Using the Carrot as the Stick: US Foreign Aid and the Effectiveness of Sanctions Threats

Early, Bryan, R; Jadoon,, Amira

doi:10.1093/fpa/orz007

Using the Carrot as the Stick: US Foreign Aid and the Effectiveness of Sanctions Threats

Early, Bryan, R;Jadoon,, Amira 2019-07-01 00:00:00 Abstract We theorize that foreign aid relationships influence both the effectiveness of economic sanctions threats and the aggressiveness of senders in imposing sanctions. Aid sanctions are generally far less costly for senders than imposing commercially oriented sanctions but can still be very costly for their targets. Being able to disrupt aid flows as part of potential sanctions enhances sender states’ credibility that they will impose painful sanctions against resisting target states. The more foreign aid a sender provides to a target state, the more successful we expect its sanctions threats to be and the more aggressive we expect the sender to be in imposing sanctions if the target resists. We test our theory using a competing risks analysis of ongoing, politically motivated sanctions threats issued by the United States from 1960–2010. Our analyses support our theory by revealing that the more foreign aid that the United States provides to target states, the more likely US sanctions threats are to succeed and the more aggressive the United States becomes in imposing sanctions. Video Abstract Video Abstract Close Introduction Economic sanctions and foreign aid alternatively rely on coercion and incentives to alter the behavior of their targets. The distinction between the two can often blur. Donors frequently provide foreign aid to incentivize recipients to change their behavior, but, once they issue aid commitments, donors can threaten recipients with the disruption or cancellation of promised aid. This type of economic coercion operates via a slightly different logic than sanctions that disrupt commercial relations based upon mutual profitability. Aid suspensions are far costlier for recipients than for donors, creating a vulnerability that donors can exploit in employing economic coercion. Aid sanctions are thus different from commercially oriented sanctions, which often can inflict comparable economic costs upon both parties. Suspending aid relationships is not necessarily without domestic political costs for senders, but cutting foreign aid will still almost always be less costly than severing trade or investment relationships. Foreign-aid relationships can thus enhance donors’ options for credibly conveying their willingness to impose costly measures when threatening targets with sanctions. Within this study, we explain how foreign-aid relationships shape both the effectiveness of sanctions threats and their senders’ willingness to follow through on them. Studying how foreign-aid flows influence the outcome of sanctions threats can provide insight into how foreign aid and sanctions policies relate to one another. Economic sanctions are compellent policies that restrict a country's government or constituents from engaging in international economic transactions to induce a change in their behavior.1 Foreign aid, in contrast, involves giving or lending foreign governments and nonstate actors money, material resources, training, or expertise. While research suggests that donors provide foreign aid to further their political interests (i.e., Drury, Olson, and Van Belle 2005), only a portion of foreign aid involves explicit quid pro quo bargains between donors and recipients. Imposing aid-related sanctions differs from simply decreasing aid to recipients (even for political reasons), as the former involves explicit demands that targets can comply with to avoid the loss of aid.2 Furthermore, donors can use the threat of foreign-aid cuts—and not just their actual reduction—to compel changes in their recipients’ behavior. The coercive nature of sanctions provides a clear demarcation between sanctioning episodes and politically motivated aid reductions (e.g., Nielsen 2013). Past works suggest that donors’ foreign-aid relationships with recipients should have a powerful effect on the outcomes of sanctions episodes at the threat stage. Morrow (1999) has argued that economic ties can enhance the range of opportunities that states have to convey credible, costly, and observable signals of their resolve during crisis bargaining. Indeed, studies on the effects of cross-border foreign direct investment and target trade dependency have shown that economic ties between sender and target states influence sanctions threats’ credibility (Kim 2013; Whang, McLean, and Kuberski 2013). The more credible and costly the prospective sanctions, the more likely target states are to capitulate during the threat stage of coercive episodes. Leveraging the TIES dataset (Morgan et al. 2014), Bapat et al. (2013, 90–92) find that sanctions threats that promise the imposition of severe costs on target states are far more effective. For states employing economic coercion, it is more cost effective to succeed at the threat stage than to succeed as a result of imposing sanctions (Drezner 2003). Therefore, states should be more attentive to the factors that make sanctions threats more effective than factors that influence the success of imposed sanctions. We theorize that existing foreign-aid relationships with target states enhance sender states’ ability to issue costly sanctions threats and increase their credibility to follow through on them. Economic sanctions and dramatic reductions in foreign-aid flows can have devastating economic consequences for aid recipients (Kharas 2008; Neuenkirch and Neumeier 2015). Donors can leverage their recipients’ vulnerability to their aid's disruption in subtle ways, making aid recipients more susceptible to economic coercion. Compared to imposing trade, investment, or financial sanctions, cutting foreign aid saves sender governments money and generates less political resentment amongst sender states’ constituents. Sender states that give substantial aid to the targets of their sanctions threats thus have enhanced credibility in conveying their willingness to follow through with imposing costly sanctions. As such, we hypothesize that target states will be more willing to give in to a sender's sanctions threats the more aid they receive from the sender. When aid recipients refuse to concede to sanctions threats alone, we expect senders to be more willing to impose sanctions against them than against non-aid recipients. We test our theory via a large-n analysis of the ongoing, politically motivated episodes of economic coercion from 1960 to 2010 in which the United States was the primary sender. As the world's most prolific sanctioner and one of the world's largest foreign-aid donors, the United States represents a critical case to be explained by our theory. We structure our study as a competing risks analysis in which the dependent variable is the yearly outcome of the sanctions threat episodes coded in one of four ways: threats persisting in a given year, ending in success, ending in the United States backing down, or resulting in sanctions being imposed. This approach allows us to distinguish between the two ways that senders can respond when their threats are not successful, by either imposing sanctions or backing down. Consistent with our theory, the quantitative analysis shows that the more aid the United States provides to the states it threatens with sanctions, the more likely those states are to capitulate at the threat stage and the more likely the United States is to impose sanctions on targets that do not concede. These observed effects are even stronger when we analyze only those coercive episodes that entailed public threats to cut foreign aid. Together, these findings indicate that US foreign aid provides American policymakers with significant leverage in employing economic coercion against the recipients of its aid. Our study contributes to the economic statecraft literature by demonstrating that foreign-aid relationships significantly influence sanctions-related behaviors. Foreign-aid relationships not only make senders’ sanctions threats more effective but also make senders more aggressive in imposing sanctions. As we demonstrate, distinguishing between failed threats that result in senders backing down versus imposing sanctions can yield theoretically valuable insights and is an approach that should be more widely adopted in leveraging the TIES dataset (Morgan et al. 2014). Our study also contributes to the broader efforts of the economic statecraft literature to understand how various forms of economic interconnectedness influence coercive behavior and outcomes within international relations (Baldwin 1985; Bapat et al. 2013; Kim 2013; Lektzian and Biglaiser 2014; Miller 2014; Barry and Kleinberg 2015; Early 2015; Peksen and Peterson 2016). If providing aid to recipient states makes them more vulnerable to economic coercion, donors could strategically invest in offering aid to states that they may want coercive leverage over in the future. In this way, foreign aid can be used as a realpolitik tool for keeping your friends close and your “frenemies” closer. In the conclusion, we discuss our findings’ policy-relevant insights for US sanctions and foreign-aid policies. Economic Coercion and Statecraft Governments employ economic statecraft, including foreign aid and sanctions, to change their external environments (Baldwin 1985). States exert pressure through the use of rewards or by inflicting pain via economic sanctions or aid cuts. States that possess asymmetric commercial relationships with their trading partners can exploit the coercive leverage that such relations offer (Hirschman 1945) and can signal their resolve more efficiently (Morrow 1999). By their very nature, the power dynamics between aid donors and recipient states comprise an imbalanced economic relationship. Most donor states give foreign aid to obtain something in return for it (Bueno de Mesquita, Smith, and Siverson 2004), as well as to maintain their relationships with allies, former colonies, and major trade partners (Alesina and Dollar 2000). Just as foreign aid can be used as a carrot to incentivize behavioral changes, its threatened withdrawal can be used as a stick to coerce recipients (Newnham 2008). Aid flows provide donors with political leverage to influence the behavior of the recipient through threatened or actual disruption of bilateral aid flows. From a donor's perspective, aid cuts may be costly and difficult to implement if political and economic interests influence donors’ foreign-aid policies with distributional consequences for domestic groups. Indeed, Milner and Tingley (2010) find that American legislators’ voting patterns on aid reflect the economic interests of their constituents. However, a closer look at their analysis suggests that such domestic interests primarily influence voting on general increases or decreases in foreign-aid appropriations rather than on aid flows involving a particular country. Furthermore, donors do not necessarily need to sever all aid to a recipient to exploit its coercive leverage but may threaten to withhold certain components, such as budget support, to induce a change in recipients’ behavior. Robertsen et al.’s (2015) analysis of Sub-Saharan African recipient states indicates frequent use of budget support aid suspensions by the UK, Netherlands, and Sweden over political, human rights, and corruption issues. They further find that while donors tend to provide more aid to their former colonies, they also suspend budget support more easily to such recipients. These studies suggest that the more recipients rely on a donor for aid, the more coercive leverage the donor possesses to compel changes in the recipients’ behavior. In practice, we often see donors use aid cuts coercively to elicit desired behavioral changes in recipient states. For example, donor countries pushed Malawi towards democratization between 1992 and 1994 by cutting 25 percent of their aid to the country to leverage political reforms. The suspension of economic aid affected Malawi quickly—resulting in currency devaluation and an annual 7.3 percent contraction of Malawi's economy (Emmanuel 2013). Additionally, the United States axed all nonhumanitarian aid to Honduras to coerce the de facto government to restore democratic rule after a coup in June 2009 (Guardian 2009). Indeed, previous work has shown that aid-dependent countries like those discussed above are particularly susceptible to economic coercion (Olson 1979). The ability to threaten aid reductions affords donors potentially significant coercive leverage over aid recipients. This is particularly applicable to bilateral aid, which is often driven by donors’ political and security interests, as compared to multilateral aid (Maizels and Nissanke 1984). Extensive use of bilateral aid can thus place donors in a powerful position to use threats of aid cuts for coercive leverage. The United States, for example, provides a large proportion of its aid bilaterally and is consistently ranked as one of the top donors globally. In 2010, 32 percent of its total aid was in the form of bilateral assistance, compared with only 7 percent for multilateral development projects (Tarnoff and Lawson 2011). Additionally, with the exception of a few years, the Unite States ranked first amongst Official Development Assistance (ODA) donors, making up 24 percent of total ODA disbursements in 2009 (Tarnoff and Lawson 2011). From a recipient's perspective, this potential leverage becomes apparent if we focus on the political and economic effects of foreign aid within recipient states. Survival-driven leaders have significant incentives to use aid to enhance their domestic political power and longevity (Bueno de Mesquita and Smith 2007; Licht 2010) or use it as a “substitute for government spending” (Kosack and Tobin 2006: 210). Governments that rely on foreign assistance to meet their citizens’ needs and maintain their rule are thus vulnerable to its potential withdrawal. This makes them beholden—at least to some extent—to the donors that provide them with aid. Focusing specifically on sanctions threats, Whang et al. (2013) argue that the efficacy of sanctions threats is conditioned by the existence and magnitude of asymmetrical economic exchange between the sender and target states. They emphasize the importance of relative economic vulnerability of the target to the disruption of their economic relationship, as opposed to an overall economic loss in the target state. States will acquiesce in the pre-imposition stage if the expected utility of compliance outweighs that of resisting the sanctions. In contrast, the sender's economic dependence on the target makes it less likely to impose sanctions if the target resists. Miller (2014) argues that the United States has used the threat of nonproliferation sanctions most effectively against states that are highly dependent upon it for their economic and security wellbeing. States that are highly dependent upon the United States are more “vulnerable” to its coercive threats according to Miller (2014, 913) because they have “more to lose from U.S. sanctions.” These works suggest that senders can exploit their targets’ vulnerability to economic disruption to obtain concessions before sanctions are imposed. This gives us a strong reason to think that foreign-aid relationships can influence the outcomes of sanctions threats. Using the Carrot as the Stick: Foreign Aid and the Effectiveness of Sanctions Threats A sender initiating a coercive episode seeks to compel its target into making the desired concessions rapidly and cost-effectively. The most efficient stage for a coercive episode to be won is at the threat stage, when the sender wagers its credibility and the prospective costs of imposing sanctions against the target's resolve. If the target state lacks resolve and views the threats as credible (Kim 2013), it will give in to the sender's demands by capitulating or negotiating concessions. On the contrary, if the target state resists the sanctions threat, the sender must decide between backing down or imposing the threatened sanctions. By backing down, the sender risks a loss of its credibility in future coercive episodes (Peterson 2013). By imposing sanctions, senders still have a chance to obtain the desired concessions but at a much higher cost than obtaining concessions during the threat phase. None of these decisions are instantaneous, either. Target states can delay their response to gain more information about the credibility of the sender state's threats. While waiting, senders can reference new information to assist their decision about the best response and explore ways of enhancing the effectiveness of their threats. Foreign Aid, Coercive Leverage, and Sanctions Threats The impact of a sanctions threat is influenced by perceptions of the credibility of the threat, its prospective costs to implement, and speculation about the future outcome should sanctions be imposed. The presence of robust foreign-aid relationships enhances the range of “opportunities” (Morrow 1999) available to senders in credibly conveying their willingness to impose costs on target states. The ability to threaten aid cuts complements senders’ abilities to threaten trade or investment sanctions. Existing foreign-aid relationships between sender and target states influence how both states perceive their positions during threat episodes. Target leaders and governments are sensitive to threatened aid cuts because their financial welfare may be directly tied to external aid. Survival driven leaders often use aid to enhance their domestic power (Bueno de Mesquita and Smith 2007; Licht 2010) and to support government spending (Kosack and Tobin 2006). The disruption of aid relationships can be economically and politically damaging for recipients, suggesting they will try to prevent their occurrence. Reflecting this, several studies indicate that donors exploit their aid relationships with recipients to obtain concessions by threatening the withdrawal of their aid (Bueno de Mesquita and Smith 2007; Newnham 2008). We argue that the more foreign aid that a sanctions-threatening donor gives to a target state, the more successful those threats will be and the more aggressive senders will be in imposing sanctions. Being able to threaten—either privately or publicly—the disruption of aid flows a target state receives allows the sender to increase the prospective costs for resisting its sanctions threats. Target states will be more sensitive to sanctions threats when they receive aid from the sender because substantial aid cuts can harm their economies. For aid recipients, finding alternative revenue sources due to unexpected budgetary shortfalls can be difficult—especially in the short term. Leaders may have to make emergency cuts to planned spending, suspend governmental services, or borrow money under duress. Indeed, Kharas (2008) has found that states suffering “aid shocks” can experience deadweight economic losses of almost 2 percent of their gross domestic products. Notably, these effects would not likely occur in isolation when aid cuts occur with other types of sanctions. Both public and private aid threats can be effective. By making public aid threats, senders can specify the dollar amounts or aid programs being jeopardized by their threats that allow targets to understand the costs of noncompliance more clearly. In some circumstances, though, senders may prefer to communicate threatened aid cuts via private channels due to the aid's political sensitivity or because they recognize targets’ reputational concerns.3 Drawing on dependency theory, Olson (1979, 485) recognized how wealthy countries could engage in more subtle and effective types of coercion using the aid they provided to less developed countries as leverage. “Modern dependency,” according to Olson, “allows effective economic coercion to be applied without much publicity and without generating a widespread nationalist response within the target-country.” This suggests that target leaders are particularly susceptible to threats that could jeopardize foreign aid, which can be communicated privately to the leadership of target states. Targeted leaders will be blamed for the hardships created by foreign-aid cuts, irrespective of whether the public is broadly aware of why those cuts were imposed. Given their self-interests in retaining power (e.g., Bueno de Mesquita et al. 2004), leaders should be much more likely to concede to sender states’ demands at the threat stage, if their governments receive a substantial amount of foreign aid from them, than allow sanctions to be implemented. As Olson (1979, 481) aptly surmises, “Dependency is indeed vulnerability.” Sanctions threats are also more credible when senders possess aid-based coercive leverage because aid cuts are less costly than commercially oriented sanctions. This argument is consistent with Morgan's (2015, 749) observation that “senders are very good at designing sanctions that are costly to the target but not to themselves.” For sender states, imposing trade and financial sanctions can involve imposing costly disruptions upon their own constituents’ commerce (Hufbauer et al. 1997). Cutting off foreign aid to target states is less likely to harm the economic well-being of domestic constituencies or result in political mobilization than commercially oriented sanctions. Disrupting foreign-aid flows may generate some domestic economic grievances, but it is unlikely to cause the same magnitude of costs as commercial sanctions or to generate the same degree of political backlash.4 Sender governments can also find cost-neutral ways of compensating adversely affected domestic constituencies by reallocating the aid money to new recipients or via other channels. Commercial interest groups often mobilize against trade-related sanctions due to the losses they impose on US businesses (Early 2015, 67). For example, the US Chamber of Commerce and National Association of Manufacturers lobbied against the imposition of sanctions against Russia after its invasion and annexation of Crimea (Jackson 2014). In 2011, the US Chamber of Commerce and fifty other trade groups lobbied Congress not to follow through on threatened legislation to sanction China for its currency manipulation practices (Agence-France Press 2011). While the domestic beneficiaries of foreign aid, like the NGO community, do exert influence over US foreign-aid allocations (e.g., Kim 2016), they tend not to pursue high-profile, well-financed campaigns to block sanctions imposition. Indeed, such efforts would run up against the US public's antipathy towards foreign aid (Milner and Tingley 2013). The political tide in the United States thus runs against interest groups seeking to block foreign-aid sanctions. This suggests that, in general, target leaders should perceive fewer domestic obstacles for the United States following through on sanctions threats that would target foreign-aid flows. The United States’ coercive behavior appears to bear this logic out. As US Senator Lindsey Graham put it, US foreign aid “is designed to create leverage so we [the US government] influence the world and not have the world just run us over” (Lesniewski 2012). The United States’ broad portfolio of foreign-aid relationships empowers it to employ economic coercion in pursuit of US foreign interests. In our sample of 192 politically motivated sanctions threat episodes with the United States as the primary sender, 80 out of the 103 cases (78 percent) in which the United States imposed sanctions also involved cutting foreign aid to target states (Morgan et al. 2014). Sanctions threats are more likely to be perceived as credible if the target state receives substantial foreign-aid flows from the sender that could be jeopardized by sanctions. As we note above, senders do not have to publicly threaten aid cuts initially to include them in the sanctions they eventually impose. As such, aid relationships are expected to influence the general efficacy of all sanctions threats cases. Based on our theory, we expect that greater aid flows will be associated with an enhanced willingness by the senders to impose sanctions if targets do not capitulate. Having leverage over target states during the threat stage will increase donors’ confidence about their ability to compel the target to capitulate by imposing sanctions. Such aggressiveness on the donor's part indicates that they expect their coercive efforts to ultimately succeed—thus highlighting the importance of distinguishing between threat failures in which senders back down versus double down on sanctioning the target. This theoretical expectation distinguishes our account from an alternative explanation that senders are bribing targets to give in to their sanctions threats. According to the latter perspective, the aid given to threatened states would lessen their costs of capitulation and, therefore, be positively correlated with threats success rather than sanctions imposition. One would expect, though, that the United States would use a bribery-based strategy to prevent having to impose sanctions. Prima facie, there is no reason why pursuing a bribery-based strategy would lead to a higher likelihood of sanctions imposition. A final implication of our theory relates to how foreign-aid relationships may affect donors’ willingness to back down from their threats. Backing down from sanctions threats can damage senders’ credibility and make their future threats less effective (Peterson 2013). Possessing foreign-aid leverage over a target could mean that donors are less likely to back down in the aggregate. Alternatively, though, having foreign-aid leverage over the target could make donors less worried about how backing down will affect their ability to coerce the target in the future—as they will still have their aid relationships to use as leverage. Thus, foreign-aid leverage that makes donor states more aggressive in imposing sanctions may also make it easier for senders to back down from unpromising sanctions threats. Given that both factors could be in play, we do not have strong theoretical predictions about the impact of foreign-aid flows on senders’ likelihoods of backing down. All else being equal, we expect that sanctions threats issued against states that receive significant quantities of foreign aid from senders should be more successful than threats issued against targets that receive little to no aid from the sender. Foreign aid from sender states raises the prospective costs of sanctions for target states and enhances their credibility of imposing such sanctions. Senders should also be more aggressive in imposing sanctions because they view their prospects of success as higher and will likely face less domestic political opposition for imposing such sanctions. Indeed, it is the enhanced willingness of senders to impose sanctions in such cases that should raise threatened states’ willingness to capitulate. Our theory gives rise to the following two hypotheses (H1 and H2): Hypothesis 1:The larger the foreign aid flows that senders provide to targets, the more likely the sender's sanctions threats will be successful. Hypothesis 2: Hypothesis 2:The larger the foreign aid flows that senders provide to targets, the more likely senders will be to impose sanctions when targets refuse to concede to the threats. Research Design We test our theory via a competing risks model capable of analyzing the factors associated with the various outcomes of US sanctions threats. Our analysis employs a sample of 192 ongoing, politically motivated sanctions threats from 1960–2010 from the TIES dataset in which the United States was the primary sender (Morgan et al. 2014).5 We focus our analysis on the United States, which is an active employer of economic sanctions and a top ranking foreign-aid donor, to control for the strength and political profile of the sender. The United States is also one of the few states that can afford to both give large amounts of foreign aid and actively employ economic coercion. We examine politically motivated sanctions because the salience of the issues involved tends to be higher.6 The unit of analysis for our dataset is the coercive episode-year, in which we code a coercive episode from the year in which the United States issues a specific sanctions threat until the episode terminates with the United States backing down from its threat, the United States imposing sanctions, or the target conceding to the sanctions threat. Our dataset is thus structured along the lines of an events history format in which yearly observations for coercive episodes are included until one of the specified termination events occur, and the case exits the dataset (Box-Steffensmeier and Jones 2004). Importantly, structuring our dataset this way allows us to include multiple observations of ongoing sanctions threats against specific target states at the same time. Via this competing risks approach, we account for the factors that influence the decisions by target or sender states that lead to terminal outcomes, such as backing down (sender), imposing sanctions (sender), or acquiescing (target). The theoretical motivation for this modeling strategy is that coercive episodes unfold over time, with both the sender and target states seeking to gain greater insights into their opponents’ resolve and the costs associated with various choices (Van Bergeijk and Van Marrewijk 1995; Early 2015). Upon the issuance of sanctions threats and prior to a terminal outcome, the persistence of sanctions threats becomes the status quo dynamic between senders and targets. Research on the consequences of sanctions demonstrates that what happens during sanctions can be more significant than how they end (Peterson and Drury 2011; Allen and Lektzian 2013). Changes in the domestic and international environments faced by leaders in both countries can influence the discrete choices they make over the course of a coercive episode. When targets acquiesce or senders back down or impose sanctions, they do so because they no longer find sanctions threats persisting as the most desirable outcome. We employ years as our temporal unit of analysis because data for our time-variant independent variables are only available at the yearly level. In order to model our competing risks analysis, we employ a multinomial logit estimator with cubic polynomials to provide a Taylor series approximation to the hazard curve. Along the line of Carter and Signorino's (2010) recommendations, we employ the number of years since a sanctions threat has been issued and the term's squared and cubed values (Time, Time2, and Time3) as our cubic polynomials. The multinomial logit analysis generates results for each possible binomial comparison between different outcomes.7 To test our theory, we are primarily concerned with assessing the impact of US Government's aid flows to target states on the comparative likelihoods of sanctions threats persisting, sanctions being imposed, or successfully ending with the target's capitulation in a given year. To account for clustering in the US-target dyads, we employ dyadic cluster-robust variance estimation as part of our analysis (Aronow, Samii, and Assenova 2015). Our results remain consistent when we employ alternative means of calculating our standard errors and alternative estimators (see supplementary files, tables A4–A6). Coding the Yearly Threat Outcomes Variable Our categorical dependent variable captures the yearly outcomes of sanctions threats for each of the 192 coercive episodes within our dataset. To code our four-category Threat Outcomes variable, we employ data on the years in which sanctions threats were ongoing and the termination of sanctions threats using information from the updated TIES dataset (Morgan et al. 2014). The baseline category of the Threat Outcomes variable is the sanctions threats persisting outcome. This is the most common yearly outcome, reflecting that threat episodes can persist in stalemates for years in some cases. Over 55 percent of the threat episodes are coded as spanning across more than one observation year. In order to identify sanctions threats that resulted in success, we followed the practices of the TIES dataset's creators in flagging successful cases as those that resulted in targets’ complete acquiescence, partial acquiescence, or a negotiated settlement (Bapat et al. 2013; Morgan et al. 2014). A threat episode was coded as ending with the United States backing down if it terminated with the sender's capitulation at the threat stage.8 The final way a threat episode could end according to our coding scheme was with the United States imposing sanctions. As Peterson (2013) has shown, the United States backing down from a sanctions threat has its own distinct set of consequences compared to imposing sanctions—highlighting why it is important to distinguish between the two alternative outcomes to success. Distinguishing between unsuccessful threat outcomes (i.e., the United States backing down vs. imposing sanctions) is central to testing the logic of our theory, as well as accurately reflecting the nature of compellent episodes (Schelling 1966). As compellent policies, sanctions episodes are considered successful if they achieve their desired outcomes, even if their senders have had to follow through on their threats. Threats that lead to sanctions being imposed do not indicate a failure of the coercive episode, since success is still possible if sanctions are imposed. Imposing sanctions versus backing down from their threats thus represent distinct alternatives for senders, with each being associated with vastly different outcomes. Confident and committed senders follow through with imposing sanctions because they think that the sanctions could work; less confident and less resolved senders will instead back down from their threats. For the period analyzed, our data indicate that 34 of the coercive episodes ended at the threat stage with the United States backing down, 101 resulted in the United States imposing sanctions, and 50 cases ended in success. This suggests that 25 percent (34/135) of our cases involving what is typically considered to be “threat failure” result in fundamentally different outcomes than the other 75 percent (101/135). Our analysis should reveal that there are significant theoretical gains to be made in distinguishing between those outcomes. Coding the Independent Variables In order to capture our dependent variable, we code US Foreign Aid as a measure of the ODA provided by the US Government to target states. ODA comprises foreign aid given by members of the OECD that matches three criteria: (1) it is donated by government agencies; (2) it is given to facilitate the “economic development and welfare of developing countries”; and (3) it is offered via concessional terms (OECD 2018). Notably, this definition excludes military aid. Consistent with our theory, we assume that the leaders of target and sender states view potential aid cuts in terms of actual monetary values rather than in proportional terms. For target states, the prospective loss of $10 million in foreign aid will create a real budgetary deficit that target leaders will have to address, irrespective of what proportion that amount of money comprises of their total economy. All else being equal, for example, both India and Rwanda would face substantially greater challenges in finding $100 million as opposed to $10 million in new aid to make up for a sanctions-induced shortfall. For leaders considering the challenges of replacing lost aid, we assume the actual amount of money being threatened matters most.9 Similarly, we think actual monetary value does the best job of operationalizing how sender states’ view aid cuts. Donor states will take into account how foreign-aid cuts will impact the domestic constituencies that benefit from them and how the aid money could be reallocated for other purposes.10 For theorizing cross-nationally about the coercive leverage that foreign-aid flows provide, we think this approach provides the best metric for capturing the degree to which target and sender leaders evaluate the prospective disruption and adverse social, political, and economic costs that aid withdrawals can inflict. We adopt Roodman's (2011) strategy of operationalizing ODA flows as net aid transfers. This measure subtracts the interest target states pay on concessional loans and the cancellation of one-off non-ODA loans that tend to be tied to aid packages given in the past years, cannot be predicted, and can skew aid flows. Operationalizing US Foreign Aid in terms of net aid transfers provides a transactional measure of the aid-flow relationship between the United States and a target state.11 The higher this value is, the greater the leverage the United States can expect to have over a given target state that it threatens with sanctions. We operationalize US Foreign Aid and our other economic variables in their real values in terms of constant 2012 dollars. The latter approach accounts for inflation and ensures that the value of the aid given by the United States is consistently measured throughout the analysis.12US Foreign Aid should have a positive effect on the likelihood of US sanctions threats succeeding and also make the United States more likely to impose sanctions if targets resist its threats. In addition to our main variable, we include a number of other variables that could influence the outcomes of coercive episodes at the threat stage. As previous studies have shown (e.g., Ang and Peksen 2007; Hufbauer et al. 2007; Bapat et al. 2013), the salience of the issue(s) at stake can have a strong impact on the likely outcome. We code High Salience Issue as a dummy variable that denotes whether the issues at stake involved efforts to contain the target's military behavior, destabilize the target's regime, address a territorial dispute, or prevent the target from acquiring dangerous military capabilities. All four issues are likely to infringe upon the target state's core political and security interests. Given the importance attributed to the participation of international organizations in coercive episodes (e.g., Drury 1998; Bapat and Morgan 2009; Bapat et al. 2013), we include a dummy variable IO Support to account for whether international organizations supported US sanctioning efforts. We include several variables to account for the characteristics of the target state. We first code a variable to denote whether the target of the US sanctions threats is a democracy. Several debates exist over whether democracies, like the United States, are less likely to sanction fellow democracies and about the ability of democracies to withstand economic coercion in comparison to authoritarian regimes (e.g., Cox and Drury 2006). We code our variable Target Democracy as a dummy variable that denotes that a state has a polity2 score of 6 or above and as a 0 otherwise using data from the Polity IV Project (Marshall and Jaggers 2009). We also control for the size of target states’ economies using a logged measure of their gross domestic products (Target Economic Size) via updated data from Gleditsch (2002). Lastly, we include controls for the relationship between the United States and the states that it is threatening with sanctions. To capture a target's susceptibility to US sanctions that would disrupt its trade flows, we include the variable Target Trade Dependence. Using current-year trade data from Barbieri and Keshk (2012), this variable codes the target's bilateral trade with the United States as a proportion of the target's total trade. We account for the political relationship between the United States and target states in several ways. First, we include the measure of political affinity using voting patterns in the United Nations General Assembly. We employ a measure that codes similarities in voting patterns coded along a continuum from –1 to 1.13 States with similar political affinities have positive scores, while dissimilar states have negative scores. As a second control, we include US Defense Pact as a dummy variable that denotes whether the United States and target state have a defense pact using Gibler (2009). According to Drezner (1999), coercive efforts should be more effective against allies since allies have fewer relative power concerns associated with giving into a sender's economic coercion. Finally, we account for the United States’ emergence as the world's lone superpower by including a dummy variable Post–Cold War in our analysis. The United States dramatically increased its use of sanctions after it became the world's lone superpower at the Cold War's conclusion (e.g., Hufbauer et al. 2007). Summary statistics and a correlation matrix are available in the supplementary files (see tables A1–A2). Discussion of the Results Table 1 presents the results yielded by our multinomial logit analysis using sanctions threats persisting as the baseline category of the yearly outcome variable. We use this as our baseline for operationally testing our hypotheses, as it is by far the most common outcome, and, conceptually, it is the yearly outcome value from which episodes transition to one of the three other terminal outcomes. Table 1 shows the results of the binomial comparisons for threats succeeding, sanctions being imposed, and the United States backing down, versus the baseline category. Model 1 includes only our theoretical variable of interest, while model 2 also includes the controls. For ease of interpretation, figure 1 displays the average marginal effects for US Foreign Aid on all four of the yearly sanctions outcome categories using model 2. It depicts the effects of US Foreign Aid when holding all other values of the variables at their observed values. Figure 1. View largeDownload slide Coefficient plot of the average marginal effects of US foreign aid on threat outcomes Figure 1. View largeDownload slide Coefficient plot of the average marginal effects of US foreign aid on threat outcomes Table 1. The effects of US foreign aid on the yearly outcomes of economic sanctions threats Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Notes: *, **, and *** denote statistical significance at the 90 percent, 95 percent, and 99 percent confidence levels, respectively, using two-tailed tests. View Large Table 1. The effects of US foreign aid on the yearly outcomes of economic sanctions threats Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Notes: *, **, and *** denote statistical significance at the 90 percent, 95 percent, and 99 percent confidence levels, respectively, using two-tailed tests. View Large With respect to our two hypotheses, the variable US Foreign Aid appears to both have a positive effect on the success of sanctions threats and also make senders more aggressive in implementing sanctions when target states resist sanction threats. In table 1, the results show that US Foreign Aid has a positive and statistically significant effect on the likelihood of sanctions threats ending in success versus persisting across all of our models. In general, and all else being equal, a one-standard-deviation increase in the constant value US Foreign Aid is associated with 1.68 greater odds of a sanctions threat ending in success as opposed to persisting. As figure 1 shows, the average marginal effects of US Foreign Aid on the likelihood of sanctions threats resulting in success are also positive and statistically significant at the 95 percent confidence level. In figure 2, we show the marginal effects of US Foreign Aid from its minimum value to $1 billion worth of net aid transfers while holding the other variables constant at their observed values. The figure illustrates a characteristic of our net aid variable (Roodman 2011) that allows for aid flows to be negative when aid recipients paid more to the United States in interest payments on aid-related loans than they received in new aid.14 This was true in roughly 9 percent of our observations.15 In 38 percent of the observations, the United States provided threat targets with a net value of 0 aid, and in 53 percent of cases, the aid values were positive. While the maximum value of US Foreign Aid was $6.7 billion in our sample, the vast majority (95 percent) of the observations involved less than $365 million in positive aid flows (see figures A7–A8). Substantively, the magnitudes of the marginal effects are small because the likelihood that threat episodes transition to a terminal success outcome in a given year is relatively low. The cumulative impact of the variable on the likelihood of success does mount over time. Figure 2. View largeDownload slide Marginal effects of US foreign aid on threat success at representative values Figure 2. View largeDownload slide Marginal effects of US foreign aid on threat success at representative values Figure 2 indicates that the average marginal effects of US Foreign Aid are positive and statistically significant at the 95 percent confidence level until the United States gives just over $400 million in net aid, at which point the variable's effect remains positive but loses statistical significance. This appears to be due to the scarcity of observations beyond this threshold (see figure A7). Examining the cases above this threshold, we identified the most significant aid outlier that resulted in a successful outcome. This case involves Israel, a leading US aid recipient. In 1992, the George H. W. Bush Administration threatened to terminate $10 billion worth of subsidized loan guarantees to force Israel to stop building new settlements in the occupied Palestinian territories. President Bush exploited Israel's reliance upon US aid to force the Israeli Government to freeze its construction of additional settlements (Friedman 1992; New York Times 1992). Our argument effectively explains the leverage wielded by the United States in this case, suggesting that this observation represents an on-the-line case that belongs in our sample. Excluding this extreme outlier case did not affect our results, but excluding all observations entailing aid flows greater than $400 million does cause the relationship to wash out (see table A19–20). Notably, none of the high-aid observations resulting in success were flagged as “least likely” outcomes in postestimation analyses.16 Overall, this suggests that high aid cases are an important part of our sample and shape our findings in ways consistent with our theory. With respect to hypothesis 2, US Foreign Aid has a positive and statistically significant effect on the likelihood of threats resulting in sanctions as opposed to persisting, according to both the results reported in table 1 and those in figure 1. Increasing US Foreign Aid by one-standard deviation increases the odds that the United States will impose sanctions during a coercive episode by 1.66 as opposed to letting the sanction threat persist. In figure 3, we show the marginal effects of US Foreign Aid on the sanctions imposition outcome. The results show that the marginal effects remain positive and statistically significant across the full range of the graph at the 95 percent confidence level—and indeed remain so until approximately $1.6 billion in positive aid flows. Once again, there are some higher-leverage cases at more extreme values that influence the results, but these are consistent with our argument and largely fall within the range in which US Foreign Aid’s effects are statistically significant. The US imposed sanctions against Israel (1982), Pakistan (1965; 1965), and India (1965; 1971) during periods in which its aid flows totaled over $1.6 billion dollars to those countries. The results are robust to the removal of the most extreme observation involving sanctions threatened against India related to its 1965 war with Pakistan. However, removing all the cases in which US Foreign Aid was greater than $1.6 billion causes the effects’ statistical significance to wash out (see tables A21–22). None of the high-aid observations appear as “least likely” cases in postestimation analyses, and the United States’ imposition of sanctions against those countries is consistent with our theory. Indeed, US policymakers publicly invoked potential aid cuts as part of their sanctions threats in each of the high aid observations—seeking to leverage the aid relationships in forcing concessions—before they followed through on their sanctions threats. While our findings do not hinge on a single case, the cases involving significant aid are both substantively important for our theory and our models’ results. Figure 3. View largeDownload slide Marginal effects of US foreign aid on sanctions imposition at representative values Figure 3. View largeDownload slide Marginal effects of US foreign aid on sanctions imposition at representative values The effects of US Foreign Aid on US policymakers’ likelihood to back down are more complicated to interpret. While table 1’s results indicate that US Foreign Aid has a positive and statistically significant effect at the 95 percent confidence level, the confidence intervals for the back down outcome overlap with zero in figure 1. Notably, the two threat episodes that ended in the United States backing down at the highest values of US Foreign Aid both involved Egypt, in 1960 and 1995. Throughout our sample period, the United States provided significant aid to Egypt both as part of Cold War politics and in order to foster peaceful relations with Israel. Excluding Egypt from our sample causes US Foreign Aid to lose its statistically significant impact on sanctions failure, but the results for the other outcomes remain unaffected (see table A7 and figures A1–A2). Our model lacks the nuance to capture how the quid pro quo bargains associated with US aid to Egypt limited its coercive leverage. Given the observed salience of the Egyptian cases, further research into the role foreign aid played in shaping US influence over the regime would be a valuable complement to our analyses. Not many of our other independent variables appear to affect threat outcomes. The variable IO Support has a positive and statistically significant effect on the likelihood of threats succeeding. Interestingly, we also found that having the support of an international organization increases the likelihood that the United States will impose sanctions. This suggests that organizationally supported sanctions threats could be more credible—and hence effective—because the United States is more likely to follow through on them. With respect to the United States backing down versus the threat persisting, the results show that Target Economic Size, US Defense Pact, and Post–Cold War have positive effects. This suggests that the United States is more likely to give up on sanctions threats that target democracies and/or allies and on threats during the post–Cold War era. The findings with respect to Target Economic Size indicate that the United States is less likely to give up on threats issued against states with large economies. Finally, the temporal controls (Time, Time2, and Time3) indicate that the longer a coercive episode lasts at the threat stage, the less likely the United States is to impose sanctions or back down—yet the more likely United States threats are to succeed. Robustness Checks While we theorize that senders exploit their aid relationships with recipients by issuing both private and public threats, we only have data on those threats that were publicly issued. By focusing on cases in which senders issued public threats to cut aid, we can ensure that our theoretical arguments successfully explain those cases in which we know donors specifically referenced aid cuts as part of their threats and had incentives to do so publicly. For example, the public threat of losing tens of millions of dollars’ worth of United States aid and trade was attributed to helping bring down Guatemala's President Jorge Serrano in 1993 after he sought to seize total power in the country. One of the first acts of his replacement, Ramiro de Leon Carpio, was to call upon the United States and other donors to rescind their sanctions threats (Golden 1993a; 1993b). Given that donors must expect to gain additional leverage from issuing their aid threats publicly, we would expect sanctions threats that explicitly invoke aid cuts to have even stronger effects. To evaluate whether the substantive effects of US Foreign Aid are stronger in the case of aid-based sanctions threats, we used the Sanctions Type Threatened variable from the TIES dataset (Morgan et al. 2014) to identify cases in which the United States solely threatened targets with the termination of foreign aid. We re-ran our original analysis using the sample of cases that only involved public aid-based sanctions threats (see table A8). The results strongly conform to our expectations. The coefficients for US Foreign Aid share the same sign and statistical significance as they did in the original analysis—but the size of the effects is larger. A one standard deviation increase in US Foreign Aid is associated with 2.34 greater odds of a sanctions threat ending in success in the aid-threat only cases. A one-standard-deviation increase in US Foreign Aid increased the odds of the United States imposing sanctions versus letting the threat persist by 2.05 in the public aid threat sample. These findings suggest that the coercive leverage provided by United States aid flows is stronger in the subset of cases where threats to aid are publicly communicated.17 We also considered the potential issue of simultaneity bias with respect to US Foreign Aid. To explore this, we ran our analysis using US Foreign Aid lagged by one year and using the value of US Foreign Aid for the year prior to the sanctions threat being issued (see tables A10–A11). The results were congruent with our main analyses. Additionally, we analyzed the implications of changing aid flows. If the positive relationship between US Foreign Aid and the success of sanctions threats is due to a bribery-based strategy rather than a coercive one, we would expect increasing aid flows to be associated with both a higher likelihood of success and a lower likelihood of imposing sanctions. We coded a variable to denote whether United States aid flows to target states increased, decreased, or remained static from the previous year. The results indicate that increasing aid flows are not associated with greater threat success but instead are associated with a higher likelihood of the United States imposing sanctions (see table A12). This runs counter to the expectations of an alternative, bribery-based account. Finally, we sought to evaluate whether our findings were driven by a subset of cases in which targets’ economies were highly dependent upon US foreign aid. We thus calculated the proportion that US foreign-aid flows comprised of each target state's economy and flagged those cases in which US aid amounted to more than 0.5 percent of the target's GDP. Rerunning our analysis without those cases (roughly 10 percent of the full sample), our results still hold (see tables A14). All these additional analyses support our theory that foreign-aid flows offer additional coercive leverage in making sanctions threats more effective and enhance senders’ aggressiveness in imposing sanctions. We also assessed whether the effects varied across the Cold War and post–Cold War eras (see table A15). Our findings show that while all other effects remain the same in the two periods, US Foreign Aid’s positive effects on the success of threats is only statistically significant in the post–Cold War period. This suggests that the coercive leverage the United States gained from its aid relationships was weaker when the recipients had a credible replacement option (i.e., the Soviet Union). Aid recipients thus appear more resistant to giving in to sanctions threats when they think a “black knight” or “aid-based sanctions buster” might be able to assist them if sanctions are imposed (Hufbauer et al. 2007; Early 2015). This sheds additional light on what influences how aid recipients respond to coercive attempts. Finally, there is a potential concern that US aid flows influence which states the United States threatens with sanctions—creating selection effect issues for our analysis. To evaluate this concern, we conducted a logit analysis on whether our US Foreign Aid variable positively influenced the likelihood of states being targeted with US sanctions (see table A23). The results indicate no such selection effects. While US aid relationships do influence the outcomes of sanctions threats, receiving greater amounts of US aid does not make recipients more likely to be targeted with coercion. Conclusion This study sought to explain how foreign-aid relationships influence the effectiveness of sanctions threats and the likelihood of sanctions being imposed. We theorized that providing foreign aid to states threatened with sanctions provides senders with greater leverage in extracting concessions. It also makes them more aggressive in imposing sanctions if sanctions threats do not elicit the desired behavior in the targeted state. Via an analysis of ongoing US sanctions threat episodes from 1960 to 2010, we found strong support for our hypotheses. Greater US foreign-aid flows to target states increase the likelihood that sanctions threats will result in success or in the United States following through with imposing sanctions when targets resist. Our findings also suggest that the positive effects of aid relationships on the success of US coercive threats are strongest in the absence of a superpower rival. Our study has several implications for research on economic coercion and foreign aid. Notably, our findings show that foreign aid constitutes an investment that provides donors with future coercive leverage in addition to more immediate quid pro quo concessions (e.g., Bueno de Mesquita and Smith 2007). This supports the argument that donors provide foreign aid more out of self-interest than altruism. Our findings also support some dependency theorists’ (e.g., Olson 1979) skepticism about the donor community's motivations in providing aid. Donors may thus provide aid strategically to recipient states that they share common interests with but also want to have leverage over in the future. For coercive episodes in which sanctions are imposed, our theory suggests that foreign-aid relationships should enhance sanctions’ likelihood of success. Our research design also presents a new approach for employing the TIES dataset (Morgan et al. 2014) that fully captures the multiple ways in which sanctions threat episodes can conclude. Building off our results, we think that additional qualitative analyses of the role aid relationships play in the use of economic coercion, especially for high aid cases the like US-Egypt relationship, is an important next step in this research agenda. While focusing on sanctions threats issued by the United States could limit our findings’ generalizability, our findings yield direct implications for US foreign policy. The number of donors that can both afford to give high levels of foreign aid and frequently employ economic coercion—like the United States—is limited. Our results suggest, though, that opportunistic donors may be able to exploit the coercive vulnerabilities associated with foreign-aid relationships on an occasional basis. For US foreign policy, our findings indicate that foreign aid buys the United States a greater amount of international influence than may have been previously thought. Indeed, our findings shed light on why US leaders may rely so heavily on economic coercion despite imposed sanctions’ poor track record of success. Even if leveraging aid relationships encourages US leaders to be highly aggressive in imposing sanctions, such aggression likely contributes to diplomatic victories before sanctions ever get imposed. The “hidden” success of US sanctions threats potentially makes up for the fact that the sanctions the United States imposes often are not that effective. Supplemental Information Supplemental information is available at the Foreign Policy Analysis data archive. Acknowledgements A previous draft was presented at the 2015 Annual Meeting of the International Studies in New Orleans, LA. We would like to thank Glenn Palmer and the comments of the anonymous reviewers for their insights and comments that contributed to this manuscript. Notes Bryan R. Early is an associate professor of political science at the University at Albany, SUNY. He is also the director of the Center for Policy Research (CPR) and the founding director of the Project on International Security, Commerce, and Economic Statecraft (PISCES). Dr. Early is an expert on economic statecraft, weapons of mass destruction, and nonproliferation issues. Amira Jadoon is an assistant professor at the Combating Terrorism Center (CTC) and the Department of Social Sciences at the US Military Academy at West Point, as well as the CTC's General John P. Abizaid Research Associate. Dr. Jadoon specializes in international security, economic statecraft, political violence, and terrorism. Footnotes 1 Governments can adopt legislation that threatens the triggered imposition of sanctions as a form of deterrent threat (see Miller 2014). Once imposed, however, sanctions are compellent in nature. 2 Sanctions packages can also entail a much broader mixture of commercial restrictions that go beyond just aid (Morgan, Bapat, and Kobayashi 2014). 3 Target leaders may not want it publicly known that they will acquiesce to threats involving aid cuts at a specific threshold (i.e., $50 million). Privately issued aid threats could be preferable in some cases, as they could make it easier for threatened leaders to acquiesce without giving away information about their vulnerability to coercion to other donors. 4 Some sender constituents can benefit from the aid packages given to target states. For example, American farmers benefit from the US food aid programs that require the purchase of US foodstuffs. Milner and Tingley (2010) show that US Congressional support for foreign aid is influenced by politicians’ domestic economic concerns; however, voting on aid flows concerning specific geostrategic issues is not particularly affected by district endowment factors and only bears a weak relationship with political economy interests. Additionally, Fleck and Kilby (2001) find no evidence that domestic economic benefits influence Congress members’ stances on foreign aid. In contrast, Hufbauer et al.’s (1997) past research has shown that trade-related sanctions result in tens of billions of dollars of lost business revenues—which translates into hundreds of thousands of lost US jobs. 5 Our dataset includes cases initiated within the TIES dataset prior to 1960 and persisted into our sample period. The TIES dataset only includes episodes initiated up to 2005 but codes outcomes through 2010. 6 We excluded coercive episodes that only involved trade or environmental issues. Trade sanctions, such as those initiated under Section 301 of the US Trade Act of 1974, have very different procedures for their imposition than politically driven sanctions. Environmental sanctions tend to involve technical more than political issues. 7 We meet the theoretical requirements of the independence of irrelevant alternatives (IIA) assumption, as each outcome is distinct from one another and there are no viable alternative outcomes. 8 We use data from the TIES dataset's “Final Outcome” variable (Morgan et al. 2014). 9 Using a measure that takes US-target aid flows as a proportion of countries’ GDPs instead of real monetary values, for example, would divorce the measure from the real budgetary impact that aid withdrawals have. The size of a country's economy does not always translate into understanding the salience of aid money to a country's leaders, especially given the range of ways they may leverage aid money to stay in power. 10 A proportional measure of the target's economic dependence on US aid would not be capable of capturing any of this information. 11 By accounting for interest payments, the net aid transfer variable can identify targets that are paying more to the US Government in interest on outstanding ODA-related loans than they are receiving in new ODA in a given year. 12 Our results also hold if we employ current-year dollar values instead (see table A3). 13 We employ Gartzke's (2010) data that interpolates for missing values from 1960 to 2008 and updated data from Voeten et al. (2013). 14 This means that the United States was receiving back more in interest payments than it was giving to the targets in new aid. 15 We obtain congruent results if we exclude the cases of negative US Foreign Aid flows (see table A13). 16 We employed Long and Freese's (2014)’s S-Post suite. 17 We also ran models that interacted public aid threats and US Foreign Aid. We found that public aid threats have positive and statistically significant effects on both threat success and sanctions imposition across a large range of US Foreign Aid; however, the confidence intervals for private aid threats overlap with the public threats for the threat success outcome. This suggests that the strongest effects of making aid threats publicly is on senders’ willingness to follow through with imposing sanctions, as they have more of their credibility at stake in those cases. Additionally, we explored interacting US Foreign Aid with High Salience Issue. We found that higher values of US Foreign Aid lead to greater threat success when sanctions are imposed for nonsalient reasons and a higher likelihood of sanctions being imposed in cases involving high-salience sanctions issues (see table A16 and figures A5–6). References Agence-France Press . 2011 . “U.S. Industry Moves to Kill China Sanctions Bill,” September 21, 2011 . Alesina Alberto , Dollar David . 2000 . “Who Gives Foreign Aid to Whom and Why?” Journal of Economic Growth 5 ( 1 ): 33 – 63 . Google Scholar Crossref Search ADS Allen Susan , Lektzian David . 2013 . “Economic Sanctions: A Blunt Instrument?” Journal of Peace Research 50 : 121 – 35 . Google Scholar Crossref Search ADS Ang Adrian , Peksen Dursun . 2007 . “When Do Economic Sanctions Work? Asymmetric Perceptions, Issue Salience, and Outcomes.” Political Research Quarterly 60 ( 1 ): 135 – 45 . Google Scholar Crossref Search ADS Aronow Peter , Samii Cyrus , Assenova Valentina . 2015 . “Cluster-Robust Variance Estimation for Dyadic Data.” Political Analysis 23 ( 4 ): 564 – 77 . Google Scholar Crossref Search ADS Baldwin David . 1985 . Economic Statecraft . Princeton, NJ : Princeton University Press . Bapat Navin , Morgan T. Clifton . 2009 . “Multilateral versus Unilateral Sanctions Reconsidered: A Test Using New Data.” International Studies Quarterly 53 ( 4 ): 1468 – 2478. Google Scholar Crossref Search ADS Bapat Navin , Heinrich Tobias , Kobayashi Yoshiharu , Morgan T. Clifton . 2013 . “Determinants of Sanctions Effectiveness: Sensitivity Analysis Using New Data.” International Interactions 39 ( 1 ): 79 – 98 . Google Scholar Crossref Search ADS Barbieri Katherine , Keshk Omar . 2012 . “Correlates of War Project Trade Data Set Codebook, Version 3.0.” http://correlatesofwar.org . Barry Colin , Kleinberg Katja . 2015 . “Profiting from Sanctions: Economic Coercion and US Foreign Direct Investment.” International Organization 69 ( 4 ): 881 – 912 . Google Scholar Crossref Search ADS Box-Steffensmeier Janet M. , Jones Bradford S. . Timing and Political Change: Event History Modeling in Political Science . Ann Arbor : University of Michigan Press , 2004 . Bueno de Mesquita Bruce , Smith Alastair , Siverson Randolph M. . 2004 . The Logic of Political Survival . Cambridge, MA : MIT Press . Bueno de Mesquita Bruce , Smith Alastair . 2007 . “Foreign Aid and Policy Concessions.” Journal of Conflict Resolution 51 : 251 – 84 . Google Scholar Crossref Search ADS Carter David B. , Signorino Curtis S. . 2010 . “Back to the Future: Modeling Time Dependence in Binary Data.” Political Analysis 18 ( 3 ): 271 – 92 . Google Scholar Crossref Search ADS Cox Dan , Drury A. Cooper . 2006 . “Democratic Sanctions: Connecting the Democratic Peace and Economic Sanctions.” Journal of Peace Research , 43 ( 6 ): 709 – 22. Google Scholar Crossref Search ADS Drezner Daniel W. 2003 . “The Hidden Hand of Economic Coercion.” International Organization 57 ( 3 ): 643 – 59 . Google Scholar Crossref Search ADS Drezner Daniel W. . 1999 . The Sanctions Paradox: Economic statecraft and International Relations . Cambridge University Press , 1999 . Drury A. Cooper . 1998 . “Revisiting Economic Sanctions Reconsidered.” Journal of Peace Research , 35 ( 4 ): 497 – 509 . Google Scholar Crossref Search ADS Drury Cooper , Olson Richard , Belle Douglas Van . 2005 . “The Politics of Humanitarian Assistance: U.S. Foreign Disaster Aid, 1964–1995.” Journal of Politics 67 : 454 – 73 . Google Scholar Crossref Search ADS Early Bryan . 2015 . Busted Sanctions: Explaining Why U.S. Economic Sanctions Fail . Stanford, CA: Stanford University Press . Emmanuel Nikolas G. 2013 . “Democratization in Malawi: Responding to International and Domestic Pressures.” African and Asian Studies 12 ( 4 ): 415 – 34 . Google Scholar Crossref Search ADS Fleck Robert , Kilby Christopher . 2001 . “Foreign Aid and Domestic Politics: Voting in Congress and the Allocation of USAID Contracts across Congressional Districts.” Southern Economic Journal , 598 – 617 . Friedman Thomas . 1992 . “Bush Tells Congress ‘No’ on Aid to Israel.” New York Times, May 18, 1992 . Gartzke Erik . 2010 . “The Affinity of Nations: Similarity of State Voting Positions in the UNGA.” http://pages.ucsd.edu/∼egartzke/htmlpages/data.html . Gibler Douglas . 2009 . International Military Alliances , 1648 – 2008 . Washington, DC : CQ Press . Gleditsch Kristian . 2002 . “Expanded Trade and GDP Data.” Journal of Conflict Resolution , 46 ( 5 ): 712 – 24 . Google Scholar Crossref Search ADS Golden Tim . 1993a . “Guatemala's Counter-coup: A Military About-Face.” New York Times, June 3, 1993 . Golden Tim . 1993b . “In Startling Shift, Guatemala Makes Rights Aide President.” New York Times, June 7, 1993 . Guardian . 2009 . “US Cuts All Non-humanitarian Aid to Honduras in Support of Zelaya.” September 2, 2009 . Hirschman Albert . 1945 . National Power and the Structure of Foreign Trade . Berkley : University of California Press . Hufbauer Gary , Elliot Kimberly , Cyrus Tess , Winston Elizabeth . 1997 . “U.S. Economic Sanctions: Their Impact on Trade, Jobs, and Wages.” Working Paper Special . Washington, DC : Peterson Institute for International Economics . Hufbauer Gary , Schott Jeffrey , Elliott Kimberly , Oegg Barbara . 2007 . Economic Sanctions Reconsidered . Washington, DC : Peterson Institute for International Economics . Jackson David . 2014 . “Business Groups Oppose Any New Sanctions on Russia.” USA Today, June 25, 2014 . Kharas Homi . 2008 . “Measuring the Costs of Aid Volatility.” Wolfensohn Center for Development Working Paper 3 . Washington, DC : The Brookings Institute . Kim Dong-Hun . 2013 . “Coercive Assets? Foreign Direct Investment and the Use of Economic Sanctions.” International Interactions 39 ( 1 ): 99 – 117 . Google Scholar Crossref Search ADS Kim Youngwan . 2016 . “How NGOs Influence US Foreign Aid Allocations.” Foreign Policy Analysis 13, no. 1 (2017): 112–32 . Kosack Stephen , Tobin Jennifer . 2006 . “Funding Self-Sustaining Development: The Role of Aid, FDI, and Government in Economic Success.” International Organization 60 : 205 – 43 . Google Scholar Crossref Search ADS Lektzian David , Biglaiser Glen . 2014 . “The Effect of Foreign Direct Investment on the Use and Success of US Sanctions.” Conflict Management and Peace Science 31 ( 1 ): 70 – 93 . Google Scholar Crossref Search ADS Lesniewski Niels . 2012 . “Lindsey Graham, Rand Paul Rivalry on Foreign Aid Grows.” Roll Call, October 10, 2012. http://www.rollcall.com/news/-218130-1.html . Licht Amanda . 2010 . “Coming into Money: The Impact of Foreign Aid on Leader Tenure.” Journal of Conflict Resolution , 54 ( 1 ): 58 – 87 . Google Scholar Crossref Search ADS Long Scott , Freese Jeremy . 2014 . Regression Models for Categorical Dependent Variables Using Stata , 3rd ed . College Station, TX : Stata Press . Maizels Alfred , Nissanke Machiko . 1984 . “Motivations for Aid to Developing Countries.” World Development 12 ( 9 ): 879 – 900 . Google Scholar Crossref Search ADS Marshall Monty , Jaggers Keith . 2009 . Polity IV Project: Political Regime Characteristics and Transitions, 1800–2007 . Arlington, VA : Polity IV Project . Miller Nicholas . 2014 . “The Secret Success of Nonproliferation Sanctions” International Organization 68 ( 4 ): 913 – 44 . Google Scholar Crossref Search ADS Milner Helen , Tingley Dustin . 2010 . “The Political Economy of US Foreign Aid: American Legislators and the Domestic Politics of Aid.” Economics & Politics 22 ( 2 ): 200 – 32 . Milner Helen V. , Tingley Dustin H. . 2013 . “Public Opinion and Foreign Aid: A Review Essay.” International Interactions 39 ( 3 ): 389 – 401. Google Scholar Crossref Search ADS Morgan T. Clifton , Bapat Navin , Kobayashi Yoshiharu . 2014 . “The Threat and Imposition of Economic Sanctions 1945–2005: Updating the TIES Dataset.” Conflict Management and Peace Science 31 ( 5 ): 541 – 58 . Google Scholar Crossref Search ADS Morgan T. Clifton . 2015 . “Hearing the Noise: Economic Sanctions Theory and Anomalous Evidence.” International Interactions 41 ( 4 ): 744 – 54 . Google Scholar Crossref Search ADS Morrow James . 1999 . “How Could Trade Affect Conflict?” Journal of Peace Research 36 ( 4 ): 481 – 9 . Google Scholar Crossref Search ADS Nielsen Richard . 2013 . “Rewarding Human Rights? Selective Aid Sanctions against Repressive States.” International Studies Quarterly 57 : 791 – 803 . Google Scholar Crossref Search ADS Neuenkirch Matthias , Neumeier Florian . 2015 . “The Impact of UN and US Economic Sanctions on GDP Growth.” European Journal of Political Economy 40 : 110 – 25 . Google Scholar Crossref Search ADS Newnham Randall . 2008 . “‘Coalition of the Bribed and Bullied?’ U.S. Economic Linkage and the Iraq War Coalition.” International Studies Perspectives 9 ( 2 ): 183 – 200 . Google Scholar Crossref Search ADS New York Times, 1992 . “Pressing for Principle in Israel.” February 26, 1992 . Olson Richard . 1979 . “Economic Coercion in World Politics: With a Focus on North-South Relations.” World Politics 31 ( 4 ): 471 – 94 . Google Scholar Crossref Search ADS Peksen Dursun , Peterson Timothy . 2016 . “Sanctions and Alternate Markets: How Trade and Alliances Affect the Onset of Economic Coercion.” Political Research Quarterly 69 ( 1 ): 4 – 16 . Google Scholar Crossref Search ADS Peterson Timothy . 2013 . “Sending a Message: The Reputation Effect of US Sanction Threat Behavior.” International Studies Quarterly 57 : 672 – 82 . Google Scholar Crossref Search ADS Peterson Timothy , Drury Cooper . 2011 . “Sanctioning Violence: The Effect of Third-Party Economic Coercion on Military Conflict.” Journal of Conflict Resolution 55 ( 4 ): 580 – 605 . Google Scholar Crossref Search ADS Robertsen Jamie , Francken Nathalie , Molenaers Nadia . 2015 . “Determinants of the Flow of Bilateral Adaptation-related Climate Change Financing to Sub-Saharan Countries: The Extent to which Environmental, Economic, Historical and Political Factors Play a Role.” LICOS Discussion Paper Series 373 . Roodman David . 2011 . “An Index of Donor Performance.” Center for Global Development. Working Paper #67 . Schelling Thomas . 1966 . Arms and Influence . New Haven, CT : Yale . Tarnoff Curt , Lawson Marian . 2011 . Foreign Aid: An Introduction to US Programs and Policy . Washington, DC: Congressional Research Service, 2011 . The Organisation for Economic Cooperation and Development (OECD). 2018 . “The Definition of Official Development Assistance.” http://www.oecd.org/dac/stats/officialdevelopmentassistancedefinitionandcoverage.htm . Van Bergeijk Peter , van Marrewijk Charles . 1995 . “Why Do Sanctions Need Time to Work? Adjustment, Learning, and Anticipation.” Economic Modeling 12 ( 2 ): 75 – 86 . Google Scholar Crossref Search ADS Voeten Erik , Strezhnev Anton , Bailey Michael . 2013 . “United Nations General Assembly Voting Data.” https://thedata.harvard.edu/dvn/dv/Voeten/faces/study/StudyPage.xhtml?globalId=hdl:1902.1/12379&studyListingIndex=1_4cbddd2916efea2ae2729fc20fda . Whang Taehee , McLean Elena , Kuberski Douglas . 2013 . “Coercion, Information, and the Success of Sanction Threats.” American Journal of Political Science 57 ( 1 ): 65 – 81 . © The Author(s) (2019). Published by Oxford University Press on behalf of the International Studies Association. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Foreign Policy Analysis Oxford University Press http://www.deepdyve.com/lp/oxford-university-press/using-the-carrot-as-the-stick-us-foreign-aid-and-the-effectiveness-of-z3n5aKuoUs

Loading next page...

References (76)

B. Mesquita, Alastair Smith (2007)
Foreign Aid and Policy Concessions
Journal of Conflict Resolution, 51
(2015)
citation_publisher=Stanford University Press, ; Busted Sanctions: Explaining Why U.S. Economic Sanctions Fail
T. Peterson, A. Drury (2011)
Sanctioning Violence
Journal of Conflict Resolution, 55
(2014)
citation_publisher=Stata Press, College Station, TX; Regression Models for Categorical Dependent Variables Using Stata
Richard Nielsen (2013)
Rewarding Human Rights? Selective Aid Sanctions against Repressive States
International Studies Quarterly, 57
(1945)
citation_publisher=University of California Press, Berkley; National Power and the Structure of Foreign Trade
B. Crane, T. Schelling (1966)
Arms and Influence
International Journal, 22
(2014)
“Business Groups Oppose Any New Sanctions on Russia.”
Young-wan Kim (2014)
How NGOs Influence US Foreign Aid Allocations
Foreign Policy Analysis, 13
B. Mesquita, Alastair Smith, Randolph Siverson, J. Morrow (2003)
The Logic of Political Survival
B. Early (2020)
Busted Sanctions
R. Olson (1979)
Economic Coercion in World Politics: With a Focus on North-South Relations
World Politics, 31
(2013)
?Economic Sanctions: A Blunt Instrument??
Journal of Peace Research, 50
(1985)
citation_publisher=Princeton University Press, Princeton, NJ; Economic Statecraft
R. Fleck, Christopher Kilby (2001)
Foreign Aid and Domestic Politics: Voting in Congress and the Allocation of USAID Contracts across Congressional Districts
Southern Economic Journal, 67
(2015)
?The Impact of UN and US Economic Sanctions on GDP Growth.?
European Journal of Political Economy, 40
H. Milner, D. Tingley (2013)
Public Opinion and Foreign Aid: A Review Essay
International Interactions, 39
(2005)
?The Politics of Humanitarian Assistance: U.S. Foreign Disaster Aid, 1964?1995.?
Journal of Politics, 67
(2010)
?Coming into Money: The Impact of Foreign Aid on Leader Tenure.?
Journal of Conflict Resolution, 54
T. Morgan (2015)
Hearing the Noise: Economic Sanctions Theory and Anomalous Evidence
International Interactions, 41
A. Drury (1998)
Revisiting Economic Sanctions Reconsidered
Journal of Peace Research, 35
(1993)
“Guatemala's Counter-coup: A Military About-Face.”
Randall Newnham (2008)
“Coalition of the Bribed and Bullied?” U.S. Economic Linkage and the Iraq War Coalition
International Studies Perspectives, 9
A. Maizels, M. Nissanke (1984)
Motivations for aid to developing countries
World Development, 12
A. Alesina, David Dollar (1998)
Who Gives Foreign Aid to Whom and Why?
Journal of Economic Growth, 5
(2012)
“Correlates of War Project Trade Data Set Codebook, Version 3.0.”
Matthias Neuenkirch, Florian Neumeier (2014)
The Impact of UN and US Economic Sanctions on GDP Growth
PSN: Other Inter-State Conflict (Topic)
Sang Kim (2015)
Does Aid Change North Korea's Foreign Policy? An Empirical Analysis Using the United Nations General Assembly Voting Data
, 25
Colin Barry, Katja Kleinberg (2015)
Profiting from Sanctions: Economic Coercion and US Foreign Direct Investment in Third-Party States
International Organization, 69
G. Hufbauer, K. Elliott, Teresa Cyrus, E. Winston (1997)
US Economic Sanctions: Their Impact on Trade, Jobs, and Wages
(2015)
?Profiting from Sanctions: Economic Coercion and US Foreign Direct Investment.?
International Organization, 69
David Lektzian, G. Biglaiser (2014)
The effect of foreign direct investment on the use and success of US sanctions
Conflict Management and Peace Science, 31
S. Allen, David Lektzian (2013)
Economic sanctions
Journal of Peace Research, 50
David Carter, Curtis Signorino (2010)
Back to the Future: Modeling Time Dependence in Binary Data
Political Analysis, 18
(1993)
“In Startling Shift, Guatemala Makes Rights Aide President.”
(1992)
“Bush Tells Congress ‘No’ on Aid to Israel.”
K. Gleditsch (2002)
Expanded Trade and GDP Data
Journal of Conflict Resolution, 46
Navin Bapat, Tobias Heinrich, Yoshiharu Kobayashi, T. Morgan (2013)
Determinants of Sanctions Effectiveness: Sensitivity Analysis Using New Data
International Interactions, 39
H. Milner, D. Tingley (2009)
The Political Economy of U.S. Foreign Aid: American Legislators and the Domestic Politics of Aid
ERN: Models of Political Processes: Rent-Seeking
(2009)
citation_publisher=CQ Press, Washington, DC; International Military Alliances
(2009)
citation_publisher=Polity IV Project, Arlington, VA; Polity IV Project: Political Regime Characteristics and Transitions, 1800?2007
(2010)
?The Political Economy of US Foreign Aid: American Legislators and the Domestic Politics of Aid.?
Economics & Politics, 22
(2004)
citation_publisher=University of Michigan Press, Ann Arbor; Timing and Political Change: Event History Modeling in Political Science
(2004)
citation_publisher=MIT Press, Cambridge, MA; The Logic of Political Survival
P. Bergeijk, C. Marrewijk (1995)
Why do sanctions need time to work? Adjustment, learning and anticipation
Economic Modelling, 12
Navin Bapat, T. Morgan (2009)
Multilateral Versus Unilateral Sanctions Reconsidered: A Test Using New Data
International Studies Quarterly, 53
G. Hufbauer (1985)
Economic Sanctions Reconsidered
(2012)
“Lindsey Graham, Rand Paul Rivalry on Foreign Aid Grows.”
A. Licht (2010)
Coming into Money: The Impact of Foreign Aid on Leader Survival
Journal of Conflict Resolution, 54
A. Drury, R. Olson, Douglas Belle (2005)
The Politics of Humanitarian Aid: U.S. Foreign Disaster Assistance, 1964–1995
The Journal of Politics, 67
(2018)
“The Definition of Official Development Assistance.”
(2016)
?Sanctions and Alternate Markets: How Trade and Alliances Affect the Onset of Economic Coercion.?
Political Research Quarterly, 69
(2007)
?When Do Economic Sanctions Work? Asymmetric Perceptions, Issue Salience, and Outcomes.?
Political Research Quarterly, 60
N. Miller (2014)
The Secret Success of Nonproliferation Sanctions
International Organization, 68
(2009)
“US Cuts All Non-humanitarian Aid to Honduras in Support of Zelaya.”
Stephen Kosack, Jennifer Tobin (2006)
Funding Self-Sustaining Development: The Role of Aid, FDI and Government in Economic Success
International Organization, 60
(2011)
citation_publisher=Washington, DC: Congressional Research Service, 2011, ; Foreign Aid: An Introduction to US Programs and Policy
(2011)
?Sanctioning Violence: The Effect of Third-Party Economic Coercion on Military Conflict.?
Journal of Conflict Resolution, 55
A. Hirschman (1946)
National Power and the Structure of Foreign Trade
J. Long, J. Freese (2005)
Regression models for categorical dependent variables using Stata, 2nd Edition
Nikolas Emmanuel (2013)
Democratization in Malawi: Responding to International and Domestic Pressures
African and Asian Studies, 12
東郷賢, 和田義郎 (2006)
Index of Donor Performance(援助供与国成績指標)について
, 54
Taehee Whang, Elena McLean, D. Kuberski (2013)
Coercion, Information, and the Success of Sanction Threats
American Journal of Political Science, 57
Dursun Peksen, T. Peterson (2016)
Sanctions and Alternate Markets
Political Research Quarterly, 69
(1966)
citation_publisher=Yale, New Haven, CT; Arms and Influence
(2010)
“The Affinity of Nations: Similarity of State Voting Positions in the UNGA.”
T. Peterson (2013)
Sending a Message: The Reputation Effect of US Sanction Threat Behavior
International Studies Quarterly, 57
Daniel Cox, A. Drury (2006)
Democratic Sanctions: Connecting the Democratic Peace and Economic Sanctions
Journal of Peace Research, 43
Trefor Morgan, Navin Bapat, Yoshiharu Kobayashi (2014)
Threat and imposition of economic sanctions 1945–2005: Updating the TIES dataset
Conflict Management and Peace Science, 31
A. Ang, Dursun Peksen (2007)
When Do Economic Sanctions Work?
Political Research Quarterly, 60
J. Morrow (1999)
How Could Trade Affect Conflict?
Journal of Peace Research, 36
(2007)
citation_publisher=Peterson Institute for International Economics, Washington, DC; Economic Sanctions Reconsidered
Dong-hun Kim (2011)
Coercive Assets? Foreign Direct Investment and the Use of Economic Sanctions
International Interactions, 39
Daniel Drezner (2003)
The Hidden Hand of Economic Coercion
International Organization, 57
Homi Kharas (2008)
Measuring the Cost of Aid Volatility
Development Economics eJournal
P. Aronow, Cyrus Samii, Valentina Assenova (2013)
Cluster–Robust Variance Estimation for Dyadic Data
Political Analysis, 23

Publisher: Oxford University Press
Copyright: © The Author(s) (2019). Published by Oxford University Press on behalf of the International Studies Association.
ISSN: 1743-8586
eISSN: 1743-8594
DOI: 10.1093/fpa/orz007
Publisher site: See Article on Publisher Site

Abstract

Abstract We theorize that foreign aid relationships influence both the effectiveness of economic sanctions threats and the aggressiveness of senders in imposing sanctions. Aid sanctions are generally far less costly for senders than imposing commercially oriented sanctions but can still be very costly for their targets. Being able to disrupt aid flows as part of potential sanctions enhances sender states’ credibility that they will impose painful sanctions against resisting target states. The more foreign aid a sender provides to a target state, the more successful we expect its sanctions threats to be and the more aggressive we expect the sender to be in imposing sanctions if the target resists. We test our theory using a competing risks analysis of ongoing, politically motivated sanctions threats issued by the United States from 1960–2010. Our analyses support our theory by revealing that the more foreign aid that the United States provides to target states, the more likely US sanctions threats are to succeed and the more aggressive the United States becomes in imposing sanctions. Video Abstract Video Abstract Close Introduction Economic sanctions and foreign aid alternatively rely on coercion and incentives to alter the behavior of their targets. The distinction between the two can often blur. Donors frequently provide foreign aid to incentivize recipients to change their behavior, but, once they issue aid commitments, donors can threaten recipients with the disruption or cancellation of promised aid. This type of economic coercion operates via a slightly different logic than sanctions that disrupt commercial relations based upon mutual profitability. Aid suspensions are far costlier for recipients than for donors, creating a vulnerability that donors can exploit in employing economic coercion. Aid sanctions are thus different from commercially oriented sanctions, which often can inflict comparable economic costs upon both parties. Suspending aid relationships is not necessarily without domestic political costs for senders, but cutting foreign aid will still almost always be less costly than severing trade or investment relationships. Foreign-aid relationships can thus enhance donors’ options for credibly conveying their willingness to impose costly measures when threatening targets with sanctions. Within this study, we explain how foreign-aid relationships shape both the effectiveness of sanctions threats and their senders’ willingness to follow through on them. Studying how foreign-aid flows influence the outcome of sanctions threats can provide insight into how foreign aid and sanctions policies relate to one another. Economic sanctions are compellent policies that restrict a country's government or constituents from engaging in international economic transactions to induce a change in their behavior.1 Foreign aid, in contrast, involves giving or lending foreign governments and nonstate actors money, material resources, training, or expertise. While research suggests that donors provide foreign aid to further their political interests (i.e., Drury, Olson, and Van Belle 2005), only a portion of foreign aid involves explicit quid pro quo bargains between donors and recipients. Imposing aid-related sanctions differs from simply decreasing aid to recipients (even for political reasons), as the former involves explicit demands that targets can comply with to avoid the loss of aid.2 Furthermore, donors can use the threat of foreign-aid cuts—and not just their actual reduction—to compel changes in their recipients’ behavior. The coercive nature of sanctions provides a clear demarcation between sanctioning episodes and politically motivated aid reductions (e.g., Nielsen 2013). Past works suggest that donors’ foreign-aid relationships with recipients should have a powerful effect on the outcomes of sanctions episodes at the threat stage. Morrow (1999) has argued that economic ties can enhance the range of opportunities that states have to convey credible, costly, and observable signals of their resolve during crisis bargaining. Indeed, studies on the effects of cross-border foreign direct investment and target trade dependency have shown that economic ties between sender and target states influence sanctions threats’ credibility (Kim 2013; Whang, McLean, and Kuberski 2013). The more credible and costly the prospective sanctions, the more likely target states are to capitulate during the threat stage of coercive episodes. Leveraging the TIES dataset (Morgan et al. 2014), Bapat et al. (2013, 90–92) find that sanctions threats that promise the imposition of severe costs on target states are far more effective. For states employing economic coercion, it is more cost effective to succeed at the threat stage than to succeed as a result of imposing sanctions (Drezner 2003). Therefore, states should be more attentive to the factors that make sanctions threats more effective than factors that influence the success of imposed sanctions. We theorize that existing foreign-aid relationships with target states enhance sender states’ ability to issue costly sanctions threats and increase their credibility to follow through on them. Economic sanctions and dramatic reductions in foreign-aid flows can have devastating economic consequences for aid recipients (Kharas 2008; Neuenkirch and Neumeier 2015). Donors can leverage their recipients’ vulnerability to their aid's disruption in subtle ways, making aid recipients more susceptible to economic coercion. Compared to imposing trade, investment, or financial sanctions, cutting foreign aid saves sender governments money and generates less political resentment amongst sender states’ constituents. Sender states that give substantial aid to the targets of their sanctions threats thus have enhanced credibility in conveying their willingness to follow through with imposing costly sanctions. As such, we hypothesize that target states will be more willing to give in to a sender's sanctions threats the more aid they receive from the sender. When aid recipients refuse to concede to sanctions threats alone, we expect senders to be more willing to impose sanctions against them than against non-aid recipients. We test our theory via a large-n analysis of the ongoing, politically motivated episodes of economic coercion from 1960 to 2010 in which the United States was the primary sender. As the world's most prolific sanctioner and one of the world's largest foreign-aid donors, the United States represents a critical case to be explained by our theory. We structure our study as a competing risks analysis in which the dependent variable is the yearly outcome of the sanctions threat episodes coded in one of four ways: threats persisting in a given year, ending in success, ending in the United States backing down, or resulting in sanctions being imposed. This approach allows us to distinguish between the two ways that senders can respond when their threats are not successful, by either imposing sanctions or backing down. Consistent with our theory, the quantitative analysis shows that the more aid the United States provides to the states it threatens with sanctions, the more likely those states are to capitulate at the threat stage and the more likely the United States is to impose sanctions on targets that do not concede. These observed effects are even stronger when we analyze only those coercive episodes that entailed public threats to cut foreign aid. Together, these findings indicate that US foreign aid provides American policymakers with significant leverage in employing economic coercion against the recipients of its aid. Our study contributes to the economic statecraft literature by demonstrating that foreign-aid relationships significantly influence sanctions-related behaviors. Foreign-aid relationships not only make senders’ sanctions threats more effective but also make senders more aggressive in imposing sanctions. As we demonstrate, distinguishing between failed threats that result in senders backing down versus imposing sanctions can yield theoretically valuable insights and is an approach that should be more widely adopted in leveraging the TIES dataset (Morgan et al. 2014). Our study also contributes to the broader efforts of the economic statecraft literature to understand how various forms of economic interconnectedness influence coercive behavior and outcomes within international relations (Baldwin 1985; Bapat et al. 2013; Kim 2013; Lektzian and Biglaiser 2014; Miller 2014; Barry and Kleinberg 2015; Early 2015; Peksen and Peterson 2016). If providing aid to recipient states makes them more vulnerable to economic coercion, donors could strategically invest in offering aid to states that they may want coercive leverage over in the future. In this way, foreign aid can be used as a realpolitik tool for keeping your friends close and your “frenemies” closer. In the conclusion, we discuss our findings’ policy-relevant insights for US sanctions and foreign-aid policies. Economic Coercion and Statecraft Governments employ economic statecraft, including foreign aid and sanctions, to change their external environments (Baldwin 1985). States exert pressure through the use of rewards or by inflicting pain via economic sanctions or aid cuts. States that possess asymmetric commercial relationships with their trading partners can exploit the coercive leverage that such relations offer (Hirschman 1945) and can signal their resolve more efficiently (Morrow 1999). By their very nature, the power dynamics between aid donors and recipient states comprise an imbalanced economic relationship. Most donor states give foreign aid to obtain something in return for it (Bueno de Mesquita, Smith, and Siverson 2004), as well as to maintain their relationships with allies, former colonies, and major trade partners (Alesina and Dollar 2000). Just as foreign aid can be used as a carrot to incentivize behavioral changes, its threatened withdrawal can be used as a stick to coerce recipients (Newnham 2008). Aid flows provide donors with political leverage to influence the behavior of the recipient through threatened or actual disruption of bilateral aid flows. From a donor's perspective, aid cuts may be costly and difficult to implement if political and economic interests influence donors’ foreign-aid policies with distributional consequences for domestic groups. Indeed, Milner and Tingley (2010) find that American legislators’ voting patterns on aid reflect the economic interests of their constituents. However, a closer look at their analysis suggests that such domestic interests primarily influence voting on general increases or decreases in foreign-aid appropriations rather than on aid flows involving a particular country. Furthermore, donors do not necessarily need to sever all aid to a recipient to exploit its coercive leverage but may threaten to withhold certain components, such as budget support, to induce a change in recipients’ behavior. Robertsen et al.’s (2015) analysis of Sub-Saharan African recipient states indicates frequent use of budget support aid suspensions by the UK, Netherlands, and Sweden over political, human rights, and corruption issues. They further find that while donors tend to provide more aid to their former colonies, they also suspend budget support more easily to such recipients. These studies suggest that the more recipients rely on a donor for aid, the more coercive leverage the donor possesses to compel changes in the recipients’ behavior. In practice, we often see donors use aid cuts coercively to elicit desired behavioral changes in recipient states. For example, donor countries pushed Malawi towards democratization between 1992 and 1994 by cutting 25 percent of their aid to the country to leverage political reforms. The suspension of economic aid affected Malawi quickly—resulting in currency devaluation and an annual 7.3 percent contraction of Malawi's economy (Emmanuel 2013). Additionally, the United States axed all nonhumanitarian aid to Honduras to coerce the de facto government to restore democratic rule after a coup in June 2009 (Guardian 2009). Indeed, previous work has shown that aid-dependent countries like those discussed above are particularly susceptible to economic coercion (Olson 1979). The ability to threaten aid reductions affords donors potentially significant coercive leverage over aid recipients. This is particularly applicable to bilateral aid, which is often driven by donors’ political and security interests, as compared to multilateral aid (Maizels and Nissanke 1984). Extensive use of bilateral aid can thus place donors in a powerful position to use threats of aid cuts for coercive leverage. The United States, for example, provides a large proportion of its aid bilaterally and is consistently ranked as one of the top donors globally. In 2010, 32 percent of its total aid was in the form of bilateral assistance, compared with only 7 percent for multilateral development projects (Tarnoff and Lawson 2011). Additionally, with the exception of a few years, the Unite States ranked first amongst Official Development Assistance (ODA) donors, making up 24 percent of total ODA disbursements in 2009 (Tarnoff and Lawson 2011). From a recipient's perspective, this potential leverage becomes apparent if we focus on the political and economic effects of foreign aid within recipient states. Survival-driven leaders have significant incentives to use aid to enhance their domestic political power and longevity (Bueno de Mesquita and Smith 2007; Licht 2010) or use it as a “substitute for government spending” (Kosack and Tobin 2006: 210). Governments that rely on foreign assistance to meet their citizens’ needs and maintain their rule are thus vulnerable to its potential withdrawal. This makes them beholden—at least to some extent—to the donors that provide them with aid. Focusing specifically on sanctions threats, Whang et al. (2013) argue that the efficacy of sanctions threats is conditioned by the existence and magnitude of asymmetrical economic exchange between the sender and target states. They emphasize the importance of relative economic vulnerability of the target to the disruption of their economic relationship, as opposed to an overall economic loss in the target state. States will acquiesce in the pre-imposition stage if the expected utility of compliance outweighs that of resisting the sanctions. In contrast, the sender's economic dependence on the target makes it less likely to impose sanctions if the target resists. Miller (2014) argues that the United States has used the threat of nonproliferation sanctions most effectively against states that are highly dependent upon it for their economic and security wellbeing. States that are highly dependent upon the United States are more “vulnerable” to its coercive threats according to Miller (2014, 913) because they have “more to lose from U.S. sanctions.” These works suggest that senders can exploit their targets’ vulnerability to economic disruption to obtain concessions before sanctions are imposed. This gives us a strong reason to think that foreign-aid relationships can influence the outcomes of sanctions threats. Using the Carrot as the Stick: Foreign Aid and the Effectiveness of Sanctions Threats A sender initiating a coercive episode seeks to compel its target into making the desired concessions rapidly and cost-effectively. The most efficient stage for a coercive episode to be won is at the threat stage, when the sender wagers its credibility and the prospective costs of imposing sanctions against the target's resolve. If the target state lacks resolve and views the threats as credible (Kim 2013), it will give in to the sender's demands by capitulating or negotiating concessions. On the contrary, if the target state resists the sanctions threat, the sender must decide between backing down or imposing the threatened sanctions. By backing down, the sender risks a loss of its credibility in future coercive episodes (Peterson 2013). By imposing sanctions, senders still have a chance to obtain the desired concessions but at a much higher cost than obtaining concessions during the threat phase. None of these decisions are instantaneous, either. Target states can delay their response to gain more information about the credibility of the sender state's threats. While waiting, senders can reference new information to assist their decision about the best response and explore ways of enhancing the effectiveness of their threats. Foreign Aid, Coercive Leverage, and Sanctions Threats The impact of a sanctions threat is influenced by perceptions of the credibility of the threat, its prospective costs to implement, and speculation about the future outcome should sanctions be imposed. The presence of robust foreign-aid relationships enhances the range of “opportunities” (Morrow 1999) available to senders in credibly conveying their willingness to impose costs on target states. The ability to threaten aid cuts complements senders’ abilities to threaten trade or investment sanctions. Existing foreign-aid relationships between sender and target states influence how both states perceive their positions during threat episodes. Target leaders and governments are sensitive to threatened aid cuts because their financial welfare may be directly tied to external aid. Survival driven leaders often use aid to enhance their domestic power (Bueno de Mesquita and Smith 2007; Licht 2010) and to support government spending (Kosack and Tobin 2006). The disruption of aid relationships can be economically and politically damaging for recipients, suggesting they will try to prevent their occurrence. Reflecting this, several studies indicate that donors exploit their aid relationships with recipients to obtain concessions by threatening the withdrawal of their aid (Bueno de Mesquita and Smith 2007; Newnham 2008). We argue that the more foreign aid that a sanctions-threatening donor gives to a target state, the more successful those threats will be and the more aggressive senders will be in imposing sanctions. Being able to threaten—either privately or publicly—the disruption of aid flows a target state receives allows the sender to increase the prospective costs for resisting its sanctions threats. Target states will be more sensitive to sanctions threats when they receive aid from the sender because substantial aid cuts can harm their economies. For aid recipients, finding alternative revenue sources due to unexpected budgetary shortfalls can be difficult—especially in the short term. Leaders may have to make emergency cuts to planned spending, suspend governmental services, or borrow money under duress. Indeed, Kharas (2008) has found that states suffering “aid shocks” can experience deadweight economic losses of almost 2 percent of their gross domestic products. Notably, these effects would not likely occur in isolation when aid cuts occur with other types of sanctions. Both public and private aid threats can be effective. By making public aid threats, senders can specify the dollar amounts or aid programs being jeopardized by their threats that allow targets to understand the costs of noncompliance more clearly. In some circumstances, though, senders may prefer to communicate threatened aid cuts via private channels due to the aid's political sensitivity or because they recognize targets’ reputational concerns.3 Drawing on dependency theory, Olson (1979, 485) recognized how wealthy countries could engage in more subtle and effective types of coercion using the aid they provided to less developed countries as leverage. “Modern dependency,” according to Olson, “allows effective economic coercion to be applied without much publicity and without generating a widespread nationalist response within the target-country.” This suggests that target leaders are particularly susceptible to threats that could jeopardize foreign aid, which can be communicated privately to the leadership of target states. Targeted leaders will be blamed for the hardships created by foreign-aid cuts, irrespective of whether the public is broadly aware of why those cuts were imposed. Given their self-interests in retaining power (e.g., Bueno de Mesquita et al. 2004), leaders should be much more likely to concede to sender states’ demands at the threat stage, if their governments receive a substantial amount of foreign aid from them, than allow sanctions to be implemented. As Olson (1979, 481) aptly surmises, “Dependency is indeed vulnerability.” Sanctions threats are also more credible when senders possess aid-based coercive leverage because aid cuts are less costly than commercially oriented sanctions. This argument is consistent with Morgan's (2015, 749) observation that “senders are very good at designing sanctions that are costly to the target but not to themselves.” For sender states, imposing trade and financial sanctions can involve imposing costly disruptions upon their own constituents’ commerce (Hufbauer et al. 1997). Cutting off foreign aid to target states is less likely to harm the economic well-being of domestic constituencies or result in political mobilization than commercially oriented sanctions. Disrupting foreign-aid flows may generate some domestic economic grievances, but it is unlikely to cause the same magnitude of costs as commercial sanctions or to generate the same degree of political backlash.4 Sender governments can also find cost-neutral ways of compensating adversely affected domestic constituencies by reallocating the aid money to new recipients or via other channels. Commercial interest groups often mobilize against trade-related sanctions due to the losses they impose on US businesses (Early 2015, 67). For example, the US Chamber of Commerce and National Association of Manufacturers lobbied against the imposition of sanctions against Russia after its invasion and annexation of Crimea (Jackson 2014). In 2011, the US Chamber of Commerce and fifty other trade groups lobbied Congress not to follow through on threatened legislation to sanction China for its currency manipulation practices (Agence-France Press 2011). While the domestic beneficiaries of foreign aid, like the NGO community, do exert influence over US foreign-aid allocations (e.g., Kim 2016), they tend not to pursue high-profile, well-financed campaigns to block sanctions imposition. Indeed, such efforts would run up against the US public's antipathy towards foreign aid (Milner and Tingley 2013). The political tide in the United States thus runs against interest groups seeking to block foreign-aid sanctions. This suggests that, in general, target leaders should perceive fewer domestic obstacles for the United States following through on sanctions threats that would target foreign-aid flows. The United States’ coercive behavior appears to bear this logic out. As US Senator Lindsey Graham put it, US foreign aid “is designed to create leverage so we [the US government] influence the world and not have the world just run us over” (Lesniewski 2012). The United States’ broad portfolio of foreign-aid relationships empowers it to employ economic coercion in pursuit of US foreign interests. In our sample of 192 politically motivated sanctions threat episodes with the United States as the primary sender, 80 out of the 103 cases (78 percent) in which the United States imposed sanctions also involved cutting foreign aid to target states (Morgan et al. 2014). Sanctions threats are more likely to be perceived as credible if the target state receives substantial foreign-aid flows from the sender that could be jeopardized by sanctions. As we note above, senders do not have to publicly threaten aid cuts initially to include them in the sanctions they eventually impose. As such, aid relationships are expected to influence the general efficacy of all sanctions threats cases. Based on our theory, we expect that greater aid flows will be associated with an enhanced willingness by the senders to impose sanctions if targets do not capitulate. Having leverage over target states during the threat stage will increase donors’ confidence about their ability to compel the target to capitulate by imposing sanctions. Such aggressiveness on the donor's part indicates that they expect their coercive efforts to ultimately succeed—thus highlighting the importance of distinguishing between threat failures in which senders back down versus double down on sanctioning the target. This theoretical expectation distinguishes our account from an alternative explanation that senders are bribing targets to give in to their sanctions threats. According to the latter perspective, the aid given to threatened states would lessen their costs of capitulation and, therefore, be positively correlated with threats success rather than sanctions imposition. One would expect, though, that the United States would use a bribery-based strategy to prevent having to impose sanctions. Prima facie, there is no reason why pursuing a bribery-based strategy would lead to a higher likelihood of sanctions imposition. A final implication of our theory relates to how foreign-aid relationships may affect donors’ willingness to back down from their threats. Backing down from sanctions threats can damage senders’ credibility and make their future threats less effective (Peterson 2013). Possessing foreign-aid leverage over a target could mean that donors are less likely to back down in the aggregate. Alternatively, though, having foreign-aid leverage over the target could make donors less worried about how backing down will affect their ability to coerce the target in the future—as they will still have their aid relationships to use as leverage. Thus, foreign-aid leverage that makes donor states more aggressive in imposing sanctions may also make it easier for senders to back down from unpromising sanctions threats. Given that both factors could be in play, we do not have strong theoretical predictions about the impact of foreign-aid flows on senders’ likelihoods of backing down. All else being equal, we expect that sanctions threats issued against states that receive significant quantities of foreign aid from senders should be more successful than threats issued against targets that receive little to no aid from the sender. Foreign aid from sender states raises the prospective costs of sanctions for target states and enhances their credibility of imposing such sanctions. Senders should also be more aggressive in imposing sanctions because they view their prospects of success as higher and will likely face less domestic political opposition for imposing such sanctions. Indeed, it is the enhanced willingness of senders to impose sanctions in such cases that should raise threatened states’ willingness to capitulate. Our theory gives rise to the following two hypotheses (H1 and H2): Hypothesis 1:The larger the foreign aid flows that senders provide to targets, the more likely the sender's sanctions threats will be successful. Hypothesis 2: Hypothesis 2:The larger the foreign aid flows that senders provide to targets, the more likely senders will be to impose sanctions when targets refuse to concede to the threats. Research Design We test our theory via a competing risks model capable of analyzing the factors associated with the various outcomes of US sanctions threats. Our analysis employs a sample of 192 ongoing, politically motivated sanctions threats from 1960–2010 from the TIES dataset in which the United States was the primary sender (Morgan et al. 2014).5 We focus our analysis on the United States, which is an active employer of economic sanctions and a top ranking foreign-aid donor, to control for the strength and political profile of the sender. The United States is also one of the few states that can afford to both give large amounts of foreign aid and actively employ economic coercion. We examine politically motivated sanctions because the salience of the issues involved tends to be higher.6 The unit of analysis for our dataset is the coercive episode-year, in which we code a coercive episode from the year in which the United States issues a specific sanctions threat until the episode terminates with the United States backing down from its threat, the United States imposing sanctions, or the target conceding to the sanctions threat. Our dataset is thus structured along the lines of an events history format in which yearly observations for coercive episodes are included until one of the specified termination events occur, and the case exits the dataset (Box-Steffensmeier and Jones 2004). Importantly, structuring our dataset this way allows us to include multiple observations of ongoing sanctions threats against specific target states at the same time. Via this competing risks approach, we account for the factors that influence the decisions by target or sender states that lead to terminal outcomes, such as backing down (sender), imposing sanctions (sender), or acquiescing (target). The theoretical motivation for this modeling strategy is that coercive episodes unfold over time, with both the sender and target states seeking to gain greater insights into their opponents’ resolve and the costs associated with various choices (Van Bergeijk and Van Marrewijk 1995; Early 2015). Upon the issuance of sanctions threats and prior to a terminal outcome, the persistence of sanctions threats becomes the status quo dynamic between senders and targets. Research on the consequences of sanctions demonstrates that what happens during sanctions can be more significant than how they end (Peterson and Drury 2011; Allen and Lektzian 2013). Changes in the domestic and international environments faced by leaders in both countries can influence the discrete choices they make over the course of a coercive episode. When targets acquiesce or senders back down or impose sanctions, they do so because they no longer find sanctions threats persisting as the most desirable outcome. We employ years as our temporal unit of analysis because data for our time-variant independent variables are only available at the yearly level. In order to model our competing risks analysis, we employ a multinomial logit estimator with cubic polynomials to provide a Taylor series approximation to the hazard curve. Along the line of Carter and Signorino's (2010) recommendations, we employ the number of years since a sanctions threat has been issued and the term's squared and cubed values (Time, Time2, and Time3) as our cubic polynomials. The multinomial logit analysis generates results for each possible binomial comparison between different outcomes.7 To test our theory, we are primarily concerned with assessing the impact of US Government's aid flows to target states on the comparative likelihoods of sanctions threats persisting, sanctions being imposed, or successfully ending with the target's capitulation in a given year. To account for clustering in the US-target dyads, we employ dyadic cluster-robust variance estimation as part of our analysis (Aronow, Samii, and Assenova 2015). Our results remain consistent when we employ alternative means of calculating our standard errors and alternative estimators (see supplementary files, tables A4–A6). Coding the Yearly Threat Outcomes Variable Our categorical dependent variable captures the yearly outcomes of sanctions threats for each of the 192 coercive episodes within our dataset. To code our four-category Threat Outcomes variable, we employ data on the years in which sanctions threats were ongoing and the termination of sanctions threats using information from the updated TIES dataset (Morgan et al. 2014). The baseline category of the Threat Outcomes variable is the sanctions threats persisting outcome. This is the most common yearly outcome, reflecting that threat episodes can persist in stalemates for years in some cases. Over 55 percent of the threat episodes are coded as spanning across more than one observation year. In order to identify sanctions threats that resulted in success, we followed the practices of the TIES dataset's creators in flagging successful cases as those that resulted in targets’ complete acquiescence, partial acquiescence, or a negotiated settlement (Bapat et al. 2013; Morgan et al. 2014). A threat episode was coded as ending with the United States backing down if it terminated with the sender's capitulation at the threat stage.8 The final way a threat episode could end according to our coding scheme was with the United States imposing sanctions. As Peterson (2013) has shown, the United States backing down from a sanctions threat has its own distinct set of consequences compared to imposing sanctions—highlighting why it is important to distinguish between the two alternative outcomes to success. Distinguishing between unsuccessful threat outcomes (i.e., the United States backing down vs. imposing sanctions) is central to testing the logic of our theory, as well as accurately reflecting the nature of compellent episodes (Schelling 1966). As compellent policies, sanctions episodes are considered successful if they achieve their desired outcomes, even if their senders have had to follow through on their threats. Threats that lead to sanctions being imposed do not indicate a failure of the coercive episode, since success is still possible if sanctions are imposed. Imposing sanctions versus backing down from their threats thus represent distinct alternatives for senders, with each being associated with vastly different outcomes. Confident and committed senders follow through with imposing sanctions because they think that the sanctions could work; less confident and less resolved senders will instead back down from their threats. For the period analyzed, our data indicate that 34 of the coercive episodes ended at the threat stage with the United States backing down, 101 resulted in the United States imposing sanctions, and 50 cases ended in success. This suggests that 25 percent (34/135) of our cases involving what is typically considered to be “threat failure” result in fundamentally different outcomes than the other 75 percent (101/135). Our analysis should reveal that there are significant theoretical gains to be made in distinguishing between those outcomes. Coding the Independent Variables In order to capture our dependent variable, we code US Foreign Aid as a measure of the ODA provided by the US Government to target states. ODA comprises foreign aid given by members of the OECD that matches three criteria: (1) it is donated by government agencies; (2) it is given to facilitate the “economic development and welfare of developing countries”; and (3) it is offered via concessional terms (OECD 2018). Notably, this definition excludes military aid. Consistent with our theory, we assume that the leaders of target and sender states view potential aid cuts in terms of actual monetary values rather than in proportional terms. For target states, the prospective loss of $10 million in foreign aid will create a real budgetary deficit that target leaders will have to address, irrespective of what proportion that amount of money comprises of their total economy. All else being equal, for example, both India and Rwanda would face substantially greater challenges in finding $100 million as opposed to $10 million in new aid to make up for a sanctions-induced shortfall. For leaders considering the challenges of replacing lost aid, we assume the actual amount of money being threatened matters most.9 Similarly, we think actual monetary value does the best job of operationalizing how sender states’ view aid cuts. Donor states will take into account how foreign-aid cuts will impact the domestic constituencies that benefit from them and how the aid money could be reallocated for other purposes.10 For theorizing cross-nationally about the coercive leverage that foreign-aid flows provide, we think this approach provides the best metric for capturing the degree to which target and sender leaders evaluate the prospective disruption and adverse social, political, and economic costs that aid withdrawals can inflict. We adopt Roodman's (2011) strategy of operationalizing ODA flows as net aid transfers. This measure subtracts the interest target states pay on concessional loans and the cancellation of one-off non-ODA loans that tend to be tied to aid packages given in the past years, cannot be predicted, and can skew aid flows. Operationalizing US Foreign Aid in terms of net aid transfers provides a transactional measure of the aid-flow relationship between the United States and a target state.11 The higher this value is, the greater the leverage the United States can expect to have over a given target state that it threatens with sanctions. We operationalize US Foreign Aid and our other economic variables in their real values in terms of constant 2012 dollars. The latter approach accounts for inflation and ensures that the value of the aid given by the United States is consistently measured throughout the analysis.12US Foreign Aid should have a positive effect on the likelihood of US sanctions threats succeeding and also make the United States more likely to impose sanctions if targets resist its threats. In addition to our main variable, we include a number of other variables that could influence the outcomes of coercive episodes at the threat stage. As previous studies have shown (e.g., Ang and Peksen 2007; Hufbauer et al. 2007; Bapat et al. 2013), the salience of the issue(s) at stake can have a strong impact on the likely outcome. We code High Salience Issue as a dummy variable that denotes whether the issues at stake involved efforts to contain the target's military behavior, destabilize the target's regime, address a territorial dispute, or prevent the target from acquiring dangerous military capabilities. All four issues are likely to infringe upon the target state's core political and security interests. Given the importance attributed to the participation of international organizations in coercive episodes (e.g., Drury 1998; Bapat and Morgan 2009; Bapat et al. 2013), we include a dummy variable IO Support to account for whether international organizations supported US sanctioning efforts. We include several variables to account for the characteristics of the target state. We first code a variable to denote whether the target of the US sanctions threats is a democracy. Several debates exist over whether democracies, like the United States, are less likely to sanction fellow democracies and about the ability of democracies to withstand economic coercion in comparison to authoritarian regimes (e.g., Cox and Drury 2006). We code our variable Target Democracy as a dummy variable that denotes that a state has a polity2 score of 6 or above and as a 0 otherwise using data from the Polity IV Project (Marshall and Jaggers 2009). We also control for the size of target states’ economies using a logged measure of their gross domestic products (Target Economic Size) via updated data from Gleditsch (2002). Lastly, we include controls for the relationship between the United States and the states that it is threatening with sanctions. To capture a target's susceptibility to US sanctions that would disrupt its trade flows, we include the variable Target Trade Dependence. Using current-year trade data from Barbieri and Keshk (2012), this variable codes the target's bilateral trade with the United States as a proportion of the target's total trade. We account for the political relationship between the United States and target states in several ways. First, we include the measure of political affinity using voting patterns in the United Nations General Assembly. We employ a measure that codes similarities in voting patterns coded along a continuum from –1 to 1.13 States with similar political affinities have positive scores, while dissimilar states have negative scores. As a second control, we include US Defense Pact as a dummy variable that denotes whether the United States and target state have a defense pact using Gibler (2009). According to Drezner (1999), coercive efforts should be more effective against allies since allies have fewer relative power concerns associated with giving into a sender's economic coercion. Finally, we account for the United States’ emergence as the world's lone superpower by including a dummy variable Post–Cold War in our analysis. The United States dramatically increased its use of sanctions after it became the world's lone superpower at the Cold War's conclusion (e.g., Hufbauer et al. 2007). Summary statistics and a correlation matrix are available in the supplementary files (see tables A1–A2). Discussion of the Results Table 1 presents the results yielded by our multinomial logit analysis using sanctions threats persisting as the baseline category of the yearly outcome variable. We use this as our baseline for operationally testing our hypotheses, as it is by far the most common outcome, and, conceptually, it is the yearly outcome value from which episodes transition to one of the three other terminal outcomes. Table 1 shows the results of the binomial comparisons for threats succeeding, sanctions being imposed, and the United States backing down, versus the baseline category. Model 1 includes only our theoretical variable of interest, while model 2 also includes the controls. For ease of interpretation, figure 1 displays the average marginal effects for US Foreign Aid on all four of the yearly sanctions outcome categories using model 2. It depicts the effects of US Foreign Aid when holding all other values of the variables at their observed values. Figure 1. View largeDownload slide Coefficient plot of the average marginal effects of US foreign aid on threat outcomes Figure 1. View largeDownload slide Coefficient plot of the average marginal effects of US foreign aid on threat outcomes Table 1. The effects of US foreign aid on the yearly outcomes of economic sanctions threats Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Notes: *, **, and *** denote statistical significance at the 90 percent, 95 percent, and 99 percent confidence levels, respectively, using two-tailed tests. View Large Table 1. The effects of US foreign aid on the yearly outcomes of economic sanctions threats Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Success vs. Persist Sanctions vs. Persist Back down vs. Persist Variables Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 US Foreign Aid 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) High Salience Issue −0.410 0.265 −1.225 (0.49) (0.47) (0.87) IO Support 1.144** 1.055* −0.104 (0.45) (0.57) (0.76) Target Economic Size 0.019 −0.144** 0.392*** (0.09) (0.07) (0.11) Target Democracy −0.425 −0.021 −1.495*** (0.40) (0.35) (0.47) Target Trade Dependence 1.721 0.528 0.418 (1.05) (0.91) (1.34) Political Affinity 0.216 −0.439 0.555 (0.66) (0.52) (0.86) US Defense Pact −0.532 0.098 1.143** (0.44) (0.40) (0.46) Post–Cold War 0.235 −0.612 1.438* (0.59) (0.38) (0.84) Time 1.063** 1.232** −2.695* −2.090 −2.482*** −2.055** (0.53) (0.55) (1.48) (1.57) (0.83) (0.93) Time2 −0.234** −0.241** 1.081* 0.883 0.458*** 0.371** (0.11) (0.12) (0.56) (0.59) (0.15) (0.17) Time3 0.010** 0.010** −0.131** −0.110* −0.022*** −0.018** (0.00) (0.00) (0.06) (0.07) (0.01) (0.01) Constant −2.783*** −3.374** 0.942 2.053 0.329 −5.301*** (0.62) (1.31) (1.07) (1.32) (0.85) (1.68) Observations 463 413 463 413 463 413 Pseudo-R2 0.077 0.128 0.077 0.128 0.077 0.128 Notes: *, **, and *** denote statistical significance at the 90 percent, 95 percent, and 99 percent confidence levels, respectively, using two-tailed tests. View Large With respect to our two hypotheses, the variable US Foreign Aid appears to both have a positive effect on the success of sanctions threats and also make senders more aggressive in implementing sanctions when target states resist sanction threats. In table 1, the results show that US Foreign Aid has a positive and statistically significant effect on the likelihood of sanctions threats ending in success versus persisting across all of our models. In general, and all else being equal, a one-standard-deviation increase in the constant value US Foreign Aid is associated with 1.68 greater odds of a sanctions threat ending in success as opposed to persisting. As figure 1 shows, the average marginal effects of US Foreign Aid on the likelihood of sanctions threats resulting in success are also positive and statistically significant at the 95 percent confidence level. In figure 2, we show the marginal effects of US Foreign Aid from its minimum value to $1 billion worth of net aid transfers while holding the other variables constant at their observed values. The figure illustrates a characteristic of our net aid variable (Roodman 2011) that allows for aid flows to be negative when aid recipients paid more to the United States in interest payments on aid-related loans than they received in new aid.14 This was true in roughly 9 percent of our observations.15 In 38 percent of the observations, the United States provided threat targets with a net value of 0 aid, and in 53 percent of cases, the aid values were positive. While the maximum value of US Foreign Aid was $6.7 billion in our sample, the vast majority (95 percent) of the observations involved less than $365 million in positive aid flows (see figures A7–A8). Substantively, the magnitudes of the marginal effects are small because the likelihood that threat episodes transition to a terminal success outcome in a given year is relatively low. The cumulative impact of the variable on the likelihood of success does mount over time. Figure 2. View largeDownload slide Marginal effects of US foreign aid on threat success at representative values Figure 2. View largeDownload slide Marginal effects of US foreign aid on threat success at representative values Figure 2 indicates that the average marginal effects of US Foreign Aid are positive and statistically significant at the 95 percent confidence level until the United States gives just over $400 million in net aid, at which point the variable's effect remains positive but loses statistical significance. This appears to be due to the scarcity of observations beyond this threshold (see figure A7). Examining the cases above this threshold, we identified the most significant aid outlier that resulted in a successful outcome. This case involves Israel, a leading US aid recipient. In 1992, the George H. W. Bush Administration threatened to terminate $10 billion worth of subsidized loan guarantees to force Israel to stop building new settlements in the occupied Palestinian territories. President Bush exploited Israel's reliance upon US aid to force the Israeli Government to freeze its construction of additional settlements (Friedman 1992; New York Times 1992). Our argument effectively explains the leverage wielded by the United States in this case, suggesting that this observation represents an on-the-line case that belongs in our sample. Excluding this extreme outlier case did not affect our results, but excluding all observations entailing aid flows greater than $400 million does cause the relationship to wash out (see table A19–20). Notably, none of the high-aid observations resulting in success were flagged as “least likely” outcomes in postestimation analyses.16 Overall, this suggests that high aid cases are an important part of our sample and shape our findings in ways consistent with our theory. With respect to hypothesis 2, US Foreign Aid has a positive and statistically significant effect on the likelihood of threats resulting in sanctions as opposed to persisting, according to both the results reported in table 1 and those in figure 1. Increasing US Foreign Aid by one-standard deviation increases the odds that the United States will impose sanctions during a coercive episode by 1.66 as opposed to letting the sanction threat persist. In figure 3, we show the marginal effects of US Foreign Aid on the sanctions imposition outcome. The results show that the marginal effects remain positive and statistically significant across the full range of the graph at the 95 percent confidence level—and indeed remain so until approximately $1.6 billion in positive aid flows. Once again, there are some higher-leverage cases at more extreme values that influence the results, but these are consistent with our argument and largely fall within the range in which US Foreign Aid’s effects are statistically significant. The US imposed sanctions against Israel (1982), Pakistan (1965; 1965), and India (1965; 1971) during periods in which its aid flows totaled over $1.6 billion dollars to those countries. The results are robust to the removal of the most extreme observation involving sanctions threatened against India related to its 1965 war with Pakistan. However, removing all the cases in which US Foreign Aid was greater than $1.6 billion causes the effects’ statistical significance to wash out (see tables A21–22). None of the high-aid observations appear as “least likely” cases in postestimation analyses, and the United States’ imposition of sanctions against those countries is consistent with our theory. Indeed, US policymakers publicly invoked potential aid cuts as part of their sanctions threats in each of the high aid observations—seeking to leverage the aid relationships in forcing concessions—before they followed through on their sanctions threats. While our findings do not hinge on a single case, the cases involving significant aid are both substantively important for our theory and our models’ results. Figure 3. View largeDownload slide Marginal effects of US foreign aid on sanctions imposition at representative values Figure 3. View largeDownload slide Marginal effects of US foreign aid on sanctions imposition at representative values The effects of US Foreign Aid on US policymakers’ likelihood to back down are more complicated to interpret. While table 1’s results indicate that US Foreign Aid has a positive and statistically significant effect at the 95 percent confidence level, the confidence intervals for the back down outcome overlap with zero in figure 1. Notably, the two threat episodes that ended in the United States backing down at the highest values of US Foreign Aid both involved Egypt, in 1960 and 1995. Throughout our sample period, the United States provided significant aid to Egypt both as part of Cold War politics and in order to foster peaceful relations with Israel. Excluding Egypt from our sample causes US Foreign Aid to lose its statistically significant impact on sanctions failure, but the results for the other outcomes remain unaffected (see table A7 and figures A1–A2). Our model lacks the nuance to capture how the quid pro quo bargains associated with US aid to Egypt limited its coercive leverage. Given the observed salience of the Egyptian cases, further research into the role foreign aid played in shaping US influence over the regime would be a valuable complement to our analyses. Not many of our other independent variables appear to affect threat outcomes. The variable IO Support has a positive and statistically significant effect on the likelihood of threats succeeding. Interestingly, we also found that having the support of an international organization increases the likelihood that the United States will impose sanctions. This suggests that organizationally supported sanctions threats could be more credible—and hence effective—because the United States is more likely to follow through on them. With respect to the United States backing down versus the threat persisting, the results show that Target Economic Size, US Defense Pact, and Post–Cold War have positive effects. This suggests that the United States is more likely to give up on sanctions threats that target democracies and/or allies and on threats during the post–Cold War era. The findings with respect to Target Economic Size indicate that the United States is less likely to give up on threats issued against states with large economies. Finally, the temporal controls (Time, Time2, and Time3) indicate that the longer a coercive episode lasts at the threat stage, the less likely the United States is to impose sanctions or back down—yet the more likely United States threats are to succeed. Robustness Checks While we theorize that senders exploit their aid relationships with recipients by issuing both private and public threats, we only have data on those threats that were publicly issued. By focusing on cases in which senders issued public threats to cut aid, we can ensure that our theoretical arguments successfully explain those cases in which we know donors specifically referenced aid cuts as part of their threats and had incentives to do so publicly. For example, the public threat of losing tens of millions of dollars’ worth of United States aid and trade was attributed to helping bring down Guatemala's President Jorge Serrano in 1993 after he sought to seize total power in the country. One of the first acts of his replacement, Ramiro de Leon Carpio, was to call upon the United States and other donors to rescind their sanctions threats (Golden 1993a; 1993b). Given that donors must expect to gain additional leverage from issuing their aid threats publicly, we would expect sanctions threats that explicitly invoke aid cuts to have even stronger effects. To evaluate whether the substantive effects of US Foreign Aid are stronger in the case of aid-based sanctions threats, we used the Sanctions Type Threatened variable from the TIES dataset (Morgan et al. 2014) to identify cases in which the United States solely threatened targets with the termination of foreign aid. We re-ran our original analysis using the sample of cases that only involved public aid-based sanctions threats (see table A8). The results strongly conform to our expectations. The coefficients for US Foreign Aid share the same sign and statistical significance as they did in the original analysis—but the size of the effects is larger. A one standard deviation increase in US Foreign Aid is associated with 2.34 greater odds of a sanctions threat ending in success in the aid-threat only cases. A one-standard-deviation increase in US Foreign Aid increased the odds of the United States imposing sanctions versus letting the threat persist by 2.05 in the public aid threat sample. These findings suggest that the coercive leverage provided by United States aid flows is stronger in the subset of cases where threats to aid are publicly communicated.17 We also considered the potential issue of simultaneity bias with respect to US Foreign Aid. To explore this, we ran our analysis using US Foreign Aid lagged by one year and using the value of US Foreign Aid for the year prior to the sanctions threat being issued (see tables A10–A11). The results were congruent with our main analyses. Additionally, we analyzed the implications of changing aid flows. If the positive relationship between US Foreign Aid and the success of sanctions threats is due to a bribery-based strategy rather than a coercive one, we would expect increasing aid flows to be associated with both a higher likelihood of success and a lower likelihood of imposing sanctions. We coded a variable to denote whether United States aid flows to target states increased, decreased, or remained static from the previous year. The results indicate that increasing aid flows are not associated with greater threat success but instead are associated with a higher likelihood of the United States imposing sanctions (see table A12). This runs counter to the expectations of an alternative, bribery-based account. Finally, we sought to evaluate whether our findings were driven by a subset of cases in which targets’ economies were highly dependent upon US foreign aid. We thus calculated the proportion that US foreign-aid flows comprised of each target state's economy and flagged those cases in which US aid amounted to more than 0.5 percent of the target's GDP. Rerunning our analysis without those cases (roughly 10 percent of the full sample), our results still hold (see tables A14). All these additional analyses support our theory that foreign-aid flows offer additional coercive leverage in making sanctions threats more effective and enhance senders’ aggressiveness in imposing sanctions. We also assessed whether the effects varied across the Cold War and post–Cold War eras (see table A15). Our findings show that while all other effects remain the same in the two periods, US Foreign Aid’s positive effects on the success of threats is only statistically significant in the post–Cold War period. This suggests that the coercive leverage the United States gained from its aid relationships was weaker when the recipients had a credible replacement option (i.e., the Soviet Union). Aid recipients thus appear more resistant to giving in to sanctions threats when they think a “black knight” or “aid-based sanctions buster” might be able to assist them if sanctions are imposed (Hufbauer et al. 2007; Early 2015). This sheds additional light on what influences how aid recipients respond to coercive attempts. Finally, there is a potential concern that US aid flows influence which states the United States threatens with sanctions—creating selection effect issues for our analysis. To evaluate this concern, we conducted a logit analysis on whether our US Foreign Aid variable positively influenced the likelihood of states being targeted with US sanctions (see table A23). The results indicate no such selection effects. While US aid relationships do influence the outcomes of sanctions threats, receiving greater amounts of US aid does not make recipients more likely to be targeted with coercion. Conclusion This study sought to explain how foreign-aid relationships influence the effectiveness of sanctions threats and the likelihood of sanctions being imposed. We theorized that providing foreign aid to states threatened with sanctions provides senders with greater leverage in extracting concessions. It also makes them more aggressive in imposing sanctions if sanctions threats do not elicit the desired behavior in the targeted state. Via an analysis of ongoing US sanctions threat episodes from 1960 to 2010, we found strong support for our hypotheses. Greater US foreign-aid flows to target states increase the likelihood that sanctions threats will result in success or in the United States following through with imposing sanctions when targets resist. Our findings also suggest that the positive effects of aid relationships on the success of US coercive threats are strongest in the absence of a superpower rival. Our study has several implications for research on economic coercion and foreign aid. Notably, our findings show that foreign aid constitutes an investment that provides donors with future coercive leverage in addition to more immediate quid pro quo concessions (e.g., Bueno de Mesquita and Smith 2007). This supports the argument that donors provide foreign aid more out of self-interest than altruism. Our findings also support some dependency theorists’ (e.g., Olson 1979) skepticism about the donor community's motivations in providing aid. Donors may thus provide aid strategically to recipient states that they share common interests with but also want to have leverage over in the future. For coercive episodes in which sanctions are imposed, our theory suggests that foreign-aid relationships should enhance sanctions’ likelihood of success. Our research design also presents a new approach for employing the TIES dataset (Morgan et al. 2014) that fully captures the multiple ways in which sanctions threat episodes can conclude. Building off our results, we think that additional qualitative analyses of the role aid relationships play in the use of economic coercion, especially for high aid cases the like US-Egypt relationship, is an important next step in this research agenda. While focusing on sanctions threats issued by the United States could limit our findings’ generalizability, our findings yield direct implications for US foreign policy. The number of donors that can both afford to give high levels of foreign aid and frequently employ economic coercion—like the United States—is limited. Our results suggest, though, that opportunistic donors may be able to exploit the coercive vulnerabilities associated with foreign-aid relationships on an occasional basis. For US foreign policy, our findings indicate that foreign aid buys the United States a greater amount of international influence than may have been previously thought. Indeed, our findings shed light on why US leaders may rely so heavily on economic coercion despite imposed sanctions’ poor track record of success. Even if leveraging aid relationships encourages US leaders to be highly aggressive in imposing sanctions, such aggression likely contributes to diplomatic victories before sanctions ever get imposed. The “hidden” success of US sanctions threats potentially makes up for the fact that the sanctions the United States imposes often are not that effective. Supplemental Information Supplemental information is available at the Foreign Policy Analysis data archive. Acknowledgements A previous draft was presented at the 2015 Annual Meeting of the International Studies in New Orleans, LA. We would like to thank Glenn Palmer and the comments of the anonymous reviewers for their insights and comments that contributed to this manuscript. Notes Bryan R. Early is an associate professor of political science at the University at Albany, SUNY. He is also the director of the Center for Policy Research (CPR) and the founding director of the Project on International Security, Commerce, and Economic Statecraft (PISCES). Dr. Early is an expert on economic statecraft, weapons of mass destruction, and nonproliferation issues. Amira Jadoon is an assistant professor at the Combating Terrorism Center (CTC) and the Department of Social Sciences at the US Military Academy at West Point, as well as the CTC's General John P. Abizaid Research Associate. Dr. Jadoon specializes in international security, economic statecraft, political violence, and terrorism. Footnotes 1 Governments can adopt legislation that threatens the triggered imposition of sanctions as a form of deterrent threat (see Miller 2014). Once imposed, however, sanctions are compellent in nature. 2 Sanctions packages can also entail a much broader mixture of commercial restrictions that go beyond just aid (Morgan, Bapat, and Kobayashi 2014). 3 Target leaders may not want it publicly known that they will acquiesce to threats involving aid cuts at a specific threshold (i.e., $50 million). Privately issued aid threats could be preferable in some cases, as they could make it easier for threatened leaders to acquiesce without giving away information about their vulnerability to coercion to other donors. 4 Some sender constituents can benefit from the aid packages given to target states. For example, American farmers benefit from the US food aid programs that require the purchase of US foodstuffs. Milner and Tingley (2010) show that US Congressional support for foreign aid is influenced by politicians’ domestic economic concerns; however, voting on aid flows concerning specific geostrategic issues is not particularly affected by district endowment factors and only bears a weak relationship with political economy interests. Additionally, Fleck and Kilby (2001) find no evidence that domestic economic benefits influence Congress members’ stances on foreign aid. In contrast, Hufbauer et al.’s (1997) past research has shown that trade-related sanctions result in tens of billions of dollars of lost business revenues—which translates into hundreds of thousands of lost US jobs. 5 Our dataset includes cases initiated within the TIES dataset prior to 1960 and persisted into our sample period. The TIES dataset only includes episodes initiated up to 2005 but codes outcomes through 2010. 6 We excluded coercive episodes that only involved trade or environmental issues. Trade sanctions, such as those initiated under Section 301 of the US Trade Act of 1974, have very different procedures for their imposition than politically driven sanctions. Environmental sanctions tend to involve technical more than political issues. 7 We meet the theoretical requirements of the independence of irrelevant alternatives (IIA) assumption, as each outcome is distinct from one another and there are no viable alternative outcomes. 8 We use data from the TIES dataset's “Final Outcome” variable (Morgan et al. 2014). 9 Using a measure that takes US-target aid flows as a proportion of countries’ GDPs instead of real monetary values, for example, would divorce the measure from the real budgetary impact that aid withdrawals have. The size of a country's economy does not always translate into understanding the salience of aid money to a country's leaders, especially given the range of ways they may leverage aid money to stay in power. 10 A proportional measure of the target's economic dependence on US aid would not be capable of capturing any of this information. 11 By accounting for interest payments, the net aid transfer variable can identify targets that are paying more to the US Government in interest on outstanding ODA-related loans than they are receiving in new ODA in a given year. 12 Our results also hold if we employ current-year dollar values instead (see table A3). 13 We employ Gartzke's (2010) data that interpolates for missing values from 1960 to 2008 and updated data from Voeten et al. (2013). 14 This means that the United States was receiving back more in interest payments than it was giving to the targets in new aid. 15 We obtain congruent results if we exclude the cases of negative US Foreign Aid flows (see table A13). 16 We employed Long and Freese's (2014)’s S-Post suite. 17 We also ran models that interacted public aid threats and US Foreign Aid. We found that public aid threats have positive and statistically significant effects on both threat success and sanctions imposition across a large range of US Foreign Aid; however, the confidence intervals for private aid threats overlap with the public threats for the threat success outcome. This suggests that the strongest effects of making aid threats publicly is on senders’ willingness to follow through with imposing sanctions, as they have more of their credibility at stake in those cases. Additionally, we explored interacting US Foreign Aid with High Salience Issue. We found that higher values of US Foreign Aid lead to greater threat success when sanctions are imposed for nonsalient reasons and a higher likelihood of sanctions being imposed in cases involving high-salience sanctions issues (see table A16 and figures A5–6). References Agence-France Press . 2011 . “U.S. Industry Moves to Kill China Sanctions Bill,” September 21, 2011 . Alesina Alberto , Dollar David . 2000 . “Who Gives Foreign Aid to Whom and Why?” Journal of Economic Growth 5 ( 1 ): 33 – 63 . Google Scholar Crossref Search ADS Allen Susan , Lektzian David . 2013 . “Economic Sanctions: A Blunt Instrument?” Journal of Peace Research 50 : 121 – 35 . Google Scholar Crossref Search ADS Ang Adrian , Peksen Dursun . 2007 . “When Do Economic Sanctions Work? Asymmetric Perceptions, Issue Salience, and Outcomes.” Political Research Quarterly 60 ( 1 ): 135 – 45 . Google Scholar Crossref Search ADS Aronow Peter , Samii Cyrus , Assenova Valentina . 2015 . “Cluster-Robust Variance Estimation for Dyadic Data.” Political Analysis 23 ( 4 ): 564 – 77 . Google Scholar Crossref Search ADS Baldwin David . 1985 . Economic Statecraft . Princeton, NJ : Princeton University Press . Bapat Navin , Morgan T. Clifton . 2009 . “Multilateral versus Unilateral Sanctions Reconsidered: A Test Using New Data.” International Studies Quarterly 53 ( 4 ): 1468 – 2478. Google Scholar Crossref Search ADS Bapat Navin , Heinrich Tobias , Kobayashi Yoshiharu , Morgan T. Clifton . 2013 . “Determinants of Sanctions Effectiveness: Sensitivity Analysis Using New Data.” International Interactions 39 ( 1 ): 79 – 98 . Google Scholar Crossref Search ADS Barbieri Katherine , Keshk Omar . 2012 . “Correlates of War Project Trade Data Set Codebook, Version 3.0.” http://correlatesofwar.org . Barry Colin , Kleinberg Katja . 2015 . “Profiting from Sanctions: Economic Coercion and US Foreign Direct Investment.” International Organization 69 ( 4 ): 881 – 912 . Google Scholar Crossref Search ADS Box-Steffensmeier Janet M. , Jones Bradford S. . Timing and Political Change: Event History Modeling in Political Science . Ann Arbor : University of Michigan Press , 2004 . Bueno de Mesquita Bruce , Smith Alastair , Siverson Randolph M. . 2004 . The Logic of Political Survival . Cambridge, MA : MIT Press . Bueno de Mesquita Bruce , Smith Alastair . 2007 . “Foreign Aid and Policy Concessions.” Journal of Conflict Resolution 51 : 251 – 84 . Google Scholar Crossref Search ADS Carter David B. , Signorino Curtis S. . 2010 . “Back to the Future: Modeling Time Dependence in Binary Data.” Political Analysis 18 ( 3 ): 271 – 92 . Google Scholar Crossref Search ADS Cox Dan , Drury A. Cooper . 2006 . “Democratic Sanctions: Connecting the Democratic Peace and Economic Sanctions.” Journal of Peace Research , 43 ( 6 ): 709 – 22. Google Scholar Crossref Search ADS Drezner Daniel W. 2003 . “The Hidden Hand of Economic Coercion.” International Organization 57 ( 3 ): 643 – 59 . Google Scholar Crossref Search ADS Drezner Daniel W. . 1999 . The Sanctions Paradox: Economic statecraft and International Relations . Cambridge University Press , 1999 . Drury A. Cooper . 1998 . “Revisiting Economic Sanctions Reconsidered.” Journal of Peace Research , 35 ( 4 ): 497 – 509 . Google Scholar Crossref Search ADS Drury Cooper , Olson Richard , Belle Douglas Van . 2005 . “The Politics of Humanitarian Assistance: U.S. Foreign Disaster Aid, 1964–1995.” Journal of Politics 67 : 454 – 73 . Google Scholar Crossref Search ADS Early Bryan . 2015 . Busted Sanctions: Explaining Why U.S. Economic Sanctions Fail . Stanford, CA: Stanford University Press . Emmanuel Nikolas G. 2013 . “Democratization in Malawi: Responding to International and Domestic Pressures.” African and Asian Studies 12 ( 4 ): 415 – 34 . Google Scholar Crossref Search ADS Fleck Robert , Kilby Christopher . 2001 . “Foreign Aid and Domestic Politics: Voting in Congress and the Allocation of USAID Contracts across Congressional Districts.” Southern Economic Journal , 598 – 617 . Friedman Thomas . 1992 . “Bush Tells Congress ‘No’ on Aid to Israel.” New York Times, May 18, 1992 . Gartzke Erik . 2010 . “The Affinity of Nations: Similarity of State Voting Positions in the UNGA.” http://pages.ucsd.edu/∼egartzke/htmlpages/data.html . Gibler Douglas . 2009 . International Military Alliances , 1648 – 2008 . Washington, DC : CQ Press . Gleditsch Kristian . 2002 . “Expanded Trade and GDP Data.” Journal of Conflict Resolution , 46 ( 5 ): 712 – 24 . Google Scholar Crossref Search ADS Golden Tim . 1993a . “Guatemala's Counter-coup: A Military About-Face.” New York Times, June 3, 1993 . Golden Tim . 1993b . “In Startling Shift, Guatemala Makes Rights Aide President.” New York Times, June 7, 1993 . Guardian . 2009 . “US Cuts All Non-humanitarian Aid to Honduras in Support of Zelaya.” September 2, 2009 . Hirschman Albert . 1945 . National Power and the Structure of Foreign Trade . Berkley : University of California Press . Hufbauer Gary , Elliot Kimberly , Cyrus Tess , Winston Elizabeth . 1997 . “U.S. Economic Sanctions: Their Impact on Trade, Jobs, and Wages.” Working Paper Special . Washington, DC : Peterson Institute for International Economics . Hufbauer Gary , Schott Jeffrey , Elliott Kimberly , Oegg Barbara . 2007 . Economic Sanctions Reconsidered . Washington, DC : Peterson Institute for International Economics . Jackson David . 2014 . “Business Groups Oppose Any New Sanctions on Russia.” USA Today, June 25, 2014 . Kharas Homi . 2008 . “Measuring the Costs of Aid Volatility.” Wolfensohn Center for Development Working Paper 3 . Washington, DC : The Brookings Institute . Kim Dong-Hun . 2013 . “Coercive Assets? Foreign Direct Investment and the Use of Economic Sanctions.” International Interactions 39 ( 1 ): 99 – 117 . Google Scholar Crossref Search ADS Kim Youngwan . 2016 . “How NGOs Influence US Foreign Aid Allocations.” Foreign Policy Analysis 13, no. 1 (2017): 112–32 . Kosack Stephen , Tobin Jennifer . 2006 . “Funding Self-Sustaining Development: The Role of Aid, FDI, and Government in Economic Success.” International Organization 60 : 205 – 43 . Google Scholar Crossref Search ADS Lektzian David , Biglaiser Glen . 2014 . “The Effect of Foreign Direct Investment on the Use and Success of US Sanctions.” Conflict Management and Peace Science 31 ( 1 ): 70 – 93 . Google Scholar Crossref Search ADS Lesniewski Niels . 2012 . “Lindsey Graham, Rand Paul Rivalry on Foreign Aid Grows.” Roll Call, October 10, 2012. http://www.rollcall.com/news/-218130-1.html . Licht Amanda . 2010 . “Coming into Money: The Impact of Foreign Aid on Leader Tenure.” Journal of Conflict Resolution , 54 ( 1 ): 58 – 87 . Google Scholar Crossref Search ADS Long Scott , Freese Jeremy . 2014 . Regression Models for Categorical Dependent Variables Using Stata , 3rd ed . College Station, TX : Stata Press . Maizels Alfred , Nissanke Machiko . 1984 . “Motivations for Aid to Developing Countries.” World Development 12 ( 9 ): 879 – 900 . Google Scholar Crossref Search ADS Marshall Monty , Jaggers Keith . 2009 . Polity IV Project: Political Regime Characteristics and Transitions, 1800–2007 . Arlington, VA : Polity IV Project . Miller Nicholas . 2014 . “The Secret Success of Nonproliferation Sanctions” International Organization 68 ( 4 ): 913 – 44 . Google Scholar Crossref Search ADS Milner Helen , Tingley Dustin . 2010 . “The Political Economy of US Foreign Aid: American Legislators and the Domestic Politics of Aid.” Economics & Politics 22 ( 2 ): 200 – 32 . Milner Helen V. , Tingley Dustin H. . 2013 . “Public Opinion and Foreign Aid: A Review Essay.” International Interactions 39 ( 3 ): 389 – 401. Google Scholar Crossref Search ADS Morgan T. Clifton , Bapat Navin , Kobayashi Yoshiharu . 2014 . “The Threat and Imposition of Economic Sanctions 1945–2005: Updating the TIES Dataset.” Conflict Management and Peace Science 31 ( 5 ): 541 – 58 . Google Scholar Crossref Search ADS Morgan T. Clifton . 2015 . “Hearing the Noise: Economic Sanctions Theory and Anomalous Evidence.” International Interactions 41 ( 4 ): 744 – 54 . Google Scholar Crossref Search ADS Morrow James . 1999 . “How Could Trade Affect Conflict?” Journal of Peace Research 36 ( 4 ): 481 – 9 . Google Scholar Crossref Search ADS Nielsen Richard . 2013 . “Rewarding Human Rights? Selective Aid Sanctions against Repressive States.” International Studies Quarterly 57 : 791 – 803 . Google Scholar Crossref Search ADS Neuenkirch Matthias , Neumeier Florian . 2015 . “The Impact of UN and US Economic Sanctions on GDP Growth.” European Journal of Political Economy 40 : 110 – 25 . Google Scholar Crossref Search ADS Newnham Randall . 2008 . “‘Coalition of the Bribed and Bullied?’ U.S. Economic Linkage and the Iraq War Coalition.” International Studies Perspectives 9 ( 2 ): 183 – 200 . Google Scholar Crossref Search ADS New York Times, 1992 . “Pressing for Principle in Israel.” February 26, 1992 . Olson Richard . 1979 . “Economic Coercion in World Politics: With a Focus on North-South Relations.” World Politics 31 ( 4 ): 471 – 94 . Google Scholar Crossref Search ADS Peksen Dursun , Peterson Timothy . 2016 . “Sanctions and Alternate Markets: How Trade and Alliances Affect the Onset of Economic Coercion.” Political Research Quarterly 69 ( 1 ): 4 – 16 . Google Scholar Crossref Search ADS Peterson Timothy . 2013 . “Sending a Message: The Reputation Effect of US Sanction Threat Behavior.” International Studies Quarterly 57 : 672 – 82 . Google Scholar Crossref Search ADS Peterson Timothy , Drury Cooper . 2011 . “Sanctioning Violence: The Effect of Third-Party Economic Coercion on Military Conflict.” Journal of Conflict Resolution 55 ( 4 ): 580 – 605 . Google Scholar Crossref Search ADS Robertsen Jamie , Francken Nathalie , Molenaers Nadia . 2015 . “Determinants of the Flow of Bilateral Adaptation-related Climate Change Financing to Sub-Saharan Countries: The Extent to which Environmental, Economic, Historical and Political Factors Play a Role.” LICOS Discussion Paper Series 373 . Roodman David . 2011 . “An Index of Donor Performance.” Center for Global Development. Working Paper #67 . Schelling Thomas . 1966 . Arms and Influence . New Haven, CT : Yale . Tarnoff Curt , Lawson Marian . 2011 . Foreign Aid: An Introduction to US Programs and Policy . Washington, DC: Congressional Research Service, 2011 . The Organisation for Economic Cooperation and Development (OECD). 2018 . “The Definition of Official Development Assistance.” http://www.oecd.org/dac/stats/officialdevelopmentassistancedefinitionandcoverage.htm . Van Bergeijk Peter , van Marrewijk Charles . 1995 . “Why Do Sanctions Need Time to Work? Adjustment, Learning, and Anticipation.” Economic Modeling 12 ( 2 ): 75 – 86 . Google Scholar Crossref Search ADS Voeten Erik , Strezhnev Anton , Bailey Michael . 2013 . “United Nations General Assembly Voting Data.” https://thedata.harvard.edu/dvn/dv/Voeten/faces/study/StudyPage.xhtml?globalId=hdl:1902.1/12379&studyListingIndex=1_4cbddd2916efea2ae2729fc20fda . Whang Taehee , McLean Elena , Kuberski Douglas . 2013 . “Coercion, Information, and the Success of Sanction Threats.” American Journal of Political Science 57 ( 1 ): 65 – 81 . © The Author(s) (2019). Published by Oxford University Press on behalf of the International Studies Association. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Journal

Foreign Policy Analysis – Oxford University Press

Published: Jul 1, 2019

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Using the Carrot as the Stick: US Foreign Aid and the Effectiveness of Sanctions Threats

Using the Carrot as the Stick: US Foreign Aid and the Effectiveness of Sanctions Threats

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Using the Carrot as the Stick: US Foreign Aid and the Effectiveness of Sanctions Threats

Using the Carrot as the Stick: US Foreign Aid and the Effectiveness of Sanctions Threats

References (76)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies