Rationalism, Strategic Trust, and International Cooperation

The literature on cooperation in international relations has been pioneered by the rationalist (earlier known as the “neoliberal institutionalist”) approach. These scholars have focused largely on “mixed-motive” situations in which short-term incentives to defect undermine potentially more fruitful long-term cooperation among states, generally because of uncertainty about intentions.Footnote ⁸ States would prefer mutual cooperation but do not believe, or cannot be certain, that others will keep their end of a cooperative bargain and fear the costs of defection. Although it is seldom if ever mentioned in the early literature, the problem posed by uncertain intentions is really a question of trust. Trust is the belief that one will not be harmed when one's interests are placed in the hands of others.Footnote ⁹ In situations of cooperation, trust is the belief that another prefers mutual cooperation to exploiting and suckering others.Footnote ¹⁰ It is critical when cooperation does not involve a simultaneous exchange of benefits, when the completion of a transaction is sequential and one party is vulnerable to defection.Footnote ¹¹ In these instances, trust is the expectation of future reciprocity.

Rationalist international relations theory rests on a particular, strategic notion of trust.Footnote ¹² In this understanding of the concept, actors trust others when they have information that leads them to believe that specific others have a self-interest in reciprocating cooperation, generally an incentive in building or sustaining a long-term, mutually beneficial relationship. Hardin calls this “encapsulated interest.”Footnote ¹³ Of course, there is also strategic distrust, the belief that another is untrustworthy based on information about that actor's intentions. This is a highly structural account of trust. Strategic trust is situational in nature. It develops when specific actors believe particular others have an incentive to honor their agreements. International relations scholars have recently begun utilizing this conception of trust explicitly.Footnote ¹⁴

All rationalist work on international cooperation is based implicitly or explicitly on creating strategic trust, altering the structure of the situation to facilitate cooperation otherwise inhibited by uncertain intentions. The mechanisms of lengthening the “shadow of the future,” generating concern for reputation, and establishing linkages between issues are all example of creating encapsulated interest and strategic trust.Footnote ¹⁵ There is less uncertainty about others' cooperative intentions if they bear larger costs for defection. International institutions facilitate all of these processes as well as provide information about state behavior directly.Footnote ¹⁶ They are created in the absence of strategic trust in order to provide strategic trust.

Diffuse Reciprocity and Multilateralism in Security Organizations: The Limits of Rational Institutionalism

Rationalist accounts of international cooperation and institutional design have great limitations. In particular, they have difficulties explaining diffuse reciprocity in which exchange occurs over a longer period of time. Actors trust strategically as long as another's cooperative behavior indicates to them that they have a vested interest in maintaining a cooperative relationship. If there is no ongoing exchange, there is no such relationship on which to base strategic trust. Strategic trust requires more specific reciprocity in which exchange takes place over a shorter time frame.

This makes it hard for rationalists to account for “qualitative” multilateralism in security institutions, in Ruggie's words “an institutional form which coordinates relations among three or more states on the basis of ‘generalized’ principles of conduct—that is, principles which specify appropriate conduct for a class of actions, without regard to the particularistic interests of the parties or the strategic exigencies that may exist in any specific occurrence.”Footnote ¹⁷ In the security arena, qualitative multilateralism involves two features—security guarantees and a commitment to peaceful conflict resolution. Both are efforts to reap the self-interested gains of cooperation by combining resources for the collective good. At least theoretically, international security organizations should be most effective at reaching outcomes that benefit their membership when they ask states to make legally binding commitments that limit their future discretion in particular cases. Security guarantees have a better deterrent effect and more success after conflict erupts if they are automatic and binding in nature, unequivocally precommitting a state in advance to the defense of others, generally regardless of the foe. Conflict resolution is most likely to nip potential problem spots in the bud if it is authoritative. That is, parties to a dispute commit in advance to always submit their disputes to arbitration or mediation, have no say in the ultimate decision, and are penalized for noncompliance or nonsubmission of disputes. Both security guarantees and conflict resolution pledges can be thought of as reciprocal exchanges. One state agrees to come to the aid of others and to allow others to mediate its disputes in exchange for others doing the same.

However, state leaders will commit to qualitative multilateralism only if they believe that states will not abandon their obligations by either refusing to abide by procedures for dispute resolution or not coming to the aid of others in case of attack. In other words, states expect reciprocity. They cannot base this expectation on strategic trust. As Ruggie notes, qualitative multilateralism requires diffuse rather than specific reciprocity.Footnote ¹⁸ States in a multilateral security arrangement cannot engage in a continuous exchange of actual benefits, only a trade of promises to generally protect the other against attack. This is because these institutions are created to deal with future scenarios unknown to the leaders at the time of their agreement, which occur infrequently and almost never simultaneously. As a consequence, encapsulated interest will not emerge and strategic trust cannot sustain cooperation. This differentiates security cooperation from political economic cooperation such as trade. A security arrangement might function on the basis of an identity of preferences, such as a mutual interest in fighting a common enemy. But if this is the case, exchange and reciprocity are not involved and trust is not an issue.

Because it places a state's fate in the hands of others to some degree, qualitative multilateralism also comes with other potential threats of opportunism in addition to abandonment. First, if security guarantees are automatic in nature, states can become ensnared in conflicts that do not serve their interests, triggered by the outbreak of hostilities on the part of any member. Security commitments might lead to moral hazard, leading states to take overly provocative acts with the expectation that others will bail them out if conflict breaks out. This is the problem of “entrapment.” Second, strong security commitments might also encourage free-riding on the defense contributions of others. Third, conflict resolution procedures allow states to exploit others by interfering in disputes in which they are not directly involved.Footnote ¹⁹ In short, qualitative multilateralism demands a lot of trust, and rationalism cannot provide it. After all, rationalism predicts cooperation and institutional creation to emerge from a lack of trust.

Social Psychology, Generalized Trust, and International Cooperation

Strategic trust is not the only type of trust, however, either conceptually or empirically. Social psychologists have long found evidence of generalized trust—a general optimism about the trustworthiness of others. Whereas strategic trust is extremely limited, confined to particular circumstances in which actors have enough relevant information about interests, generalized trust is not tailored to individual circumstances. It serves as an inclination to trust independent of specific information about any particular other.Footnote ²⁰ Generalized trust cannot be based on information about all prospective interaction partners as this would be impossible to collect.Footnote ²¹ As Uslaner writes, whereas the etymology of strategic trust is “A trusts B to do X,” the etymology of generalized trust is “A trusts” or “A is trusting.”Footnote ²² A general belief that others are largely trustworthy or untrustworthy, independent of the particular partner or situation one faces immediately, indicates a dispositional quality of individuals that reflects a particular ideological view of the world. It is an attribute of the truster, not the relationship with or even the characteristics of the specific target of distrust or trust.Footnote ²³

Social psychologists have found overwhelming evidence that generalized trust, based on assessments before an experiment begins of participants' general expectations of how others will play experimental games, improves cooperation levels both in dyadic interactions but also in all kinds of public goods and commons dilemmas. Individuals placed in the same structural situations, with the same incentives and receiving the same feedback, are more likely to contribute to a public good or restrain from consumption of a common resource if they believe before the experiment that others will do so as well, that is, if they believe that the other has “assurance” preferences.Footnote ²⁴ Generalized trusters begin interactions with cooperation, even without specific information about the trustworthiness of others. Given that subjects have no prior knowledge of those with whom they are interacting and they cooperate to varying degrees within the same incentive structure, this cannot be strategic trust. During games, generalized trusters are more likely to give others the benefit of the doubt, allowing for some level of defection at least for a time in an effort to elicit cooperation. Generalized trust has an effect across multiple structural situations, inducing more cooperation than rationalists would expect even in prisoners' dilemmas. Generalized distrust matters as well. Those who lack generalized trust defect even in assurance situations in which both are left better off by cooperation. Whereas generalized trust leads to more cooperation than rationalists would expect, its opposite leads to less.Footnote ²⁵

Generalized trust is never total. It can easily exist simultaneously with the strategic distrust of a particular other. It is best thought of as explaining relative differences between individuals. Although below I refer to generalized trusters and nontrusters, this is mere shorthand and not meant to convey an absolute dichotomy. Generalized trusters are also not altruistic or gullible. In fact it has been found that they are better predictors of others' types after interaction. Generalized trusters defect in the face of consistent noncooperation. They expect reciprocity, even if only over the longer term.Footnote ²⁶

Because trust of the generalized variety cannot be based on specific information about others, it must be a kind of “moralistic” trust.Footnote ²⁷ Rotter defines this variety of trust as the “generalized expectancy held by an individual that the word, promise, oral or written statement of another individual or group can be relied on.”Footnote ²⁸ Generalized trusters believe that intentions and behavior reflect traits of the trustee, rather than the situation. When one trusts moralistically, one is making judgments about the inherent trustworthiness or lack of trustworthiness of others.Footnote ²⁹

In the context of strategic interaction, generalized trust is the belief that others will generally feel morally bound to reciprocate cooperation. Those subjects identified as generalized trusters before experiments tend to attribute behavior by others during experiments to moral characteristics and believe that honesty will have a greater effect on the level of cooperation of others than those who lack generalized trust. Their own level of cooperation increases much more sharply against players identified as moral than does that of those who lack generalized trust.Footnote ³⁰

As a form of “social capital,” generalized trust promotes cooperation even in highly uncertain situations deemed inhospitable to collaboration by rationalism.Footnote ³¹ Those who are more trusting expect less opportunism. Cooperation is less worrisome when others are generally expected to reciprocate and live by their commitments. Although the nature of social capital and its precise definition are contentious issues, I use the term merely to describe a resource of individuals, groups, or societies that serves to promote higher levels of cooperation that leave everyone better off.

A belief that others are generally trustworthy provides the confidence needed to cooperate even when exchange occurs over a long period of time. Generalized trust facilitates the diffuse reciprocity often necessary for cooperation. Where generalized trust is lacking, actors will insist on more specific reciprocity, since they are less certain that others will honor their agreements.Footnote ³² Therefore, in the relationship between specific reciprocity and strategic trust on the one hand and diffuse reciprocity and generalized trust on the other, cause and effect are reversed. Whereas strategic trust follows from specific reciprocity, diffuse reciprocity over time follows from generalized trust.Footnote ³³ In international relations, generalized trust should promote qualitative multilateralism.

Generalized trust also facilitates cooperation with those about whom one has little information, when strategic trust does not exist. According to Yamagishi and Yamagishi, trust is a “springboard” in uncertain situations to leap into the “outside world,” “emancipating” individuals from the secure confines of stable relationships and allowing them to seek other cooperative partners with whom there might be greater gains.Footnote ³⁴ In international relations terms, generalized trust should promote broader organizations with more members, or “quantitative” multilateralism.Footnote ³⁵

Those who lack generalized trust fall back either on strategic trust or what Uslaner calls particularized trust, moralistic trust of a particular other. Uslaner describes it as trust in others who are like us, based on a shared identity. The etymology is “A trusts B.”Footnote ³⁶ In this way particularized trust is different from generalized trust, which facilitates trust of outsiders rather than creating or reinforcing trust within groups. Studies show that less trusting individuals rely more on committed relationships instead of seeking out potentially more beneficial relationships with others they do not know.Footnote ³⁷

Generalized trusters can even cooperate at greater levels without any kind of coercive protection. This reveals a paradox. Generalized trust likely makes state leaders more willing to commit to hierarchy since the creation of a higher authority involves placing one's fate in another's hands. However, generalized trusters are relatively less likely to see the need for this transfer of control because they are more trusting. They are more inclined to believe, ceteris paribus, that agreements that rationalists might consider “cheap talk” will suffice for cooperation. A lack of generalized trust, in contrast, creates a felt need for hierarchy to protect against opportunism because others are regarded as more untrustworthy and likely to defect. Yet without generalized trust, state leaders cannot make such a commitment to hierarchy in the first place. Hierarchy requires a solution to a second-order trust problem that only generalized trusters, if they see the need, can solve. Institutions, however, are still necessary, even for generalized trusters, because generalized trust is never total. Not everyone is peaceful. Pure anarchy is not sufficient. Generalized trust serves as a source of social capital to construct an anarchical society, a system of constraints placed on states that facilitates cooperation, even while the international system remains anarchic.

Research Design

On the basis of this theoretical review, it is possible to draw the following, contrasting rationalist and social psychology-inspired hypotheses concerning prominent features of institutional design highlighted by rationalists.Footnote ³⁸ Rationalists would expect states to commit to qualitatively multilateral institutions only if they can design other aspects of the institution in a way that reduces the problems of trust, uncertainty, and opportunism inherent in this particular institutional design. This leads to my first rationalist hypothesis (RH):

Reducing membership allows for easier monitoring of behavior. States are better able to predict voting patterns in conflict resolution procedures and potential conflict scenarios, reducing the uncertainty that works against multilateral cooperation. There is a greater chance for specific reciprocity in a smaller organization because interests can be traded off. States are also more likely to have overlapping and identical preferences so that reciprocity is not necessary at all.Footnote ³⁹ In other words, state preferences for qualitative and quantitative multilateralism are inversely related.

This solution allows for limitations on all types of opportunism by taking control out of the hands of those who might act opportunistically. The argument that distrust leads states to delegate control over implementation of policy to circumvent defection is a common one in the rationalist literature. The greater the problem of distrust, the more hierarchy needed to solve it.Footnote ⁴⁰ This is the direct descendant of the Leviathan.

If, for whatever reason, states cannot reduce the number of members or impose hierarchy, rationalism leads one to expect that states will instead promote two other institutional features.

States can water down their security guarantees or refuse to make any commitments in advance so as to reduce concerns about entrapment. To avoid exploitation, conflict resolution can be diluted by making submission and adjudication contingent on the agreement of parties to the dispute, or not specifying any penalty for noncompliance. Most obviously, states can insist on a veto for all matters affecting their sovereignty, whether it be taking action to enforce a security guarantee or consenting to conflict resolution or accepting its results.Footnote ⁴¹ All of these steps reduce qualitative multilateralism by removing the general nature of commitments and making them more discretionary.

This can be done through the inclusion of withdrawal mechanisms that allow states to drop out of any cooperative arrangement if their interests are not served or by limiting the term of commitment.Footnote ⁴²

A key distinction between the social-psychological argument and rationalism is that the former expects variation in the same structural circumstances whereas the latter, relying as it does on strategic trust, cannot explain why decision makers in the same position with the same information would make different choices in situations of security cooperation.Footnote ⁴³ Generalized trusters will expect less opportunism and therefore endorse different forms of international cooperation than nontrusters. This leads to my first social psychological hypothesis (SPH):

Generalized trust provides the basis for the diffuse reciprocity on which qualitative multilateralism is based. It reduces worries about abandonment, exploitation, free-riding, and entrapment endemic to this institutional form.

Nontrusters will, seemingly paradoxically, be the most convinced that hierarchy is necessary to ensure compliance with obligations but be the most resistant to endorse such a design.

Generalized trusters will have more confidence in others in general and have more faith that multilateral institutions will work. Consequently, they should favor larger multilateral organizations with more members that add to the deterrent and pacifying effect. Nontrusters will prefer unilateralism or limited cooperation with a smaller number of other states who have a demonstrated record of trustworthiness, shared identity, and/or closely overlapping interests.

Fearing opportunism less, generalized trusters will more be more likely to promote the benefits of greater certainty and assurance by agreeing to organizations with longer terms and fewer provisions for opting-out. Nontrusters will prefer flexible and nononerous opt-out provisions.

By analyzing the internal politics of international institution-building in the United States, it is possible to hold constant other structural influences that have nothing to do with trust and more precisely pinpoint the influence of variation in dispositions to trust. Generalized trust, however, is difficult to measure. I have combed through statements, speeches, memoirs, diaries, and biographies to assemble relevant passages that reveal any information about the core beliefs of the most important participants in the process, those around whom the main coalitions for and against multilateralism formed.

I also rely on proxies. Generalized trust is manifested, I argue, in concerns about opportunism in international relations—exploitation, entrapment, abandonment, and free-riding. These are the indirect measures of the independent variable that lead individuals to their dependent variable—their preferred degree of cooperation and form of institutional design. I am looking for patterns. Assessments of the likelihood of each kind of opportunism should not vary systematically across individuals in the same structural situation according to rationalism, whereas they will in a social psychological account. Concern for one type of opportunism should not necessarily be correlated with concern about others if structure and situation drives the process, but this will be the case if ideological dispositions to trust are more important.

One might still object that concern for opportunism could be a post hoc rationalization of a position based on a different set of interests, or a measure of views about the nature of foreign affairs independent of generalized trust. Therefore, I also look for correlations between the previous measures and domestic policy positions so as to have a marker of generalized trust independent of the phenomenon under study—international cooperation.

All of the leading theorists in political psychology agree that a general sense of threat and fear is central for explaining the adoption of rightist political views. Duckitt and his colleagues argue that the right has a “motivational goal” of security, driven by a belief that the world is a dangerous place in which others are out to harm them.Footnote ⁴⁴ Feldman claims that rightist ideology is a reflection of a pessimistic view of human nature, consistent with a longstanding observation about the nature of the right.Footnote ⁴⁵ Jost and colleagues call this the “existential motive,” a common denominator in the right that they find in their remarkable effort to synthesize the findings of hundreds of studies on the psychological correlates of political ideology with eighty-eight different samples from a multitude of countries.Footnote ⁴⁶ All are capturing the same core aspect of rightist thinking—that others cannot be trusted.

Generalized trust appears to provide a central pillar of the deep underlying ideological structure that divides left and right and from which specific policy positions emerge. It explains the right's endorsement of strong law-and-order policies and traditional morality domestically and a strong military to defend from international threats. The more trusting left, in contrast, adopts more libertarian positions on law and order, civil rights, and personal expression and a relatively more antimilitarist foreign policy position.Footnote ⁴⁷

This remarkable parallel between studies in social and political psychology allows one to use domestic policy positions as a proxy measure for generalized trust. More conservative members should be the least trusting, more liberal the most trusting individuals. For this I use Poole and Rosenthal's D-Nominate scores, which are themselves proxy measures of ideology based on voting behavior, as proxy measures for generalized trust in individuals.Footnote ⁴⁸ To the extent that ideology correlates with party, the relationship between political ideology and trust also leads one to expect that there should be significant partisan divisions in American domestic politics over international cooperation and institutional design. Qualitative evidence reveals that the party of the right in the United States, the Republican Party, has consistently been the party of order, tradition, and a strong defense.Footnote ⁴⁹ Still, party is only a proxy. Liberal Republicans and conservative Democrats should depart in predictable ways from the party line.

None of these three measures—statements on generalized trust, concerns about opportunism, and domestic policy positions—are ideal by themselves, but they command greater confidence if they are systematically correlated, considering nothing else would seem to explain that pattern. This is a triangulation strategy.

The League of Nations

Even before the United States entered World War I, President Woodrow Wilson began a crusade to establish a League of Nations around the concept of collective security. Although Wilson had the full backing of his Democratic Party throughout the entire process, the president pursued his foreign policy agenda almost completely alone. He brought back from Paris a covenant based largely on his vision, requiring members of the League to submit matters of dispute to a variety of conflict-resolution mechanisms before any recourse to force, or else be subject to automatic economic sanctions on the part of all other League members. Members would be forced to abstain from consideration of their own cases. The covenant also specified that an attack on any member was essentially a declaration of war on them all, and pledged that all members would assist in guaranteeing the territorial integrity and political independence of any victim. This was a moral obligation; the exact contributions would be worked out in each individual case but could include military force.

Such an organization had to rest on diffuse reciprocity, and Wilson understood as much. Of the security guarantee, Wilson said, “When all unite to act in the same purpose, all act in the common interest, and are free to live their own lives under a common protection.”Footnote ⁵⁰ Wilson also recognized that the Americans might sometimes lose in conflict-resolution proceeding, and pledged they would “take our medicine.”Footnote ⁵¹ Qualitative multilateralism required some loss of state discretion to reap the gains of cooperation. “Some of our sovereignty would be surrendered,” Wilson admitted. “Without some sacrifice,” the enterprise could not succeed, he said,Footnote ⁵² a sentiment he consistently expressed in public and private.Footnote ⁵³

Wilson's position is a puzzle for rationalism. Lake argues that the U.S. objective interest in collective security after World War I was limited at best; it had an interest in a mechanism to ensure peace, particularly in Europe, but not at any price as the United States still had a viable strategy of unilateralism, given the nature of military technology and the nation's relative geographic isolation.Footnote ⁵⁴ Even though the United States had an interest in deterring future global conflict, any binding commitments could lead to entrapment or free-riding. The only theoretical solution was the imposition of some sort of hierarchy to prevent opportunism, but U.S. relative power would not allow this. The default should have been, according to Lake, unilateralism.Footnote ⁵⁵ The risk of opportunism only increased with large numbers, not only in terms of the security guarantee but also the possibility of exploitation in third-party conflict resolution with forced abstention. Yet Wilson wanted the league to be universal in order to mobilize a greater deterrent power.Footnote ⁵⁶ He even successfully fought off the British to ensure that the League Council, its most important body, would consist of more than simply the great powers.Footnote ⁵⁷

The standard answer to this theoretical puzzle is that Wilson was an “idealist.”Footnote ⁵⁸ I agree, but this begs the question: what is an idealist? Wilson was not a pacifist; after all he had led the United States on a moral crusade into World War I and endorsed coercive sanctions to guarantee peace. His vision was not based on altruistic concern and a cosmopolitan identity, but on a sense of American self-interest in a new era of interdependence.Footnote ⁵⁹ I maintain that Wilson was led to believe collective security would work based on his sense of generalized trust. Wilson had an optimistic view of human nature. He once privately advised his daughter that “most people are fundamentally good—of that I am sure. Don't let a few cheap and dishonest ones hurt you.”Footnote ⁶⁰ Wilson's generalized trust was evident in his domestic political agenda. Wilson was one of the great liberal reformers in American history. His progressive beliefs were based on his acceptance of the moral nature of the American masses.Footnote ⁶¹ This was a consistent theme of the era on the part of all Democratic leaders.Footnote ⁶² In keeping with the generalized nature of his trust, Wilson believed this was true universally, beyond U.S. borders.Footnote ⁶³

Wilson's generalized trust found indirect expression in his lack of concern about opportunism of all kinds in his proposed League. In terms of abandonment, he stressed that states would meet their commitments.Footnote ⁶⁴ Wilson claimed, even privately, that he did “not think such a refusal [to comply with the security guarantee] would likely often occur.”Footnote ⁶⁵ Entrapment was not a worry for Wilson either. He argued the League would help solve that particular problem of opportunism, not exacerbate it. Turning George Washington's farewell address on its head, Wilson claimed that the League was not an entangling but rather a “disentangling alliance … which would disentangle the peoples of the world from those combinations in which they seek their own separate and private interests and unite the people of the world to preserve the peace of the world upon a basis of common right and justice.”Footnote ⁶⁶ Collective security would also lessen exploitation. In the League, “nations with one accord adopt the doctrine of President Monroe as the doctrine of the world: that no nation should seek to extend its polity over any other nation or people, but that every people should be left free to determine its own polity, its own way of development, unhindered, unthreatened, unafraid.”Footnote ⁶⁷

Generalized trust helps explain the paradox that Wilson, while he asked more of the United States than others in his country were willing to provide, did not see the need for a strong institution to guarantee the meeting of that obligation. The sense of moral obligation would suffice, both for the United States and other members of the League. For Wilson, collective security was “a very grave and solemn moral obligation.” Even though such a moral obligation was “binding in conscience only, not in law,”Footnote ⁶⁸ Wilson stressed in a closed-door meeting that a “moral obligation is of course superior to a legal obligation, and, if I may say so, has a greater binding force.” “There is a national good conscience in such a matter,” he expressed to dubious senatorsFootnote ⁶⁹ and in other instances in private correspondence with his collaborators.Footnote ⁷⁰ Some sort of international state, therefore, was not necessary. Wilson wrote that he thought the League was “a matter of moral persuasion more than a problem of juridical organization.”Footnote ⁷¹ This belief in moral obligations is reflected in Wilson's choice of name for the League's constitution, a “covenant.”Footnote ⁷²

Wilson's sense of generalized trust led him to take a position anomalous for rationalists. However, his position also elicited one of the most heated debates in the history of U.S. foreign policy. The influence of traditional isolationists, those who eschewed any political engagement particularly with the European great powers, was marginal.Footnote ⁷³ Wilson's primary antagonists were a group of “conservative internationalists” located exclusively in the Republican Party.Footnote ⁷⁴ Unlike the handful of remaining isolationists, they accepted U.S. interest in engagement with the political and military affairs of the wider world, but they opposed collective security. Led by Henry Cabot Lodge, the Republican majority leader and chair of the Senate Foreign Relations Committee, their primary difference with Wilson was over opportunism. Rationalists have difficulty explaining radically opposed visions for U.S. security in the same structural circumstances.

Opponents of the League were systematically more concerned about opportunism of all kinds, both in public and private. Their primary fear was the entrapment that might result from any security guarantee. Lodge asked, “Will it not be worthwhile to pause a moment before we commit ourselves to an army of 500,000 men, to be held ready for war at the pleasure of other nations in whose councils we shall have but one vote if we are true to the President's policy of the equality of nations?”Footnote ⁷⁵ Conservative internationalists also were concerned that other states would exploit U.S. interests in the League's conflict-resolution institutions.Footnote ⁷⁶ They were convinced that states, including the United States, would abandon their pledges when a crisis ensued. Publicly, Lodge complained of “too many and too Utopian proposals … and too difficult obligations.”Footnote ⁷⁷

This indirect expression of a lack of generalized trust was complemented by a direct one, seen most clearly in these individuals' pessimistic statements about human nature, both in international relations and in life in general.Footnote ⁷⁸ Lodge cautioned, “We must deal with human nature as it is and not as it ought to be.”Footnote ⁷⁹ Because human nature was conflictual, international relations are marked by discord. Lodge declared, “There has been pretty constant fighting in this unhappy world ever since the time when history begins its records, and in speaking of lasting peace in terms of history we can only speak comparatively.”Footnote ⁸⁰ Lodge expressed these same sentiments long before he became a leading Republican politician.Footnote ⁸¹ Similarly, Republican Senator Philander Knox said, “One must be visionary indeed to suppose that the heterogeneous peoples of the earth could so completely overcome human nature as to combine now in the real internationalism of a world State or even in a league involving a great catalogue of unnatural self-restraints. Such conceptions to-day [sic] touch rather the postulates of religion than the facts of statesmanship.”Footnote ⁸²

In a final marker of a lack of generalized trust, these conservative internationalist opponents of the League were among the most conservative members in the U.S. Senate.Footnote ⁸³ The most prominent League opponents—Lawrence Sherman, Philander Knox, George Moses, Frank Brandegee, Lodge, and Warren Harding—all ranked in the top fifteen in their party in terms of their D-Nominate scores. The party's ideological program was marked by a general pessimism. In an exhaustive review of the history of American party ideology, Gerring writes, “Perhaps more than any other value, order—and its antithesis, anarchy—defined … Republican ideology” at this time.Footnote ⁸⁴ This was natural because Republicans “viewed human nature with an abiding mistrust,” and feared that without order, society would generate into chaos.Footnote ⁸⁵

Fear of opportunism led the conservative internationalists toward the seeming paradox of arguing that the League required more force, yet asked too much of its members at the same time. In a closed-door meeting with Wilson, for which a transcript is now available, Senator Harding asked skeptically that “if there is nothing more than a moral obligation on the part of any member of the league, what avail [the security guarantee]?” He predicted that others will “take advantage of the [moral] construction that you place upon these articles.”Footnote ⁸⁶ The League could not work on the basis of “verbal adherence to general principle…. You cannot make effective a league of peace, ‘supported by the organized force of mankind,’ by language or high-sounding phrases,” declared Lodge.Footnote ⁸⁷ The senator said that if the League was to be more than an “exposition of vague ideals,” it must have “authority to issue decrees and force to sustain them.”Footnote ⁸⁸ At another point, he explained, “There is no halfway house to stop at…. The system must be either voluntary or there must be force behind the agreement.”Footnote ⁸⁹ Yet consistent with a lack of generalized trust, the conservatives never made any such hierarchical proposals.

Conservative internationalists instead fell back on cooperation with a smaller group with whom the Americans shared interests, experience, and even identity.Footnote ⁹⁰ Knox said, “The league of nations that now challenges our solicitude is the league of nations of which we are now a member—the glorious present alliance of the many powers with whom we are now fighting as a league to enforce and to maintain peace from disturbance by the German menace.” He suggested that, “Out of the present alliance … it would seem possible to perpetuate the league we have … as a league for one single purpose of enforcing peace.”Footnote ⁹¹ This suggested a particular form of organization, a concert of great powers not unlike that formed after the Napoleonic wars that would enforce the peace for the world.Footnote ⁹² However, this arrangement would not put the United States at any risk of opportunism. Knox advocated “a permanent committee for consultation.”Footnote ⁹³ The concert of nations would be informal and ad hoc in nature without any general obligations or even voting procedures. It “entangles us in no way,” reassured Knox.Footnote ⁹⁴ Conservatives opposed both qualitative and quantitative multilateralism.

In response to Wilson's proposed covenant, Lodge, in consultation with other senators and Republican Party luminaries, proposed a set of four reservations to the ratification instrument that would restore American sovereignty and remove concerns about opportunism. An exemption for the Monroe Doctrine would prevent exploitation of the United States by countries by freeing the United States of the obligation to submit matters of vital interest inside its hemisphere to the dispute-resolution procedures of the League. Disavowing any American precommitment to the security of member states removed the threat of entrapment. Withdrawal provided a further layer of protection by allowing the United States to exit at any point in time. A restriction of the League's jurisdiction to anything considered domestic in nature demonstrated the common denominator of conservative objections—a desire to maintain untrammeled American sovereignty given the potential for opportunism by other states.Footnote ⁹⁵

The reservations passed along ideological lines in both the Foreign Relations Committee and the full Senate.Footnote ⁹⁶ However, Wilson proved unwilling to compromise, particularly on the all-important moral obligation of Article X. Republicans would not vote for the League without reservations, the Democrats would not vote for the League with them. The treaty went down to defeat in the Senate in both forms because some degree of bipartisan support was necessary given the need for a two-thirds vote to pass a treaty in the chamber. The United States never joined the League. Cooper has compiled a table of the percentage of times that individual senators voted with the Wilson administration's position.Footnote ⁹⁷ The data includes ninety votes, including amendments and reservations, all of which aimed at weakening the treaty. When I correlate this percentage score with D-Nominate scores acting as an indirect measure for generalized trust, the result is striking—a correlation of 0.83.Footnote ⁹⁸ Rationalism cannot account for these divisions, which indicate variations in levels of generalized trust.

The United Nations

Generalized trust was dealt a blow by the experience of the World War II. Moral obligations alone had clearly failed to stop Adolph Hitler. References to the benign nature of humankind were almost unheard of in light of Nazi transgressions. A common refrain was that any new international organization needed more teeth to better deter future aggressors.Footnote ⁹⁹ A consistent theme in the literature on the United Nations is that the Democrats were so chastened by the failure of the League to prevent aggression in the interwar years that they preferred a more realistic alternative that more closely guarded American sovereignty.Footnote ¹⁰⁰ This is most evident in the preference for traditional diplomacy and great power consultation of President Franklin D. Roosevelt, which contrasted sharply with Wilson. President Roosevelt believed that peace could be maintained by institutionalizing the wartime alliance and creating a concert of the great powers, whose power would serve as an overwhelming deterrent to any future aggressor.Footnote ¹⁰¹

However, generalized trust was much more resilient than most accounts allow. Core beliefs rarely undergo fundamental re-evaluation. While FDR supported a great power concert, the major players in foreign affairs in his administration had a vision almost identical to Wilson.Footnote ¹⁰² Three officials consistently stand out as the most important in all accounts of the UN's creation—Secretary of State Cordell Hull, his trusted aide Leo Pasvolsky, and Undersecretary of State Sumner Welles. Because the two latter individuals were never elected officials, operating largely privately and producing mostly dry policy memos, it is difficult to elicit their levels of generalized trust. However, it is known that all three were devoted “Wilsonians” committed to collective security.Footnote ¹⁰³ In his term as a U.S. senator from Tennessee, Hull was its tenth most liberal member, with a voting record that stands 0.9 standard deviations toward the left of the political spectrum from the party mean, based on D-Nominate scores.

Unlike Wilson, Roosevelt was remarkably uninvolved in the planning for the League's successor, leaving it almost exclusively to the State Department. This allowed these generalized trusters to situate the four-policemen idea within a broader framework more consistent with qualitative and quantitative multilateralism. Pasvolsky in particular was the driving force behind the Draft Constitution, the first blueprint for what would become the United Nations, and almost completely neglected in historical accounts. In accordance with FDR's wishes, the four policemen enjoyed pride of place, constituting an Executive Committee with “responsibility in matters of international security.” However, in most other ways, the document was a return to Wilsonian collective security. If disputes threatened to lead to a “breach, or imminently threatened breach, of the peace between nations,” a broader Executive Council would request the parties to restore the position before the onset of conflict. States failing to comply were “presumed to intend a violation of the peace of nations” at which time the council “shall apply all the measures necessary to restore or maintain the peace.” The use of “shall” was telling, since it was not left up to the council's discretion whether it would act. This was almost an automatic collective security guarantee, and much stronger than the one included in the covenant.Footnote ¹⁰⁴

The guarantee was all the more striking in light of the Draft Constitution's voting provisions for the council. In instances of enforcement of the peace, decisions would require a two-thirds majority vote of the council including only three quarters of the Executive Committee, who would be permanent members. This meant that on the most important issues great powers would not have a veto over matters concerning the use of their own armed forces.Footnote ¹⁰⁵ The combination of these provisions risked entrapment.

From the beginning, planners recognized this, but believed that these limitations on sovereignty were necessary to ensure the success of the organization. The need for unanimity had paralyzed the League. The earliest plans in the State Department stressed the necessity of some “derogation” of sovereignty to permit decisive, automatic, and rapid action by the institution.Footnote ¹⁰⁶ Democratic officials embraced qualitative multilateralism.

These more trusting Wilsonians also wanted a more inclusive organization than the concert proposed by FDR and set about to broaden the role played by other states.Footnote ¹⁰⁷ Quantitative and qualitative multilateralism went together. The next draft of a constitution for a postwar organization, the “Staff Charter” of August 1943, strengthened the security guarantee and left the voting procedure intact but eliminated the Executive Committee altogether, leaving only an Executive Council.Footnote ¹⁰⁸ Generalized trusters also preferred a larger Executive Council and eventually won over Roosevelt to a council of eleven, with five permanent and six rotating members.Footnote ¹⁰⁹ Hull had made Roosevelt into a “universalist,” concludes Campbell.Footnote ¹¹⁰

Up to this point, the State Department had done little canvassing of the Senate. Although it is frequently maintained that World War II marked a decisive break in American foreign policy, bringing about a bipartisan consensus in favor of American multilateralism, this is hardly the case.Footnote ¹¹¹ Republicans embraced the same forms of cooperation that protected American unilateralism as they had more than twenty years before. While the Democratic administration was generally contemplating an institution with significant multilateral elements, Republicans in the same structural circumstances were advocating a different vision than the Democrats. Rationalism cannot account for this.

In the fall of 1943, the Republican Advisory Council began to formulate an official position. No one was more important in this meeting than Senator Arthur Vandenberg of Michigan, who was also the chairman of the Foreign Relations Committee. The Republican Senate steering committee claimed that his “views respond more completely to the composite judgment and conscience of Republicans in this international field than do those of any other Senators.”Footnote ¹¹² Vandenberg was deeply skeptical of collective security, expecting abandonment and doubting the power of moral obligations: “I doubt whether any hard and fast international contracts looking toward the automatic use of cooperative force in unforeseeable emergencies ahead will be worth any more, when the time comes, than the national consciences of the contracting parties when the hour of acid test arrives,” he wrote privately.Footnote ¹¹³ However, Vandenberg was also deeply opposed to any sort of hierarchy, such as an international police force or elements of a world state. The Senator saw himself as guaranteeing the “continuance of the American Flag over the Capitol.”Footnote ¹¹⁴ “I think we must maintain our own sovereignty in the final analysis,” he wrote in his diary.Footnote ¹¹⁵ He exhibited the same seeming paradox as his conservative predecessors twenty years before.

On the basis of Vandenberg's drafts, the Republican conference at Mackinac endorsed a resolution in favor of only “responsible participation by the United States in postwar co-operation organization among sovereign nations to prevent military aggression and to attain permanent peace with organized justice in a free world.”Footnote ¹¹⁶ The party completely marginalized the only three politicians pushing for more dramatic plans based on pooling sovereignty—Wendell Willkie, Harold Stassen and Warren Austin—who were, not incidentally, by far the most liberal members of the party domestically.Footnote ¹¹⁷ Vandenberg noted that the sovereignty argument has great importance to “conservative portions of the party.”Footnote ¹¹⁸ Austin's small victory was that the party platform even mentioned an “organization” at all.

It was at this point that the Roosevelt administration, recognizing the need for two-thirds Senate support of a constitution for a future organization, began to water down its more ambitious, qualitatively multilateral plans with an eye toward winning the support of less trusting and more unilateralist conservatives. For them, this was the true lesson of 1919, not the failure of collective security but the need to manage the domestic political process before the treaty came before the full Senate.Footnote ¹¹⁹ With domestic politics in mind, the Democratic administration inserted a great power veto for enforcement matters in their new planning document and eliminated the automaticity of the security guarantee.Footnote ¹²⁰ Hull wrote in his memoirs that in light of the League experience, he realized that only with the veto could they gain congressional approval, even if he himself was disappointed with the fact.Footnote ¹²¹ He noted that the key argument used against the League was that it might entrap the United States in others' conflicts.Footnote ¹²²

By late April 1944, the nature of the international organization was largely set, laid out in a paper entitled the “Possible Plan for a General International Organization.”Footnote ¹²³ It closely resembled the eventual form of the United Nations, essentially fusing a concert-like arrangement with a universal security organization. States would commit not to use force to resolve disputes. The great powers, as permanent members of the council with a veto, would consider potential flashpoints in the international system, deciding what to do on a case-by-case basis, not general rules. Not surprisingly, when Hull shared the draft with Republicans in a bipartisan committee of Senators designed to ensure congressional acquiescence, Vandenberg was extremely pleased that it protected against entrapment, avoided any type of international hierarchy, and envisioned a prominent role for a few great powers.Footnote ¹²⁴ However, action required not only great power unity but also the consent of other states as well. The process was also to be formal in nature, based on voting rather than informal consultation.

One remaining issue was the “absolute veto,” whether the ability of the permanent members to block action both in peaceful settlement and enforcement action could apply to their own conflicts with other members. Administration policy at the time was to oppose such a provision.Footnote ¹²⁵ A rationalist account of the creation of the United Nations might suspect that American consideration of restricting the application of the great power veto was based on an assessment of likely alignment of voting patterns after the war, that is, an early prognosis of the Cold War. It is theoretically possible that the United States predicted that any danger in giving up an American veto over enforcement by the United Nations was made up for by the gain in preventing Soviet obstruction, particularly as it could count on cooperation from the other permanent members, particularly Great Britain and China.

Of course by 1948 Soviet Union vetoes would restrict the Security Council from taking action in a myriad of situations. However, these American plans predate deep suspicions of Soviet intentions. The top-secret discussions within the State Department concerning both the veto and the absolute veto were framed around the general problem of how to create an organization more effective than the League of Nations, without any specific mention of potential problem countries. The faction lobbying against the absolute veto for the great powers was led by the Wilsonian Pasvolsky and others who had opposed the great power veto as a whole earlier in the planning process. The group argued that such a provision would undermine the universal nature of the institution, making it de facto less quantitatively multilateral and more of a concert of great powers. More important, however, was the fear that such a veto would hamstring the organization by requiring unanimity.Footnote ¹²⁶

The limited support for the absolute veto in the Democratic administration was driven by ratification considerations, an anticipation of Republican objections, particularly on the part of the State Department liaison to Congress.Footnote ¹²⁷ There was not even consideration at that point that the British and Soviets might not go along with such a plan. When the American diplomat Bowman conferred with his British colleague on the issue, he “admitted that they thought only of the Senate & had never discussed the effect on Commonwealth or Latin America.”Footnote ¹²⁸ Hull, as both a Wilsonian and a former senator who knew the extent of Republican reservations, was torn, but ultimately opposed the absolute veto.Footnote ¹²⁹

When strategic distrust about the Soviet Union did begin to emerge, due to the Soviets' very insistence on maintaining an absolute veto and their actions in Poland and other parts Eastern Europe, it had the effect of reinforcing the determination of those American officials, the vast majority of the delegation, who had opposed the absolute veto from the beginning. Hull's position, along with Pasvolsky's, hardened.Footnote ¹³⁰ Strategic distrust and generalized trust pushed in the same direction, toward restrictions on great power prerogatives, although the latter was present before the former.

However, ultimately generalized distrust trumped both strategic distrust and generalized trust. Where Soviet behavior reinforced the already existing inclinations of Democrats to restrict the veto, it led to a tension in the Republican position as strategic distrust came into conflict with the general pessimism that had led conservatives to support the veto. Even though it was Republicans who expressed most skepticism about the reliability of the Soviets, Vandenberg could not consent to any significant limitation on the veto.Footnote ¹³¹ Soviet insistence on an absolute veto, while it worked against Democratic preferences for the United Nations, actually served conservative ends. As Vandenberg explained, “the irony of the situation is that the greater the extent of the ‘veto,’ the more impossible it becomes for the new League to involve America in anything against our own will.”Footnote ¹³² However, Soviet actions had the effect of softening Republican objections enough so that they could agree to a compromise that would remove the absolute veto over issues of peaceful settlement but reinstate it on questions of enforcing the peace.Footnote ¹³³ Hull called this deal, which the Soviets eventually agreed to and became enshrined in the UN Charter, the “absolute minimum of what we could accept.”Footnote ¹³⁴ It was necessary to ensure great power unity, but also to get the charter through the Senate. In this way, the great power veto was overdetermined.

The administration resisted pressure from Vandenberg to mention a right of withdrawal as a principle in the charter. However, the U.S. government did not consent to the British desire for an explicit statement forbidding withdrawal, conceived with the idea of keeping the United States in the United Nations. Officials feared this would raise hackles in the Senate.Footnote ¹³⁵ The delegation sought to play it down the middle, arguing to the Senate that such a right was understood even if it was not stated, but not mentioning it so as not to raise the alarm internationally of a return to isolation and thereby undermine multilateralism.Footnote ¹³⁶ Vandenberg concluded “it's not worth a row.”Footnote ¹³⁷

Given the fact that the administration anticipated and incorporated the more unilateral preferences of nontrusters in their drafting of the charter, its passage in the Senate was assured and a quantitative analysis of voting patterns therefore reveals nothing. It was ratified with only two dissenting votes. However, a comprehensive reading of the ratification debate reveals patterns completely consistent with the social psychological argument. Only Republicans stressed that the retention of sovereignty was a virtue that allowed them to support the treaty.Footnote ¹³⁸ They were consistently more pessimistic in their basic beliefs about international affairs and human nature.Footnote ¹³⁹ Democrats, and almost only Democrats, lamented the hindrance that might be posed by the great power veto and the lack of sovereignty transfer. Almost exclusively they stressed how interdependence necessitated cooperation to secure national interests.Footnote ¹⁴⁰ The only exceptions to these clear partisan differences were liberal Republicans such as Warren Austin, George Aiken, and Howard Alexander Smith, three of the four most liberal Republican senators according to their D-Nominate scores, or those conservatives such as Albert Hawkes and Alexander Wiley who expressed a sense of general optimism about international cooperation that contrasted sharply with others in their party.Footnote ¹⁴¹ I can find only one obvious exception to this general ideological pattern in the entire UN debate—the opposition of mainstream conservative Senator Henry Bridges to the veto.Footnote ¹⁴²