-1944, Roosevelt considered Brazil for a permanent UN Security Council seat, largely to bolster geographic representativeness. But Brazil was not consulted about its own possible inclusion, a clear signal that the US did not seriously consider Brazil to be a great power. In fact, Brazil harbored doubts about the Dumbarton Oaks plan, especially the proposed permanent-member veto. When Brazilian President Getúlio Vargas decided to pursue a permanent Security Council seat, the idea divided the Brazilian foreign ministry. Many senior diplomats had been chastened by Brazil’s withdrawal from the League of Nations in 1926, when Germany was given permanent seat and Brazil was shunned. This experience shaped Brazilian preferences for a multilateral regional system with precedence over global arrangements that privileged great powers. Brazil’s wartime Foreign Minister Oswaldo Aranha emphasized Brazil’s privileged role as Pan-American champion and interlocutor between the US and Spanish America in an autonomous inter-American system.Footnote 91
The inter-American conference convened at Chapultepec Castle in Mexico City in early 1945, with an agenda focused on the plans for world and regional institutions. US preparations were spearheaded by Rockefeller and Under Secretary of State Joseph Grew, whose roles in antecedent Pan-American cooperation laid the foundation for a cross-national, pro-regional coalition.Footnote 92 Brazil, Mexico, and most Latin American states emphasized preferences for a strong and relatively independent regional security system layered over Pan-American institutions. In preparation for Chapultepec, Brazil’s position was that “the Security Council of the future world organization will never intervene, except in the rarest exceptions, in our hemisphere…. The decisions of the Security Council would be executed in the American continent by its own states.”Footnote 93 The Brazilian proposal converted principles of nonintervention to preserve the autonomy of the inter-American system. Nonintervention was central to what Ruggie calls “principles of ordering relations” among states. Small states were great advocates, too. Uruguay argued that the regional agenda should include “a pact for mutual guarantee of political independence and territorial integrity of the American nations, complemented by the elements that exist in that sense in American Law and their due coordination with the system of world security.”Footnote 94 Antecedents shaped the range of choices in the post-war juncture and molded actors’ preferences. The great power centrism of Dumbarton Oaks and Latin America’s uneven experiences with the League of Nations suggested that Latin Americans would have less influence on global decisions than they would at the regional level. The Latin American proposals framed prewar regional antecedents as successes to build upon, linking arguments for regional precedence, nonintervention, and greater multilateralism.
Chapultepec: An Inter-American High Point
The meeting at Chapultepec was a result of Latin American concerns that the inter-American system was losing precedence as the postwar order took shape, in contrast with its growth during previous decades.Footnote 95 The Latin Americans argued for a rejuvenated inter-American system characterized by greater prominence and multilateralism. Brazil argued, perhaps wishfully, that the Dumbarton Oaks draft “foresees and therefore gives prestige to regional understandings, of which none has the prestige, constancy, organic nature, or permanent juridical content of the Inter-American system.” The Brazilian proposal noted the Treaty of Versailles’ reference to the Monroe Doctrine and “implicitly this [inter-American] system.” In a concrete proposal for layering, Brazil argued that American countries should create a permanent inter-American commission of foreign ministers to “oversee the good execution of the conventions, agreements, or treaties elaborated by various Pan-American conferences.” The Brazilians suggested building on and combining two existing institutions: Pan-American diplomatic meetings and the consultative Governing Board of the Pan-American Union. Brazil’s proposed commission would have prerogative over inter-American matters, reducing UN influence and enhancing the Latin American voice.Footnote 96 Speaking to assembled delegates, Brazilian Foreign Minister Pedro Leão Velloso stressed the evolutionary progression of the inter-American “machinery” that stood ready to help “safeguard the peace and security of the world” in cooperation with the United Nations, converting regional agreements into a multilateral component of global order.Footnote 97 Brazilian preferences for autonomous and multilateral regional security institutions were shaped by investments in past institutions and expectations of continued returns. Others shared this assessment. Latin American diplomats noted that the global system was an uncertain experiment—an allusion to the failed League of Nations—while the inter-American system was a bona fide success. Paraguay argued, “the Pan-American system is so real and concrete that no American State could renounce its benefits, which have been achieved in over fifty years of joint efforts.”Footnote 98 Antecedent institutions had emerged gradually; now, the Latin Americans argued, they should be transformed and expanded—not discarded in favor of new, uncertain, global institutions.
The Act of Chapultepec, adopted by all American nations except Argentina on March 6, 1945,Footnote 99 enshrined principles that favored multilateralism: equality, nonintervention, and collective security. It also called for building upon prior arrangements with a “treaty establishing procedures whereby such threats or acts may be met” by measures ranging from breaking diplomatic relations to economic sanctions to the “use of armed force to prevent or repeal aggression.”Footnote 100 In addition, it determined that decisions would be taken through the existing pau governing board, converting it into a political body and raising its profile. As the meeting adjourned, several Latin American states “declared expressly that the Pan-American system is compatible with the aims, purposes, and objectives of the Organization of the United Nations, and that, in consequence, it shall continue functioning autonomously.”Footnote 101 The foundations of postwar multilateral security institutions emerged at Chapultepec as the Axis threat waned and before concerns about a Soviet threat in the Americas emerged. Multilateralism did not hinge on US expectations of great power status or shared burdens; references to identity were situated in antecedents of Pan-Americanism.
San Francisco: Regional Visions in Doubt
If Chapultepec demonstrated inter-American agreement, the conference in San Francisco illustrated the postwar juncture’s contingency. Latin Americans left Chapultepec believing that the United States would continue to expand its commitments to diplomatic consultation, nonintervention, regional defense, and peaceful dispute settlement.Footnote 102 The United States expected strong Latin American support for its positions on the new United Nations organization in San Francisco. But the comity was disrupted, first over Latin American insistence on including Argentina in the meetings. Despite domestic discontent about the inclusion of an Argentine government often depicted in the US media as an Axis sympathizer, the US delegation relented to Latin American pressure. More serious was the renewed clash over regional versus global predominance in security governance. US advocates of global institutional preeminence dominated the San Francisco preparations. Pasvolsky’s plan granted UN dominance over regional bodies with “the nations in the inter-American system obligated to seek permission and risk a veto before they could take any defensive action.”Footnote 103 Roosevelt’s death just two weeks before the opening of the conference, along with Secretary of State Cordell Hull’s earlier illness and resignation, exacerbated the uncertainty.
As the conference began, the new Truman administration—populated largely with holdovers from Roosevelt’s government—was divided over global versus regional security arrangements and whether the US should multilateralize its commitments in the Western Hemisphere. Many US defense figures wanted to preserve unilateral freedom of action. Secretary of War Henry Stimson complained, “The Dumbarton proposals have practically wiped out the unilateral character of the Monroe Doctrine and places our use of the Monroe Doctrine, in case of enforcement by arms, at the mercy of getting the assent of the Security Council to be created at San Francisco.” Stimson recognized the weight of Pan-American antecedents, noting that agreements from “various conferences with South American Republics during the past twelve years” under the Good Neighbor Policy “put serious obstacles in the path of the [unilateral] exercise of the Monroe Doctrine.”Footnote 104 General George Marshall and Navy Secretary James Forrestal also favored reinforcing US unilateral prerogatives in Latin America.Footnote 105 Ultimately, US attachment to the Monroe Doctrine provided impetus for a grand bargain with Latin America.
In response to such concerns, Pasvolsky argued that the Security Council veto preserved the Monroe Doctrine and US unilateral rights, making a regional clause unnecessary. He rejected Latin American pressures for regional security multilateralism as inimical to global order: “If we open up the Dumbarton Oaks Proposals to allow for regional enforcement action on a collective basis, the world organization is finished.” Pasvolsky sought to escape such antecedents. Despite the Act of Chapultepec, he argued that the US was “not now obligated permanently” by its regional commitments to Latin America.Footnote 106 The US delegation worried that other American states could take actions under Chapultepec without US approval. Stettinius agreed: “While recognizing the strategic importance of inter-American solidarity, we should not allow ourselves to be compelled to adopt a position contrary to our own view of national interest.”Footnote 107 Despite defense concerns, the global institutional advocates initially dominated the US discussions.
But antecedents shaped the preferences of other actors who emphasized the perceived benefits of layering and converting the existing institutional architecture to more multilateral forms. Countering Pasvolsky, the Latin Americans wanted to preserve the regional system and threatened a walkout from San Francisco. Latin American diplomats then exploited US divisions by appealing to advocates of Pan-American institutions.Footnote 108 They contacted Rockefeller and a bedridden Hull, and pressed US Senator Arthur Vandenburg, a delegation member, to include a reference to the Chapultepec accords in the San Francisco treaty.Footnote 109 Colombian Foreign Minister Alberto Lleras Camargo (who became president in August 1945) and Cuba’s formidable Ambassador Guillermo Belt demanded a public presidential declaration of support for the inter-American system as embodied in Chapultepec and for holding conferences to formalize antecedents of inter-American cooperation.Footnote 110 To bridge the divide, US delegate Harold Stassen proposed a “collective self-defense” approach to link the global organization and regional pacts, formulated as an “inherent right of self-defense, either individual or collective.” The response to Latin American pressure led to UN Article 51.Footnote 111
Latin Americans sought to preserve and to formalize hemispheric diplomatic consultation in a context where US interests were suddenly global. Pro-regional US diplomats also expected continued returns from inter-American cooperation. US Ambassador in Brazil Adolf Berle noted—and opposed—in October 1945 “one current of opinion which by and large would like to end the regional agreement which has served us well.”Footnote 112 Inter-American solidarity had extended US influence and dampened opposition. A memorandum to President Truman in late 1945 noted Latin American concerns that the new administration, “[did] not realize that continental solidarity is essential from the political, human, and economic point of every country in the western hemisphere, including the United States.” But praising solidarity was distinct from believing the United States and Latin American shared a collective identity, a view few US policymakers held. The same memorandum evinced condescending essentialism that casts doubts on a collective identity explanation, saying that public statements “would be very welcome by the Latins. They are a touchy and emotional people who like to be catered to and patted on the back.”Footnote 113 The language, common in the period, echoed long-standing tropes in US foreign policymaking about Latin American distinctiveness and inferiority. The subordinate treatment was perceived by Latin Americans, too. Brazilian delegate Bertha Lutz, previously pro-US in orientation, left the San Francisco conference stung by “the second-class status of being a ‘Latin American.’” In May 1945, she privately wrote of “being patronized” by the US and British delegates and complained of treatment “so inconsiderate of the Latin Americans that they are all feeling humiliated and sore.” Certain ties of history bound the United States and Latin America, but for most US foreign policy elites, race, religion, and social development created distinct identities.
Collective Security and the Road to Rio
When the United States waffled on its regional commitments at San Francisco, Latin Americans pressured Truman for a conference to convert antecedent defense agreements into formal, multilateral institutions. This pressure eventually led to the Inter-American Conference for the Maintenance of Continental Peace and Security, held in Petrópolis, outside Rio de Janeiro, in August and September 1947. Later seen as the encapsulation of US dominance, the resulting Rio Treaty emerged from a Latin American diplomatic victory.
The Rio conference was originally scheduled to be held shortly after San Francisco, but worsening US relations with Argentina’s President Juan Perón, including State Department attempts to sway Argentinian elections by tying Perón to fascists, postponed the conference. Brazil initially endorsed the delay, as did the pau, agreeing that Argentina had not met inter-American commitments and therefore could not join a conference based on them.Footnote 114 But soon Latin American ambassadors noted “rumors” that the US wanted to drop the regional defense pact.Footnote 115 By April 1946, Brazil began highlighting “the just aspirations of the American Republics to conclude the Inter-American Treaty of Mutual Assistance in the briefest possible period and without fracturing continental unity.”Footnote 116 During the delay, the US military advocated for limited defense burden-sharing: a continental plan for weapons standardization and “for granting of military rights, facilities, etc. whenever military enforcement action was necessary.”Footnote 117 The plan failed, however, as many in Latin America and at the State Department rejected arms spending as needless and dangerous.Footnote 118 This rejection highlights the secondary importance of burden-sharing and the common perception of low external threats.
Brazil’s first draft of the Rio Treaty in September 1945 advanced a strong version of inter-American security multilateralism, making responsive measures largely obligatory and requiring signatories to facilitate and aid the passage of collective military forces.Footnote 119 Brazilians argued against replicating the top-heavy Security Council in the Americas because it would be “contrary to the traditional bases of the Inter-American System.”Footnote 120 Instead, Brazil favored broad-based multilateral institutions patterned after inter-American antecedents. Conversely, US planning for the conference in Rio, initially chaired by Pasvolsky, emphasized that inter-American responses would be “executed subject to” the Security Council.Footnote 121 UN Article 51 had not ensured an autonomous regional security system.
With the meeting in sight, Brazil successfully limited the agenda to a security treaty—with economic and other proposals delayed until a subsequent conference in Bogota.Footnote 122 One proposal offered for the Rio agenda, but soon postponed, was Mexican Foreign Minister Jaime Torres Bodet’s call for an inter-American “constitutive charter” in April 1947. Torres Bodet noted the pau’s history as “the oldest regional body” and the need to modernize “a series of conventions and resolutions [that] can be difficult to consult and are at times of uncertain contractual value.”Footnote 123 By the end of June, eighteen American republics expressed agreement on the need for a clearer, multilateral structure building on previous institutions.Footnote 124 Dispute resolution was paramount for Mexico, which sought to layer over existing arbitration practices to create security institutions to curtail inter-American aggression.Footnote 125 Shortly after the Rio Treaty’s approval, Torres Bodet told a US audience that the pact would carry out “a noble tradition of continental solidarity and common defense of the American peoples …as old as the independence of our Republics.”Footnote 126 He sought to build on that accomplishment with his proposed inter-American charter.
Ultimately, negotiation of the collective security agreement went smoothly, in part because US preferences had grown more compatible with what Latin Americans had advocated since 1944—greater regional autonomy. US-Soviet competition played a role in changing US preferences, but antecedents structured the range of available, mutually acceptable choices. By 1947, the United States had reduced its attachment to the Security Council’s centrality and sought explicit inter-American commitments, as opposed to only consultations, to “assist” in response to an attack.Footnote 127 An attack “imposes common obligations immediately,” the US opined—a more expansive interpretation of Article 51.Footnote 128 The US left the definition of aggression open to include “subversion or political attack,” in a nod to regional concerns about communism and wartime fears of fifth-column fascists.Footnote 129 In Latin America, the United States was more concerned about neofascism in the form of Argentina’s Perón than about a Soviet-backed revolution; Latin Americans highlighted communist threats more often than their US counterparts in diplomatic conversations in 1947.Footnote 130
Despite preconference concerns that “the treaty will not bind the United States without its consent,”Footnote 131 the United States accepted a binding two-thirds vote in the event of aggression. Secretary of State Marshall internally cited the US “desire to make the treaty as effective as possible” as the reason for bowing to Latin American preferencesFootnote 132 expressed in the Mexican plan presented to the pau.Footnote 133 The Latin American majority preference for a binding solution that obligated the US without a veto converted antecedents of sovereign equality and nonintervention to a new postwar context. Nineteen American nations adopted the collective defense formula on August 30, 1947, a victory for regional arrangements under the UN framework and a precedent for nato.Footnote 134
An Inter-American Constitution
The Ninth International Conference of American States, held in Bogota from March 30 through May 2, 1948, was the culmination of efforts to formalize and multilateralize antecedents. In the period 1944 to 1948, both the global and Latin American domestic political context had changed. The breakdown of US-Soviet cooperation fed US concerns about international communism, although US policymakers still considered Latin America a secure and secondary zone. In many Latin American countries, domestic politics took a rightward and authoritarian turn. However, the change in inter-American security preferences was minimal; for Latin America, surging anticommunism was a return to the historical norm. In fact, the Bogota conference suffered disruption when Jorge Eliécer Gaitán, a populist Colombian presidential candidate, was murdered near the conference site; riots erupted and some delegation members and their families fled, and Colombia veered into violence and turmoil. The expansive violence would spur a military takeover in 1953. Even so, cia concerns about communist activities at the summit did not include fears of a leftist uprisingFootnote 135 and ultimately, the events in Bogota did not alter the outcome of the proceedings.
The Bogota conference focused on inter-American political and legal instruments that included and went beyond security governance. Plans for an inter-American constitutional order originated with Mexico’s concerns about uncertain US commitment. Torres Bodet remained its most forceful advocate, emphasizing the need to build “on all of the juridical elements of Pan-Americanism that have been defined and refined in the light of experience.”Footnote 136 Without pressing external threats and despite a limited ability for burden-sharing,Footnote 137 US and Latin American leaders sought to increase and institutionalize inter-American cooperation, which had been ad hoc before the war. Mechanisms of layering and conversion produced rapid changes in this critical juncture.
Brazil and Mexico sought a stronger, more independent inter-American secretariat in which Latin Americans would have a greater voice. When Chile proposed a rotating council, Brazil invoked regional antecedents to oppose it because “it would lack the tradition, unity, means of information, etc.” Mexico concurred, saying it “had not come to the conference with the motive of retreating from the conquests reached by Pan-Americanism.”Footnote 138 Brazilian and Mexican preferences were shaped by the perceived advantages of layering on antecedents. At Latin America’s urging, and particularly Mexico’s, the conference drew on antecedents for a Declaration of the Rights and Duties of American States that averred “faith in the principal of continental solidarity and proclaim[ed] their unshakeable loyalty to the Inter-American system.”Footnote 139 The conference recreated previous institutions under the oas structure and passed a treaty on dispute settlement as the peaceful equivalent to Rio’s security alliance.Footnote 140 The new institutions strengthened legal restraints on unilateral intervention while keeping the United States engaged in the region.
Antecedents shaped the regional security arrangements that emerged from the postwar juncture, especially in the Americas, where external threat and burden-sharing were notably absent. By shaping the range of choices and preferences, antecedents helped to produce multilateral security institutions in the region. Earlier experiences led Latin Americans to expect increasing returns from institutionalized consultation and decision-making—with policy feedback shaping layering. For the United States, building on the antecedents that had ensured cooperation in World War II seemed a prudent course. Layering on and converting existing institutions were perceived as entailing lower setup costs and offering greater certainty while limiting the range of compatible choices. Actors who had been engaged in antecedent cooperation formed coalitions to defend and expand it; they presented regional multilateralism as less costly and less uncertain than alternative arrangements, such as US unilateralism or solely global security institutions. The Pan-American Union, which was itself a site of conversion and layering in the new order, aided coordination and helped to address contentious situations like Argentina’s participation. Although antecedents could be instrumentalized in a form of rhetorical entrapment, they also shaped preferences, choices, and outcomes in the permissive conditions of the critical juncture. The often incremental processes of layering and conversion took on new importance in the rapid restructuring of the inter-American system during the critical juncture, as diplomats reshaped previous antecedents for new purposes and a new context.
Conclusions
This article examines the emergence of multilateral security institutions in the Americas after World War II despite the absence of causes highlighted in the literature. I empirically expand the “no nato in Asia” debate and challenge its most influential conclusions. My explanation builds on and contributes to hi scholarship by examining how antecedents shape institutional outcomes during critical junctures. In so doing, my work provides new tools for understanding the emergence of regional institutions, highlighting how processes associated with incremental change—conversion and layering—occur during critical junctures. These tools facilitate a growing dialogue between hi and ir. The value of hi for explaining key phenomena in international relations goes beyond illustrating how path dependence shapes the continuity of international institutions; it also can help to explain processes of change. I particularly highlight how better understanding the dynamics of critical junctures and mechanisms of change allows scholars to systematize the treatment of history in their explanations of the emergence and evolution of international institutions.
A full exploration of the creation of nato and the failure of the proposed Pacific Pact and seaTO is beyond this article’s scope and is well covered in the literature.Footnote 141 But it is worth mentioning the three regions’ distinct antecedents as they relate to divergence in the form of regional security arrangements. There is broad agreement on the absence of multilateral security arrangements in Asia during the postwar period. If identity and history are competing, the absence of US-Asian security multilateralism is congruent with both factors (see Table 1). In the inter-American system, however, identity and history diverge. Although US policymaker biases toward Asians that Hemmer and Katzenstein emphasize find echoes in the paternalistic, patronizing, and racist views of Latin Americans in the 1940s, historical antecedents were on Latin America’s side. US relations with Asia lacked the earlier structures and practices that gradually characterized US-Latin American relations beginning at least with the 1889 First International Conference of American States in Washington. In the mid-twentieth century, actors perceived the recreation of earlier Pan-American arrangements as beneficial, which shaped their preferences to maintain and expand them. US historical ties with Asia were more limited and did not include substantial, shared historical antecedents with a region heavily shaped by European colonialism. Asian diplomats could not appeal to antecedents, nor did they benefit from perceptions of shared identity. Bilateral and initially ad hoc models dominated in that region.
For nato’s emergence, both identity and antecedents support multilateralism; further study is needed to untangle the two in causal processes. At a glance, antecedents appear to be important. According to one historian, nato’s “antecedents can be traced back to the Knox Resolution and to the abortive French security treaty of 1919.”Footnote 142 Within the living memory of many policymakers, the United States and Western European states had cooperated in the First World War, the Paris conference of 1919, the Dawes Plan of 1924, and through economic—and often nominally private—cooperation in European post-WWI reconstruction. These events had continued importance for influential Atlanticists in the US foreign policy establishment; even many League of Nations opponents favored strong trans-Atlantic defense ties.Footnote 143 Despite US rejection of the League, coordination continued in agreements like the 1930 London Naval Conference. The most evident antecedents include cooperation during both world wars. Although North Atlantic cooperation was dominated by relations between Britain, the US, and Canada, especially after mid-1940, broader antecedents exist, such as Norwegian Foreign Minister Trygve Lie’s 1940 proposal for North Atlantic security. As in the Americas, a host of activity followed World War II but preceded nato’s emergence.Footnote 144 Douglas Brinkley underscores the role of transatlantic antecedents in shaping US Secretary of State Dean Acheson’s approach to North Atlantic security cooperation.Footnote 145 Antecedents predisposed actors to multilateral agreements in the critical juncture of the late 1940s and conditioned how they responded to a new threat, providing material for rapid institutional construction.
External threat is considered the key explanation for nato, but it is ambiguous in explaining weaker US-Asian multilateralism. In the Americas, multilateral security institutions emerged despite low perceptions of external threat. A 1947 CIA report dedicated only one page to Latin America: “In Latin America Soviet objectives are limited and negative…. With the conclusion of the Treaty of Rio de Janeiro, however, United States interests in Latin America appear to be reasonably secure.”Footnote 146 Latin Americans were more concerned with internal communist threats than was the Truman Administration. In Brazil, home of the largest communist party in the region, government repression went against the US ambassador’s advice.Footnote 147 Regionally, the US opposed formal anticommunist agreements despite calls for such inter-American proclamations from Latin Americans.
Although the presence of great powers to share security burdens does not explain the emergence of security cooperation in the Americas, it may be quite important for understanding subsequent institutional evolution. nato evolved toward more multilateralism, although one could read too much egalitarianism backward into its founding moment, when even the British depended on US economic and military assistance.Footnote 148 John Baylis writes that nato “[signaled] the end of British ambitions for a more independent world role.”Footnote 149 Only later in the Cold War period, with recovery, could the European powers provide some balance within multilateral nato. US commitments grew more expansive—an illustration of historical institutionalists’ point that institutional paths often diverge from founders’ intentions. By comparison, the inter-American system drifted away from multilateralism in practice, though not in form, as asymmetries worsened with the intensification of the Cold War.
The findings presented in this article challenge scholars of regionalism to more clearly distinguish between broad treatments of identity and the more specific implications of historical antecedents of regionalism. Hemmer and Katzenstein’s explanation highlights collective identity, and it is reasonable to ask whether historical antecedents are only a component of identity. Certainly, collective identity relates to history. But my explanation emphasizes something more specific. The United States had history with China and Japan, for example, but none of it constituted a shared historical antecedent of regionalism. The possibilities for rapid layering or conversion were minimal. When questions of global and regional order were open for renegotiation after World War II, the presence of historical antecedents shaped the range of feasible choices and favored divergent outcomes in different regions. Pan-American institutions, summits, and practices served as important models. The pau offered a focal point, a physical site, and established procedures for discussions between Latin American and US diplomats. Many Latin Americans perceived benefits from deepening and expanding regional cooperation in a new global context. At the 1947 Rio conference, Colombian Foreign Minister Alberto Lleras Camargo “made an extensive comparison between the inter-American system and UN pointing out that UN is only [an] experiment and weak, experiencing ‘continuous friction with reality, very similar in appearance to failure.’”Footnote 150 Antecedents provided important benefits to the United States, too, giving the country legitimacy, a geostrategic reserve, and close commercial ties. Many US and Latin American actors expected continued benefits.
The case, crucial to inter-American institutional creation, should invite greater reflection on the nature of early-Cold War US hegemony in the Americas, which is often treated as monolithic, “crudely imperial,” and therefore of secondary importance to understanding the emergence of US-led international orders.Footnote 151 Although the possibility of coercion cast a shadow, the emerging postwar inter-American order was also characterized by negotiated bargains that incorporated Latin American leaders’ interests, even at the height of US power. In the postwar critical juncture, Latin American diplomats and some supportive US policymakers drew on antecedents to shape and advance plans for multilateral security institutions. These included mechanisms of restraint. Despite many Latin American concerns about the United States’ ultimately interventionist nature, Latin American diplomats cited the Monroe Doctrine and US-led Pan-Americanism in support of a grand bargain that would extend and institutionalize US engagement while restricting unilateralism.Footnote 152 In earlier summits, the American states had committed themselves to juridical equality, nonintervention, peaceful settlement of disputes, and diplomatic consultation. Post-World War II, they refashioned earlier institutions and formalized those commitments. US adherence to these principles would remain incomplete in the new inter-American system, as it always had been. Latin Americans realized that. But in a variation of nato’s first Secretary General Lord Ismay’s quip about the organization, the inter-American system remained Latin America’s best bet to keep the United States in (regional politics), out (of internal affairs), and down (proscribed from intervening), all at once.
Acknowledgements
Earlier versions of this article were presented at the University of Oxford, University College London, Universidad de los Andes, Centro de Investigación y Docencia Económicas, Pontificia Universidad Católica de Chile, and at annual conventions of the International Studies Association, Latin American Studies Association, and Swiss Political Science Association. Thanks to the participants in those discussions, three anonymous reviewers and the editors of World Politics, and to Christopher Hemmer, Louise Fawcett, Max Paul Friedman, Giovanni Mantilla, Galia Press-Barnathan, Martin Binder, Maria Koinova, Vincenzo Bove, and Carsten-Andreas Schulz for helpful comments.
Funding
Research for this article was conducted under grants from the British Academy/ Leverhulme Trust, Truman Library Institute, and Fulbright US Scholars Program.
Author
Tom Long is an associate professor in the Department of Politics and International Studies at the University of Warwick and an affiliated professor at the Centro de Investigación y Docencia Económicas, Mexico City. His research explores the international relations of the Americas and the dynamics of international power asymmetries. He is author of Latin America Confronts the United States: Asymmetry and Influence (2015). He can be reached at T.Long.1@warwick.ac.uk.
