Why the United States Supports International Enforcement for Some Treaties but not for Others

Under what conditions should we expect the United States to support international enforcement of treaties? We hypothesize that U.S. support is most likely for treaties where international enforcement will cause considerable (desired) behavioral change by other countries but little (undesired) behavioral change by the United States. Similarly, U.S. support is least likely for treaties where international enforcement will generate the converse effects. In developing this hypothesis, we derive specific conditions under which we should expect U.S. benefits of international enforcement to outweigh U.S. costs (or vice versa). We also provide empirical examples. Finally, we consider three alternative explanations of U.S. views on international enforcement—concern for U.S. sovereignty, desire to prevent infringements on U.S. constitutional protection of individual rights, and the usefulness of international enforcement as a domestic commitment device. We discuss these alternative explanatory factors’ relationship to our own hypothesis.


Introduction
To be effective, an international agreement must satisfy three conditions (Barrett, 2003).First, all of the (major) countries concerned must participate.Second, the participating countries' commitments must be sufficiently deep to be able to solve or at least significantly reduce the problem the agreement aims to alleviate.Finally, the agreement must achieve high compliance rates.To fulfill these conditions, the agreement may need to restructure the parties' incentives, and a potentially powerful way to do this is to incorporate provisions for international enforcement.As used here, enforcement refers to the threat or actual use of material consequences1 to enhance treaty compliance, treaty participation, or both.
Examples of agreements that contain such provisions include the 1995 WTO agreements, the 1989 Montreal protocol on substances that deplete the ozone layer, and the 1992 Chemical weapons convention.The 1997 Kyoto protocol, too, includes an enforcement system.In contrast, Kyoto's successor, the 2015 Paris agreement, does not contain any provisions for enforcement.In this respect, Paris resembles many other environmental agreements, such as the 1999 Gothenburg protocol and the 1985 Helsinki protocol.
Two main factors influence whether an international agreement will come to include international enforcement.The first is whether the agreement has a need for enforcement, which is often linked to situation structure.While an agreement aiming to solve a coordination problem provides no incentive to be noncompliant or to withdraw, an agreement aiming to solve a collaboration problem might posit a strong incentive for defection.2Thus, it is primarily agreements which aim to solve a collaboration (dilemma-type) problem that have a need for enforcement (see e.g., Koremenos, 2007).
Second, even when an agreement has a need for enforcement, it may be difficult or even impossible to muster the required support among the negotiating parties for incorporating enforcement provisions (Hovi & Sprinz, 2006;Underdal, 1980;Ward, Grundig, & Zorick, 2001).The negotiation and design of international agreements (including their system of enforcement) are often based on the consensus principle, which provides a veto to each country (or at least to each major country) in the negotiations.Thus, erecting institutions for international enforcement typically requires support from each of the (major) countries involved.
However, the conditions under which (major) countries can be expected to support-or opposeinternational enforcement remains poorly understood.With a key focus on five cases of international treaty enforcement, we seek to contribute to filling this void by identifying the main conditions under which one particularly important country-the United States-is likely to support (or oppose) international treaty enforcement. 3e assume that the key determinant of U.S. support for (or opposition to) international enforcement is whether the United States expects to derive net benefits from such enforcement.Based on this general hypothesis, we then identify the more specific conditions under which we should expect the United States to support (or oppose) international enforcement and exemplify these expectations with observations from the five cases in focus. 4To complement our analysis, we also discuss three other explanations of U.S. support (or opposition) to international enforcement.Although the focus here is on the United States, the inherent logic is arguably generic, a point we return to in the concluding section.
Our discussion is primarily deductive in nature, but also draws significantly on data from 14 interviews with practitioners, observers and experts.The interviewees were selected on the basis of their experience with and/or expertise on U.S. foreign policy in general and the five core cases in particular. 5In the text, interviewees are referenced by number as listed in the Annex (e.g., I9 refers to interviewee number 9).The Annex also lists the names and affiliations of those interviewees who permitted such listing.
We proceed as follows.Following a review of the relevant literature, we develop our hypothesis concerning the circumstances under which the United States would be likely to support international enforcement.Next, based on this hypothesis, we identify and exemplify more specific conditions under which we should expect the United States to support (or oppose) international enforcement.Then, we discuss the relationship between our own hypothesis and three other explanations.Finally, we conclude.Before we embark on our analysis, however, we need to specify in more detail what we mean by U.S. support of international enforcement.

U.S. Support for International Enforcement
When we say that the United States supports international enforcement, we mean that it signals a willingness during the negotiations to submit to such enforcement, provided other countries do the same.When we talk about the United States' support for international enforcement, we actually conflate several distinct decision points that involve different U.S. government actors, notably the executive branch/president and the Senate:6 1) The executive decides to enter into treaty negotiations; 2) the negotiation team, appointed and instructed by the executive, decides to support inclusion of an enforcement system in the treaty; 3) the president decides to sign the treaty; 4) the president decides to send the treaty to the Senate for its advice and consent; and 5) the Senate decides to adopt a resolution of ratification.In our analysis, "U.S. support" for international enforcement refers to the views and decisions of the president's administration, i.e., the executive (decision points 1-3).This choice means that when U.S. ratification requires that the Senate provide its advice and consent,7 U.S. "support" for international enforcement may or may not entail that the United States actually submits to such enforcement; indeed, it may not even eventually become a party to the treaty concerned.8

Literature Review
To the best of our knowledge, no previous scholarly work has directly addressed the research question under consideration here.However, we draw on-and contribute to-four strands of related research.
The first strand seeks to explain why countries create and empower institutions for international enforcement or dispute settlement (e.g., Helfer, 2006;Helfer & Slaughter, 2005;Posner & Yoo, 2005a, 2005b; see also Guzman, 2002).For example, offering a theory of "constrained independence", Helfer (2006) argues that countries create and delegate authority to international courts and tribunals (ICs)9 to enhance the credibility of their own international commitments.According to him, countries also seek to prevent ICs from overreaching through the use of formal, structural, political, and discursive control mechanisms (Helfer, 2006). 10n contrast, Posner andYoo (2005a, 2005b) use a principal-agent framework to explain when countries will comply with IC rulings.According to them, adjudication does not provide incentives for compliance; rather, it merely adds information.Instead, they argue, reputation effects and fear of retaliation constitute the main incentives for compliance.
The theories offered by Helfer and by Posner and Yoo focus on explaining why countries collectively create ICs, how they collectively prevent ICs from overreaching, and why and when they choose to comply (or not to comply) with IC adjudication.However, these theories are less helpful for explaining whether a particular country (such as the United States) individually will support a particular IC or a particular enforcement institution of another type (e.g., a compliance committee for an international environmental agreement).To explain the emergence and persistence of an international enforcement institution it does not suffice to establish a collective motive for this institution; one must also establish that each major country has an individual motive to support it.
Much of this work is obviously relevant for our paper.For instance, with reference to U.S. policies on ICs, Romano notes that "American attitudes and behaviors toward international courts are highly contextual, changing between courts or dispute settlement procedures and between issues" (2009, p. xix).He further notes that case studies on human rights regimes suggest "that the United States conceives of these bodies mostly as a oneway road-that is, as tools to influence the conduct of other nations, rather than instruments to affect internal change" (Romano, 2009, p. xxi).Both of these observations can be understood in light of our hypothesis that expected costs and benefits motivate the United States' support for, or opposition to, international enforcement.Nevertheless, these scholars' work differs from ours concerning the main focus.In particular, their work does not seek to compare and contrast U.S. motives for supporting (or opposing) international enforcement across different treaties.Moreover, and perhaps partly as a result of this more narrow focus, these scholars' work does not try to develop a general hypothesis concerning the circumstances under which the United States might be expected to support international enforcement.
The third strand seeks to explain compliance with international agreements.Much research in this strand relates to the controversy between the "enforcement school" (e.g., Barrett, 2003;Downs & Jones, 2002;Downs, Rocke, & Barsoom, 1996) and the "managerial school" (e.g., Chayes & Chayes, 1993, 1995) over the effect of enforcement on compliance.Some scholars also aim to build bridges between rationalist theories (the enforcement school) and constructivist theories (the managerial school) concerning the determinants of compliance (e.g., Checkel, 2001).Other important work in this strand aims to explain compliance with treaties-particularly human rights treaties-that lack enforcement with material consequences (Simmons, 2000(Simmons, , 2009)), or considers how selection effects concerning treaty participation influences compliance (Simmons & Hopkins, 2005;von Stein, 2005).
We share these scholars' interest in international enforcement; however, their work also differs from ours concerning the main focus.While scholars working in strand three focus on the determinants of compliance, we focus on the political feasibility of international enforcement, particularly the determinants of a single country's support (or lack of support) for such enforcement.
A second difference has to do with the fact that several studies in strand three focus on human rights treaties without material consequences for noncompliance. 11In contrast, our hypothesis is developed specifically for treaties that include such consequences, and we leave it for future research to consider whether this hypothesis also holds for treaties that only include provisions for adjudication without any material consequences for noncompliance or nonparticipation. 12inally, a fourth strand of relevant research focuses on the relationship between institutional design and state behavior.For example, Moe shows how institutions and their design are not only expressions of cooperation but also of power (Moe, 2005), while Koremenos, Lipson and Snidal (2001) identify enforcement as one of three main channels through which institutional design might influence such behavior.
The present paper contributes to this fourth strand in two ways.First, our hypothesis helps explain how a country's support (or lack of support) of an international enforcement institution depends on the design of the institution concerned.Second, it highlights the conditions under which a powerful country-the United States-will support some enforcement institutions but not others.

A Simple Hypothesis Concerning U.S. Support for International Enforcement
Underlying most (if not all) treaties is the idea that a mutual exchange of deep commitments will generate net benefits for all member countries. 13However, assuming that other member countries commit deeply and fulfill their parts of the bargain, a given country might benefit even more if it either fails to make deep commitments or fails to fulfill some or all of the deep commitments it does make.Indeed, the possibility that a country might seek to gain a free ride on other countries' efforts provides an important motivation to create international enforcement institutions which can entice member countries to make deep commitments and fulfill them.
We proceed on the assumption that the United States will support international enforcement of a given treaty if (and only if) such enforcement generates net expected benefits for the United States.14Given this assumption, U.S. support for international enforcement of a treaty will depend on the balance of the benefits and costs that the United States expects to incur from such enforcement.U.S. benefits primarily derive from international enforcement's influence on other countries' behavior.The more international enforcement can be expected to change these other member countries' behavior in the desired direction (which may partly depend on the depth of these other members' commitments), the greater the expected benefits of such enforcement for the United States.
Similarly, U.S. costs primarily derive from the international enforcement institution's influence on the United States' own behavior.The more the United States expects international enforcement to influence U.S. behavior in a direction that the United States dislikes,15 the larger the expected costs of such enforcement for the United States.
Table 1 illustrates how U.S. support for international enforcement will vary, depending on the extent to which the United States expects such enforcement to influence its own and other countries' behavior.
First, consider the two cells off the main diagonal.For these two cells, it is straightforward to predict the U.S. position concerning international enforcement.The United States would likely (strongly) support international enforcement if its own behavior may be expected to be largely independent of such enforcement, while other members' behavior may be expected to be substantially influenced by it (see the bottom-left cell in Table 1).
Conversely, the United States would likely (strongly) oppose international enforcement if it had reason to expect that such enforcement would influence its own behavior substantially and in costly ways, while influencing other countries' behavior only moderately (see the topright cell in Table 1). 16or cases that fall in the two cells on the main diagonal, predicting the U.S. position on international enforcement is less straightforward.In either cell, the costbenefit balance could be either positive or negative; thus, whether the United States would support international enforcement might be in doubt.
However, the two cells on the main diagonal differ in at least one important respect.If the effect of enforcement is expected to be modest for all countries (the bottom-right cell) enforcement will be largely pointless.Thus, in the bottom-right cell, we expect the Table 1.U.S. support for and opposition to enforcement by enforcement's expected effect on its own and other countries' behavior.

Substantial Modest
Expected effect of international enforcement on U.S. behavior

Substantial
Moderate support or opposition, depending on the benefit-cost balance

Modest
Strong support Indifference United States to be largely indifferent to international enforcement.
In contrast, if the effect of international enforcement is expected to be substantial for all countries (the topleft cell), enforcement might make a substantial difference to the treaty's effectiveness.Thus, both U.S. benefits and U.S costs might be substantial.If the United States expects its substantial benefits from international enforcement to slightly outweigh its substantial costs, it would then likely moderately support such enforcement.Conversely, if the United States expects its substantial costs from international enforcement to somewhat outweigh its substantial benefits, it would likely moderately oppose such enforcement.
We now develop these initial considerations in more detail and consider how they might throw light on U.S. support (or lack of support) of international enforcement.In doing so, we use the five treaties that constitute this paper's main focus as examples.

When Would the United States Support International Enforcement?
Table 1 suggests that the United States would support international enforcement under (at least) four sets of circumstances.
First, the United States would support international enforcement of treaties for which 1) the United States has only shallow commitments, while 2) some or all other countries have deep commitments, and 3) the United States expects enforcement to enhance these other countries' compliance substantially.Compliance with shallow commitments is by definition nearly costfree; thus, international enforcement under these circumstances would likely entail few (if any) costs for the United States.Simultaneously, the United States would likely benefit substantially from the positive influence of international enforcement on other members' compliance.
Consider the 1989 Montreal Protocol.The construction of this treaty arguably followed a more general U.S. practice concerning international environmental agreements, namely "to act first at home, and then to build on that approach at the international level" (Purvis, 2004, p. 175).Thus, Montreal largely extended to other countries regulations which were similar to those which had already been adopted by the United States.U.S. commitments were, therefore, shallow (they required little policy change); in contrast, commitments were deeper for other countries, which had not yet-when the treaty was being negotiated-introduced regulations similar to those required by Montreal.
Thus, the inclusion of trade restrictions to enforce participation in and compliance with the Montreal Protocol was supported by the United States; indeed, it was largely based on a U.S. proposal (Benedick, 1991, p. 91).As one of our interviewees (I6) put it, "The United States expected it would comply with the freeze and phase-out.Moreover, U.S. business wanted control with the substitutes market.Enforcement was important to us because we had to make sure there was no leakage of banned CFCs from parties in noncompliance or from non-parties.The best way to do this was trade restrictions, so we pursued a ban on trade with non-parties and countries in noncompliance".
The trade restrictions might have positively influenced other countries' willingness to participate in and comply with the Montreal Protocol.Indeed, anecdotal evidence suggests that they have induced some countries to accede to the treaty (after having come to recognize the drawbacks of being excluded from Western markets) and that they have also improved compliance (Brack, 2003, p. 220; see also Aakre, Helland, & Hovi, 2016, p. 1320).Since the trade restrictions could be expected to enhance participation and compliance while having little, if any, influence on U.S. policies, it is hardly surprising that the United States so eagerly supported their inclusion.
Second, we should also expect the United States to support international enforcement of treaties in which the United States 1) has a deep commitment, yet 2) expects to be fully compliant independent of international enforcement, while 3) expecting international enforcement to substantially enhance other countries' compliance.The reason the United States expects to be fully compliant even without enforcement might originate from normative political factors (e.g., political pressure from other countries or from domestic environmental groups) or from institutional features (e.g., the possibility that NGOs would take domestic legal action).
Consider the Kyoto Protocol.In the environmental politics literature, it is commonly held that if the United States ratifies an environmental treaty, it will also comply with it.Most (if not all) of our interviewees seemed to share this view.To name only one example, I4 said, "The United States is often compliant anyway with the treaties we sign and ratify, so the United States will benefit from enforcement of other countries' commitments".
According to Article 6 of the U.S. Constitution, "all Treaties made…under the Authority of the United States, shall be the supreme Law of the Land".17Hence, if U.S. authorities ratify a treaty yet fail to comply with its regulations, they can be sued by, say, an environmental NGO.Such resort to domestic courts to enforce international regulatory standards is far more common in the United States than in, say, European countries (Brunnée, 2004, p. 640).As a result, U.S. negotiators will often be more reluctant than other countries' negotiators to accept stringent international commitments (Wiener, 2003, p. 647).Other things being equal, however, the possibility of domestic enforcement also makes it more likely that the United States would comply with the commitments it does accept.
Hence, during the negotiations over Kyoto's enforcement system, the U.S. administration had reason to believe that-even if the United States were to ratify Kyoto-international enforcement would largely make a difference for other countries' compliance.The strong U.S. support for strict enforcement of Kyoto is therefore understandable.As I11 commented: "If the United States ratifies a treaty, it will likely obey it; that is part of the reason why we will often push for strict enforcement, to ensure that others will also comply". 18To the extent that enforcement would enhance other countries' compliance with Kyoto, the United States would benefit both in terms of reduced global warming and in terms of there being fewer detrimental economic competition effects.
Third, we should also expect the United States to support international enforcement for treaties in which 1) the United States and other countries have deep commitments, 2) the United States expects enforcement to enhance those other countries' compliance, and 3) the United States is either exempt from enforcement or able to block enforcement measures against itself.Again, enforcement of treaties where these conditions hold will promote other countries' compliance (and thus entail positive benefits for the United States), while having little influence on U.S. behavior (and thus entailing few, if any, costs).
Consider the CWC, which gives the Organization for the Prohibition of Chemical Weapons (OPCW) the authority to suspend a noncompliant Party's privileges under the treaty and to recommend sanctions based upon the collective action of the other States Parties (Arti-cle XII).In cases of "particular gravity", the OPCW may also consult with the UN Security Council to request harsher sanctions or even military action if necessary.
Assuming that international enforcement would enhance other States Parties' compliance with the CWC, the United States would benefit because of the reduced risk of exposure to hazardous chemicals for U.S. personnel engaged in U.S. military action abroad.
Concerning U.S. costs from enforcement, it is important to emphasize that U.S. noncompliance with the CWC was a real possibility; indeed, according to one estimate, the United States would not meet the deadline that expired on 31 December 2012. 19Moreover, the Convention's enforcement system does not enable U.S. prevention of all kinds of enforcement against the United States.In particular, the United States may be unable to prevent the OPCW from suspending U.S. privileges under the Convention (such as the right to vote and the right to request an "on-site challenge inspection" of any facility or location controlled by one of the other States Parties).It may also be unable to prevent the OPCW from recommending voluntary sanctions against the United States.However, as a permanent member of the UN Security Council, it can veto use of the OPCW's most potent enforcement measures-mandatory sanctions or even military action imposed through a UN Security Council resolution.In this respect, the United States differs significantly from most other States Parties to the CWC.
The role of the UN Security Council also protects the United States against abuse of the CWC enforcement system.According to I11, such protection was a major concern for the United States: The defense department had no strong interest in chemical weapons, so there was relatively little controversy on the substance.The biggest controversy concerned two points: 1) there should be effective enforcement and monitoring, and 2) there should be protection against abuse of the enforcement mechanisms.
Finally, we should also expect the United States to support international enforcement for treaties in which 1) the United States and other countries have deep commitments, 2) the United States expects enforcement to enhance its own and other countries' compliance, provided that 3) the United States believes that the additional benefits it will derive from the increase in other countries' compliance will outweigh its own additional costs of compliance.
Consider the 1995 creation of the WTO, which significantly strengthened the world trade dispute-settlement mechanisms, compared to previous GATT arrangements (Bello, 1996).The new mechanisms entail, among other things, that a consensus is now required to reject a panel report.Hence, a country can no longer veto a complainant's request for permission to enact countermeasures against it.
The new mechanisms provoked considerable controversy in the United States.Major concerns included the fear that the Dispute Settlement Understanding (DSU) might "threaten U.S. sovereignty and undermine the effectiveness of section 301" (Dunoff, 2009).Ultimately, however, the U.S. administration concluded that the benefits outweighed the costs and the United States became a member from 1 January 1995.
Stricter enforcement of WTO treaties could be expected to cause increased adherence to WTO regulations by other countries, which would entail beneficial competition effects for the United States.On the other hand, it could also entail increased costs, by enhancing U.S. compliance.According to I3, the U.S. administration expected the benefits to outweigh the costs: "It was clear [during the Uruguay Round negotiations] that the United States would win more than it would lose [from strict WTO enforcement], because the United States was [already] complying more than most other countries did".Similarly, I1 said, "The United States would be less likely to violate than other countries and could use the new enforcement system to 'lock in' these other countries".
The views of I3 and I1 mirror those expressed by Bello and Holmer (1994).Stating that the advantages of the DSU would outweigh any U.S. disadvantages, they emphasize that "the United States likely will…be a plaintiff in WTO dispute settlement proceedings at least as often as it proves to be a defendant".They go on to argue that "as the world's largest exporter, the United States has at least as much interest in the international trading system and the WTO as any other nation on earth" (Bello & Holmer, 1994, pp. 1102-1103).
A similar view was also expressed by the Office of the U.S. Trade Representative in the Executive Office of the President of the United States: To ensure that the United States secures the full benefits of the WTO Agreements, the United States sought and obtained a strong, binding and expeditious dispute settlement process for the WTO….As a result, under the WTO we have better enforcement of U.S. rights and more certainty that our trading partners will abide by the rules and open their markets to American exports.(as cited in Bacchus, 2003, p. 440) It is worth noting that the United States has a reasonably good track record concerning compliance with adverse WTO dispute settlement rulings (Bown, 2005;Wilson, 2007).It thus seems that the U.S. views concerning the DSU in the mid-1990s have continued to prevail in the years following the U.S. accession to the WTO treaties.

When Would the United States Oppose International Enforcement?
The logic underlying Table 1 suggests that the United States would decline to support international enforcement under at least two sets of circumstances.
First, we should expect the United States to oppose international enforcement if 1) the United States has deep commitments that are 2) only partly under the control of U.S. authorities, while 3) most or even all other member countries have only shallow commitments.International enforcement may then be expected to entail significant costs for the United States.In addition, international enforcement would require little behavioral change for other countries and would thus produce few (if any) benefits for the United States.Under these circumstances, international enforcement would likely produce net costs for the United States.
Consider the ICC.According to Posner and Yoo (2005b, p. 970), the ICC's members "consist mainly of states who do not expect that their citizens will commit war crimes or human rights violations on foreign soil".If their account is correct, the ICC will require little or no behavioral change by these member countries and will, therefore, impose few if any costs on them.Thus, the benefits provided by the ICC for the United States are likely to be small or even nonexistent.In contrast, the United States and other major powers that "foresee a need to engage in significant military action" (Posner & Yoo, 2005b, p. 970) might well face substantial costs by submitting to international enforcement of war crimes.In addition, the ICC might entail a risk of political prosecution, perhaps even more so than in the case of the CWC.
The desire to avoid these costs helps explain why the United States has declined to submit to ICC enforcement.In the words of I8, "As a great power, the United States is special, with worldwide reach.It is therefore uniquely at risk all over the world".I8 added that there is "real fear among conservatives that the ICC would be used, not only against dictators, but also against the United States, for political reasons".Similarly, I5 said that "as a global power we would be more likely to be exposed to prosecution".I5 added: "As reflected in President Bush's unsigning of the Rome Statute, we might be an attractive target for certain other countries".I11 went even further: The objection was that, rightly or wrongly, you might get political prosecution, such as prosecution of Kissinger for the bombing of Cambodia or prosecution of Obama for the bombing of Libya.Sure, had the UN Security Council been in control, we would likely have agreed to participate in the ICC.But there was widespread fear of political prosecution.People said things like: "If you like Kenneth Starr, you are going to love the ICC." Second, and finally, we should also expect the United States to oppose international enforcement of treaties in which 1) the United States and other countries have deep commitments, 2) the United States expects enforcement to enhance its own and other countries' compliance, but 3) the United States expects that the benefits it will derive from the increase in other countries' compliance will be outweighed by the increase in its own compliance costs.None of the five treaties that constitute this paper's main focus would immediately seem to fit this possibility, at least not at the time when the treaties were negotiated.Following the current protectionist wave in U.S. politics, however, the U.S. calculus concerning international enforcement of the WTO agreements might change, perhaps even tipping the benefit-cost balance to replace U.S. support for international enforcement by U.S. opposition.

Other Explanations
What is the relationship between our hypothesis and other possible explanations of U.S. support for and opposition to international enforcement?In this section, we consider three such other explanations.

Concern for U.S. Sovereignty
Practically all our interviewees mentioned sovereignty issues as being key to understanding U.S. positions on international enforcement.Several interviewees argued that when international enforcement is perceived to be associated with sovereignty issues, it may prevent U.S. support for international enforcement even in cases where such enforcement would entail net U.S. benefits in terms of behavioral change.For instance, I7 argued that our hypothesis "omits an important non-monetary cost: U.S. sovereignty.Ensuring U.S. sovereignty massively outweighs the material or security gain that could be achieved [through an international enforcement system]".Similarly, pointing to the Senate's significant role in the ratification process, I8 commented that "members of the Senate do not base their decisions on rational analysis of costs and benefits; rather, they base their decisions on politics and sovereignty concerns.The administration often faces a big opposition on giving up sovereignty".With specific reference to arms control treaties, I1 stated that "it is a question of how intrusive inspections are, but they will involve sovereignty concessions".
Most treaties require states to give up sovereignty; hence, a key question is why sovereignty issues would bar U.S. support for some international enforcement systems but not for others.Responding to this question, interviewees pointed-directly or indirectly-to the extent to which agreements are clearly defined or openended.This distinction was particularly highlighted by I7, who emphasized that "the United States would…be leery towards treaties with open-ended obligations".Explaining the point, I7 invoked the example of the Arms Trade Treaty: "Key terms in this treaty are interpreted differently; there is no consensus on key terms.Indeed, many core terms cannot be defined precisely.This implies an open-endedness that the United States should not be a party to".
Several interviewees identified a similar problem with the United Nations Convention on the Law of the Sea (UNCLOS).I3 explained: Some in the Senate are concerned about signing the United States up to compulsory dispute settlement under the Convention.The Convention excludes from dispute settlement those disputes that involve 'military activities', but some fear that a tribunal might not honor the U.S.'s assessment of what constitutes a U.S. 'military' activity.
Open-ended agreements are perceived as a sovereignty issue because they imply that "countries do not know what they sign on to" ( I7).For open-ended agreements, international enforcement "would mean that obligations can be designed by third parties" (I7).I2 said that "U.S. decision makers generally dislike the idea of having a committee decide what the United States should do": Similarly, I14 stated that: treaties that give authority to a foreign agency without Senate advice and consent would not be [perceived to be] in the interest of the American people.A widespread view is that sovereign states should not be dictated to by some foreign power or bureaucracy.
Two treaties in particular were often mentioned by interviewees-the Convention on the Rights of Persons with Disabilities (the Disabilities Convention) and UNC-LOS.Interestingly, these treaties share some common features.One is that they do not include international enforcement (in the sense of this paper). 20The Disabilities Convention's Committee on the Rights of Persons with Disabilities monitors compliance; however, it cannot impose any material consequences on noncompliant countries. 21Similarly, although UNCLOS has an elaborate mechanism for peaceful dispute settlement, including various arbitration panels and the International Tribunal on the Law of the Sea, it includes no material consequences to enforce these bodies' verdicts. 22 second shared feature is that existing U.S. policies are largely in line with both treaties' provisions. 23ith regard to UNCLOS, Borgerson contends, "While the United States treats most parts of the convention as cus-tomary law, it remains among a handful of countries…to have signed but not yet acceded to the treaty" (2009, p. 3).Similarly, with regard to the Disabilities Convention, the report that accompanied the Letter of Transmittal (from the President to the Senate) declared that "the United States would be able to implement its obligations under the convention using its vast existing networks of laws affording protection to persons with disabilities.Therefore, no new legislation would be required to ratify and implement the convention". 24The letter further stated that "the provisions of the convention are not self-executing, and thus would not be directly enforced by U.S. courts or of itself [sic] give rise to individually enforceable rights". 25overeignty issues may be particularly prone to be decisive in situations where enforcement's effect on both one's own and others' behavior is expected to be modest (see the bottom-right cell in Table 1).In such situations, a country's position concerning international enforcement-and indeed concerning the treaty itselfwould likely be determined by factors such as principles, and sovereignty might be one such principle.In these two cases, however, sovereignty concerns did not prevent the U.S. administration from supporting the treaties.The (renegotiated) UNCLOS was signed by the Clinton administration in June 1995 and the Disabilities Convention was signed by the Obama administration in June 2009. 26n the Senate, on the other hand, the Disabilities Convention raised concerns over sovereignty.For instance, Senator Inhofe (R-OK) stated: I do oppose the United Nations Convention on the Rights of Persons with Disabilities because I think it does infringe upon our sovereignty, establishing an unelected United Nations bureaucratic body called the Committee on the Rights of Persons with Disabilities and a Conference of States Parties.These unelected bureaucratic bodies would implement the treaty and pass so-called recommendations that would be forced upon the United Nations and the United States if the United States is a signatory. 27n response to this argument and to similar arguments from other senators, Senator Kerry (D-MA) remarked: the Senator mentioned the question of a committee being created, and sometimes committees make recommendations outside of the purview of something.That may be true.But when have words, I ask the Senator-when have words or suggestions that have no power, that cannot be implemented, that have no access to the courts, that have no effect on the law of the United States and cannot change the law of the United States-when has that ever threatened anybody in our country? 28t thus seems that sovereignty may be a decisive factor in cases where both the material costs and the benefits of international enforcement are modest.In the case of the Disabilities Convention, U.S. benefits of international enforcement would be marginal, whereas the political cost of pursuing the issue in the Senate could be significant, which may contribute to explaining why the administration chose not to pursue the advice and consent of the Senate further.
In contrast, when material costs and benefits are substantial, sovereignty appears less likely to be decisive.For example, sovereignty clearly played a role in the debate over U.S. participation in the WTO agreements; however, ultimately concerns over sovereignty gave way to pecuniary considerations.Hence, the United States acceded to the WTO treaties and supported international enforcement of them.Similarly, sovereignty concerns did not prevent the United States from ratifying the Montreal Protocol or the CWC.Nor were concerns about sovereignty a central argument against U.S. ratification of Kyoto; rather, the Senate's resistance was largely based on concerns about Kyoto's likely effect on the U.S. economy.
Thus, our hypothesis facilitates mapping the conditions under which we should expect the United States to accept restraints on its sovereignty.In particular, our hypothesis suggests that the United States would accept such restraints if 1) other countries also accept restraints on their sovereignty and 2) the U.S. restraints are largely formal whereas other countries' restraints are substantive (thereby influencing these countries' behavior in a way that benefits the United States).However, as emphasized by some of our interviewees, open-ended commitments might make it difficult to foresee whether U.S. restraints on sovereignty will prove substantive or largely formal.

Desire to Prevent Infringements on U.S. Constitutional Protection of Individual Rights
One interviewee (I5) suggested that the United States is particularly reluctant to join international enforcement systems for treaties involving individual rights.I5 said that "the United States has problems with international enforcement in cases where individuals are involved".Using the ICC as an example, I5 continued: In the United States, domestic statute is required for criminal prosecution.There are many constitutional protections of individual liberties and criminal en-forcement raises some of those core constitutional protections.Taking away someone's liberty is a very serious matter and that makes the ICC different from international enforcement more generally.
For treaties with enforcement systems that influence individual rights, the costs (i.e., the potential violation of U.S. citizens' constitutional rights) might outweigh the benefits of joining the treaty.A case in point is the ICC, whose jurisdiction implies potential prosecution of individual U.S. citizens in situations where U.S. courts have dismissed the case.One interviewee (I11) maintained that "although ICC prosecution of U.S. personnel would likely be rare [because of U.S. domestic enforcement], there was the possibility of political prosecution.And political cases are exactly the cases we are unlikely to pursue domestically" (emphasis added).
To determine the conditions under which a concern for individual-rights weighs more heavily for U.S. support or opposition to international enforcement than material costs and benefits do, we would need to consider treaties for which the desire to protect U.S. citizens' constitutional rights points in one direction, whereas material costs and benefits point in the opposite direction.For example, we could consider treaties for which the individual-rights explanation points in the direction of U.S. opposition to international enforcement, while U.S. net benefits derived from the influence of enforcement on countries' behavior point in the direction of U.S. support.
We do not exclude the possibility that such treaties actually exist; however, none of the treaties that constitute our main focus satisfy this criterion.In particular, concerning the ICC, the individual-rights hypothesis points in the same direction as our hypothesis does: Both lead us to expect the United States to be reluctant to submit to international enforcement concerning war crimes.Thus, determining which of these hypotheses best explains U.S. nonparticipation in the ICC must be left for future research.

Usefulness of International Enforcement as a Domestic Commitment Device
Referring to U.S. participation in the WTO, some interviewees mentioned that international enforcement can be useful for U.S. authorities as a domestic commitment device.For example, seeing U.S. participation in the WTO even as an instrument to curb U.S. protectionism, I16 emphasized how the risk of penalization could be used in the domestic debate on protectionist measures: "Domestic commitments are important in the case of the WTO.U.S. authorities can say, 'If we do this, then we will be penalized'.So the WTO enforcement system can be used domestically to counter calls for protectionism".This point was also made by I7: "The U.S. turn to free trade was pri-marily to curb U.S. protectionism, which only could be achieved if everyone else also pursued free trade".
The domestic-commitment argument is well known from the literature on trade policy.In the GATT/WTO legal system, it is generally considered desirable to settle disputes through agreement among the parties to the dispute.Nevertheless, Hudec, Kennedy and Sgarbossa (1993) find that almost 50% of the disputes in the 1948-1990 period ended in a ruling rather than in a negotiated settlement.According to them, the reason is political: It may be that defendant governments find it difficult to settle once the complaint is launched.The political costs of agreeing to modify or remove a trade barrier can be quite high.It may be better to fight and lose in a lawsuit because then the unpleasant corrective action can be blamed squarely on GATT law.(Hudec et al., 1993, p. 8) Incorporating this function of enforcement into the calculus underlying our hypothesis would mean adding political benefits, thereby strengthening the overall balance between U.S. benefits and U.S. costs.For enforcement systems that essentially influence other countries' behavior while having little effect on U.S. behavior (bottom-left cell in Table 1), taking this function into account will therefore simply add to the already strongly positive U.S. material net benefits.Moreover, for enforcement systems that essentially influence U.S. behavior (top-right cell in Table 1), the domestic-commitment argument will improve an otherwise negative balance; however, it must carry very high weight to be able to outweigh strongly negative material net benefits.In contrast, for enforcement systems having a significant impact on the United States' and other countries' behavior (top-left cell in Table 1), the domestic-commitment argument might-in some cases-plausibly cause a (slightly) negative U.S. material benefit-cost balance to become positive when the domestic commitment effect is also taken into account.
The difference between these three types of case helps explain why the domestic-commitment argument is typically linked to international trade and the WTO treaties (which we have previously placed in the upperleft cell in Table 1), while rarely (if ever) being mentioned for the other treaties that constitute our main focus. 29hus, just as in the case of the sovereignty argument, our hypothesis seems helpful for determining the conditions under which the domestic-commitment argument may be expected to influence the U.S. position concerning international enforcement.

Conclusions
In this paper, we have proposed a simple hypothesis concerning the circumstances under which the United States would likely support international enforcement of treaties.According to this hypothesis, a key determinant is whether the United States can reasonably expect international enforcement to generate net U.S. benefits.Net U.S. benefits are most likely for treaties where international enforcement will largely influence other countries' behavior (in the direction desired by the United States), and least likely for treaties where international enforcement will largely influence U.S. behavior.
Based on this hypothesis, we identified several conditions under which we should expect the United States to support (or oppose) international enforcement and compared these expectations to observations from five cases of international enforcement.The analysis indicates that our hypothesis indeed provides a plausible explanation for why the United States supports international enforcement for the WTO treaties, the Montreal Protocol, and the CWC, why it also supported international enforcement in the Kyoto negotiations, and why it does not participate in the ICC.
Our analysis indicates, moreover, that in situations where material costs and/or benefits are substantial, a concern for these costs and benefits are likely to determine the U.S. response.In situations where material costs and benefits are modest, however, sovereignty concerns will likely determine the U.S. response.Moreover, in situations where U.S. benefits only outweigh U.S. costs by a fairly narrow margin, the desire to use international enforcement as a domestic commitment device may play an important role.
Our hypothesis thus seems to offer a simple yet fruitful baseline for explaining U.S. views on international enforcement.In addition to explaining the U.S. position on international enforcement of the treaties mentioned above, it also provides guidance for demarcating the conditions under which factors such as sovereignty and the usefulness of international enforcement as a domestic commitment device might play a decisive role for this position.
To what extent may our hypothesis be expected to hold also for other countries than the United States?Clearly, the logic underlying Table 1 is generic: The change that international enforcement causes in other countries' behavior tends to generate benefits, while the change it causes in a country's own behavior tends to generate costs.However, the circumstances under which enforcement will generate substantial benefits and only moderate costs (or vice versa) will likely vary from one country to another.First, domestic political and legal institutions differ across countries.For example, few other countries offer possibilities for domestic enforcement of their own treaty compliance comparable to those existing in the United States.Similarly, few other countries can expect to influence the design of a treaty as much as the United States influenced the design of (say) the Montreal Protocol.Also, only four other countries hold veto power in the UN Security Council.
Second, as a rule, major powers such as the United States will likely be able to obtain indemnity from en-forcement more often than other countries will.Exceptions exist, however.Consider the Kyoto Protocol, where enforcement applies only to Annex I (developed) countries.While initially reluctant to accept international enforcement, developing countries turned into strong supporters of such enforcement once it became clear that it would not apply to them (Werksman, 1996, p. 95).An important factor making the developing-country exemption from international enforcement feasible was Kyoto's classification of the parties into Annex B (developed) and non-Annex B (developing) countries, with only the former having binding emissions limitation commitments.
In conclusion, our hypothesis provides a promising starting point even for explaining other countries' views on international enforcement.Thus, it should represent a step forward for mapping the conditions under which the incorporation of enforcement measures in treaties is likely to be politically feasible.However, in applying our hypotheses to other countries, one should always be careful to consider the international position and domestic institutions of the country under consideration, as well as the specifics of each treaty.