Abstract

When international institutions obligate their members to impose economic sanctions against a target state, how much do those sanctions obligations actually impact their members' behavior? To date, the consensus view has treated all international institutions as if they are equally capable of making multilateral sanctioning efforts more effective. Building upon the enforcement theory of sanctions cooperation, we instead theorize that the ability of international institutions to constrain their members from engaging in spoiler behaviors degrades the larger they are. We hypothesize that sanctions obligations imposed by smaller-sized institutions are more effective at preventing their members from becoming extensive trade-based sanctions busters than those imposed by larger ones. We test our hypothesis via a quantitative analysis of how the involvement of five different international institutions in sanctioning efforts influenced their members' likelihoods of sanctions-busting. We find that only the smaller-sized institutions we examine appear capable of constraining their members from undercutting sanctioning efforts. Notably, we find no evidence that the United Nations' sanctions actually prevent its members from sanctions-busting.

When a country is sanctioned, what role do international institutions play in affecting whether other countries in the world support the sanctioning effort or undercut it? While common wisdom suggests that international institutions play a major role in enhancing the effectiveness of economic sanctions,2 variation appears to exist in the extent of their contributions. For instance, when the United Nations (UN) General Assembly called upon its members to impose economic sanctions against South Africa for its policy of Apartheid in 1962, the measure appeared to enjoy widespread support. Yet a substantial number of member countries, including the United States, United Kingdom, and Germany, still actively engaged in trade-based sanctions-busting on South Africa's behalf. In contrast, when the Arab League imposed sanctions against Egypt following the Camp David agreement in 1978, none of its members undercut the sanctioning effort. Since the presence of even a small number of extensive trade-based sanctions busters can dramatically undermine the chances of sanctioning efforts being successful (Early 2011), understanding which international institutions can effectively prevent their members from engaging in such spoiler behaviors is of significant importance. In this paper, we develop an explanation for why the sanctions imposed by some international institutions appear far more effective than others at preventing their members from becoming trade-based sanctions busters.

The literature on economic sanctions has mainly focused on how international institutions can contribute to the success of multilateral sanctioning efforts (e.g., Drury 1998; Drezner 2000; Hufbauer, Schott, Elliot, and Oegg 2007; Bapat and Morgan 2009). According to these perspectives, international institutions help solve the cooperation challenges associated with imposing multilateral sanctions that involve large numbers of states. To date, however, scholars have treated international institutions as homogenous entities that have homogenous effects on international cooperation irrespective of the size of the sanctions coalitions they are responsible for marshalling. Since international institutions vary in size from 3 to 193 states, the potential challenges individual institutions face in coordinating, monitoring, and enforcing their sanctioning efforts can vary widely depending upon their size. The depth of the sanctions commitments that international institutions obtain from their members may also differ based on their size and memberships. This suggests that a more nuanced explanation is needed of how differences in the international institutions involved in imposing economic sanctions can affect their members' willingness to cooperate with sanctioning efforts.

In this study, we focus on explaining the effects that international institutions' participation in sanctioning efforts have on whether states engage in extensive trade-based sanctions-busting on a target's behalf. Often, third-party states can do far more damage to sanctioning efforts by actively undercutting them than they can contribute to their success by participating in them (Early 2011). International institutions' ability to prevent states from sanctions-busting can be a crucial contributor to a sanctioning effort's potential success. Building off of Daniel Drezner's (2000) enforcement theory, we develop a theoretical framework that explains how the size of international institutions contributes to the quality of cooperation their participation provides to sanctioning efforts. We argue that smaller international institutions can obtain deeper sanctions commitments from their members and more effectively monitor and police those commitments than larger international institutions. As such, the sanctions imposed by small international institutions should be far more effective at preventing their members from becoming trade-based sanctions busters than those imposed by larger-sized institutions.

We evaluate our theory via a quantitative analysis of the factors that affect states' likelihoods of sanctions-busting. Using data on trade with sanctioned states from 164 sanctions episodes and on the most active international institutions involved in imposing sanctions (the Arab League, European Community/European Union [EC/EU], Organization of American States [OAS], Organization of African Unity [OAU], and UN), we analyze the conditional effects of institutional sanctions on states' likelihoods of sanctions-busting. Consistent with our theory's expectations, our analysis reveals that only the smaller-sized EC/EU and Arab League proved capable of preventing their members from becoming extensive trade-based sanctions busters. Notably, we find no evidence that UN sanctions constrain its members from engaging in trade-based sanctions-busting.

Our study makes several important contributions. Foremost, our findings indicate that not all international institutions equally contribute to the effectiveness of sanctioning efforts—at least in the realm of preventing spoiler behaviors. More generally, they also support the conclusion that larger-sized multilateral sanctioning coalitions are not always more effective (Miers and Morgan 2002). Our findings suggest that scholars may need to rethink how international institutions actually contribute to the effectiveness of sanctioning efforts. They also add to the ongoing debate over the tradeoffs between obtaining broad but shallow versus narrow but deep cooperation via international regimes. From a practical perspective, policymakers seeking international support for their sanctioning efforts can benefit from knowing which institutions' participation is most useful in preventing sanctions-busting. Our findings also raise interesting issues related to forum-shopping behavior, as policymakers could be/should be focusing their efforts on recruiting the institutional support of only the institutions whose involvement can best contribute to their sanctioning efforts' effectiveness.

Our paper proceeds as follows. We begin by discussing the existing literature on economic sanctions and the role international institutions can play in affecting whether third-party states undercut or support sanctioning efforts. Next, we present our theory on the role of institutional size in affecting the efficacy of institutionally imposed sanctions. We then present our research design, conduct our empirical analysis, and discuss our findings with respect to each of the international institutions we evaluate. Our conclusion discusses the theoretical and policy implications of our findings.

Economic Sanctions, Spoiler Behaviors, and International Institutions

Economic sanctions are restrictions that governments place upon their constituents' ability to engage in commercial or financial transactions with a target entity in order to alter its behavior. They succeed when they impose high enough costs that their targets prefer to bear the costs of acquiescing to their senders' demands than to allowing the sanctions to persist. As a coercive policy tool, economic sanctions can be employed in support of or as an alternative to the use of military force. Economic sanctions can be very costly due to the disruptions they inflict on participants' commerce and their other negative externalities. Since many international institutions do not have the authority to authorize their members' use of military force, economic sanctions often constitute the strongest coercive policy they are capable of enacting. The overarching track-record of economic sanctions suggests many obstacles exist to their success: depending upon the source one consults, politically motivated economic sanctions only appear to achieve their goals about 23–34% of the time (Hufbauer et al. 2007; Bapat and Morgan 2009:1082).

Given the limited effectiveness of economic sanctions, much of the sanctions literature has focused on the factors associated with why they succeed or fail. This work has shown that a number of factors, including the type of government a target state possesses, the costs the sanctions impose on their target, the type of sanctions imposed, their objective, and the length of time they persist, can affect sanctions outcomes (e.g., Hufbauer, Schott, and Elliott 1990; Drury 1998, 2005; Drezner 2000; Allen 2005; Ang and Peksen 2007; Hufbauer et al. 2007; Lektzian and Souva 2007; Bapat, Heinrich, Kobayashi, and Morgan 2013). Of particular salience to our inquiry is the work that has explored the role played by third-party states3 in either undercutting the senders' sanctioning efforts or cooperating with them.

Research on the effects of states undercutting sanctioning efforts has shown that spoiler states can have a profound effect on sanctions outcomes. Sanctions busters are states that dramatically increase their trade and/or foreign assistance with target states after they have been sanctioned (Hufbauer et al. 2007; Early 2009, 2011). Much of the early work on this topic explored the effects of foreign assistance provided by so-called “black knights,” but this research yielded few consistent findings (e.g., Hufbauer et al. 1990; Drury 1998; Lektzian and Souva 2007). In contrast, recent work focusing on the role played by sanctions-busting trade has shown that it has a major impact on sanctions outcomes. McLean and Whang (2010), for example, find that when a target's largest pre-sanctions trading partners increase their trade with it following sanctions, the sanctions are less likely to succeed. In examining the impact of extensive trade-based sanctions busters, Early (2011) demonstrates that the likelihood sanctions succeed dramatically declines as the number of third-party states providing targets with extensive amounts of sanctions-busting trade rises. As his research shows, only a handful of sanctions busters can make it highly unlikely that a sanctioning effort will succeed.

Understanding the motives that drive sanctions-busting trade is vital to being able to determine how much of that support a target state is likely to receive. Commercial motivations are at the heart of what drives most private actors to engage in sanctions-busting trade, whereas government interests in sanctions-busting are shaped by both the commercial and political benefits it can provide. Early (2009, 2012) has shown that most of the sanctions-busting trade conducted by third-party states is driven by commercially motivated firms and traders. This means that the more profitable the trade opportunities with target states are, the more likely third-party sanctions busters will emerge to exploit those trade opportunities (also, see Kaempfer and Lowenberg 1999; Drezner 2000). Whereas most of the actual third-party trade conducted with target states is conducted by the private sector, individuals and firms respond to governmental policies that can make their sanctions-busting trade more or less profitable. They can also lobby governments to adopt favorable policies that support their trade with target states, or lobby to prevent their governments from actively joining sanctioning efforts. This suggests that firms with an interest in sanctions-busting will tend to concentrate in third-party venues that offer them the most profitable, reliable means of exploiting the trade opportunities vis-à-vis a particular target state (Early 2009).

The potential role that third-party states can play in undermining sanctions is one of the primary motivations that sender states have in seeking multilateral cooperation for their sanctioning efforts (Martin 1992:3, 4). Intuitively, one would expect that multilateral sanctions should be more effective than unilateral sanctions because the costs they impose upon target states should be greater. Yet, one of the puzzling findings to arise from early research using the Hufbauer et al. (1990) data set is that unilateral sanctions appeared more effective than multilateral sanctions. A number of explanations have been leveled to explain these findings, many of which focus on the role played by international institutions in influencing the impact of multilateral cooperation (e.g., Drury 1998; Kaempfer and Lowenberg 1999; Drezner 2000; Miers and Morgan 2002; Bapat and Morgan 2009).

The spatial theory (Bapat and Morgan 2009) and enforcement theory (Drezner 2000) have emerged as two of the leading explanations of how international institutions affect whether multilateral cooperation helps or hinders sanctioning efforts. Both theories contend that international institutions affect sanctions' success by placing constraints upon their members' ability to undercut sanctioning efforts and presume that the costlier sanctioning efforts are to their targets the more likely they are to succeed. They also both treat international institutions as if they are all equally effective in preventing spoiler behavior. Each theory, though, emphasizes different causal mechanisms by which institutions could influence their members' behavior, leading them to make different predictions about institutions' exact effects on sanctions' success.

The spatial theory views sanctioning episodes as a form of hostile negotiations (Miers and Morgan 2002; Bapat and Morgan 2009). According to this perspective, as the issues under dispute between the sender and target states grow more complex, it becomes increasingly difficult for multilateral sanctions coalitions to remain united compared to unilateral sanctioning efforts. When participant states undercut their coalition partners (e.g., by engaging in sanctions-busting trade with them), it can undermine their bargaining position vis-à-vis the target and negatively impact the sanctions' chances of being successful. Bapat and Morgan (2009) argue that the involvement of international institutions in multilateral sanctioning efforts can help coalition members coalesce around a united bargaining position when multiple issues are at stake vis-à-vis target states, thus contributing to the success of multilateral sanctioning efforts. Bapat and Morgan (2009) find support for their spatial theory-based hypotheses via their analysis,4 but they do not evaluate the direct effects that institutional involvement have on state behavior.

The enforcement theory emphasizes the fact that participating in sanctioning efforts is costly for both governments and their constituents (Drezner 2000). Consistent with the discussion of sanctions-busting trade above, it presumes that the greatest threat to the success of sanctions is that profit-driven actors will undercut sanctioning efforts by exploiting the lucrative commercial opportunities they create. Summarizing the challenges to cooperation inherent in multilateral sanctioning efforts, Drezner (2000: 83) argues that

[P]ressures to defect from the agreed-upon sanctions should be more acute as the size of the sanctions coalition increases. As the number of cooperating states increases, so do the economic rents that could be accrued for defecting. Monitoring also becomes more difficult as the number of states increases. The importance of each secondary sender decreases as the number of senders increases. Therefore, the temptation to defect would be particularly acute at high levels of international co-operation.

Drezner's account emphasizes the dual challenges of preventing sanctions-busting caused by private actors circumventing their governments' sanctions and governments formally reneging from their commitments to participate in sanctioning efforts (Drezner 2000:84–86). He argues that institutions can help prevent sanctions-busting conducted by private actors via the enhanced monitoring and enforcement capabilities they provide, making sanctions-busting more difficult and costly. To prevent partner governments from reneging or “backsliding” on their sanctions commitments, Drezner theorizes that international institutions can be used to provide states with side-payments and offer reassurance that fellow participants are following through on their commitments. On the basis of this account, Drezner hypothesizes that multilateral sanctions that are supported by international institutions should be significantly more successful than multilateral sanctions efforts without them, which in turn should also be less successful than unilateral sanctions. Drezner's empirical analysis supports those hypotheses, but he also never evaluates the direct implications of his theory on state behavior.

The existing literature thus indicates that third-party sanctions busters can significantly undermine the effectiveness of sanctioning efforts and assumes international institutions contribute to sanctioning efforts by inhibiting such spoiler behaviors. Whether all or even some international institutions inhibit spoiler behaviors is actually an open empirical question, though. Evaluating the microlevel effects of institutionally imposed sanctions can thus offer valuable insights into the causal mechanisms underpinning the leading explanations for how international institutions contribute to sanctions' effectiveness.

Why Size Matters: Institutionally Imposed Sanctions and the Prevention of Sanctions-Busting

In developing our theory of why only certain international institutions appear effective at preventing sanctions-busting, we build on the core arguments developed within Drezner's (2000) enforcement theory but extend them in important ways. We argue that the ability of international institutions to prevent their members from sanctions-busting degrades as their size increases, because (i) larger institutions face difficulties in imposing deep, costly sanctions due to their memberships' greater heterogeneity; (ii) they confront greater monitoring and enforcement challenges; and (iii) they are less able to compensate adversely affected members in order to prevent them from backsliding. Our account thus addresses the role played by institutional size in affecting sanctioning commitments and the monitoring and enforcement of those commitments.

International Institutions' Role in Preventing their Members from Sanctions-Busting

The extent to which countries can support or undercut sanctioning efforts exists on a continuum. Third-party governments can shallowly cooperate with sanctioning efforts by imposing weak sanctions that they weakly/moderately enforce or deeply cooperate with sanctioning efforts by imposing severe sanctions that they assiduously enforce. On the other hand, they can undercut sanctioning efforts by allowing or even encouraging their constituents' trade with the target to increase in response to the commercial opportunities created by the sanctions. States that are not the sanctions' primary senders can thus respond to a sanctioning effort by helping it, hindering it, or remaining indifferent to it.

Assuming that leaders of third-party states must balance their priority of staying in power domestically along with their countries' foreign policy prerogatives,5 it can be very difficult to convince the leaders of states whose constituents can substantially profit from sanctions-busting trade to restrict such relationships. For third-party leaders, the costs of imposing and enforcing economic sanctions grow larger the greater the disruptions they inflict and the higher the opportunity costs of refraining from sanctions-busting are. Beyond the material losses that participating in sanctioning efforts can cause, adversely affected constituents can politically mobilize against leaders that support the costly sanctions policies. This creates tensions for third-party leaders that must balance between various sources of external pressure to cooperate with sanctioning efforts versus their constituents' interests in undercutting them. As Early (2012) has demonstrated with respect to allies of sender states, third-party leaders tend to give preference to their constituents' commercial interests in sanctions-busting over their foreign policy interests in supporting sanctioning efforts when their constituents can substantially profit from sanctions-busting. Because the costs of imposing and enforcing economic sanctions can be so significant for third-party leaders, the primary backers of sanctioning efforts often turn to the use of coercion or other institutionally based methods of recruiting and enforcing multilateral cooperation for their sanctioning efforts (Martin 1992).

International institutions can become involved in sanctioning efforts when sanctions' primary senders seek to use institutions as vehicles for obtaining broader multilateral support for their sanctioning efforts or when sanctions emerge organically from within an institution's membership. We view international institutions as being reflective of at least some subset of their members' interests, which can vary from institution to institution. In the UN, for example, both the General Assembly and the Security Council have been involved in obligating their members to impose sanctions. While the sanctions imposed by the UN Security Council reflect a smaller subset of countries' interests within the institution, its sanctions carry with it much stronger obligations to comply than those passed by the General Assembly Resolutions (Hufbauer et al. 2007). Members of an institution may also be coerced into participating in sanctioning efforts by their primary backers (Martin 1992). The United States, for example, actively coerced numerous OAS members to vote in support of institutional sanctions against Cuba in 1962. Once in place, sanctions obligations can often be maintained despite significant objections from portions of an institution's membership. For instance, a significant number of UN members advocated for the removal of the UN-imposed sanctions against Iraq during the late 1990s, but the United States was able to block those efforts within the Security Council.

Institutionally imposed sanctions thus represent an at least partially collective interest within an institution's membership for obtaining the sanctions' goals vis-à-vis the target state, but also serve to influence the behavior of members that would otherwise not independently cooperate with sanctioning efforts on their own accord. It is this latter group that sanctions busters emerge from and, in turn, the group that institutionally imposed sanctions could have the most significant impact upon. Most of the time, only a small number of countries within an institution will possess the commercial and political profiles to be highly profitable sanctions-busting venues. Yet, since only one or two trade-based sanctions busters can undermine the effectiveness of sanctioning efforts, the ability of international institutions to limit the emergence of these spoilers constitutes one of the most significant contributions they can make to sanctioning efforts' success. For these reasons, our theory focuses on international institutions' ability to prevent this type of spoiler state from emerging.

While the leading theories (e.g., Drezner 2000; Bapat and Morgan 2009) argue that institutions make multilateral sanctioning efforts more effective by helping countries overcome the challenges of cooperating with large numbers of actors,6 we expect that coalition size still poses a significant impediment to sanctions cooperation even when it is supported by international institutions. Smaller international institutions should be better able to obtain deep, binding commitments from their members to impose sanctions than larger institutions. The broader and more diverse a set of states seeking to reach a multilateral cooperative agreement is, the shallower the agreements they are apt to reach (Downs, Rocke, and Barsoom 1998). Just because institutions are small does not mean that their members will share consensus views with respect to employing sanctions: the smaller institutions are, though, the more homogenous their members' preferences are apt to be.7 To get approval for sanctions in larger, more diverse institutions, sanctions advocates within the institution may have to settle for weak, watered-down sanctions obligations with little monitoring or oversight. Recalcitrant states in larger institutions may also disingenuously agree to support the adoption of sanctions with the expectation that they will not actually have to comply with them.8 All else being equal, we expect that sanctions imposed by smaller-sized international institutions will tend to entail deeper, more constraining sanctions obligations than those imposed by larger-sized institutions, and their members will tend to be less internally divided over the issue.

The effectiveness of the monitoring and enforcement mechanisms by which international institutions can prevent extensive trade-based sanctions-busting by their members will decline as the number of members they have rises. Preventing private sector sanctions-busting by members of a sanctions coalition poses significant monitoring and enforcement challenges that only grow larger the more states participate in a multilateral sanctions coalition. Sanctions-busting traders are adept at masking the nature of their transactions, working through middle-men, and using creative ways of financing such transactions. They can also respond flexibly to crackdowns and readily shift their activities to states in which oversight and enforcement is more lax (Early 2012). Even when institutionally supported sanctioning efforts have possessed robust forces capable of monitoring a target state's land and sea borders, such as the UN-backed sanctions against Serbia and Iraq, significant sanctions-busting can still take place (Andreas 2005; Duelfer 2005). Beyond interdicting sanctions-busting transactions, holding perpetrators legally accountable for their activities is made more difficult by the fact that nearly all institutionally imposed sanctions require domestic implementing legislation or regulations to take effect within participating countries. Whereas the obligation to impose sanctions may come from an international body, the implementation, monitoring, and enforcement of the sanctions occur primarily at the national level (see Morgan and Bapat 2003). The constituents of countries with significant incentives to sanctions-bust will not only pressure their governments to allow their noncompliance but will also discourage activities that allow for that noncompliance to be monitored. If participant governments are unwilling to enforce their sanctions obligations on their own citizens, international institutions can do little to punish perpetrators on their own. All this suggests that, even in best-case scenarios, international institutions face substantial challenges in preventing private sector sanctions-busting from taking place.

The challenges associated with monitoring private sector sanctions-busting are apt to overwhelm the capabilities of international institutions the larger they become. These challenges are significant both when the institution's bureaucracy is charged with monitoring sanctions compliance, such as with UN Sanctions Committees, and when monitoring relies on self-reporting by member states. If we assume that the resources provided to sanctioning efforts do not grow commensurately with each additional member that an international institution has, the resources available to monitor each state's compliance will grow scarcer the larger it is. Furthermore, the challenge of organizing and managing monitoring activities grows more complex the more countries an institution is tasked with monitoring, especially if member states lack the domestic capacity to monitor their own citizens' compliance with sanctions obligations. Even tracking self-reporting by member states can be difficult when large numbers of states are involved, as ensuring that all states provide accurate, up-to-date information and analyzing that data are time- and labor-intensive endeavors. Identifying noncompliant members, especially if they seek to mask their sanctions-busting, will thus grow harder the larger an international institution is.

Cost-sensitive sanctions-busting traders recognize that their likelihood of being caught and punished for violating institutionally imposed sanctions is lower the larger the international institution that imposes them is. This is especially true if their host governments are not committed to enforcing them. Particularly in the absence of institutionalized monitoring and enforcement efforts to ensure member compliance, governments will often be subject to less individual pressure to fully enforce their sanctions obligations the larger the institution that imposed them is. At the firm-level, sanctions-busting traders also have more options for switching their activities to other noncompliant member states in larger international institutions than smaller ones if their host government cracks down on them. There are more likely to be higher numbers of weakly committed states in large institutionalized sanctioning coalitions, giving them more options for shifting their sanctions-busting activities to fellow member states as opposed to nonmembers. Firms also have less chance of having their sanctions-busting activities being observed and reported by larger institutions, further reducing the risks associated with flouting their home governments' institutional sanctions obligations.

The ability of international institutions to prevent governmental backsliding via side-payments also declines the larger an institution is. Since international institutions often have scant resources to devote to compensating members for their participation, competition for those scarce resources will be greater in larger institutions. The UN Charter, for example, has a provision built into Article 50 of Chapter VII that allows states to seek compensation for damages done by UN-imposed sanctions. In practice, the use of this provision has been stymied by the challenges entailed in assessing appropriate levels of compensation, coordinating its distribution, and the lack of sufficient resources to compensate all the states that deserve it. During the UN sanctions imposed against Iraq (1990–2003), 21 states petitioned for compensation. This far outstripped the UN's available resources. Following up on their requests, all 21 states joined in a letter to the Security Council indicating that the UN's commitments to support the damages its sanctions caused “have not evoked responses commensurate with the urgent needs of the affected countries” (UN 1996:29). Given that sanctions-busting firms can also just shift their activities away from states that increase their enforcement activities, larger institutions also face greater challenges in using targeted side-payments to gain the compliance of recalcitrant participants because those targets can continuously shift.

Finally, the role international institutions can play in reassuring states of their peers' compliance is likely to have the strongest effects in small institutions. To the extent that members are complying, states should be far better able to ascertain that fact when they only have to evaluate the behavior of a small pool of states—especially when the institutional sanctions do not include monitoring provisions (e.g., Olson 1965:45). And, given that the likelihood that members will sanctions-bust in larger institutions is higher than in smaller ones, this mechanism is apt to be less relevant in larger-sized institutions. For this reason and all of the ones discussed above, we hypothesize that economic sanctions imposed by smaller-sized international institutions are more effective at preventing their members from engaging in extensive trade-based sanctions-busting than those imposed by larger institutions.

Research Design

To test our hypothesis, we conduct a quantitative analysis of the factors affecting the likelihood that third-party states will become extensive trade-based sanctions busters on behalf of target states. We construct a time-series cross-sectional data set using 164 sanctions episodes imposed from 1950 to 2002 in the Hufbauer et al. (2007) data set. Yearly observations of the directed dyadic relationship between target and third-party states are our unit of analysis.9 Our coding method denotes the aggregate conditions created by multiple sanctions episodes being in place against a target state in a given year rather than coding separate observations for each sanctions episode. We restrict our analysis to those observations in which third parties and target states had active trading relationships.10 The dependent variable of our analysis is a dichotomous variable that denotes whether a third party engaged in extensive sanctions-busting on a target state's behalf in a given year. We conduct our analysis using binomial probit estimation with standard errors clustered by the target country for each of the five international institutions we analyze. To account for temporal dependence, we also include variables to account for the number of years when no sanctions-busting took place and its squared and cubed values (No Bust, No Bust2, No Bust3).11 In the sections below, we describe how we coded our dependent variable, our institutional membership and sanctions variables, and the controls we included to account for potentially confounding factors.

Coding Sanctions-Busting Behavior

We adopt the same approach taken in previous studies to denote that a third-party state engaged in extensive sanctions-busting trade on a target state's behalf (Early 2011). Two primary criteria are used to identify sanctions-busting behavior: a significant spike in the amount of bilateral trade conducted between a third-party and target state and that trade's occurrence in sufficient quantities to substitute for the target's trade losses experienced due to the sanctions. To account for the first criterion, separate variables were created to account for the change in the current year import and export flows between target and third-party states relative to their values from the previous year.12 Observations were flagged in which either import or export flows increased by more than 5%. Next, observations were identified in which the combined bilateral trade conducted by a target with a third party constituted at least 5% of the target's total yearly trade. To identify observations of sanctions-busting behavior, we flagged those observations in which both criteria were met and then flagged all subsequent years in which the respective import or export values continued to grow and the third party's share of the target's total trade remained above 5%. Our Sanctions-Busting variable denotes those observations in which the sanctions-busting criteria we set were met versus not (0 = No; 1 = Yes). Consistent with previous studies' findings (Early 2009), our coding procedures indicate that sanctions-busting occurred in roughly 2.6% of the observations for which trade data were available.

Coding Institutional Membership and Institutional Sanctions

We seek to account for the state-level impact of institutionally imposed economic sanctions by accounting for whether a given international institution obligates its members to impose sanctions and whether a given state is a member of that institution. Controlling for the particular characteristics and membership profile of an international institution is essential to assessing whether its imposition of sanctions alters the baseline set of behaviors that can be expected of its members in the absence of sanctions. To identify the international institutions involved in particular sanctions episodes, we consulted the list of primary senders and cooperating institutions compiled by Hufbauer et al. (2007) as part of their sanctions data set. Because the Hufbauer et al. data were collected at the sanctions episode level, the data set does not provide information on the specific years that participating institutions obligated their members to impose sanctions against target states. As such, we conducted individual case inquiries into each international institution listed as participating in sanctions episodes. We found that while some cooperating international institutions listed by Hufbauer et al. created specific obligations for their members to impose sanctions against target states, others only offered moral support for the sanctions or imposed nonsanctions related penalties.

Using the information provided in Hufbauer et al.'s supplemental case histories and a wide collection of other resources, we coded new variables to denote the specific years in which individual international institutions adopted policies that created an obligation for their member states to impose sanctions. Of the institutions identified in the data set, the UN, EC/EU, OAS, OAU, and Arab League imposed sanctions the most actively. Institutions like ASEAN, the Commonwealth of Nations, and MERCOSUR sometimes supported sanctioning efforts, but they rarely obligated their members to actually impose sanctions. Focusing on the top-five sanctioning institutions provides us with 64 distinct episodes of international institutions obligating their members to participate in sanctioning efforts, broken down as follows: EC/EU (30 episodes), UN (22 episodes), OAS (four episodes), OAU (four episodes), and the Arab League (four episodes). Given the prevalence of cases involving the EC/EU and UN, the results concerning these institutions should be accorded the most evidentiary weight. Temporally speaking, it is also notable that these institutions became significantly more active in imposing sanctions after the Cold War's conclusion than during it, especially the EC/EU and UN. To operationalize our institutional sanctions variables, we code dichotomous variables for whether each institution imposed sanctions obligations upon its members against a target state in a given year (0 = no; 1 = yes). Overall, approximately 16% of the third-party states in the data set we analyze are subject to institutional sanctions obligations imposed by the Arab League, EC/EU, OAS, OAU, or UN.

To account for states' institutional affiliations, we code variables to denote each state's membership status with respect to the five institutions listed above. Importantly, being party to an international institution can exercise independent effects on how countries respond to sanctioning efforts. Institutional memberships can influence country's domestic and foreign policies and influence their capabilities and the political and commercial networks they have access to. As our theory focuses on, the size of a particular institution will also condition the effects of the sanctions obligations it imposes. As Figure 1 illustrates, membership in these institutions changed over time but distinctions exist with respect to their general sizes. Membership in the UN was most extensive, followed by the OAU, the OAS, the Arab League, and EC/EU. Over the periods analyzed, these institutions averaged the following number of members: EC/EU (9.7), Arab League (15.4), OAS (27.0), OAU (46.9), and UN (127.6). These figures indicate that the EC/EU and Arab League are comparatively smaller than the other institutions, while the UN is significantly larger than the rest of the institutions.

The Membership Counts of the Five International Institutions, 1945–2004

The Membership Counts of the Five International Institutions, 1945–2004

Recognizing that institutional size exists on a continuum, we designate the EC/EU and Arab League as constituting smaller-sized institutions and the OAS, OAU, and UN as large institutions.13 Controlling for institution-specific effects, we expect that the small institutions will be best able to constrain their members from sanctions-busting when they impose sanctions while the larger institutions will be unable to prevent it. In terms of its size, the UN represents the comparatively best test of our prediction that sanctions imposed by larger-sized institutions do not effectively prevent their members from sanctions-busting. The sizes of the OAS and OAU are not as definitively large as the UN, but we still expect them to be less capable of preventing trade-based sanctions-busting than the Arab League and EC/EU. The (albeit limited) variation we have in each institutional category should provide greater confidence that institutional size is responsible for the differences in institutional effectiveness we hypothesize about if our predictions are correct.

Table 1 illustrates the four different conditional categories we use to test our hypothesis on the effects of institutionally imposed sanctions on states' sanctions-busting behavior, controlling for whether or not states are members of a given institution. To evaluate our theory's hypothesis, we seek to test the conditional effect of each individual institution's sanctions, controlling for whether states are members of them or not. By controlling for membership-specific effects (Member State/Institutional Sanctions versus Member State/No Institutional Sanctions), we can test exactly how much more or less likely an institution's members will be to become trade-based sanctions busters if the institution imposes sanctions obligations. Testing the effects of each institution individually, rather than pooling them together, allows us to control for their specific membership-profiles and evaluate whether the size of an institution influences the impact of their sanctions.

Table of the International Institutions Membership and Sanctions Imposition Interaction

  Is the State a Member of the Particular International Institution? 
  Yes No 
Did the Particular International Institution Participate in the Sanctioning Effort? Yes II. Member State/Institutional Sanctions I. NonMember State/Institution Sanctions 
 No III. Member State/No Institutional Sanctions IV. NonMember State/No Institutional Sanctions 
  Is the State a Member of the Particular International Institution? 
  Yes No 
Did the Particular International Institution Participate in the Sanctioning Effort? Yes II. Member State/Institutional Sanctions I. NonMember State/Institution Sanctions 
 No III. Member State/No Institutional Sanctions IV. NonMember State/No Institutional Sanctions 

The prevalent alternative perspective within the literature does not view the size and/or characteristics of international institutions as meaningfully impacting their effectiveness. To assess whether this view is capable of observing the impact that institutional sanctions have on states' likelihood of sanctions-busting, we code a dummy variable that captures only whether a country is a member of an institution imposing sanctions (Member State/Institutional Sanctions versus all other categories). This variable only accounts for whether states are subject to an institutional obligation to impose sanctions against the target state.

Control Variables

To account for the effects of other factors, we include a number of control variables that are known to influence sanctions-busting behavior. Drawing on the liberal-based explanation of sanctions-busting trade, we include a series of variables to account for the commercial ties and profiles of the third-party and target states (Early 2009, 2012). To account for the size of third-party and target states' economies (LnGDPT and LnGDP3), we include the logged values of both countries' current year GDP figures (Gleditsch 2008). We would expect that the larger a target's economy is, the less it will need other states to sanctions-bust on its behalf. Conversely, the larger a third party's economy is, the more capable it will be of engaging in sanctions-busting when an opportunity arises. To account for the extent of a third party's commercial ties with the target (Trade Share3T), we employ a trade share variable that denotes the proportion that bilateral trade between the target and third party constitutes of the total yearly trade conducted by the third party (Barbieri et al. 2008). This variable has been shown to have a potent positive effect on third-party states' likelihood of sanctions-busting on behalf of a particular target state. Lastly, we include a variable to account for the third-party states' degree of trade openness (Trade Openness3), which has been shown to positively affect states' likelihood of sanctions-busting. This variable is coded as the third party's total trade divided by its GDP. All four variables are lagged one year to ensure their exogeneity.

Additionally, we include a variable to account for the political ties between third parties and target states. As previous research has shown, allies are more likely to sanctions-bust on each others' behalf than other states (Early 2009, 2012). We thus code a variable to denote whether the target and third-party states shared a defense pact alliance (Defense Pact3T) in a given year (Leeds et al. 2002). We also included controls for whether the governments of the target and third-party states were democratic or not (DemocracyT and Democracy3) and the effects of joint possession of democratic institutions (Joint Democracy3T) using data from the Polity IV Data Set.14 We expect the joint presence of democratic institutions within the target and third parties to have a positive effect on sanctions-busting.

Lastly, we include a number of controls for the target and third-party states' geographic relationship, the type of sanctions employed, and their duration. Third parties sharing close borders with sanctioned state are more likely to engage in smuggling and black market forms of sanctions-busting that cannot be captured in legitimate, recorded trade figures (Early 2009). Counterintuitively, this implies that neighboring states will be less likely to be observed as being major sanctions busters using measures of legitimate, recorded trade flows. As such, we include a dummy variable Neighbors3T to denote whether countries are separated by 12 miles of water or less (Stinnett, Tir, Schafer, Diehl, and Gochman 2002). Additionally, we also accounted for distance (LnDistance3T) using the logged number of miles that separate the two states (and Bennett and Stam 2000). To denote the severity of the sanctions imposed (Severe Sanctions), we code the combined presence of export, import, and financial sanctions against a target state during a given year as constituting comprehensive, harsh sanctions. This variable is coded as 1 if such sanctions are in place and 0 otherwise, using data from Hufbauer et al. (2007). Lastly, we included a count variable to measure the number of years that sanctions have been in place continuously against the target state (Years Sanctioned).

Discussion of the Results

In Table 2, we present the results of our analysis on the impact that institutionally imposed economic sanctions have on whether states engage in extensive trade-based sanctions-busting. Models 1–5 present the institution-specific results using our categorical variables that capture the conditional effects of each institution's sanctions, controlling for institutional membership. Model 6 employs the pooled independent variable that only codes whether states are party to an institution that has imposed sanctions against a target. We indicate which categories are used as the null baselines for comparison in each model, which for the first five models is Member/No Institutional Sanctions. Our hypothesis thus predicts that Member/Institutional Sanctions should have a negative impact in the Arab League and EC/EU models and no effect in the OAS, OAU, and UN models. We discuss the results for each institution and then discuss their overarching implications in comparison with the results from the pooled analysis. Overall, our results appear quite supportive of our hypothesis that smaller international institutions are more effective at preventing sanctions-busting than larger institutions.

Binomial Probit Estimation of Trade-Based Sanctions-Busting with Clustered Standard Errors

 (1) (2) (3) (4) (5) (6) 
 AL EC/EU OAS OAU UN All IIs 
Member/Institutional Sanctions N/A −0.26*** (0.08) 0.46 (0.59) −0.02 (0.18) −0.03 (0.15) −0.05 (0.12) 
Member/No Institutional Sanctions Null Null Null Null Null Null 
NonMember/No Institutional Sanctions −0.24* (0.13) −0.32*** (0.05) 0.30*** (0.07) 0.24** (0.11) 0.16 (0.10) Null 
NonMember/Institutional Sanctions −0.18 (0.21) −0.42*** (0.10) 1.17*** (0.25) 0.71*** (0.26) 0.55** (0.26) Null 
LnGDPT −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) 
LnGDP3 0.35*** (0.02) 0.34*** (0.02) 0.36*** (0.02) 0.34*** (0.02) 0.35*** (0.02) 0.35*** (0.02) 
Trade Share3T 0.30*** (0.09) 0.30*** (0.09) 0.29*** (0.09) 0.29*** (0.09) 0.30*** (0.09) 0.29*** (0.09) 
Trade Openness3 0.16*** (0.02) 0.15*** (0.02) 0.16*** (0.02) 0.17*** (0.02) 0.16*** (0.02) 0.16*** (0.02) 
Democracy T −0.10 (0.11) −0.13 (0.11) −0.09 (0.11) −0.12 (0.10) −0.11 (0.11) −0.11 (0.11) 
Democracy3 0.22*** (0.06) 0.06 (0.06) 0.19*** (0.06) 0.14** (0.06) 0.17*** (0.06) 0.18*** (0.06) 
Joint Democracy3T 0.23* (0.12) 0.24* (0.12) 0.20 (0.12) 0.27** (0.11) 0.23** (0.12) 0.23* (0.12) 
Defense Pact3T 0.43*** (0.11) 0.45*** (0.10) 0.56*** (0.11) 0.41*** (0.10) 0.44*** (0.10) 0.43*** (0.10) 
Severe Sanctions 0.03 (0.09) 0.02 (0.09) −0.01 (0.09) −0.03 (0.10) 0.02 (0.09) 0.03 (0.09) 
Years Sanctioned 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 
LnDistance3T −0.27*** (0.04) −0.25*** (0.05) −0.26*** (0.05) −0.27*** (0.04) −0.27*** (0.04) −0.27*** (0.04) 
Neighbors3T −1.15*** (0.33) −1.00*** (0.34) −1.13*** (0.34) −1.16*** (0.32) −1.17*** (0.33) −1.17*** (0.33) 
No Bust −0.41*** (0.04) −0.41*** (0.04) −0.42*** (0.04) −0.41*** (0.04) −0.41*** (0.04) −0.41*** (0.04) 
No Bust2 19.43*** (2.69) 19.14*** (2.69) 19.77*** (2.75) 19.21*** (2.77) 19.39*** (2.67) 19.42*** (2.71) 
No Bust3 −0.27*** (0.06) −0.27*** (0.06) −0.28*** (0.06) −0.27*** (0.06) −0.27*** (0.06) −0.27*** (0.06) 
Constant −4.29*** (0.65) −4.08*** (0.64) −4.94*** (0.65) −4.41*** (0.63) −4.42*** (0.65) −4.36*** (0.65) 
Pseudo R2 0.46 0.47 0.47 0.46 0.46 0.46 
Observations 78,140 78,341 78,341 75,740 78,341 78,341 
 (1) (2) (3) (4) (5) (6) 
 AL EC/EU OAS OAU UN All IIs 
Member/Institutional Sanctions N/A −0.26*** (0.08) 0.46 (0.59) −0.02 (0.18) −0.03 (0.15) −0.05 (0.12) 
Member/No Institutional Sanctions Null Null Null Null Null Null 
NonMember/No Institutional Sanctions −0.24* (0.13) −0.32*** (0.05) 0.30*** (0.07) 0.24** (0.11) 0.16 (0.10) Null 
NonMember/Institutional Sanctions −0.18 (0.21) −0.42*** (0.10) 1.17*** (0.25) 0.71*** (0.26) 0.55** (0.26) Null 
LnGDPT −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) −0.08*** (0.02) 
LnGDP3 0.35*** (0.02) 0.34*** (0.02) 0.36*** (0.02) 0.34*** (0.02) 0.35*** (0.02) 0.35*** (0.02) 
Trade Share3T 0.30*** (0.09) 0.30*** (0.09) 0.29*** (0.09) 0.29*** (0.09) 0.30*** (0.09) 0.29*** (0.09) 
Trade Openness3 0.16*** (0.02) 0.15*** (0.02) 0.16*** (0.02) 0.17*** (0.02) 0.16*** (0.02) 0.16*** (0.02) 
Democracy T −0.10 (0.11) −0.13 (0.11) −0.09 (0.11) −0.12 (0.10) −0.11 (0.11) −0.11 (0.11) 
Democracy3 0.22*** (0.06) 0.06 (0.06) 0.19*** (0.06) 0.14** (0.06) 0.17*** (0.06) 0.18*** (0.06) 
Joint Democracy3T 0.23* (0.12) 0.24* (0.12) 0.20 (0.12) 0.27** (0.11) 0.23** (0.12) 0.23* (0.12) 
Defense Pact3T 0.43*** (0.11) 0.45*** (0.10) 0.56*** (0.11) 0.41*** (0.10) 0.44*** (0.10) 0.43*** (0.10) 
Severe Sanctions 0.03 (0.09) 0.02 (0.09) −0.01 (0.09) −0.03 (0.10) 0.02 (0.09) 0.03 (0.09) 
Years Sanctioned 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 0.04*** (0.01) 
LnDistance3T −0.27*** (0.04) −0.25*** (0.05) −0.26*** (0.05) −0.27*** (0.04) −0.27*** (0.04) −0.27*** (0.04) 
Neighbors3T −1.15*** (0.33) −1.00*** (0.34) −1.13*** (0.34) −1.16*** (0.32) −1.17*** (0.33) −1.17*** (0.33) 
No Bust −0.41*** (0.04) −0.41*** (0.04) −0.42*** (0.04) −0.41*** (0.04) −0.41*** (0.04) −0.41*** (0.04) 
No Bust2 19.43*** (2.69) 19.14*** (2.69) 19.77*** (2.75) 19.21*** (2.77) 19.39*** (2.67) 19.42*** (2.71) 
No Bust3 −0.27*** (0.06) −0.27*** (0.06) −0.28*** (0.06) −0.27*** (0.06) −0.27*** (0.06) −0.27*** (0.06) 
Constant −4.29*** (0.65) −4.08*** (0.64) −4.94*** (0.65) −4.41*** (0.63) −4.42*** (0.65) −4.36*** (0.65) 
Pseudo R2 0.46 0.47 0.47 0.46 0.46 0.46 
Observations 78,140 78,341 78,341 75,740 78,341 78,341 

(Notes. *, **, ***denote statistical significance using two-tailed tests at the 90%, 95%, and 99% confidence intervals, respectively.)

Model 1: The Arab League

The Arab League appears to have been remarkably effective at obtaining its members' compliance with its sanctions. No Arab League members engaged in sanction-busting when the institution imposed sanctions. As a perfect predictor of nonbusting, we cannot conduct probabilistic comparisons of Member/Institutional Sanctions versus Member/No Institutional Sanctions. Yet, the finding that no Arab League members undercut their institutional sanctioning obligations strongly supports our first hypothesis. Looking more closely at the cases the Arab League was involved in appears to support our theory of why its sanctions were effective at preventing its members from sanctions-busting. Most of the sanctioning efforts of the Arab League pertained to Israel-related issues, which have enjoyed strong support among its relatively homogenous membership. The institution was able to obtain deep commitments from its members with regard to the sanctions it has imposed and hold its members accountable to them—even on the sanctions imposed against Egypt following the Camp David Accords. Considering the close commercial ties that numerous League members had with Egypt, the strong compliance that the Arab League was able to obtain for its sanctions is quite noteworthy.

Our analysis also reveals that Arab League members are more likely to sanctions-bust than other countries when the institution does not participate in sanctioning efforts. The negative and statistically significantly sign on NonMember/No Institutional Sanctions indicates that members of the Arab League are more likely to become trade-based sanctions busters than nonmembers when their institution does not participate in sanctioning efforts. When the Arab League does not participate in sanctioning efforts in a typical scenario,15 its members appear 90% more likely to engage in sanctions-busting than nonmember states. Looking more closely at the sanctions-busting conducted by Arab League members, it appears as if Saudi Arabia, Iraq, and the United Arab Emirates were the countries most actively involved in it. While all three countries are large fossil fuel exporters, the UAE also has an international reputation for being involved in sanctions-busting trade. The Arab League's ability to constrain its members from sanctions-busting suggests that the institution's sanctions notably constrained its members' behaviors.

Model 2: The European Community/Union

While members of the EC/EU appear to have very strong sanctions-busting orientations, the institution proved capable of constraining its members from doing so when it imposed sanctions. Controlling for the specific effects of being an EC/EU member, the negative and statistically significant effect of Member/Institutional Sanctions indicates that the EC/EU's imposition of sanctions obligations causes a marked reduction in its members' likelihoods of sanctions-busting. In a typical scenario, the EC/EU's imposition of sanctions diminishes its members' likelihoods of sanctions-busting by approximately 39%. Considering that the EC/EU imposed more sanctions than any other institution in our study, this finding has particularly salient implications. The EC/EU's ability to constrain its members from sanctions-busting strongly supports our hypothesis that small institutions can best overcome the challenges of enforcing sanctions.

The negative and statistically significantly sign on NonMember/No Institutional Sanctions indicates that the EC/EU's members appear to have a strong predilection for sanctions-busting if the institution does not support sanctioning efforts. When the EC/EU does not participate in sanctioning efforts, a state is 86% more likely to sanctions-bust if it is an EC/EU member compared to not. Importantly, these effects hold even though the model controls for the effects of economic size, trade openness, regime type, and the strength of a third party's commercial ties to target countries. These findings are consistent with those of previous studies (Yang, Askari, Forrer, and Teegen 2004) and reflect both the strong commercial interests and competitiveness of the EC/EU's members. In the UN-backed sanctioning effort against Rhodesia in 1965, for example, the European Community “not only initially failed to take any action to implement the relevant Security Council Resolutions (‘SCRs’) imposing sanctions but actually included Rhodesia on the list of countries with which trade was liberalized.”16 Lastly, the negative and statistically significant results regarding NonMember/Institutional Sanctions indicate that non-EC/EU members are less likely to sanctions-bust during EC/EU sanctions than EC/EU members when the EC/EU does not impose sanctions. This once again appears to reflect the proclivity of EC/EU members to sanctions-bust compared to other states when they are not constrained by institutional sanctions.

Model 3: The Organization of American States

Consistent with our hypothesis, Member/Institutional Sanctions' lack of statistical significance indicates that the OAS's imposition of sanctions appears to have no effect on its members' likelihoods of sanctions-busting. As a large, heterogeneous institution compromised of most of the countries in the Western hemisphere, the institution had difficulties effectively marshaling sufficient resources to monitor and enforce its members' compliance. In the absence of our theory, the institution's ineffectiveness could be puzzling given the United States' dominant role within organization and that all the OAS-imposed sanctions were championed by the US Government. The OAS's involvement in sanctioning efforts, though, appears to have done little to constrain sanctions-busting among its members.

When the OAS does not participate in sanctioning efforts, the positive and statistically significant findings with respect to NonMember/No Institutional Sanctions indicate that OAS members are significantly less likely to engage in sanctions-busting than nonmembers. OAS members in a typical situation are 60% less likely than nonmembers to engage in sanctions-busting when the institution does not impose sanctions. The positive and statistically significant sign on NonMember/Institutional Sanctions indicates that nonmembers of the OAS are more likely to sanctions-bust during OAS sanctions than its members are when the institution does not impose sanctions. These two results are the opposite of those found in the EC/EU cases.

Model 4: The Organization of African Unity

The nonsignificance of Member/Institutional Sanctions in the OAU model once again supports our hypothesis that large institutions are ineffective at preventing their members from sanctions-busting. It is notable that nearly all of the sanctions episodes the OAU involved itself in concerned African countries. While OAU members may have supported the imposition of sanctions in principle, they were often unwilling or unable to impose them when the opportunity costs were high. In the case of the sanctions imposed against Rhodesia and South Africa, many of the OAU members were still heavily dependent on the target countries for their own economic well-being (Blumenfeld 1991; Strydom and Huaraka 2004). The institution's large size and heterogeneous membership posed significant obstacles to effective coordination of sanctioning efforts. It also lacked the ability to effectively monitor its members' compliance and provide supplemental assistance to countries that needed support in order to comply. All this supports our theory.

The results further indicate that NonMember/No Institutional Sanctions and NonMember/Institutional Sanctions have positive and statistically significant effects on states' likelihood of sanctions-busting. These findings are similar to those from the OAS cases and are the opposite of the EC/EU cases. These results indicate that, in a typical scenario, the OAU's members are 57% less likely to engage in sanctioning-busting than nonmembers when the institution does not participate in sanctioning efforts. They also indicate that non-OAU members are more likely to sanctions-bust during OAU sanctions than OAU members are when the institution does not impose sanctions. Given that the memberships of the OAS and OAU were far less developed than the EC/EU's membership during the periods we analyzed, these cross-institutional differences in their sanctions-busting profiles make sense and demonstrate the importance of controlling for institution-specific factors in our analysis.

Model 5: The United Nations

The UN has a nearly universal membership and is the largest international institution we analyze—providing the strongest test of our hypothesized nonrelationship. The nonsignificance of Member/Institutional Sanctions suggests that, in general, the UN's sanctions cannot constrain its members from sanctions-busting. For policymakers, this potentially calls into question whether seeking the UN's support for sanctioning efforts is worthwhile. Even when broad political consensus appears to exist for UN sanctions, examples abound of member states choosing to ignore their institutional sanctions obligations. In the case of the UN sanctions imposed against Saddam Hussein's regime in Iraq (1991–2003), France, Jordan, Syria, and Turkey, all formed lucrative trading relationships with the country in violation of the UN's sanctions (Duelfer 2005:48). During the UN sanctions imposed against Rhodesia (1965–1979), the country was still able to receive vital petroleum imports from the subsidiaries of US and British oil companies operating out of South Africa (Rowe 2001:133–134). And more recently, the UN Security Council sanctions imposed against the Qaddafi Regime in Libya during 2011 were undercut by Turkey, Kenya, and a host of other African countries (Richter 2011). In the best of scenarios, the sheer size of the UN and the heterogeneity of its members' interests make the monitoring and enforcement of its sanctioning obligations overwhelmingly difficult. That is not to say that UN sanctions never make valuable contributions to sanctioning efforts; rather, our results solely indicate that they generally cannot prevent one of the most problematic behaviors linked to sanctions' failure.

The membership-specific effects of the UN are less salient given the near universal membership the institution has tended to possess since its creation. The insignificance of NonMember/No Institutional Sanctions suggests that the membership effects of being party to the UN are not nearly as strong as they are with respect to the regional institutions we evaluate.17 Similar to the OAS and OAU cases, though, NonMember/Institutional Sanctions is positive and has a statistically significant effect. The paucity of non-UN members that remain in the international system, however, suggests that this finding carries little contemporary relevance.

Comparing the Results across Institutions and to the Pooled Analysis

Our analysis demonstrates that meaningful variation exists in the ability of institutionally imposed sanctions to prevent their members from engaging in trade-based sanctions-busting and supports our hypothesis that institutionally imposed sanctions are less effective when imposed by large international institutions. The smaller-sized EC/EU and Arab League are the only two international institutions whose sanctions imposed meaningful constraints upon their members' sanctions-busting activities. Other than its size, the Arab League's institutional characteristics appear much more similar to the OAS and OAU than the EC/EU. Yet, even though the Arab League's bureaucratic capacity lagged significantly behind that of the EC/EU, its sanctions obligations still imposed meaningful constraints upon its members' sanctions-busting behavior. Consistent with our expectations, the larger institutions we analyzed (the OAS, OAU, and UN) did not appear capable of preventing their members from engaging in sanctions-busting. The fact that, despite their significant differences, the constraining impact of institutionally imposed sanctions broke down in these three cases suggests that a common factor, such as their size, was responsible. Involving larger institutions may mean greater numbers of participants in sanctioning efforts (e.g., Martin 1992), but the inability of those institutions to prevent defections by their members appears to undercut their usefulness.

As a final test of our hypothesis and method of analysis, we evaluate whether the traditional approach of treating all sanctions imposed by international institutions homogenously can capture any sort of aggregate impact on states' likelihood of sanctions-busting. We use a pooled dummy variable in Model 6 that denotes whether or not a country was part of an international institution that imposed sanctions against a target (Member/Institutional Sanctions) and consider all the other categories part of the null. The results from Model 6 show that this traditional approach yields statistically insignificant results, and the same holds true if we run separate, disaggregated analyses for sanctions imposed by small and large institutions. This suggests that it is crucial to assess the effects of institutional sanctions on members' behavior in ways that control for the characteristics of the institution imposing the sanctions. Doing so allows observers to isolate the cases in which institutional sanctions have a salient impact and when they do not. In the absence of controlling for the profiles of OAS and OAU states, for example, one could mistakenly attribute their members' comparatively low-levels of sanctions-busting to the fact that the institutions participated in the sanctioning efforts. As Downs, Rocke, and Barsoom (1996) have aptly observed, good news about compliance is not necessarily good news about cooperation—and this appears to ring true with respect to these cases. Our theory and novel analytical approach thus provide insights into the differential degree to which international institutions are capable of preventing spoiler behaviors that past approaches were not able to capture.

Additional Findings

The findings regarding our control variables are almost completely consistent with prior studies, providing additional support for the liberal explanation of the determinants of sanctions-busting (Early 2009, 2012). The variables LnGDP3, Trade Share3T, Trade Openness3 are all positive and statistically significant across all five models. The variables involving the regime types of the target and third parties (DemocracyT, Democracy3, and Joint Democracy3T) require a conditional interpretation but are also consistent with our expectations. The results indicate that when the third party is not a democracy, the presence of democratic institutions in the target does not increase the likelihood of sanctions-busting occurring. Yet when the target is not a democracy, democratic institutions within a third party do increase the likelihood of that state sanctions-busting. The only anomaly occurs in the model involving the EC/EU in which Democracy3 loses its significance. This is likely because being a democracy is a requirement for EC/EU membership. The joint presence of democratic institutions in both countries positively affects the likelihood of sanctions-busting occurring. The effects of Defense Pact3T are also positive and statistically significant, while LnGDPT has the negative and statistically significant sign expected across all of the models.

The geographic and sanctions variables also behaved largely according to our expectations. As expected, both LnDistance3T and Neighbors3T are negative and statistically significant across all the models. Somewhat surprisingly, Severe Sanctions does not have a statistically significant effect. Following up on this finding, we also tested whether the goals for which the sanctions were imposed might have instead influenced third-party states' sanctions-busting behavior. We thus replaced the Severe Sanctions variable with dummy variables for whether the sanctions imposed against targets sought to disrupt their foreign military activities or cause a regime change (Hufbauer et al. 2007). Neither of the variables was statistically significant or impacted the significance of any of the other variables. Sanctions-busting thus appears to take place when all varieties of sanctions are imposed. The length of time that sanctions are in place against a target state (Years Sanctioned), though, has a positive and statistically significant effect. The long-term trend with respect to the no sanctions-busting controls is negative (that is, the negative sign of No Bust3), indicating that the longer third-party states go without sanctions-busting on behalf of a target state, the less likely they are to do so.

Conclusion

This study has sought to explain the effects of institutionally imposed economic sanctions upon states' likelihoods of undercutting sanctioning efforts. We theorized that international institutions' ability to prevent their members from sanctions-busting degrades the larger they become. In analyzing five different international institutions' involvement in 164 sanctions episodes, we found that sanctions obligations imposed by smaller-sized international institutions can constrain their members from extensively sanctions-busting, but those imposed by larger-sized institutions cannot. Our findings suggest that smaller-sized international institutions appear the most capable of meaningfully contributing to sanctioning efforts.

Our findings offer a number of important theoretical contributions to the study of economic statecraft and international institutions. Our analysis shows that the ability of international institutions to prevent costly defections by their members degrades the larger they become. This supports the broader view that there is a meaningful tradeoff between breadth and depth in multilateral cooperative efforts (e.g., Downs et al. 1996). Our findings also suggest that the microfoundations of existing theories that seek to explain institutions' contributions to multilateral sanctioning efforts' success may need to be revisited. International institutions appear not to be nearly as effective at overcoming the challenges associated with large-scale cooperation as previously thought. We also expect that institutional size is not the only factor that can cause variation in the efficacy of institutionally imposed sanctions. A productive path for future research would be to explore how other institutional characteristics, such as their bureaucratic resources and institutional autonomy, can influence the efficacy of institutional sanctions. As a first-cut effort, though, our study has demonstrated the perils of assuming that all institutions equally contribute to sanctioning efforts and provides a starting point for understanding why differences exist in their effectiveness. Lastly, our findings suggest that future research should explore forum-shopping behavior in terms of which international institutions sanctions-backers seek the support of for their sanctioning efforts in light of differences in their effectiveness. Obtaining a better understanding of why particular international institutions get involved in sanctioning efforts will complement and potentially add to our insights into the consequences of their involvement.

From a policy perspective, our analysis has revealed a number of potentially salient insights. Foremost, we have discovered that the UN's involvement in sanctioning efforts has no appreciable effect on its members' likelihood of sanctions-busting. While having the support of the UN may yield other side-benefits, the institution's failure to constrain its members from undercutting the sanctioning efforts it supports is notable. This suggests that policymakers in sender states should think twice about whether the costs of gaining the UN's support for a sanctioning effort are worth the effort. Secondly, we have found that members of the EC/EU have been avid sanctions busters, but that the EC/EU's participation in sanctioning efforts can constrain its members from doing so. A potentially important caveat to extending our findings to contemporary sanctioning efforts is that our analysis took place prior to the EU's major expansion in 2004. The array of coordination and decision-making problems the EU has faced since its influx of new members suggests that the body's effectiveness at preventing sanctions-busting may have declined to some extent. Overall, our advice to policymakers seeking multilateral cooperation for their sanctioning efforts is to seek the support of smaller institutions that can build and maintain a stronger consensus for sanctioning a target than the broader but shallower cooperation of larger international institutions.

Notes

2
For example, see Martin 1992; Drury 1998; Drezner 2000; Bapat and Morgan 2009.
3
The term “third-party states” refers to all other countries in the world besides the sanctions' primary senders and their targets.
4
Bapat and Morgan (2009) also employ the Threat and Imposition of Economic Sanctions Data Set in their analysis instead of the previously employed Hufbauer et al. (2007) data set.
5
For example, see Bueno de Mesquita, Smith, and Siverson (2004).
6
See also Olson (1965).
7
The size of an institution thus partially proxies for its level of homogeneity, but other factors certainly influence the amount of homogeneity that exists with respect to member preferences on sanctions issues.
8
It is also possible that large institutions may be less likely to impose sanctions in any form than small institutions, due to the greater challenge of obtaining sufficient support for such policies among their larger memberships.
9
We thus exclude observations that involve the states designated by Hufbauer et al. (2007) as primary senders from our analysis. We treat sanctions imposed solely by international institutions without the support of specifically designated primary sender as not having primary senders.
10
We consider active trading relationships as those in which bilateral trade is >0, which limits our analysis to the commercially relevant body of observations.
11
Carter and Signorino 2010.
12
To code this variable, we used current year trade flow data expressed in US dollars from the Barbieri, Keshk, and Pollins (2008). Importantly, these data only capture legitimate, recorded trade flows.
13
The literature on collective action has not yielded a generalizable “rule of thumb” for the absolute distinction between a small group versus a large group; instead, “smaller” and “larger” are used as relational descriptors of where observations comparatively relate to one another on a size continuum in a given context (e.g., see Isaac, Walker, and Williams 1994). Given the intense monitoring capabilities required to ensure compliance with sanctions obligations, we would expect the ability of international institutions to detect and report noncompliant behavior and hold member states accountable for it will decline as institutions grow significantly larger. There is obviously an ambiguous gray area dividing smaller-sized and larger-sized institutions, and other factors may be relevant as well.
14
We consider democracies to be countries with polity2 scores of >6 (Marshall and Jaggers 2004).
15
For all the illustrative predicted probabilities calculated in this section, we employ the same typical scenario with respect to all other factors in the model. We do employ the mean LnGDP3 and Trade Openness3 values calculated specifically for each institution's membership, though.
16
Bethlehem (2004:128).
17
In several other model specifications, however, being party to the UN when it does not impose sanctions does exercise a weak, negative effect on states' likelihood of sanctions-busting.

References

Allen
Susan
. (
2005
)
The Determinants of Economic Sanctions Success and Failure
.
International Interactions
 
31
(
2
):
117
138
.
Andreas
Peter
. (
2005
)
Criminalizing Consequences of Sanctions: Embargo Busting and Its Legacy
.
International Studies Quarterly
 
49
:
335
360
.
Ang
Adrian
Peksen
Dursun
. (
2007
)
Why Do Economic Sanctions Work?
Political Research Quarterly
 
60
(
1
):
135
145
.
Bapat
Navin
Heinrich
Tobias
Kobayashi
Yoshiharu
Morgan
T. Clifton
. (
2013
)
Determinants of Sanctions Effectiveness: Sensitivity Analysis Using New Data
.
International Interactions
 
39
:
79
98
.
Bapat
Navin
Morgan
T. Clifton
. (
2009
)
Multilateral Versus Unilateral Sanctions Reconsidered: A Test Using New Data
.
International Studies Quarterly
 
53
(
4
):
1075
1094
.
Barbieri
Katherine
Keshk
Omar
Pollins
Brian
. (
2008
)
Correlates of War Project Data Set Codebook, Version 2.0
 . Available at http://correlatesofwar.org. (Accessed October 1, 2008.)
Bennett
D. Scott
Stam
Allan
. (
2000
)
EUGene: A Conceptual Manual
.
International Interactions
 
26
:
179
204
.
Bethlehem
Daniel
. (
2004
)
The European Union
. In
National Implementation of UN Sanctions
 , edited by
Gowlland-Debbas
Vera
Tehindrazanarivelo's
Djacoba Liva
.
Leiden, The Netherlands
:
Brill Academic Publishers
.
Blumenfeld
Jesmond
. (
1991
)
Economic Interdependence in Southern Africa: From Conflict to Cooperation
 .
New York
:
St. Martin's Press
.
Bueno de Mesquita
Bruce
Smith
Alastair
Siverson
Randolph M.
(
2004
)
The Logic of Political Survival
 .
Cambridge
:
MIT Press
.
Carter
David
Signorino
Curtis
. (
2010
)
Back to the Future: Modeling Time Dependence in Binary Data
.
Political Analysis
 
18
(
3
):
271
292
.
Downs
George
Rocke
David
Barsoom
Peter
. (
1996
)
Is the Good News about Compliance Good News about Cooperation?
International Organization
 
50
(
3
):
397
.
Downs
George
Rocke
David
Barsoom
Peter
. (
1998
)
Managing the Evolution of Multilateralism
.
International Organization
 
52
(
2
):
397
419
.
Drezner
Daniel
. (
2000
)
Bargaining, Enforcement, and Multilateral Sanctions: When Is Cooperation Counterproductive?
International Organization
 
54
(
1
):
73
102
.
Drury
A. Cooper
. (
1998
)
Revisiting Economic Sanctions Reconsidered
.
Journal of Peace Research
 
35
(
4
):
497
509
.
Drury
A. Cooper
. (
2005
)
Economic Sanctions and Presidential Decisions
 .
New York
:
Palgrave MacMillan
.
Duelfer
Charles
. (
2005
)
Comprehensive Report of the Iraq Survey Group
 .
Washington, DC
:
Central Intelligence Agency
.
Early
Bryan R.
(
2009
)
Sleeping with Your Friends' Enemies: An Explanation of Sanctions-Busting Trade
.
International Studies Quarterly
 
53
(
1
):
49
71
.
Early
Bryan R.
(
2011
)
Unmasking the Black Knights: Sanctions Busters and Their Effects on Sanctions Success
.
Foreign Policy Analysis
 
7
(
4
):
381
402
.
Early
Bryan R.
(
2012
)
Alliances and Trade with Sanctioned States: A Study of U.S. Economic Sanctions, 1950–2000
.
Journal of Conflict Resolution
 
56
(
3
):
547
572
.
Gleditsch
Kristian
. (
2008
)
Expanded Trade and GDP Data, v5.0
 .
University of Essex
. Last Updated 25 February, 2008. Available at http://privatewww.essex.ac.uk/∼ksg/exptradegdp.html. (Accessed October 1, 2008.)
Hufbauer
Gary
Schott
Jeffrey
Elliot
Kimberly
Oegg
Barbara
. (
2007
)
Economic Sanctions Reconsidered
 ,
3rd
ed.
Washington, DC
:
Peterson Institute for International Economics
.
Hufbauer
Gary
Schott
Jeffrey J.
Elliott
Kimberly A.
(
1990
)
Economic Sanctions Reconsidered: History and Current Policy
 ,
2nd
ed.
Washington, DC
:
Institute for International Economics
.
Isaac
R. Mark
Walker
James
Williams
Arlington
. (
1994
)
Group Size and the Voluntary Provision of Public Goods: Experiment Evident Utilizing Large Groups
.
Journal of Public Economics
 
54
:
1
36
.
Kaempfer
William H.
Lowenberg
Anton D.
(
1999
)
Unilateral Vs. Multilateral Sanctions: A Public Choice Perspective
.
International Studies Quarterly
 
43
:
37
58
.
Leeds
Brett Ashley
Ritter
Jeffrey
Mitchell
Sara McLaughlin
Long
Andrew
(
2002
)
Alliance Treaty Obligations and Provisions, 1815–1944
.
International Interactions
 
28
(
3
):
237
260
.
Lektzian
David
Souva
Mark
. (
2007
)
An Institutional Theory of Sanctions Onset and Success
.
Journal of Conflict Resolution
 
51
:
848
.
Marshall
Monty G.
Jaggers
Keith
. (
2004
)
Polity IV Project: Political Regime Characteristics and Transitions, 1800–2003
 .
College Park
:
University of Maryland
.
Martin
Lisa
. (
1992
)
Coercive Cooperation
 .
Princeton
:
Princeton University Press
.
McLean
Elena
Whang
Taehee
. (
2010
)
Friends or Foes? Major Trading Partners and the Success of Economic Sanctions
.
International Studies Quarterly
 
54
(
2
):
427
447
.
Miers
Anne
Morgan
Clifton
. (
2002
)
Multilateral Sanctions and Foreign Policy Success: Can Too Many Cooks Spoil the Broth?
International Interactions
 
28
(
2
):
117
136
.
Morgan
T. Clifton
Bapat
Navin
. (
2003
)
Imposing Sanctions: States, Firms, and Economic Coercion
.
International Studies Review
 
5
(
4
):
65
79
.
Olson
Mancur
. (
1965
)
The Logic of Collective Action
 .
Cambridge
:
Harvard University Press
.
Richter
Paul
. (
2011
)
Attempt to Freeze Moammar Kadafi's Assets Hits Resistance
.
Los Angeles Times
  (April 24, 2011).
Rowe
David
. (
2001
)
Manipulating the Market
 .
Ann Arbor
:
University of Michigan Press
.
Stinnett
Douglas M.
Tir
Jaroslav
Schafer
Philip
Diehl
Paul F.
Gochman
Charles
. (
2002
)
The Correlates of War Project Direct Contiguity Data, Version 3
.
Conflict Management and Peace Science
 
19
(
2
):
58
66
.
Strydom
Hennie
Huaraka
Tunguru
. (
2004
)
South Africa and Namibia
. In
National Implementation of UN Sanctions
 , edited by
Gowlland-Debbas
Vera
Tehindrazanarivelo's
Djacoba Liva
.
Leiden, The Netherlands
:
Brill Academic Publishers
.
United Nations (UN)
. (
1996
)
Report of the Security Council Committee Established by Resolution 661
 .
New York
:
United Nations
(August 26).
Yang
Jiawen
Askari
Hossein
Forrer
John
Teegen
Hildy
. (
2004
)
U.S. Economic Sanctions: An Empirical Study
.
The International Trade Journal
 
28
(
1
):
23
62
.

Author notes

1
A previous version of this paper was presented at the 2011 Annual Meeting of the American Political Science Association in Seattle, Washington. We thank Navin Bapat, William Boettcher, Amanda Carroll, Michael Cobb, Gary Hufbauer, Douglas Stinnett, and the faculty of the Political Science Department at North Carolina State University for their comments on previous drafts.