The role and welfare rationale of secondary sanctions: A theory and a case study of the US sanctions targeting Iran

Abstract

This paper investigates the role and the welfare rationale of secondary sanctions using a game theoretic framework and a case study of the US sanctions targeting Iran. Existing literature on secondary sanctions focuses either on the sender–third party or the sender–target relations, and fails to address the interdependency of the three players’ strategies. An integrated approach allows us to examine the conditions under which the secondary sanction succeeds in coercing the third party to participate in a sanction campaign against a target. I argue that it acts as a commitment device for the third parties that value target compliance but find it too costly to voluntarily participate in the sanctions when the target complies at a suboptimal level. Despite the coercive nature, secondary sanction can be welfare improving for them. The framework provides an explanation of the successful outcome of the recent US secondary sanctions targeting Iran.

Keywords

Credibility of punishment threats secondary sanctions US sanctions against Iran welfare analysis of secondary sanctions

Introduction

In July 2015, after 20 months of negotiations, the P5+1 and Iran reached a Joint Comprehensive Plan of Action limiting Tehran’s nuclear ability in return for terminating the implementation of all nuclear-related economic and financial sanctions. The multilateral sanctions regime erected between 2007 and 2013 has been assessed to have played a pivotal role in bringing Iran out to the negotiation table (Maloney, 2015; Nephew, 2015). While the UN Security Council Resolutions (UNSCRs) have provided legal and political grounds for individual states to administer their citizens and firms doing business with Iran (O’Sullivan, 2010: 12), the US secondary sanctions since 2010 have influenced the extent of regulation by threatening to restrict the third parties’ US market access if they conduct certain transactions with the Islamic Republic (Katzman, 2015: 35). US secondary sanctions have been heavily criticized for their extraterritoriality and were once a source of serious tension between the USA and its allies (Patterson, 2013). The enforcement was close to nil before 2010, which often raised questions of their effectiveness (Clawson, 2010). The recent success associated with the Iran sanctions regime, however, has led the policy makers in Washington to look into secondary sanctions as a possible foreign policy tool to deal with “bad” actors and to resolve difficult conflicts (Maloney, 2015).

In light of renewed interest, this paper examines the mechanism through which a secondary sanction induces multilateral cooperation for a sanction campaign and greater target compliance, and provides a welfare argument for its use in foreign policy. The work builds on a strand of literature studying the importance of sustaining multilateral cooperation for sanction success (Bapat and Morgan, 2009; Drezner, 2000; Early, 2012; Hufbauer et al., 2007; Kaempfer and Lowenberg, 1999; Martin, 1992, 1993; McLean and Whang, 2010). The secondary sanction is an instrument of coercion to bolster the multilateral alliance: while the leading sender pressures the target of behavioral change, it also threatens the third parties to cease certain transactions with the target in case of its low compliance. Existing research has addressed the secondary sanction from the point of view of the sender focusing on the sender–third party (Martin, 1992, 1993) or sender–target dynamics (McLean and Whang, 2010), taking the actions of the target or the third party to be given. With an integrated framework of all three parties I turn my attention to the third party and its sanction dynamics with the sender and the target. Since almost all the states become “third parties” under a secondary sanction threat and they play a critical role in sanction success (Early, 2012; McLean and Whang, 2010), a reinterpretation of the secondary mechanism from their point of view is warranted. Faced with a secondary sanction threat, under what conditions is the third party willing to participate in a multilateral sanction campaign? For the third party, are secondary sanctions welfare improving or worsening?

I contribute to the literature developing a framework that describes the non-symmetrical coercive interaction among a sender, a target, and a third party in an infinitely repeated, complete information trade sanctions game. I identify two subgame perfect Nash equilibria of the game—a Backfill Equilibrium and a Multilateral Cooperation Equilibrium—and argue that the credibility of the secondary sanction threats set the two equilibria apart. In the Backfill Equilibrium, the threat of punishment is not credible and the third party will continue to trade with the target and substitute for the sender in the target’s market or “backfill,” which the target takes into account in offering a low compliance level. In the Multilateral Cooperation Equilibrium, a credible secondary punishment threat great enough to coerce the third party to participate and small enough so that the sender is willing to execute helps to establish credibility that the third party will punish the target for low compliance and convinces the target to comply at a higher level. Based on the developed framework, I examine how possible substitutions by third parties that are not receivers of the secondary sanction and willing to exploit opening trade opportunities owing to executed sanctions change the workings of the secondary sanction and the multilateral sanction outcome.

A case study of the US secondary sanctions targeting Iran from 1996 is presented as an illustration of the framework. When the Iran Libya Sanctions Act of 1996 (ILSA) went into effect, the USA failed to establish credibility that it would carry out the punishment threats: the EU, then Iran’s largest trade partner, strongly objected, and the USA relented by means of presidential waiver. The USA’s valuation of Iran’s compliance was not high enough to antagonize its closest ally. Foreign investment in Iran’s energy sector thrived and Iran did not acquiesce: the situation corresponds to the Backfill Equilibrium outcome. Secondary sanction threats since the Comprehensive Iran Sanctions, Accountability, Divestment Act of 2010 (CISADA), on the other hand, have been accepted to be credible. The USA not only was able to communicate effectively its increased valuation of Iran’s compliance, but also expanded the scope of sanctions to include Iran’s key industries. Doing so was expected to raise the probability of Iran’s greater compliance, which made it worthwhile for the USA to incur higher costs in punishing the third parties for non-cooperation. The CISADA also raised the cost of non-cooperation for the third party states by increasing the punishment levels as well as strengthening the monitoring and enforcement mechanisms. By brokering deals between the third party states of concern and those that could potentially substitute Iran as their trade partner, the USA at the same time pushed down the cost of cooperation. Major trade partners of Iran have joined the USA, affecting Iran’s willingness to negotiate and agree on an accord. The situation corresponds to the Multilateral Cooperation Equilibrium outcome.

Since the deal was only struck in July 2015, previous literature on multilateral sanctions as well as the US secondary sanctions targeting Iran has focused on why the USA failed to coerce Iran to comply: lack of ability to secure multilateral cooperation, importance of the issues at stake for both the USA and Iran such as terrorism, nuclear proliferation and weapons of mass destruction, and uncertainty about each side’s resolve are some of the explanations that were provided (Drezner, 2000; Langlois and Langlois, 2010). I contribute to the sanctions literature on Iran by analyzing why it worked for CISADA when it did not for ILSA: changes in the key parameters related to the valuations of Iran’s compliance as well as the scope of sanctions are identified and their effects are analyzed.

From a policy perspective, the natural question that arises is whether the use of secondary sanction should be tolerated even though it is an “extraterritorial extension of sanctioning state’s jurisdiction that impinges on the rights of neutral states to regulate their own citizens and companies” (Meyer, 2009: 932). If the secondary mechanism can be welfare improving from the point of view of the third parties in the aggregate, the use may be better defended. Based on the developed framework, I show that a credible secondary sanction threat acts as a commitment device for the third parties that cannot commit to participate in the sanction campaign initiated by the sender even though higher compliance by the target is welfare enhancing. The third parties’ enhanced commitment to punish low compliance effectively increases the expected costs of low compliance, which convinces the target to comply at a higher level. Despite the coercive nature, secondary sanctions can be welfare enhancing.

The paper is organized as follows. I first discuss the existing literature on secondary sanctions and multilateral cooperation, and make a case for a new framework. I then describe a game theoretic model of secondary sanctions followed by a welfare analysis. The case of US secondary sanctions against Iran is presented, and the last section concludes.

Case for a new framework of secondary sanctions

Although multilateral cooperation has often been associated with sanction success (Elliott, 1998; Gilpin, 1984; Martin, 1992, 1993; McLean and Whang, 2010), sustaining one is not an easy task. The potential senders not only have to sacrifice existing economic, political and military ties with the target but also have to give up possible gains from opportunities of backfill owing to ongoing sanctions (Drezner, 2000; Early, 2012; Martin, 1992, 1993). Some commitment devices that could secure multilateral cooperation are institutions that could facilitate monitoring, information exchange and enforcement, and issue-linkages using side payments or punishments.¹ Secondary sanction, with a clear objective of coercing the third parties to partake in a sanction campaign, is one type of issue-linkage attempting to deal with the enforcement problem.

Outcomes of the secondary mechanism have varied. In the case of the Arab League Boycott of Israel, although there were states that reduced trade with Israel, countries such as the USA and France adopted legislation to forbid their citizens from complying (Neff, 1990: 147). After the Helms Burton Act of 1996, US firms doing business with Cuba were replaced by ones from Belgium, Canada, France, and Germany (Hufbauer et al., 1997; Yang et al., 2009). The EU and others have stood adamantly against the extraterritorial application of the ILSA 1996, while they have actively partaken in the multilateral sanction campaign since the CISADA 2010.

Martin (1992, 1993) points to credibility of sanction threats as being central in facilitating cooperation. Martin (1992) argues that the third party determines the extent of cooperation after gauging the leading sender’s commitment credibly signaled by greater self-imposed costs in punishing the target, since it demonstrates the administration’s ability to overcome domestic opposition (Martin, 1992: 37). The theoretical model involving the sender and the third party focuses on their decisions to sanction the target together or independently. Martin (1993) argues that the credibility of a secondary sanction threat can be established by increasing domestic and international audience costs that the sender will bear for reneging on threats. International institutions raise international audience costs. She provides a dyadic model with a sender and a third party, and shows that high audience costs lead to greater commitment by the sender, which convinces the third party to partake in the multilateral sanction. Both do not address the target’s strategies given secondary sanctions. McLean and Whang (2010) explore the role that the third parties play in multilateral sanction success and empirically and theoretically show that if the major trading partners of the target cooperate with the sender, the sanction is more likely to succeed. They construct a model that captures the dynamics of the sender and the target, sweeping the effects of the third party’s action under the sanction costs borne by the sender and the target: the third party’s participation is not a decision variable. Early (2012), on the other hand, focuses on the condition under which the third party participates in a multilateral sanction campaign. He shows empirically that when there are high commercial benefits for the third parties trading with the target as well as defense pacts between them, the third parties are less likely to cooperate with the sender. Atzili and Pearlman (2012) look at non-symmetrical triadic coercive interaction among two states and a non-state actor where a state coerces another to deter it from supporting a non-state actor. They find that the stronger the targeted regime is, the better control it may have over the non-state actor and thus the more likely it is that the deterrent action will succeed.

As seen above, existing theoretical research has focused on one of the three dyadic sanctions related to the secondary sanction—the sender–third party, the sender–target, and the third party–target sanctions, taking the action of the player that is left out as given. Such an approach makes it easy for us to apply the many existing frameworks that describe how a sanction works. The credibility of sanction threats (Martin, 1993; Schelling, 1960), general sanction costs to the sender as well as the receiver (Eaton and Engers, 1992, 1999; Langlois and Langlois, 2010; Martin, 1993; Smith, 1996), domestic audience costs (Dorussen and Mo, 2001; Schelling, 1960), interest groups (Kaempfer and Lowenberg, 1999), patience (Eaton and Engers, 1992), international organizations (Drezner, 2000; Martin, 1993) and existing economic ties with other states (McLean and Whang, 2010) are some of the factors that have been theoretically demonstrated to affect whether the sender is able to convince the receiver to comply. The bottom line is that the sender will threaten with, carry out, or at times choose to give up sanctions, and the receiver will choose whether to acquiesce and if so the timing of it if and only if the benefits of doing so outweigh the costs. Uncertainties in the players’ willingness to give in and the payoff parameters as well as different bargaining processes and status quos of the sanction game shape the players’ strategies and result in diverse sanction outcomes.

Most relevant to this work is the literature on credibility and sanction costs. These two subjects are interrelated since a necessary condition for a sanction threat to be credible is that the sender has to be willing to incur the costs of sanctioning the receiver. In general, it is accepted that the greater the cost the receiver incurs owing to sanctions, the more likely it is to acquiesce. Although I acknowledge the strategic importance of sanction costs (Baldwin, 1985: 107) that could potentially signal one’s resolve and convince one’s counterpart to acquiesce (Langlois and Langlois, 2010), I do not consider them in my model.

Introducing a third party in a dyadic setting may change the dynamics of the two players and the sanction outcome. A successful secondary sanction threat on a third party effectively raises the cost of sanctions for the target and could push it to accept a deal that was previously not agreeable. The cost effect on the sender, however, is ambiguous. Depending on the relations among them, the sender may not be incurring greater costs in sending out threats against both the third party and the target. For instance, if the target relies heavily on the third party, the third party relies heavily on the sender and the sender relies heavily on the target, for the sender, threatening the third party to sanction the target could turn out to be less costly than sanctioning the target by itself aiming to coerce a certain level of target compliance. A three-player model as presented here with the dependency structure explicitly configured can be useful in capturing such sanction dynamics and the possible outcomes.

Model

To focus on the bare bones dynamics of how the secondary sanction works within the multilateral sanction campaign, I develop a simple, infinitely repeated complete information game of trade sanctions² involving three nation states³—R, S, and T: S is the sender, R is the third party,⁴ and T is the target of the multilateral sanction campaign. To the best of my knowledge, all of the secondary sanctions other than the Arab League’s boycott of Israel have been initiated by the USA. This is not a surprise because only secondary sanctions sent out by an entity that has economic leverage over others would actually have bite. Thus, assigning specific roles to R, S, and T seems reasonable. The model is a modification and an extension of the sanction frameworks introduced in Eaton and Engers (1992) and Martin (1993) in line with the work of Basu (1986) that studies three-party non-symmetrical coercive interaction.⁵

I am interested in a situation where the compliance of the target is not a once-and-for-all decision but something that has to be negotiated and monitored repeatedly (Eaton and Engers, 1992): even when the sender accepts, the game goes on to the next stage. I assume the sanctions game to be infinitely repeated, since it is realistic to assume that the decision-makers do not know when the game will actually end when they make a move (Osborne and Rubinstein, 1994: 135). The compliance level can be conceptualized in terms of a range of objectives that the sender wants to achieve through a multilateral sanction campaign, from drawing the target out to the negotiation table to a full-fledged change in policy of the target. The costs experienced by the target increase with the level of compliance. The costs may be due to domestic politics (Iida, 1993; Mo, 1994) or policy implementation. The sender also incurs audience cost when it decides not to sanction the target given a compliance level (Martin, 1993). The sender values target compliance more than the third party.

The model explicitly addresses the possibility of substitution between the sender and the third party in the target’s market. When the sender sanctions the target and threatens the third party to induce it to join in, rather than acquiescing, the third party can choose to continue to trade with the target and “backfill” the void made by the sender in the target’s market. This assumption is in line with empirical evidence: after the Helms Burton Act of 1996, US firms doing business with Cuba were replaced by ones from Belgium, Canada, France, and Germany (Yang et al., 2009).

Existing studies in the literature agree that sanctions alone rarely bring about desired change in the policy of the target country. The necessary condition for a sanction to be successful, however, is that the cost of defiance (sanctioning) has to be greater than the cost of compliance (not sanctioning) from the point of view of the receiver (sender) (Hufbauer et al., 2007: 50). I also assume that it is the relative payoffs that players compare when making a decision. To focus on pure strategies, I assume that when the expected payoffs are the same between accepting and rejecting an offered compliance level, the sender chooses to accept it. Similarly, when the expected payoffs are the same for trading and not trading with the target, the third party chooses to trade.

Sequence of moves

The timeline of the stage game is as follows. The game starts with the target T offering to comply at a certain level k_i. The sender chooses to accept k_i and trade, {T}, or reject k_i and not to trade, {N}. When the sender S chooses to accept and trade, then the stage game ends. When the sender chooses to sanction the target, the third party R decides whether to trade {t} or not to trade {n} with the target. If the third party chooses to trade with the target, the sender chooses a punishment level $p \in [0, 1]$ by which it reduces its trade with the third party. The stage game is depicted in Figure 1.

Figure 1.

Game tree of a secondary sanction game.

Payoffs

The total expected payoff of player j without any sanction imposed is represented as U_j for j = R, S, T. The payoffs depend on the gains from trade from trading with one another and a compliance level $k_{i} > 0$ by the target. When the target complies at k_i, the sender obtains a payoff of γk_i where γ > 1, and the third party, βk_i, where $- 1 \leq β \leq 1$ : the sender values target compliance more than the third party and the third party may gain negative payoffs from target compliance. The sender is also subject to domestic audience cost L(k_i) when it accepts k_i and trades. We assume L′(k_i) < 0 and L′′(k_i) > 0 so that as the compliance level goes up, domestic audience cost decreases at an increasing rate. We also assume that k_l represents the minimum compliance level that the sender will not sanction the target for: $k_{l} \equiv {k_{j} | γ k_{j} - L (k_{j}) + U_{S} = U_{S} (1 - a_{ST})}$ . The left-hand side of the equation in the parentheses is the sender’s payoff for accepting the target compliance at k_j, while the right-hand side is its payoff for rejecting the target’s proposed compliance level and sanctioning. The target incurs the cost of g(k_i) complying at k_i, where g(k_i) > 0, g′(k_i) > 0, and g′′(k_i) > 0: a unit increase in the compliance level becomes more costly as the level increases. Thus, when the target cooperates at k_i and the sender accepts, there are no sanctions and their payoffs are U_T−g(k_i) and γk_i –L(k_i) +U_S, respectively. We assume that when both the sender and the third party credibly threaten the target to comply, the target complies at a level k_T such that it is greater than the level k_l required by the sender’s domestic audience cost and there will be no sanctions for the period.

The coefficient α_jk satisfying 0 ≤α_jk < 1 captures the extent to which j’s total trade depends on k. Although we are concerned with the dynamics among R, S, T, there are more than three players in the world: we assume α_jk+α_jl < 1, for j, k, l = R, S, T and j≠k, j≠l, k≠l. The third party’s possibility or the ability to backfill is represented by a coefficient of substitution, δ, which satisfies 0 ≤δ≤ 1. When the third party cannot substitute for the sender, δ = 0, and when it fully substitutes for the sender, δ = 1. The additional payoff for the third party from backfill depends on δ and the size of the market α_TS U_T that can be captured owing to executed sanctions by the sender. The payoff from backfill B(δ, α_TSU_T) increases with δ and α_TS: the greater the probability of substitution and the greater the void from ongoing sanctions, the greater the expected payoff from backfill. We assume that B′′(δ) < 0: as the extent of substitution increases, the benefit from additional unit of backfill decreases. The punishment threat in the secondary sanction is represented with $p \in [0, 1]$ , by which the sender reduces its trade with the third party. Suppose the target offers a compliance level k_i and the sender decides to sanction. If the third party chooses to maintain its trade relations with the target as well as to backfill despite the secondary sanction threat, the payoffs of R, S, and T are $U_{R} (1 - p α_{RS}) + B (δ, α_{TS} U_{T})$ , $(1 - α_{ST} - p α_{SR}) U_{S}$ and $(1 - α_{TS} (1 - δ)) U_{T}$ , respectively. If the target complies at k_i and the sender accepts and decide not to sanction, then the payoffs of R, S, and T are $U_{R} + β k_{i}$ , $γ k_{i} + U_{S} - L (k_{i})$ , and U_T–g(k_i), respectively.

Equilibrium

Backfill Equilibrium

I solve for a Subgame perfect Nash equilibrium strategy profile such that every player’s strategy is a best response to that of others in every subgame. We suppose that the three players discount the future payoffs with discount factors θ_S, θ_T and θ_R, all greater than 1/2: the discount factor can be interpreted as each player’s perceptions of the probability that the game will continue to the next period. There are many Subgame perfect Nash equilibria of an infinitely repeated game. We introduce two: a Backfill Equilibrium where there is no agreement on target compliance, the sender sanctions, and the third party continues its trade with the target and backfill, and a Multilateral Cooperation Equilibrium where the target complies at a level that satisfies the sender under the multilateral and secondary sanction threats.

In the Backfill Equilibrium, while the sender’s threat to punish the target for low compliance is credible owing to domestic audience costs L(k_i), its threat to punish the uncooperative third party is not. The target expects the third party to continue trading and to pick up the contracts left by the sender when the sender sanctions. There is no agreement reached regarding the target’s compliance level since the target is not willing to comply at the minimum level k_l that the sender’s domestic audience requires of the target. Thus, the sender sanctions the target and the third party trades and backfills. The sender does not punish the third party for non-cooperation because it is too costly to do so, and knowing this, the third party will not partake in the sanctions. Given the sender’;s and the third party’;s expected actions, it is optimal for the target to offer a compliance level that will lead the sender to sanction. The equilibrium strategy profile is characterized in Proposition 1 following Abreu (1998) and a sketch of the proof is in the Appendix.

Proposition 1. The Backfill Equilibrium strategy profile consists of the following, where $k_{l} \equiv {k_{j} | γ k_{j} - L (k_{j}) + U_{S} = U_{S} (1 - α_{ST})}$ and $g^{- 1} ((α_{TS}) (1 - δ) U_{T}) < k_{l}$ :

Target offers any $k_{i} \in [0, k_{l})$ .

Sender trades for all compliance level k_i≥k_l and sanctions for k_i < k_l. If the third party trades and backfills, the sender does not punish.

The third party trades and backfills whenever the sender sanctions the target.

We see that for a sender with domestic audience that is favorable to the target (low domestic audience costs), and that relies more on the target, it is willing to accept a lower compliance level offered.

Multilateral Cooperation Equilibrium

In the Multilateral Cooperation Equilibrium, the credibility of the sender’s threat to punish the third party’s non-cooperation convinces the target to comply at a higher level. Here, I explore the conditions that are necessary to sustain multilateral cooperation while players use one kind of “forgiving grim trigger”⁶ strategy that resembles how the secondary sanction within a multilateral sanction campaign works in reality. In this specific Multilateral Cooperation Equilibrium strategy profile, when the target offers to comply at a level that is lower than satisfactory, the sender and the third party both sanction for a single period.⁷ If the third party continues to trade and backfill while the sender sanctions for a compliance level less than satisfactory, then the sender punishes the third party for a single period. When the sender fails to punish the target’s low compliance or the third party’s non-cooperation, from the next period on there is no agreement on a compliance level: the target rejects compliance at a level of the sender’s satisfaction and the third party will not cooperate with the sender. Since both the sender and the third party can credibly commit to sanction the target for low compliance at this multilateral cooperation equilibrium, the target complies at a minimum level k_t that the sender and the target will not sanction. The compliance level k_t depends on how much the target relies on the sender and the third party. The equilibrium strategy profile is characterized below. The proof is in the Appendix.

Proposition 2. Suppose the secondary sanction threat p satisfies p = p_E where

\frac{α_{RT}}{α_{RS}} + \frac{B (δ, α_{TS} U_{T})}{α_{RS} U_{R}} \leq p_{E} \leq \frac{(γ k_{t} - L (k_{t}) + α_{ST} U_{S}) θ_{S}}{α_{SR} (1 - θ_{S}) U_{S}}

The following strategy profile then constitutes a Multilateral Cooperation Equilibrium that results in a compliance level k_t by the target where $k_{t} \equiv g^{- 1} ((α_{TS} + α_{TR}) U_{T})$ and $k_{l} \equiv {k_{j} | γ k_{j} + U_{S} - L (k_{j}) = U_{S} (1 - α_{ST})}$ .

Equilibrium path: play according to the sequence (k_t, T).

The target specific punishment path (which applies whenever the target offers a compliance level less than k_t): the sender and the third party sanction the target for a single period and return to the equilibrium sequence (k_t, T).

Third party specific punishment path (which applies whenever the third party trades with the target and backfills after the sender sanctions): the sender punishes the third party by p_E and returns to the equilibrium sequence (k_t, T).

The sender specific punishment path (which applies whenever the sender accepts k_l≤k_i < k_t or fails to punish an uncooperative third party by p_E): from the next period on, the sequence (k_b, N, t, p = 0), where $k_{b} \in [0, k_{l})$ is played indefinitely.

The condition under which the third party participates in the multilateral sanction is

\frac{α_{RT}}{α_{RS}} + \frac{B (δ, α_{TS} U_{T})}{α_{RS} U_{R}} \leq p_{E} \leq \frac{(γ k_{t} - L (k_{t}) + α_{ST} U_{S}) θ_{S}}{α_{SR} U_{S} (1 - θ_{S})}

On the left, which we denote as p _E, is the minimum punishment threat that the sender should threaten the third party with to compel the third party to participate in sanctioning the target. It is the ratio of the sum of the third party’s expected payoffs of trading with the target and backfilling to the expected payoff of maintaining the trade relations with the sender. As the probability of backfill increases, the third party’s reliance on the target increases, and that on the sender decreases, p _E has to be higher to convince the third party to cooperate. As the target relies more on the sender, p _E increases: the expected payoff from backfill is higher from the third party’s point of view and thus the punishment from the secondary sanction would have to be greater in order to induce participation.

On the right, which we denote as ${\bar{p}}_{E}$ ,⁸ is the maximum amount of punishment threat that the sender is willing to impose on the third party. For the third party, a punishment threat beyond ${\bar{p}}_{E}$ is not credible since the costs exceed the benefits for the sender to carry out the punishment threat. It is the ratio of the sum of the sender’s expected payoff of punishing the uncooperative third party to the sender’s expected payoff of ignoring the third party’s non-cooperation and maintaining trade relations. The sender is willing to punish the uncooperative third party at a greater level when the target relies more on the third party. So does the sender that values target compliance more, relies on the third party less, relies on the target more, and discounts the future less.⁹

When the maximum amount of credible punishment is less than the minimum amount of punishment required to induce the third party to participate in the sanction against the target ( ${\bar{p}}_{E} < {\underline{p}}_{E})$ , then the secondary sanction mechanism described above will no longer be able to pressure the target to comply at the satisfactory level (k_t). Knowing this, the target will offer a compliance level below k_l because it is evident that the third party will continue to trade and backfill while the sender will carry out the sanction threats. Figure 2 summarizes how the range of secondary sanction threats leading to multilateral cooperation changes depending on the sender and the third party relations with the target. The dotted line depicts ${\underline{p}}_{E}$ and the solid line ${\bar{p}}_{E}$ .

Figure 2.

The range of punishment threats of an effective secondary sanction.

The third parties with low reliance on the target (low α_RT) will join the multilateral sanction since the minimum punishment required is less than the maximum the sender is willing to exert. For those with greater reliance on the target (high α_RT), the minimum punishment required is higher, which may be above the level that the sender can credibly threaten the target with. For the senders with greater reliance on the third parties (high α_SR), their credible punishment threat is less because punishing the third parties for non-cooperation becomes more costly. If it is tied with a high backfill possibility or a low reliance of the third party on the sender so that ${\underline{p}}_{E}$ shifts up or ${\bar{p}}_{E}$ shifts down, there may exist an α_SR* beyond which the secondary sanction mechanism fails to rally cooperation.

The minimum required as well as the maximum credible secondary sanction threats have to be greater to convince a target with greater reliance on the sender (high α_TS) to comply at a higher level. This is because when the target relies heavily on the sender and the sender is expected to sanction the target, the third party expects that there will be more business opportunities opening up owing to the sender’s absence. Thus, it would require a greater punishment threat from the sender in order for the third party to cooperate. From the point of view of the sender, greater reliance of the target means a higher level of compliance possible if the secondary sanction succeeds. The sender is willing to incur the cost of punishing the third party, which raises the maximum credible threat. The bottom left graph depicted in Figure 2 reflects a scenario where a unit increase in target reliance on the sender α_TS results in a greater increase in the backfill payoff of the third party than the sender’s payoff from higher compliance by the target. In other words, when the target relies on the sender much, the sender may not be willing to punish the third party to induce cooperation because it is too costly to convince the third party to forgo substantial backfill opportunities.

For a third party with greater possibility of backfill (a greater substitution coefficient δ), the minimum required punishment is higher because the benefit of backfilling increases. There may exist δ^∗, beyond which the third party declines to cooperate because backfilling benefits are too great to let go. If the sender’s reliance on the third party is very low, then the maximum credible level of punishment threat is higher than the minimum required punishment threat for all δ. Then regardless of the expected payoff of backfill, the third party will decide to cooperate.

Discussion

Impact of possible substitution by others

In the game provided in the previous section, we only allow for substitution between the sender and the third in the target’s market. In reality though, there are other third parties that are not receivers of the secondary sanction that are willing to take advantage of opening trade opportunities with the sender, the target, and the third party owing to executed sanctions. For exposition purposes, I use the term “the fourth party” to distinguish them from the third party that is the receiver of the secondary sanction. We discuss in this subsection the impact the fourth has on the workings of the secondary sanction and the multilateral sanction outcome. Based on the necessary condition that we have derived for a Multilateral Sanction Equilibrium,

\frac{α_{RT}}{α_{RS}} + \frac{B (δ, α_{TS} U_{T})}{α_{RS} U_{R}} \leq p_{E} \leq \frac{(γ k_{t} - L (k_{t}) + α_{ST} U_{S}) θ_{S}}{α_{SR} U_{S} (1 - θ_{S})}

we can reach the following conclusions.

First of all, the existence of a fourth party will lower the sender costs of secondary sanctions, since the sender can substitute the trade with the fourth party for that with the third party. This will increase the maximum punishment ${\bar{p}}_{E}$ that it is willing to impose on the non-cooperating third party.

For the third party, the existence of a fourth party that it can substitute the sender with lowers the cost of the secondary sanction, effectively increasing the minimum secondary sanction threat to coerce the third party to participate, ${\underline{p}}_{E}$ , to go up. When the third party can substitute the target with a fourth party, the cost of participating in the multilateral sanction campaign goes down, which lowers the minimum secondary sanction threat ${\underline{p}}_{E}$ required to pressure the third party to participate go down.

When the target can substitute the third party with a fourth party, then the cost of multilateral sanctions will go down for the target, and the compliance level that the multilateral cooperation front consisting of the sender and the third party can push for will be lower. The sender’s maximum punishment it is willing to impose on the uncooperative third party ${\bar{p}}_{E}$ goes down. Now if it becomes lower than ${\underline{p}}_{E}$ , there will be no multilateral sanction front.

Welfare impact of secondary sanctions

Can a secondary sanction, despite its coercive nature, be welfare improving from the point of view of a third party? If it turns out that the third party would have participated without a punishment threat from the sender, the secondary sanction can be considered to be welfare neutral. If the third party only participates under the secondary sanction threat and its payoff increases when it does, the secondary sanction is welfare improving. If the third party only participates under the threat and its payoff decreases when it does, the secondary sanction is welfare worsening.

We first note that there may be third parties that would voluntarily participate in the infinitely repeated game without a secondary sanction threat. For them, their valuation of target compliance is high enough to forgo expected returns from continuing trade and backfilling for higher levels of target compliance in future periods. Knowing this, the target believes that the third party will participate in the multilateral sanction campaign, which results in a higher compliance level. More formally, the situation corresponds to the stage game in Figure 1 ending with the third party deciding whether to join in the multilateral sanction campaign or not. Suppose the third party’s valuation of target compliance β that makes third party ambivalent between trading with the target and joining the multilateral sanction campaign without a secondary sanction threat is β_m. There exists a subgame perfect Nash equilibrium strategy profile where the third parties with β≥β_m commit to participate in the multilateral sanction campaign and such expectation will result in a higher compliance level by the target. For such third parties, secondary sanction is welfare neutral since they would have participated in the multilateral sanction campaign without one.

I argue that there exists β_s<β_m such that third parties with valuation of target compliance greater (less) than β_s are better (worse) off with secondary sanctions imposed. A proof is given in the Appendix. The intuition is that with binding secondary sanctions, the cost of trading and backfilling increases and the minimum β (or β_s) that makes participating in the multilateral sanction campaign worthwhile decreases: with greater costs for continued trade imposed owing to secondary sanctions, the third parties that would not have participated voluntarily would be better off joining the multilateral sanction campaign against the target.

For the third parties with valuations of target compliance that satisfies β_s≤β ≤ β_m, even though they are better off when the target complies at a higher level, they cannot commit to voluntarily participate because once the sender has decided to sanction the target, they are better off continuing trade with the target and backfilling. Knowing this, the target does not comply at a higher level, which results in a sanction by the sender and a continued trade and backfill by the third party. This, however, is suboptimal for the third parties because their valuations of the target compliance are high enough such that a higher compliance level would result in higher payoffs. By increasing the cost of continued trade and backfill, the secondary sanction acts as a commitment device for the third parties to sanction the target, ultimately raising the cost of low compliance of the target and leading it to comply at a higher level. The third party is better off vis-à-vis the situation without secondary sanctions. Note here that if the third party expects to gain more from backfilling, it will have to value target compliance more in order for its welfare to be enhanced owing to the secondary sanction.

On the other hand, the third parties with β < β_s participating in the multilateral sanction campaign under secondary sanction threats are worse off since they do not value target compliance much. Being forced into a multilateral sanction campaign decreases their welfare.

US triadic sanctions against Iran

ILSA and the Backfill Equilibrium

Since mid-1995, several executive orders that prohibited US domestic firms from trading with or investing in Iran were issued by the US presidents to penalize and restrain Iran’s illicit nuclear proliferation activities, support for international terrorism, and arms exports (later, human rights abuses). Having the third largest proven petroleum reserves in the world, Iran’s petroleum sector generates approximately 20% of its total GDP and 80% of its total exports. As Iran invited foreign investments in its energy sector in 1995, the USA turned to foreign states and firms for cooperation to halt the country’s energy sector development and economic growth. There was increasing evidence that the EU firms were replacing the US counterparts that had exited. The US signed into law the ILSA (or ISA¹⁰), targeting firms that invested more than $20 million a year in Iran’s energy sector. The sanction punishment included bans on Export–Import bank assistance, the issuance of export licenses, loans from US financial institutions, and procurement of goods and services by the US government.

The punishment threat, however, turned out not to be credible. The EU, the largest trade partner of both the USA and Iran then, strongly opposed to the extraterritorial application of the US domestic law and threatened to file a complaint before the World Trade Organization. Since 1992, the EU had been pursuing a policy of “critical dialogue,” cooperating with Washington on the exports of arms and nuclear and dual-use technology but maintaining civilian trade and investment (Rodman, 2001; 185). In the end, the USA backed down and waived penalizing a project that was determined to be a violation of the ILSA. The EU decided not to file any complaints and promised to collaborate with the USA on further anti-proliferation and anti-terrorism efforts (Katzman, 2015: 20).

The outcome of the US–EU conflict over ILSA was a signal to the rest of the world that the USA was not committed enough to carry out the sanction threats (relatively low γ) (Bahgat, 2003; 120). For most of the international community, Iran’s nuclear program and support of terrorist organizations were not a major concern (low β). Participating in the sanction and letting go of the existing trade relations and potential business opportunities especially in the energy sector (high α_RT and δ) were too costly. Once it was clear that no sanction punishments would be carried out, it was only rational that they did not participate. Empirically, there were a few cases in which foreign firms and states publicly acknowledged the deterrent impact of ILSA right after its implementation, but the effect was at most transient (Rodman, 2001; 187). Table 1 shows major investments and development projects by foreign state-owned firms over US$1 billion in Iran’s energy sector since 1999. Despite active business, from 1996 until September 2010, there were no violation determinations.¹¹

Table 1.

Post-1999 major investments/major development projects in Iran’s energy sector (over US$1 billion)

Date	Field/project	Company/status	Value (US$ billion)
February 1999	Doroud (oil)	Total (France)/ENI (Italy)	1
July 2000	Phase 4 and 5, South Pars (gas)	ENI	1.9
June 2001	Darkhovin (oil)	ENI	1
September 2002	Phase 9 + 10, South Pars (gas)	LG Engineering and Construction Corp.	1.6
October 2004	Yadavaran (oil)	Sinopec (China)	2
December 2006	North Pars Gas Field (offshore gas). Includes gas purchases	China National Offshore Oil Co.	16
February 2007	Phase 13, 14—South Pars (gas)	Royal Dutch Shell, Repsol (Spain)	4.3
July 2007	Phase 22, 23, 24—South Pars (gas)	Turkish Petroleum Company (TPAO)	12
December 2007	Golshan and Ferdowsi onshore and offshore gas and oil fields and LNG plant	Petrofield Subsidiary of SKS Ventures (Malaysia)	15
February 2008	Lavan Field (offshore natural gas)	PGNiG (Polish Oil and Gas Company, Poland)	2
April 2008	Iran’s Kish gas field	Oman (co-financing of project)	7
June 2008	Resalat Oilfield	Amona (Malaysia). Joined in June 2009 by CNOOC and another China firm, COSL.	1.5
March 2009	Phase 12 South Pars (gas)	Oil and Natural Gas Corp. of India, Oil India Ltd, India Oil Corp. Ltd in 2007; Sonanagol (Angola) PDVSA (Venezuela)	8 from Indian firms/1.5 from Sonangol/0.78 from PDVSA
August 2009	Abadan refinery	Sinopec	6
October 2009	South Pars Gas Field—Phases 6–8, Gas Sweetening Plant	G and S Engineering and Construction (South Korea)	1.4
November 2009	South Pars: Phase 12—Parts 2 and 3	Daelim (South Korea)—Part 2; Tecnimont (Italy)—Part 3	4
February 2010	South Pars: Phase 11	CNPC (China)	4.7
December 2011	Zagheh Oil Field	Tatneft (Russia)	1

Source: Reproduced from table 5 of Katzman (2015: 54–58).

In sum, the lack of credibility of the secondary sanction threat led the third parties not to cooperate with the US sanctions against Iran. The USA’s valuation of Iran’s compliance was not high enough to antagonize the EU, its major ally. Almost all third party states chose to continue trade with Iran and started to make use of new opportunities, substituting for the US firms that exited owing to unilateral sanctions. Foreign investment in Iran’s energy sector thrived and Iran and the USA were not able to come up with an agreement. Owing to ongoing unilateral sanction and subsequent backfills by the third parties, the USA kept losing its economic leverage over Iran. The world was at a Backfill Equilibrium

CISADA and the Multilateral Cooperation Equilibrium

In 2002, with the exposure of two undeclared nuclear facility sites, the Iranian nuclear program became one of the most high-profile issues on the international agenda (Patterson 2013: 138). After the IAEA concluded that Iran was not complying with the Comprehensive Safeguards Agreement in 2005, a series of UNSCRs—1696 (2006), 1737 (2006), 1747 (2007), and 1803 (2008)—followed, condemning Iran for its uranium enrichment programs and demanding that Tehran verify its compliances with the IAEA Board of Governor’s Resolution GOV/2006/14. The UNSCRs focused on hampering Iran’s ability to acquire equipment, technology, and finances for nuclear enrichment and major arms (such as ballistic missiles) development from the member states. In 2009, the fear of Iran getting closer to developing the capability of producing a nuclear weapon grew (IAEA, 2009: 7). The revelation of a covert uranium enrichment facility at Qom in September followed by Iran’s rejection of a UN-brokered deal designed to process its nuclear fuel abroad¹² led the USA and the EU to push for another round of sanctions through the UN. The UNSCR 1929 (2010),¹³ although the strongest yet, did not include any energy sector-related sanctions that the USA and the EU had lobbied for¹⁴ and China and Russia so resisted (Hurst, 2012: 547).

As the next step, the USA strengthened its secondary sanctions. On 1 July 2010, President Obama signed into law an amendment of ISA, the CISADA, which restricted foreign companies’ exports of refined petroleum to Iran and the sales of equipment or services that could help Iran to produce or import refined petroleum.¹⁵ Multiple executive orders¹⁶ as well as Acts followed, expanding the scope of sanctions and raising the level of punishment against uncooperative third parties. While in the original version of ILSA there was neither a time limit for investigation and determination of sanction violations nor a guideline to determine which third party to waive the sanctions and how long it would be waived for,¹⁷ the CISADA obligated the administration to begin investigation when there was “credible information” about a potential violation, and made it mandatory to determine violations in 180 days. Subsequent laws amended the CISADA to set up mechanisms for the US Congress to oversee the administration’s investigations for violations (Katzman, 2015: 13). Determinations of sanctions became more regular. On 25 May 2011, the US government issued its first seven sanction determinations,¹⁸ and on 12 January 2012, three additional firms were determined to have violated CISADA.

Armed with the CISADA, the US government persuaded other nation states to impose sanctions against Iran in addition to the UNSCR 1929: The EU,¹⁹ Japan, South Korea, Norway, Canada, and India announced their independent sanctions soon after. The UNSCR 1929 had provided a stronger political and legal basis for multilateral cooperation. The contents and development processes of unilateral sanctions on Iran since 2010, however, show that the actual measures imposed by each state were carefully fine-tuned through rounds of formal and informal talks with the US government in accordance with the USA’s secondary sanctions.²⁰ For instance, identifying Iran as a jurisdiction of “primary money laundering concern” under section 311 of the US Patriot Act, President Obama signed the National Defense Authorization Act for Fiscal Year 2012 into law on 31 December 2011 to impose financial sanctions to foreign financial institutions owned or controlled by the government of a foreign country that engages in financial transactions with the Central Bank of Iran for the sale or purchase of petroleum or petroleum products to or from Iran (H.R. 1540-351). The US President could provide a renewable waiver of 120 days to countries that had significantly reduced the volume of crude oil imports from Iran.

As the law went into effect on 1 July 2012, the EU, which used to import 20% of the total Iranian crude oil exports, stopped importing crude oil from Iran entirely. The five largest buyers of Iranian oil—China, India, Japan, South Korea, and Turkey—among others by then had already dramatically reduced their purchases in order to get waivers for the financial sanctions (“Asian nations slash Iran oil imports,”Financial Times, 22 March 2012). China, South Korea, and Japan, which collectively imported about 65% of the total Iranian oil, cut their oil imports from Iran by 40, 16, and 34%, respectively. Turkey also cut its oil purchase by 20% after being warned by Washington to do so (“Turkey to cut Iran oil imports, bows to U.S. pressure,” Reuters, 30 March 2012), but was further pressed to follow with a further cut in six months to get an extension on the waiver that was issued (“Turkey eyes further 10 pct cut in Iran oil buys,” Reuters, 4 July 2012). The USA at the same time brokered deals between the Asian countries and the Gulf Cooperation Council countries such as the UAE, Kuwait, and Saudi Arabia so that Iranian oil could be substituted (Kleine-Ahlbrandt, 2010). Iran’s crude oil production that was 2.8 million barrels/day in July 2011 was reduced to 0.94 million barrels/day in July 2012, resulting a sharp fall in oil export revenues from US$9.8 billion to US$2.9 billion (Coredesman et al., 2014: 21). International oil majors such as Total (France), ENI (Italy), INPEX (Japan), and Royal Dutch Shell (Britain and the Netherlands) pledged to close down their businesses with Iran.

Increasing economic isolation seems to have affected Iran’s willingness to negotiate. Crude oil exports used to take up more than 80% of total oil exports and more than 60% of government revenues in Iran: a sharp drop in oil revenues and its isolation in the international financial system significantly reduced Iran’s foreign exchange reserves, leading to a collapse of Iranian Riyals and inflation. Under economic duress, Iranians elected the most moderate candidate, Hassan Rouhani, as the president in 2013 who pledged dialogue with the USA and the rest of the world.^21,22 On 24 November 2013, Iran and P5+1²³ signed the “Joint Plan of Action” in Geneva, which required Iran to freeze or reverse progress at all of Iran’s major nuclear facilities as well as to allow a marked increase in monitoring by the IAEA. In return, Iran was to receive temporary relief on some economic sanctions.²⁴ In July 2015, the P5+1 and Iran reached a Joint Comprehensive Plan of Action that obligates Iran to fully cooperate with the international community regarding its nuclear programs and the EU and the USA to terminate the implementation of all nuclear-related economic and financial sanctions.²⁵

How does the Multilateral Sanction Equilibrium²⁶ developed in the previous section map to what we have seen after the CISADA? Increasing concerns of Iran developing the capability of producing a nuclear weapon can be translated to increased valuations of target compliance for the USA as well as the third parties (higher γ, β). With much concern about Iran’s alleged nuclear proliferation activities, the USA was committed in having a strong multilateral coalition. If there were third party states that ignored the secondary sanctions, they could substitute for those that participated and effectively reduce the cost of the multilateral sanctions for Iran. If so, not only would it be hard for the multilateral coalition to pressure Iran into greater compliance, but also the USA itself may not be able to afford the costly secondary sanctions, which in effect would reduce the credibility of the sanction threats.

Visiting Iran’s major trade partners before and after the passage of CISADA, the USA communicated its commitment. It also institutionalized the monitoring and enforcement processes that raised the domestic audience costs and pushed the US government to commit to punish non-cooperation. In the CISADA, the scope was extended to include petroleum products and later on crude oil trade expanding the set of sanctionable third parties: if the secondary sanctions were to be effective in coercing them, Iran would be under greater pressure to comply at a significantly higher level than before. Even though the expected cost that would have to be incurred by the USA for punishing non-cooperation was greater since more third parties were subjected to secondary sanctions (greater α_SR), Iran’s reliance on the goods subject to punishment was substantial (much greater α_TR); countries such as China, Japan, South Korea, and India that Iran conducted trade with were important economic partners of the USA but the levels of dependency on trade of petroleum and petroleum products were so high for Iran that the expected gain (a higher level of compliance by Iran) from a successful sanction campaign was much greater than the cost that the USA would have to incur if it were to punish non-cooperation. The USA could now credibly raise the level of punishment threats against the third parties. The USA also tried to lower the third party cost of joining in the multilateral sanctions by brokering deals between oil producing states in the Gulf and the states that had to reduce oil imports from Iran. Such reduction of cost of secondary sanctions for third parties lowered the minimum secondary punishment threats to coerce them to join the multilateral sanction campaign, inducing more third parties to participate.

Almost all of the third parties of concern joined in the sanction campaign against Iran after 2010. Not only were the secondary sanction threats credible, but they were high enough to induce third parties to cooperate, ultimately contributing to pressuring Iran to come to the negotiation table and agree on an accord, reaching a Multilateral Cooperation Equilibrium.

What can we say about the welfare impacts of CISADA? The welfare of the third parties such as the EU that were very much concerned with Iran’s alleged nuclear proliferation activities and voluntarily participated in the sanction campaign since 2010 (higher β) is likely to have improved owing to CISADA because without it coercing the other states to join the campaign, it might have taken a longer time for Iran to come to the negotiation table or Iran might not have accepted an agreement with such rigid restrictions on its nuclear activities.

The welfare of the third party states such as India, Japan, South Korea that were moderately concerned with Iran’s compliance is likely to have improved as well. These are the states that would not have voluntarily participated in the multilateral sanction campaign without a biting secondary sanction threat even though they would have been better off with a higher level of compliance by Iran. This is because once Iran offers a suboptimal compliance level and the US initiates a sanction campaign, as a follower, the best response for them is to continue their trade with Iran and to backfill for the participating states. Knowing this, Iran will not comply at a higher level. With CISADA imposed, however, the cost of trading with Iran significantly increased for them: the returns were no longer great enough to make up for being sanctioned by the USA as well as for the loss of utility from a suboptimal level of Iran’s compliance. Owing to secondary sanctions, the third party states could credibly threaten Iran with their participation in the multilateral sanction campaign and Iran’;s costs of low compliance were effectively raised. CISADA resulted in a higher compliance level and led to improved welfare for the third parties that were coerced into participating in the sanction campaign.

For third party states that do not have much concern regarding Iran’s compliance or may even derive negative utility from it, not only would they not have voluntarily participated in the sanction campaign against Iran without a punishment threat from the USA, but with CISADA successfully coercing all third parties to participate, its welfare would worsen since an increase in Iran’s compliance would push down the payoffs.

Given such welfare consequences for different group of third party states, if most of the third parties consider Iran’s greater compliance as the first best, then the secondary sanctions that paved the way to Iran’;s higher compliance have been welfare improving for the third parties as a whole. If most of them do not consider Iran’s greater compliance as the first best, then the secondary sanctions have been welfare worsening.

Conclusion

This paper studies the structure of secondary sanctions through a simple game theoretic framework and a case study and analyzes their effectiveness as a mechanism to overcome the commitment and enforcement problems that arise in sustaining multilateral sanction campaigns. I explore the conditions under which the secondary sanction can coerce the third parties to participate and its welfare consequences.

I have identified a Backfill Equilibrium and a Multilateral Cooperation Equilibrium which are consistent with the outcomes of the US secondary sanctions targeting Iran. After the ILSA, the world was in a Backfill Equilibrium where there was no agreement reached on how much Iran should comply with the demands of the USA, the USA sanctioned and the others continued to trade with Iran. After the CISADA and a series of secondary sanctions, the world is now at a Multilateral Cooperation Equilibrium with an agreement on Iran’s nuclear non-proliferation. The USA raised the cost of non-cooperation for the third party states by increasing the level of punishments for non-compliance as well as strengthening monitoring and enforcement. It also reduced the cost of cooperation for the third party states by brokering deals between them and crude oil exporting states so that they could substitute Iranian oil. Both efforts effectively lowered the minimum required punishment threat for the third party to participate in the multilateral sanction campaign. The heightened concern for Iran’s alleged nuclear proliferation activities as well as a greater expected compliance level owing to the enlarged scope of sanctions in case of success raised the stakes of a strong multilateral sanction front and led the USA to willingly bear the costs of carrying out the secondary sanctions. As the USA successfully coerced more third party states to cooperate, lower compliance became more costly to Iran because it was not able to find other states to trade with. The secondary sanctions led Iran to agree to a higher compliance level.

With the developed framework, I was able to show that if for the third parties that moderately or more value Iran’s compliance, regardless of whether they willingly participated in the multilateral sanction campaign or not, in the end, they must have become better off: the USA actually did a service to the third parties by coercing them to participate in the multilateral sanction campaign. For those third parties that do not value Iran’s compliance, then owing to the US secondary sanctions they must have become worse off.

Footnotes

Appendix

Acknowledgements

A preliminary version of the paper was published as a KIEP working paper 12–03. I thank Kaushik Basu, Myeonghwan Cho, Ram Dubey, Benjamin Ho, Jeongmeen Seo, Chrysostomos Tabakis, the seminar participants at IPES 2011 as well as four anonymous reviewers for constructive comments. All remaining errors are mine.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Abreu

(1998) On the theory of infinitely repeated games with discounting. Econometrica56(2): 383–396.

Atzili

Pearlman

(2012) Triadic deterrence: Coercing strength, beaten by weakness. Security Studies21(2): 301–335.

Bahgat

(2003) American Oil Diplomacy in the Persian Gulf and the Caspian Sea. Gainesville, FL: University Press of Florida.

Baldwin

(1985) Economic Statecraft. Princeton, NJ: Princeton University Press.

Bapat

Morgan

(2009) Multilateral versus unilateral sanctions reconsidered: a test using new data. International Studies Quarterly53: 1075–1094.

Basu

(1986) One kind of power. Oxford Economic Papers38(2): 259–282.

Cirone

Urpelainen

(2013) Trade sanctions in international environmental policy: Deterring or encouraging free riding? Conflict Management and Peace Science30(4): 309–334.

Clawson

(2010) U.S. sanctions. In: Wright

(ed.), The Iran Primer; Power, Politics, and U.S. Policy. Washington, DC: USIP Press, pp. 115–118.

Cordesman

Gold

Coughlin-Schulte

(2014) Iran—Sanctions, energy, arms control, and regime change. Report of the CSIS Burke Chair in Strategy. Center for Strategic and International Studies. Available from: http://csis.org/files/publication/140122_Cordesman_IranSanctions_Web.pdf (accessed 6 January 2015).

10.

Dorussen

(2001) Ending economic sanctions: Audience costs and rent-seeking as commitment strategies. Journal of Conflict Resolution45(4): 395–426.

11.

Drezner

(2000) Bargaining, enforcement and multilateral sanctions: when is cooperation counterproductive? International Organization54(1): 73–102.

12.

Early

(2012) Alliances and trade with sanctioned states: A study of U.S. economic sanctions, 1950–2000. Journal of Conflict Resolution56(3): 547–572.

13.

Eaton

Engers

(1992) Sanctions. Journal of Political Economy100(2): 899–928.

14.

Eaton

Engers

(1999) Sanctions, some simple analytics. American Economic Review89(2): 409–414.

15.

Elliott

(1998) The Sanction glass: Half full or completely empty? International Security23(1): 50–65.

16.

EU (2015) Joint Statement by EU High Representative Federica Mogherini and Iranian Foreign Minister Javad Zarif Switzerland. Available from: http://eeas.europa.eu/statements-eeas/2015/150402_03_en.htm (accessed 10 April 2015).

17.

Gilpin

(1984) Structural constraints on economic leverage: Market-Type systems. In: McCormick

Bissell

(eds), Strategic Dimensions of Economic Behavior. NewYork: Praeger, pp. 105–128.

18.

Hatlebakk

(2002) A new and robust subgame perfect Nash equilibrium in a model of triadic power relations. Journal of Development Economics68(3): 225–232.

19.

Hufbauer

Kimberly

Tess

Winston

(1997) US economic sanctions: Their impact on trade, jobs, and wages. Working Paper Special April1997. Peterson Institute for International Economics.

20.

Hufbauer

Schott

Kimberly

Oegg

(2007) Economic Sanctions Reconsidered, 3rd edn.Washington, DC: Institute for International Economics.

21.

Hurst

(2012) Obama and Iran. International Politics49: 545–567.

22.

Iida

(1993) When and how do domestic constraints matter? Two level games with uncertainty. Journal of Conflict Resolution37(3): 403–426.

23.

IAEA (2009) Implementation of the NPT Safeguards Agreement and relevant provisions of Security Council resolutions 1737 (2006), 1747 (2007), 1803 (2008) and 1835 (2008) in the Islamic Republic of Iran. GOV/2009/74, pp. 1–7. Available from: https://www.iaea.org/sites/default/files/gov2009-74.pdf (accessed 2 February 2015).

24.

Kaempfer

Lowenberg

(1999) Unilateral versus multilateral international sanctions: A public choice perspective. International Studies Quarterly43(1): 37–58.

25.

Katzman

(2015) Iran sanctions. CRS Report for Congress. Available from: http://www.fas.org/sgp/crs/mideast/RS20871.pdf (accessed 10 April 2015).

26.

Kim

(2010) Realism and liberalism in economic sanctions: An analysis of South Korea’s sanctions on Iran in 2010. In: SAIS US–Korea 2010 Yearbook, pp. 35–51. Available from: http://uskoreainstitute.org/wp-content/uploads/2011/05/SKim_YB2010.pdf. (accessed 5 January 2015).

27.

Kleine-Ahlbrandt

(2010) Why China drags its feet on UN sanctions against Iran. In: Un-sanctioning Iran: What the Nuclear Deal Means for the Future of Sanctions. International Crisis Group. Available from: http://www.crisisgroup.org/en/regions/asia/north-east-asia/china/kleine-ahlbrandt-why-china-drags-its-feet-on-un-sanctions-against-iran.aspx (accessed 20 September 2015).

28.

Langlois

(2010) Costly interference: A game theoretic analysis of sanctions. Peace Economics, Peace Science and Public Policy16(1): 1–34.

29.

Maloney

(2015) Un-sanctioning Iran: What the nuclear deal means for the future of sanctions. In: Markaz Middle East Politics and Policy. Available from: http://www.brookings.edu/blogs/markaz/posts/2015/08/03-iran-sanctions-future-nuclear-deal-maloney (accessed 20 September 2015).

30.

Martin

(1992) Coercive Cooperation: Explaining Multilateral Economic Sanctions. Princeton, NJ: Princeton University Press.

31.

Martin

(1993) Credibility, costs, and institutions: cooperation on economic sanctions. World Politics45(3): 406–432.

32.

McLean

Whang

(2010) Friends or foes? Major trading partners and the success of economic sanctions. International Studies Quarterly54: 427–447.

33.

Meyer

(2009) Second thoughts on secondary sanctions. University of Pennsylvania Journal of International Law30: 905–968.

34.

(1994) The logic of two-level games with endogenous domestic coalitions. Journal of Conflict Resolution38(3): 402–422.

35.

Naqvi

Wemhoner

(1995) Power, coercion, and the games landlords play. Journal of Development Economics47(2): 191–205.

36.

Neff

(1990) Friends but No Allies: Economic Liberalism and the Law of Nations. New York: Columbia University Press.

37.

Nephew

(2015) A sanctions approach to “Plan B” for the Iran nuclear problem. Center on Global Energy Policy Issue Brief March 2015. Available from: http://energypolicy.columbia.edu/sites/default/files/energy/Issue%20Brief_A%20Sanctions%20Approach%20to%20%22Plan%20B%22%20for%20the%20Iran%20Nuclear%20Problem_March%202015.pdf (accessed 25 September 2015).

38.

Osborne

Rubinstein

(1994) A Course in Game Theory. Cambridge, MA: MIT Press.

39.

O’Sullivan

(2010) Iran and the great sanctions debate. The Washington Quarterly33(4): 7–21.

40.

Patterson

(2013) EU Sanctions on Iran: The European political context. Middle East Policy20(1): 135–146.

41.

Rodman

(2001) Sanctions Beyond Borders: Multinational Corporations and U.S. Economic Statecraft. Lanham, MD: Rowman & Littlefield.

42.

Schelling

(1960) The Strategy of Conflict. Cambridge, MA: Harvard University Press.

43.

Sherrill

(2014) Why Hassan Rouhani won Iran’s 2013 presidential election. Middle East Policy21(2): 64–75.

44.

Smith

(1996) The success and use of economic sanctions. International Interactions21(3): 229–245.

45.

UN Security Council (2006) United Nations Security Council Resolution 1737. Available from: http://www.un.org/ga/search/view_doc.asp?symbol=S/RES/1737(2006) (accessed 15 March 2015).

46.

UN Security Council (2008) United Nations Security Council Annual Report S/2008/839. Available from: http://www.un.org/ga/search/view_doc.asp?symbol=S/2008/839 (accessed 15 March 2015).

47.

UN Security Council (2009) United Nations Security Council Annual Report S/2009/688. Available from: http://www.un.org/ga/search/view_doc.asp?symbol=S/2009/688 (accessed 15 March 2015).

48.

Yang

Askari

Forrer

Zhu

(2009) How do US economic sanctions affect EU’s trade with target countries? World Economy32(8):1223–1244.