Designing foreign policy

Abstract

The literature on economic sanctions has long studied sender countries’ policymaking as a simple choice between imposing sanctions to extract concessions from the targeted country and doing nothing. We depart from this simplifying assumption and analyze sanctions as a multifaceted foreign policy instrument. We argue that senders design sanction policies in response to policy preferences of two domestic constituencies. Voters expect a response to an international dispute in the form of some policy, such as economic sanctions; hence, the sender’s policymakers seek to demonstrate their competence in foreign affairs by imposing sanctions. Once the policymakers announce the use of sanctions, special interest groups that stand to experience economic losses when this foreign policy is implemented pressure the policymakers to choose sanction measures limiting such losses. As a result, the policymakers design sanction policies to include measures that will be less detrimental to special interest groups. We test our theoretical argument using the Threat and Imposition of Sanctions data and show that, while pressures from public opinion increase the likelihood of sanctions, special interest groups that benefit from the relationship with the target country are associated with a lower probability of the use of sanction measures that would impose substantial costs on domestic interest groups.

Keywords

economic sanctions foreign policy human rights public opinion special interest groups targeted sanctions

The nearly universal consensus in the sanctions literature is that sanctions are not effective. However, this ineffectiveness is not uniform – while most sanctions do not appear to have any significant effect on targeted countries’ behavior, a few cases demonstrate that this foreign policy can in fact succeed in extracting policy concessions. The most recent illustration of the ability of sanctions to coerce is the agreement that was reached by the permanent members of the UN Security Council and Germany, on the one hand, and Iran, on the other. For the first time in almost a decade Iran has agreed to stop some of its uranium enrichment activities. While many policymakers remain skeptical of Iran’s willingness to abide by the terms of the agreement in the long run, few deny that it was the impact of economic sanctions that made the agreement possible in the first place. What is the difference between successful and failed applications of the same coercive policy?

We argue that sanctions that succeed are deliberately designed by sanctioning governments to be sufficiently costly for the target. Sanction costs determine the coercive potential of sanctions: the target has no incentives to offer concessions to the sender and modify or abandon its policies unless the sender’s punitive response changes the target’s preferences in favor of such concessions. Thus, in Iran’s case, sanctions included a broad range of economically damaging measures implemented by the USA and the EU, such as import and export restrictions, financial sanctions, and travel bans. Therefore, we consider the design of sanctions to be the key to their outcomes.

To explain determinants of sanction design, we examine sanctions as a sender government’s response to another country’s controversial actions. When governments choose their response, they are usually constrained by domestic actors’ preferences and domestic political institutions. In particular, democratic regimes empower voters and make them a relevant domestic constituency in foreign policy decisionmaking. Another important domestic actor that has incentives to influence the government’s choice of coercive measures is special interest groups, such as exporters, because of distributional consequences generated by sanctions.

After a brief overview of the relevant research on sender governments’ use of sanctions and domestic political factors that affect the governments’ decisions, we develop our theoretical argument that analyzes sanctions as a complex foreign policy tool that is rationally designed by sender governments constrained by domestic politics. We then link domestic actors’ preferences to the choice of specific sanction measures by sender countries. Voter expectations of action lead to the policymakers’ decision to turn to sanctions when the sender country is confronted with a target country’s controversial behavior. When the policymakers face the pressure from special interest groups that would suffer significant losses as a result of interrupted economic exchanges with the target, the policymakers design sanctions to address the groups’ concerns. In this case, the policymakers tailor the sanction package to include measures that aim to minimize special groups’ economic losses and, hence, policymakers’ political costs. Our analyses yield robust evidence of the empirical relationships predicted by our theoretical framework: when voters are informed and hence relevant from the standpoint of foreign policymaking, their preferences increase the probability of sanction imposition, while special interest groups’ preferences have a direct influence on the choice of specific sanction measures, such as targeted sanctions, aid cuts, and export restrictions, as well as the use of high-impact sanctions. We conclude by discussing the implications of our findings for studies of sanction effectiveness and suggesting directions for future research.

Domestic politics and sanction policies

Economic sanctions are a foreign policy tool that is used to extract concessions from targeted countries. A naïve model of sanction policy assumes that this policy is designed to maximize the probability of success. However, this model fails to explain why sanctions are ineffective in many cases: the consensus in the existing research is that sanctions are at best only moderately successful (Hufbauer et al., 2008; Morgan & Schwebach, 1997; Pape, 1997; Drury, 1998). Subsequent scholarship has sought to explain the continued use of this foreign policy despite its low success rate. One prominent theoretical approach has emphasized sender governments’ interest in signaling their resolve to target states prior to sanction imposition and hence the apparent ineffectiveness of sanctions has been attributed to selection effects in sender–target interactions at the threat stage (Drezner, 2003; Eaton & Engers, 1992; Lacy & Niou, 2004; Morgan & Miers, 1999; Smith, 1995). While this explanation is plausible, a recent study finds no evidence of signaling effects of sanction threats (Whang, McLean & Kuberski, 2013).

A different body of the sanction research has departed from the unitary state assumption and begun examining the role of domestic politics in sanction use. This development shares the focus on domestic actors’ policy preferences with other areas of IR research; for example, studies of trade policies (Grossman & Helpman, 2001; Hiscox, 2002) and exchange rates (Frieden, 1991; Bearce, 2010) incorporate policy preferences of relevant domestic actors into models of governments’ policy choices. These studies, in particular, identify two important types of domestic actors in democratic countries – voters and special interest groups – that affect their governments’ policymaking.

While scholars debate how much public opinion influences politicians’ decisionmaking in foreign affairs of democratic countries (Ginsberg, 1986; Brace & Hinckley, 1993; Stimson, 1991; Powlick & Katz, 1998), voters’ evaluation of their policymakers’ performance in foreign affairs is part of the overall job performance assessment and, hence, is likely to shape policymakers’ behavior.¹ When voters evaluate the performance of elected officials, voters are primarily guided by their concerns about the collective national interest (Kiewiet & Kinder, 1981; Mutz, 1992; Mansfield & Mutz, 2009).² Existing IR research indicates that democratic leaders are concerned about voters’ evaluation of leaders’ performance when it comes to the use of military force (MacKuen, 1983; Ostrom & Simon, 1985; Ostrom & Job, 1986; Fordham, 1998). Similarly, Whang (2011) shows that the use of economic sanctions is driven by domestic political considerations – sanctions boost presidential approval ratings, even though sanctions may or may not succeed in achieving their stated goals. These findings have been qualified by scholars who have identified voter awareness as a critical intervening factor: we should only expect voters to constrain foreign policymaking when voters know about contested issues and international disputes, that is, when voters are informed (Grossman & Helpman, 1996) and are likely to expect their policymakers to adopt appropriate foreign policies. In fact, studies of trade issues show that, when voter awareness is high, voter opinion plays an important role in influencing policy positions of public officials in democracies (Fordham & McKeown, 2003; Kono, 2008). Sobel (2001) suggests that voter awareness is high when it comes to domestic economy or politics; on the other hand, voters largely overlook foreign affairs unless their country is directly threatened or actively involved in an international dispute.

In the case of special interest groups, there is a strong incentive to pay keen attention to foreign policies that can have negative distributional consequences for these groups. Kaempfer & Lowenberg (1988, 1992) note that sanctions are government interventions that generate economic distortions which increase the income of some domestic groups at the expense of others. Previous research has found that policymaking in fact reflects domestic groups’ concerns about distributional consequences of foreign policies. For instance, Kastner (2007: 670) suggests that country leaders’ utility from implementing policies that affect international trade ‘depends in part on the trade policy preferences of their core constituents’; therefore, the leaders will avoid ‘taking actions detrimental to foreign commerce’ if their constituents oppose these actions.

Designing sanction policy

How do these domestic policy preferences, structured by domestic political institutions, affect the government’s decision to use sanctions and the design of the sanction policy? When the government considers its response to another country’s pursuit of a controversial policy, the decisionmaking process can be disaggregated into two stages. First, the sender government chooses to respond or take no action. Second, if the government makes a decision to act, the policy response can then be designed to include a set of specific measures. Therefore, we study sanction imposition using this two-stage approach: the choice to use sanctions is followed by sanction design, that is, selection of specific instruments that will apply coercive pressure on the target.³

To develop a domestic-level explanation of sanction policies, we start with the proposition that sanctions are a policy that politicians develop under the pressure of societal demands. We also assume that leaders are driven by the desire to remain in office, which incentivizes them to remain attuned to domestic actors’ policy preferences. Domestic actors often have divergent preferences in terms of their countries’ policies, especially when such policies generate tangible costs or benefits that are distributed unequally among different actors. Specifically, sanction policies divide domestic actors along the pro-sanction/anti-sanction line.⁴ Those who stand to benefit from the policy support it, whereas those who will bear the costs after the policy is implemented oppose its adoption.

The lack of response becomes costly for democratic leaders when voters expect and demand action, because in addition to the costs imposed by the target’s policy, the sender’s leadership will be held accountable for doing nothing. We adopt the view that in democracies, leaders cannot ignore public opinion in foreign affairs if they want to keep domestic support for actions overseas. In addition, elections serve to enhance leader accountability, thereby creating incentives for the leaders to demonstrate in their policymaking that they are responsive to voters’ demands. Since voters expect their country’s leaders to take some steps to address an international dispute, rather than do nothing, we consider the voters’ dispute awareness as a determinant of the importance of voter preferences. When voters are not aware of the disputed issue, policymakers do not need to take voter preferences into account and may be less likely to adopt a sanction policy in the absence of voters’ demand for it. In fact, frequently, bureaucratic branches (e.g. the Office of Foreign Assets Control of the US government) decide on the imposition and extension of sanctions without any public attention. When, on the other hand, an international dispute has generated significant public awareness, and consequently voters expect a policy response, the policymakers take action by announcing the use of economic sanctions. Yet, when it comes to the specifics of announced policies, the public ‘leaves it up to the political leadership … to decide on the appropriate course of action’ (Knecht, 2011: 28). In sum, voters want to see a policy response to an international dispute, but do not have strong or consolidated preferences regarding the nuances of their government’s response.

Once the sender’s policymakers have decided to adopt the sanction policy, and voters have been appeased, the sender government needs to respond to another set of demands – demands from special interest groups that pressure the policymakers to avoid restricting economic exchanges with the target country. There is a variety of measures that senders can consider, including limits on financial relations and/or aid flows, suspensions of economic agreements, measures targeting specific individuals or groups in the target country, multilateral measures, and even offers of positive inducements. Corresponding target costs range from low to high. At the same time, these measures generate costs for the sender economy. While extant research points out that such costs are generally low,⁵ sanction costs are not distributed equally across the entire economy and all domestic actors. Often, some actors bear a larger share of costs, while others may in fact receive benefits. Therefore, domestic groups that expect to be negatively affected by distributional consequences of sanctions have powerful incentives to pressure their governments to minimize such consequences. Special interest groups have an advantage in organizing to convey their preferences, as well as their intensity, to policymakers because such groups can overcome collective action problems relatively easily due to their smaller numbers. Economic interest groups also tend to have resources to contribute to political campaigns and lobby the government for the adoption of favored policies. If politicians choose a policy that is contrary to these groups’ preferences, their financial contributions will decrease or stop. This possibility creates incentives for the policymakers to remain responsive to special interest groups.

We rely on the sectoral model of trade policy preferences to examine how specific sanction instruments affect domestic interest groups.⁶ The model predicts what types of economic cleavages will arise domestically due to changing levels of international trade, and thus informs our expectations regarding the sources of interest group opposition that policymakers may experience as they consider imposing economic sanctions against the target country.⁷ The sectoral model of trade policy preferences assumes that factors of production are sector specific, that is, returns to factors of production depend on returns generated by the economic sector in which the factors are employed. Generally, when trade levels increase, factors specific to export-oriented sectors benefit in terms of real returns; factors employed in import-competing industries, on the other hand, receive lower returns. The pattern of domestic losses and gains reverses when trade decreases, as is the case when the sender imposes trade restrictions.

We generalize the insights of the sectoral model to include any coercive instruments utilized by the sender that could result in restricted economic exchanges with the target. In particular, when export-oriented sectors depend on the target’s market, they are likely to suffer losses after sanction initiation and have an interest in opposing any types of sanctions that will directly or indirectly reduce exports to the target. Similarly, we expect industries that rely on intermediate inputs purchased from the target’s companies to oppose economic sanctions, especially if the sender government seeks to restrict imports. If, on the other hand, the trade relationship between the sender and target is primarily based on imports of goods from the target, the sender’s import-competing industries are positioned to gain from sanctions, especially when they take the form of import restrictions. The import-competing sector is, therefore, not expected to object to the government action. In fact, sanctions should be favored by the import-competing sector because restrictions on trade with the target are essentially a protectionist policy. In sum, the set of specific measures chosen and implemented by the sender against the target will include restrictions that are favored by special interest groups, and exclude restrictions that would be met with opposition.

Empirical implications

Our theoretical argument suggests that sender countries’ policymakers are most sensitive to the pressure of public opinion when they consider whether they should follow up on the threat to impose sanctions. When voter awareness is high, we expect policymakers to be motivated to show their voters that the country’s leadership is prepared to do something about the target’s objectionable behavior. Therefore, sanctions should be more likely when voters pay attention to the dispute with the target country.

Voter Awareness Hypothesis: When voter awareness of an international dispute is high, the likelihood of sanctions increases.

The effect of interest groups’ preferences on sender countries’ policymaking manifests itself when we examine specific sanction instruments chosen by senders. Our hypotheses focus on one set of interest groups – export-oriented industries – because the size and strength of this sector can be easily captured empirically by gauging export flows. Import-competing groups, on the other hand, present an empirical challenge: import flows do not accurately represent the size of this sector because some imports do not generate competition in the sender country. Such imports serve as intermediate inputs; alternatively, sender companies may not produce similar or substitute goods. Given that we do not have a valid measure of the size of import-competing sectors, we omit the discussion of our theoretical expectations for these groups.

One sanction instrument that can be chosen in response to export-oriented sectors’ preferences is targeted sanctions, that is, actions taken by the sender to punish select individuals as opposed to damaging the entire economy of the target country.⁸ We argue that policymakers have an incentive to use targeted sanctions not merely for humanitarian reasons, that is, to avoid collateral damage that will be imposed on vulnerable populations, but due to the support for this instrument from an important domestic constituency.⁹ Export-oriented groups oppose measures that damage their economic relations with the target country, and these groups’ opposition should make policymakers attempt to limit the scale of this damage. When targeted sanctions, rather than all-out economic warfare, are used as a sanction instrument, the interests of the sender’s economic groups are less likely to suffer because the target country’s economy will not experience significant disruptions and most economic activities can continue as usual.

Targeted Sanctions Hypothesis: As the sender country’s export-oriented sector trading with the target country increases, the likelihood of targeted sanctions increases as well.

We can also identify instruments that senders will avoid in formulating their sanction policies when special interest groups oppose sanctions. For instance, we expect export and aid sanctions to be less likely when such opposition is present. Since export-oriented groups suffer from restrictions on economic exchanges with the target, we expect that their opposition to sanctions should make policymakers reluctant to hinder exports to the target. Therefore, as the size and, consequently, political influence of interest groups that export goods to the target grows, the use of export controls becomes less likely.

Export Sanctions Hypothesis: As the sender country’s export-oriented sector trading with the target country increases, the likelihood of export sanctions declines.

Similarly, the use of aid cuts is expected to meet opposition from export-oriented industries in the sender country. Without such opposition, termination of foreign aid is an attractive sanction instrument because this measure is cheap and easy to implement. Many developing countries rely on aid as an important source of their budget revenues: 2009 World Development Indicators data show that an average aid recipient derives 24% of its central government expense, or 7.4% of its gross national income, from official development assistance (ODA).

Despite the attractiveness of aid cuts as a sanction measure, aid reductions can hurt the interests of the sender country’s exporters. A significant share of bilateral foreign assistance comes in the form of tied aid, that is, the donor requires the recipient to spend this aid on purchases of goods and/or services from the donor country. A 2009 OECD report states that, even though the share of tied aid has declined between 1999 and 2007, fully tied ODA still makes up 24%, and other types of financial assistance, such as technical cooperation, tend to have a greater proportion of tied aid than ODA. The USA, the sender country in a large number of sanctions episodes in the TIES dataset, reports that 57% of its bilateral aid is fully tied (Clay, Geddes & Natali, 2009). If the sender terminates aid flows, its exporters stand to lose the opportunity to sell their goods and/or services to the target without any competition. As the size of the exporting sector grows, we expect the sender to be less likely to choose aid termination as a sanction measure.

Aid Sanctions Hypothesis: As the sender country’s export-oriented sector trading with the target country increases, the likelihood of aid sanctions declines.

Policymakers’ response to pressures from interest groups is also reflected in the overall impact of the sanction package chosen by the sender, that is, the scale of economic costs that the sanctions are designed to deliver. The sender’s sanction policy implemented against the target may in practice mean measures that are widely different in the scale of damage that the target economy will experience. The chosen measures may have only a limited impact: for instance, suspended economic agreements are unlikely to cause severe and widespread damage to the target’s economy. However, when the sender wishes to impose significant costs on the target’s economy, more punishing sanction strategies are available, such as embargoes and blockades. When special interest groups expect to suffer significant losses as a result of sanctions, policymakers implement a less damaging sanction policy to appease the special interest groups. Such a policy imposes lower costs on the special interest groups as it is less disruptive for the economic relationship with the target. Therefore, we expect that, as the size of interest group opposition grows, the target’s economy will bear lower sanction costs.

Target Costs Hypothesis: As the sender country’s export-oriented sector trading with the target country increases, the target country’s sanction costs decrease.

Data, methods, and variables

Data

We rely on the updated version of the Threat and Imposition of Sanctions (TIES) dataset, 1945–2005, for sanctions information (Morgan, Bapat & Kobayashi, 2013), and bilateral trade data from the Correlates of War Project for information on sender–target trade flows (Barbieri, Keshk & Pollins, 2008). For each event, the TIES dataset identifies the actors (e.g. sender countries and their political institutions responsible for sanction policies, such as the legislature, the executive, or a bureaucratic branch, and target countries and individuals or groups targeted by sanctions), the types/intensity of sanctions (e.g. targeted sanctions or aid sanctions), and issues at stake (e.g. modification of environmental policies or territorial disputes).¹⁰ After merging TIES data with information on bilateral trade flows, we are able to investigate how democratic policymakers respond to conflicting preferences of voters and interest groups in designing sanction policies.

The unit of observation is the directed dyad-year.¹¹ While the TIES dataset has 1,412 sanction episodes, the number of dyad-year observations in which there is an ongoing episode of threatened/imposed sanctions is 3,030 in our dataset. The number of observations declines when we apply the following two modifications to the data to be able to test our hypotheses. First, we drop cases of collective sanctions against EEC/EU countries. Second, because our theory notes that voters represent one of the two pillars of societal preferences for sanction policies, we limit the sample to include only democratic sender countries. We define a country as democratic when its Polity IV score is greater than or equal to 7. Finally, due to data availability limitations on one of our voter awareness measures (Social globalization), which is only available for 1970–2010, the resulting number of observations is 1,223 at the selection stage and 399 at the outcome stage.¹²

Methods

We rely on a two-stage process to evaluate our hypotheses. At the selection stage, the sender country’s policymakers decide whether or not to impose sanctions against the target. Our theoretical framework requires testing the Voter Awareness Hypothesis at this stage: regardless of the type of actual sanction measures chosen, voters’ demand for their leaders to do something about the target’s objectionable behavior should shape the policymakers’ decision to initiate sanctions. Thus, sanctions are more likely when voters pay attention to the dispute with the target. The next step – the outcome stage – is conditional on the selection stage, that is, sanctions are imposed. While the sender country’s policymakers choose to initiate sanctions to satisfy voters, the outcome stage allows the policymakers to address the demands of the second type of domestic actors – special interest groups. The policymakers do so by choosing measures that limit sanction costs imposed on the interest groups. Thus, we test the Target Costs Hypothesis, Targeted Sanctions Hypothesis, Export Sanctions Hypothesis, and Aid Sanctions Hypothesis at the outcome stage. While the policymakers have an incentive to choose sanctions to appease voters (selection stage), they also want to protect interest groups by implementing specific sanction measures that are less costly to these domestic groups (outcome stage). Thus, our theory and hypotheses suggest that a two-stage model with sample selection, that is, bivariate probit with sample selection, should be the appropriate estimation method to test both levels of hypotheses for voters and special interest groups. We run four Heckman models depending on the different kinds of specific sanction measures. To account for the possibility of correlation within each sender’s sanction episodes (especially when one sender is responsible for multiple sanction cases), we use robust standard errors with the cluster option. In addition, we report results of robustness checks based on univariate probit models for all five hypotheses, and based on ordered logit using Target costs for the Target Costs Hypothesis. Here, since we seek to analyze how sanction policies are designed in terms of the target’s costs once the sender country’s policymakers have decided to use sanctions, when we test the Targeted Sanctions Hypothesis, Export Sanctions Hypothesis, Aid Sanctions Hypothesis, and Target Costs Hypothesis, we only include observations in which sanctions are actually imposed, that is, threat-only observations are excluded.

Dependent variables

The Voter Awareness Hypothesis, as stated earlier, is tested in the selection equations of the Heckman models. The selection variable, Sanctions, takes the value of 1 if sanctions are imposed, and 0 otherwise. For the outcome-stage hypotheses, that is, the Targeted Sanctions Hypothesis, Export Sanctions Hypothesis, Aid Sanctions Hypothesis, and Target Costs Hypothesis, we rely on four binary dependent variables, capturing the sender’s choice of specific sanction measures. Specifically, Targeted sanctions takes the value of 1 if the Threatened targeted interest variable in TIES takes a value between 2 and 5, which represents sanctions aiming at ‘regime leadership’, ‘business interest’, ‘political interest’, or ‘military’. Targeted sanctions equals 0 if Threatened targeted interest is ‘general’ and, therefore, the sender seeks to impose sanction costs on the entire target population. Export sanctions equals 1 if Sanction type in TIES is 1, 2, 4, or 5, because these categories include export restrictions, and 0 otherwise. Aid sanctions equals 1 if Sanction type in TIES is 1, 2, 5, or 7 – sanctions in these categories aim to reduce aid flows to a target country – and 0 otherwise. Target costs (binary) is based on Target economic costs in TIES, and takes the value of 1 if the target suffers ‘major’ or ‘severe’ costs, and 0 otherwise.¹³ For a robustness check using ordered logit, we rely on a measure that has more nuanced information on target costs than the binary variable: Target costs is a four-category ordinal variable that ranges from 0 to 3. It modifies the Target economic costs variable of TIES so that Target costs equals 1 if the impact of sanctions on a target country is ‘minor,’ 2 if ‘major’, and 3 if ‘severe’. When no information is available regarding the impact of sanctions, Target costs equals 0.

Explanatory variables

To test the Voter Awareness Hypothesis in the selection equation, our theory requires creating a measure that captures whether voters are aware of international disputes. We do not assume that voters always know about disputes; disputed issues may not be well publicized or critical enough to draw attention. Not surprisingly, it is not a trivial exercise to measure the extent to which voters are aware of a crisis that results in sanctions. To make sure that our analysis is not sensitive to a particular measure of voter awareness, we create two variables operationalizing different mechanisms that lead to voter awareness: one is based on Sanction identity in TIES, and the other on Social globalization of the KOF Index of Globalization (Dreher, 2006). Sanction identity identifies the leading institution in the sender country that initiates sanctions. These institutions include bureaucracy, legislature, executive/government, judiciary, etc. We expect voters to be more aware of disputes when legislative or executive branches, as opposed to bureaucracy or judiciary, suggest sanctions as a policy choice. This greater awareness stems from the fact that politicians elected to the legislative and executive branches have incentives to publicize their work and/or their opponents’ positions to affect election outcomes. Thus, our first variable, Voter awareness identity, equals 1 if Sanction identity indicates legislature, executive or government, and 0 otherwise.¹⁴

Our second voter awareness variable, Social globalization, measures how much access voters may have to information about events in international politics, and is based on Social globalization that is part of KOF Index of Globalization. While Voter awareness identity gauges whether the survival of policymakers imposing sanctions depends on voters, which would generate incentives to publicize sanction use, Social globalization measures voters’ ability to access information about the sender–target dispute through international channels. This measure follows Keohane and Nye’s conceptualization of the social dimension of globalization, which is expressed as the spread of ‘ideas, information, and images’ and movements of people (Keohane & Nye, 2000: 4). We expect voters that have better access to information to be more likely to learn about international disputes, which should result in voters’ demands for a policy that would address the issue. Social globalization ranges from 0 if the country is not globalized on this dimension, to 100 if fully globalized.¹⁵ The two voter awareness variables, Voter awareness identity and Social globalization, are included at the selection stage to examine whether they have a positive effect on the decision to initiate sanctions.¹⁶

Another explanatory variable directly linked to our theoretical argument is Export sector opposition. We rely on the Correlates of War Project for trade data.¹⁷ Note that all exporters in the sender country, whether they export intermediate input goods or goods for final consumption, have the common interest in opposing trade restrictions. Therefore, we are not likely to observe divergent reactions from different subsectors of export-oriented industries as we anticipate in the case of import sanctions. Rather, domestic opposition from exporting groups will intensify when the sender has exposure in its export relationship with the target. Export sector opposition is a continuous variable measuring total exports from the sender to the target, divided by the sender’s GDP (logged). As values of Export sector opposition increase, the sender’s exports to the target, as a share of the sender’s GDP, increase, and hence we expect domestic opposition to rise because economic costs will be greater.

The remaining regressors represent domestic/international political and economic factors that previous research suggests can influence sanction decisions of sender countries: Target democracy, Capability ratio, Alliance, Contiguity, and GDPPC. ¹⁸

Data analysis

Tables I and II display our estimation results. To test the Voter Awareness Hypothesis, these tables use Voter awareness identity and Social globalization to measure the likelihood that voters know about international conflicts. For the hypotheses linking interest group preferences and specific sanction choices, Table I provides estimation results for the Target Costs Hypothesis and Targeted Sanctions Hypothesis, whereas Table II reports estimation results for the Export Sanctions Hypothesis and Aid Sanctions Hypothesis.

To summarize the main findings briefly, our results provide strong empirical support for all of our hypotheses. On the voter side, voter awareness is associated with a higher likelihood of sanction initiation that demonstrates policymakers’ active leadership and readiness to do something in response to the target’s controversial actions. As voters are more likely to have information regarding international disputes, policymakers tend to opt for economic sanctions, as positive and significant coefficients of Voter awareness identity and Social globalization across all sanction imposition models in Tables I and II suggest. On the interest group side, the main explanatory variable, Export sector opposition, is statistically significant at conventional

Table I.

Target costs and targeted sanctions results

Outcome stage DV =	Target costs			Targeted sanctions
	Probit	Heckman probit	Ordered logit	Probit	Heckman probit
Export sector opposition	−0.121* (0.057)	−0.145* (0.040)	−0.178* (0.036)	0.102* (0.041)	0.139* (0.053)
Alliance	0.099 (0.118)	−0.326 (0.209)	−0.086 (0.154)	0.094 (0.133)	−0.205 (0.285)
Contiguity	−0.172* (0.044)	−0.192* (0.067)	−0.265* (0.077)	0.073 (0.051)	0.097* (0.043)
GDPPC	−0.000* (0.000)	−0.000 (0.000)	−0.000* (0.000)	0.000* (0.000)	0.000^† (0.000)
Capability ratio	0.234* (0.076)	0.293* (0.056)	0.166* (0.045)	0.008 (0.036)	0.130^† (0.069)
Target democracy	−0.279 (0.171)	−0.523* (0.076)	−0.084 (0.152)	−0.020 (0.115)	−0.148* (0.074)
Constant	−2.028* (0.855)	−1.941* (0.635)		0.707 (0.765)	2.128* (1.055)
cut1			−1.993 (1.067)
cut2			2.609* (0.721)
cut3			3.973* (0.687)
Selection stage DV = Sanctions
Voter awareness identity	1.424* (0.118)	1.423* (0.118)		1.424* (0.118)	1.355* (0.094)
Social globalization	0.017* (0.005)	0.017* (0.004)		0.017* (0.005)	0.023* (0.004)
Alliance	0.332* (0.054)	0.349* (0.061)		0.332* (0.054)	0.329* (0.053)
Contiguity	−0.040 (0.033)	−0.039 (0.034)		−0.040 (0.033)	−0.040 (0.035)
GDPPC	−0.000* (0.000)	−0.000* (0.000)		−0.000* (0.000)	−0.000* (0.000)
Capability ratio	−0.151* (0.036)	−0.154* (0.037)		−0.151* (0.036)	−0.150* (0.041)
Target democracy	0.002 (0.052)	0.006 (0.051)		0.002 (0.052)	0.025 (0.043)
Constant	−1.409* (0.258)	−1.429* (0.252)		−1.409* (0.258)	−1.657* (0.252)
Rho		−0.554*			−0.964*
Number of observations	399	1,220	395	399	1,220
(for probit selection)	1,223			1,223
Log-likelihood	−125.853	−541.285	−297.944	−266.610	−618.047
(for probit selection)	−490.873			−490.873
Chi-squared	2,789.31	2,462.710	268.300	78.720	57.800
(for probit selection)	7,199.560			7,199.560

^† p < 0.1, *p < 0.05.

levels in the expected directions. In particular, as the size of the export sector increases and the degree of sanction opposition grows, policymakers tend to resort to the use of targeted sanctions but refrain from using aid or export sanctions. Moreover, as export sector opposition increases, overall target costs decline.

Table I reports three sets of empirical results testing the Target Costs Hypothesis and two sets of results testing the Targeted Sanctions Hypothesis with robust support for both hypotheses, regardless of the estimation method used. While these hypotheses pertain to the outcome stage of our analysis, that is, the effects of interest group preferences on the choice of specific sanction measures, the selection stage deals with the Voter Awareness Hypothesis. First, the coefficients on Voter awareness identity and Social globalization are positive and significant in all specifications in Table I. When voters know about international disputes, the voters’ demand for policymakers to do something about the target’s behavior increases the likelihood of sanction imposition.¹⁹ Our results based on Heckman probit suggest that the probability of sanction initiation increases from 0.11 to 0.59 when sanction decisions are made by elected officials who have an interest in publicizing their active policymaking and, hence, informing voters, that is, when we change values of Voter awareness identity from 0 to 1 in our calculations of predicted probabilities. Similarly, the probability of sanction imposition increases from 0.02 when Social globalization is at its lowest level, to 0.29 for the maximum value of this voter awareness variable, while the remaining regressors are held at their mean values.

Table II.

Export and aid sanctions results

Outcome stage DV =	Export sanctions		Aid sanctions
	Probit	Heckman probit	Probit	Heckman probit
Export sector opposition	−0.136^† (0.069)	−0.134* (0.027)	−0.178^† (0.096)	−0.290* (0.106)
Alliance	−0.127 (0.087)	−0.477* (0.051)	−0.241* (0.113)	−0.441* (0.171)
Contiguity	0.040 (0.061)	0.092 (0.068)	0.083 (0.103)	0.077 (0.106)
GDPPC	−0.000* (0.000)	−0.000 (0.000)	−0.000* (0.000)	−0.000* (0.000)
Capability ratio	0.009 (0.078)	0.023 (0.040)	0.250* (0.098)	0.105 (0.096)
Target democracy	−0.735* (0.174)	−0.637* (0.126)	−0.885* (0.218)	−0.926* (0.165)
Constant	−2.419* (1.051)	−2.078* (0.464)	−2.623* (1.255)	−3.812* (1.555)
Selection stage DV = Sanctions
Voter awareness identity	1.424* (0.118)	1.425* (0.116)	1.424* (0.118)	1.424* (0.117)
Social globalization	0.017* (0.005)	0.017* (0.005)	0.017* (0.005)	0.017* (0.005)
Alliance	0.332* (0.054)	0.348* (0.061)	0.332* (0.054)	0.348* (0.060)
Contiguity	−0.040 (0.033)	−0.040 (0.034)	−0.040 (0.033)	−0.040 (0.034)
GDPPC	−0.000* (0.000)	−0.000* (0.000)	−0.000* (0.000)	−0.000* (0.000)
Capability ratio	−0.151* (0.036)	−0.155* (0.038)	−0.151* (0.036)	−0.155* (0.038)
Target democracy	0.002 (0.052)	0.005 (0.052)	0.002 (0.052)	0.003 (0.052)
Constant	−1.409* (0.258)	−1.436* (0.263)	−1.409* (0.258)	−1.431* (0.263)
Rho		−0.400*		−0.313*
Number of observations	399	1,220	399	1,220
(for probit selection)	1,223		1,223
Log-likelihood	−171.599	−580.696	−135.652	−560.432
(for probit selection)	−490.873		−490.873
Chi-squared	759.010	4,482.700	1,668.790	6,050.790
(for probit selection)	7,199.560		7,199.560

^† p < 0.1, *p < 0.05.

Table III.

Predicted probabilities of sanction policy choices

	Export sector opposition
	Min	Mean	Max
Targeted sanctions	0.000	0.212	0.891
Target costs	0.578	0.071	0.003
Export sanctions	0.742	0.221	0.032
Aid sanctions	0.996	0.387	0.005

The table reports conditional (on selection) predicted probabilities, that is, Pr(Outcome DV = 1 | Selection DV = 1), based on Heckman probit models presented in Tables I and II.

Second, while the Targeted Sanctions Hypothesis, Export Sanctions Hypothesis, and Aid Sanctions Hypothesis focus on the choice of specific measures by policymakers in sender countries, the first three columns of Table I explore overall implications of our theory for target countries. The Target Costs Hypothesis is tested using univariate probit, bivariate probit with sample selection (that is, Heckman probit), and ordered logit. In all three models, Export sector opposition decreases the economic costs that the target suffers as a result of sanctions, and these results are significant at the 5% level. For example, the Heckman probit result implies that while the sender country decides to use sanctions at the selection stage to satisfy the public demand for take-charge leadership, the sender does not impose sanctions that can inflict maximum costs on the target as interest group opposition to sanctions increases in the sender country. That is, at the outcome stage, when politicians seek to meet interest group demands, these demands to minimize the damage to the sender–target economic relationship are critical in shaping the policymakers’ choice of sanction measures, as the negative and significant effects of Export sector opposition on Target costs indicate. For example, as Table III demonstrates, the target economy suffers substantial costs with the probability of 0.6 when Export sector opposition is at its lowest level, but at the maximum level of opposition, the probability of high costs is only 0.003.²⁰ Consequently, sanctions are not designed to cause substantial damage to the target’s economy when domestic interest groups exert pressure on the sender country’s policymakers. Finally, the selection parameter, rho, is estimated to be –0.554 and passes statistical significance at the 5% level, which implies that we should correct for non-random sample selection when we analyze the use of sanctions and sanction design.

Third, we report results from the tests of the Targeted Sanctions Hypothesis in the fourth and fifth columns of Table I, based on univariate probit and Heckman probit models, respectively. The rho parameter is negative and significant, which indicates the presence of selection effects. The estimated effect of export sector opposition on the probability of targeted sanctions is positive and reaches statistical significance. As the export relationship between the sender and target grows, export industries present a more powerful opposition force to the possibility of disruptions of their economic transactions. Policymakers in the sender country react to this opposition by tailoring sanctions, that is, by choosing more limited sanctions and focusing on the target’s leadership rather than seeking to cripple the target’s entire economy. Specifically, our result shows that, at the average level of export opposition, the predicted probability of targeted sanctions is 0.2 (see Table III). This probability increases to 0.9 when the Export sector opposition variable takes its maximum value.

Table II displays our empirical findings for the Export Sanctions Hypothesis and Aid Sanctions Hypothesis. We report univariate and Heckman probit models for each hypothesis. Since the rho parameter gains statistical significance in both export and aid sanction cases, we focus on interpreting the bivariate results. The negative coefficients of Export sector opposition in both models imply that, as the value of Export sector opposition increases, the sender country is less likely to impose sanctions that require aid reductions or export restrictions. Again, the substantive significance of these results can be evaluated by calculating predicted probabilities of aid cuts and export sanctions. Table III shows that, when we vary Export sector opposition from its lowest to the highest level, the probability of export restrictions declines from 0.742 to 0.032, while the probability of aid reductions drops from 0.996 to 0.005.

Our results also confirm the Voter Awareness Hypothesis, as the coefficients on Voter awareness identity and Social globalization are positive and statistically significant in all columns of Table II.²¹ When policymakers of the sender country’s institutions initiating sanctions are subject to electoral competition for their survival, these policymakers are more sensitive to public opinion regarding their handling of international disputes than their counterparts in sanctioning institutions that bear no electoral scrutiny by voters. Thus, when elected officials lead the sanction policy, the voters are more likely to be informed about sanctions cases. The voters are also more informed when their country is open to flows of information and social exchanges with other countries. Under these conditions, policymakers are more likely to impose sanctions as a cost-effective way to demonstrate their leadership in foreign affairs.

In sum, these results present strong support for our theory. When voters are informed about the conflict with the target country, policymakers respond to public expectations of action by initiating sanctions. However, when special interest groups are likely to be hurt by sanctions, policymakers impose sanctions but seek to mitigate the impact by designing sanctions to minimize the costs to the interest groups. Thus, voter demands pressure policymakers to do something about the target’s behavior and lead to the use of sanctions, while special group demands make policymakers tailor the details of the sanction policy to exclude measures that are, directly or indirectly, costly to interest groups.²²

Several other explanatory variables yield statistically significant results in our analyses. First, we find that democratic targets receive policy responses different from those aimed at non-democracies. In all models of Tables I and II, Target democracy does not affect the probability that the sender imposes sanctions. However, when it comes to implementation, the sender mostly seeks to avoid using costly measures against democratic targets. That is, policymakers in the sender country are less likely to choose more damaging sanctions, as the negative association between Target democracy and Target costs in Table I shows, and are less likely to impose export or aid restrictions, as the negative relationships between Target democracy, on the one hand, and Export sanctions and Aid sanctions, on the other hand, suggest. At the same time, democratic targets are less likely to experience targeted sanctions. Together, these results indicate that democratic senders utilize sanctions as a symbolic expression of disapproval against democratic targets, mostly seeking to appeal to public opinion rather than cause any damage to targets’ economies. When dealing with non-democracies, on the other hand, democratic senders attempt to maximize the economic impact of their sanction policies. Selection stage results also suggest that affluent senders are more reluctant to use sanctions and if such senders do sanction, they do not seek to maximize the economic damage to the target: in all models, GDPPC has a negative and statistically significant effect on sanction initiation; in the outcome stage, high-income senders tend to avoid export and aid sanctions, while preferring targeted sanctions.

Measures of geopolitical factors yield several interesting results as well. We find that allies are more likely to experience sanctions, but the sender is less likely to employ export or aid restrictions as sanction measures (Table II) when the sender and target countries are formally allied. These results suggest that senders may expect allies to be less likely to resist, but seek to avoid undermining allies’ capabilities by limiting access to exports or foreign aid. Finally, the sender–target capability ratio has a negative and significant effect on sanction initiation across all models, and a positive and significant impact on the likelihood of substantial sanction costs imposed on the target (Table I). As power asymmetry increases, the sender may be more likely to receive concessions prior to sanction imposition; hence, the need to impose sanctions may decline. Yet, if sanctions are initiated, such a sender has a greater capacity to damage the target’s economy.

Conclusion

This article addresses one of the longstanding puzzles in sanctions research: while some sanctions succeed, most fail, and yet governments continue relying on this coercive policy in international disputes. We argue that the choice to impose sanctions and the design of sanction policies represent the government’s response to pressures from domestic constituencies with different policy preferences. Therefore, we depart from the standard approach viewing this foreign policy as a single coercive tool and from the assumption that the use of sanctions is best understood as a strategy adopted by unitary actors (i.e. sender countries) in international bargaining. We consider various measures that senders may implement as part of a broad sanction policy. We also recognize different domestic actors in sender countries that shape this policy. From this perspective, policymakers in sender countries respond to domestic policy preferences and design sanction policies to satisfy these actors’ divergent preferences. Policymakers respond to voters’ demand for action by announcing the decision to impose economic sanctions, thereby addressing the public’s expectation of a response to a target country’s objectionable behavior when voter awareness is high. However, in the face of interest group opposition, the sender’s policymakers have an incentive to select measures that appease special interest groups by mitigating the impact on the economic ties with the target country.

The empirical evidence presented in this article lends robust support to our argument. We analyze the use of specific sanction measures, such as targeted sanctions, export restrictions, and aid cuts, as well as the overall sanction impact on the target country’s economy. Our results suggest that senders’ sanction measures are indeed affected by voters’ and special interest groups’ preferences. More specifically, we find that when the level of voter awareness is high, policymakers are more likely to impose sanctions to punish the target for its unacceptable actions. However, we find that as the size of the export sector increases and, consequently, the degree of sanction opposition from special interest groups grows, the sender is more likely to use targeted sanctions but is reluctant to reduce aid flows, restrict exports, or utilize any measures that could generate significant target costs. In sum, our empirical findings demonstrate that policymakers in democratic sender countries indeed design sanction policies to satisfy both political and economic constituencies.

Our analyses have two important implications for the study of sanction use and sanction effectiveness. First, the results reported in this article indicate that the sender country’s sanction policies are not always designed to maximize their effectiveness. When policymakers select sanction measures that are less detrimental to special interest groups, the unintended consequence is that the instrumental effectiveness of sanction policies in extracting concessions from the target declines. This negative effect is due to the well-established correlation between the target’s sanction costs and sanction success. Therefore, the policymakers’ attempt to address interest groups’ concerns may be partly responsible for the low success rate of economic sanctions. Second, despite the unimpressive success record, policymakers continue using economic sanctions because the policymakers in democratic sender countries have an incentive to impose sanctions to appease voters. Voters demand a response when they are informed about the target’s controversial policy, and the policymakers offer such a response by announcing sanctions. Previous studies show that voters’ approval of policymakers increases just as a result of something being done, that is, adopting a sanction policy as opposed to doing nothing. Therefore, pressures from different constituencies play an important role not only in explaining the choice of specific sanction measures by democratic policymakers, but also in shaping outcomes of economic sanctions. These results suggest that future research needs to take into account the domestic politics of economic sanctions and consider the complexity of this foreign policy in developing explanations of the use and effects of economic coercion.

Footnotes

Replication data

The online appendix, dataset, and replication file for the empirical analysis in this article can be found at .

Acknowledgments

We would like to thank our anonymous reviewers, the editor, Clifton Morgan, David Bearce, and panelists at the 2012 annual meeting of the International Studies Association for their valuable feedback. We are also grateful to Mitch Radtke for outstanding research assistance.

Funding

This work was in part supported by the National Research Foundation of Korea (Grant NRF-2013S1A3A2055081).

Notes

References

Baldwin

Robert

(1985) The Political Economy of US Import Policy. Cambridge, MA: MIT Press.

Barbieri

Katherine

Keshk

Omar

Pollins

Brian

(2008) Correlates of War Project Trade Data Set Codebook (http://correlatesofwar.org).

Bearce

David

(2010) Democracy and international exchange rate regime commitments. Unpublished manuscript (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1917021).

Brace

Paul

Hinckley

Barbara

(1993) Follow the Leader: Opinion Polls and the Modern Presidents. New York: Basic.

Clay

Edward

Geddes

Matthew

Natali

Luisa

(2009) Untying aid: Is it working? An evaluation of the implementation of the Paris Declaration and of the 2001 DAC recommendation of untying ODA to the LDCs. Copenhagen: Danish Institute for International Studies.

Cortright

David

Lopez

George

(2002) Smart Sanctions. Lanham, MD: Rowman & Littlefield.

Destler

I Mac

(1992) American Trade Politics. Washington, DC: Institute for International Economics.

Dreher

Axel

(2006) Does globalization affect growth? Applied Economics 38(10): 1091–1110.

Drezner

Daniel W

(2003) The hidden hand of economic coercion. International Organization 57(3): 643–659.

10.

Drury

A Cooper

(1998) Revisiting economic sanctions reconsidered. Journal of Peace Research 35(4): 497–509.

11.

Eaton

Jonathan

Engers

Maxim

(1992) Sanctions. Journal of Political Economy 100(5): 899–928.

12.

Fordham

Benjamin

(1998) Partisanship, macroeconomic policy, and US uses of force, 1949–1994. Journal of Conflict Resolution 42(4): 418–439.

13.

Fordham

Benjamin

McKeown

Timothy

(2003) Selection and influence: Interest groups and congressional voting on trade policy. International Organization 57(3): 519–549.

14.

Frieden

Jeffry

(1991) Invested interests: The politics of national economic policies in a world of global finance. International Organization 45(4): 425–451.

15.

Ginsberg

Benjamin

(1986) The Captive Public. New York: Basic.

16.

Grossman

Gene

Helpman

Elhanan

(1996) Electoral competition and special interest politics. Review of Economic Studies 63(2): 265–286.

17.

Grossman

Gene

Helpman

Elhanan

(2001) Special Interest Politics. Cambridge, MA: MIT Press.

18.

Haberler

Gottfried

(1936) The Theory of International Trade. London: Hodge.

19.

Hainmueller

Jens

Hiscox

Michael

(2010) Attitudes toward highly skilled and low-skilled immigration. American Political Science Review 104(1): 61–84.

20.

Harrod

Roy

(1939) An essay in dynamic theory. Economic Journal 49(193): 14–33.

21.

Hiscox

Michael

(2002) Commerce, coalitions, and factor mobility: Evidence from congressional votes on trade legislation. American Political Science Review 96(3): 593–608.

22.

Holsti

Ole

(1992) Public opinion and foreign policy. International Studies Quarterly 36(4): 439–466.

23.

Hufbauer

Gary

Schott

Jeffrey

Elliott

Kimberly

Oegg

Barbara

(2008) Economic Sanctions Reconsidered. Washington, DC: Institute for International Economics.

24.

Kaempfer

William

Lowenberg

Anton

(1988) The theory of international economic sanctions. American Economic Review 78(4): 786–793.

25.

Kaempfer

William

Lowenberg

Anton

(1992) International Economic Sanctions. Boulder, CO: Westview.

26.

Kastner

Scott

(2007) When do conflicting political relations affect international trade? Journal of Conflict Resolution 51 (4):664–688.

27.

Keohane

Robert

Nye

Joseph

(2000) Introduction. In: Nye

Joseph

Donahue

John

(eds) Governance in a Globalizing World. Washington, DC: Brookings Institution Press, 1–44.

28.

Kiewiet

Roderick

Kinder

Donald

(1981) Sociotropic politics. British Journal of Political Science 11(2): 129–161.

29.

Knecht

Thomas

(2011) Paying Attention to Foreign Affairs: How Public Opinion Affects Presidential Decision Making. University Park, PA: Pennsylvania State University Press.

30.

Kono

Daniel

(2008) Does public opinion affect trade policy? Business and Politics 10(2): 1–21.

31.

Lacy

Dean

Niou

Emerson MS

(2004) A theory of economic sanctions and issue linkage: The roles of preferences, information, and threats. Journal of Politics 66(1): 25–42.

32.

Lavergne

Réal

(1983) The Political Economy of US Tariffs. New York: Academic.

33.

Lippmann

Walter

(1922) Public Opinion. New York: Macmillan.

34.

MacKuen

Michael

(1983) Political drama, economic conditions, and the dynamics of presidential popularity. American Political Science Review 27(2): 165–192.

35.

Mansfield

Edward

Mutz

Diana

(2009) Support for free trade: Self-interest, sociotropic politics, and out-group anxiety. International Organization 63(3): 425–457.

36.

Morgan

T Clifton

Miers

Anne C

(1999) When threats succeed: A formal model of the threat and use of economic sanctions. Paper presented at the Annual Meeting of the American Political Science Association, Atlanta, GA.

37.

Morgan

T Clifton

Schwebach

Valerie L

(1997) Fools suffer gladly: The use of economic sanctions in international crises. International Studies Quarterly 41(1): 27–50.

38.

Morgan

T Clifton

Bapat

Navin

Kobayashi

Yoshiharu

(2013) Threat and Imposition of Sanctions (TIES) data 4.0 users’ manual (http://www.unc.edu/∼bapat/TIES.htm).

39.

Mutz

Diana

(1992) Mass media and the depoliticization of personal experience. American Journal of Political Science 36(2): 483–508.

40.

Nooruddin

Irfan

Peyton

Autumn

(2010) Dynamics of influence in international politics: The ICC, BIAs, and economic sanctions. Journal of Peace Research 47(6): 711–721.

41.

Ohlin

Bertil

(1933) Interregional and International Trade. Cambridge, MA: Harvard University Press.

42.

Ostrom

W Charles

Jr Job

Brian L

(1986) The president and the political use of force. American Political Science Review 80(2): 541–566.

43.

Ostrom

W Charles

Jr Simon

Dennis

(1985) Promise and performance: A dynamic model of presidential popularity. American Political Science Review 79(2): 334–358.

44.

Pape

Robert

(1997) Why economic sanctions do not work. International Security 22(2): 90–136.

45.

Powlick

Philip

Katz

Andrew

(1998) Defining the public opinion/foreign policy nexus. Mershon International Studies Review 42(1): 29–61.

46.

Shagabutdinova

Ella

Berejikian

Jeffrey

(2007) Deploying sanctions while protecting human rights. Journal of Human Rights 6(1): 59–74.

47.

Smith

Alastair

(1995) The success and use of economic sanctions. International Interactions 21(3): 229–245.

48.

Sobel

Richard

(2001) The Impact of Public Opinion on US Foreign Policy since Vietnam. Oxford: Oxford University Press.

49.

Stimson

James

(1991) Public Opinion in America. Boulder, CO: Westview.

50.

Stolper

Wolfgang F

Samuelson

Paul A

(1941) Protection and real wages. Review of Economic Studies 9(1): 58–73.

51.

Viner

Jacob

(1937) Studies in the Theory of International Trade. New York: Harper.

52.

Whang

Taehee

(2011) Playing to the home crowd? Symbolic use of economic sanctions in the United States. International Studies Quarterly 55(3): 787–801.

53.

Whang

Taehee

McLean

Elena V

Kuberski

Douglas

(2013) Coercion, information, and the success of sanction threats. American Journal of Political Science 57(1): 65–81.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB