The Double-Edged Sword of Carrots: Exploring Their Simultaneous Effects on Threat and Imposition Stages of Economic Sanctions

Abstract

This article investigates the effectiveness of incentive-based policies, or “carrots,” during the threat and imposition stages of sanctions, with a focus on addressing potential endogeneity arising from self-selection processes in designing sanctions strategy. Successful sanctions require the sender’s sustained commitment until the target concedes, demonstrating a consistent policy stance and a readiness to bear associated costs. Carrots utilized during the threat stage represent a tit-for-tat approach, complementing the sender’s consistent policy stance of conditionally combining both carrots and sticks, and potentially reducing the perception of bluffing, thereby increasing target compliance probability. However, introducing carrots during the imposition stage may be misinterpreted as a sudden shift towards appeasement, undermining policy coherence and indicating a lack of capacity or willingness to enforce sanctions. This perception could lead the target to await sanctions removal, resulting in a failed outcome. Following the mitigation of potential sender self- selection into carrots, the findings emphasize the importance of combining carrots with sanctions threats. Nonetheless, caution is advised when using carrots as a substitute for ineffective punitive strategies, as adverse effects may arise.

RESUMEN

Este artículo investiga la efectividad de las políticas basadas en incentivos, o “zanahorias”, durante las etapas de amenaza e imposición de sanciones, con un enfoque consistente en abordar la posible endogeneidad que surge de los procesos de autoselección durante el diseño de la estrategia de sanciones. El éxito de las sanciones requiere el compromiso sostenido por parte del emisor hasta que el destinatario ceda, ya que esto demuestra una postura política coherente y una disposición a asumir los costes asociados. Las zanahorias que se utilizan durante la etapa de amenaza representan un enfoque de ojo por ojo, que complementa la postura política coherente por parte del emisor consistente en combinar, condicionalmente, tanto las zanahorias como los palos (castigos), y reduce potencialmente la percepción en lo referente a faroles, aumentando así la probabilidad de cumplimiento del objetivo. Sin embargo, la introducción de zanahorias durante la etapa de imposición puede interpretarse, erróneamente, como un cambio repentino hacia la conciliación, lo que menoscaba la coherencia de las políticas e indica una falta de capacidad o de voluntad para hacer cumplir las sanciones. Esta percepción podría llevar al destinatario a esperar el levantamiento de las sanciones, lo que provocaría un fracaso a nivel de resultado. Tras reducir el posible sesgo de la autoselección de las zanahorias por parte del agente que impone las sanciones, nuestras conclusiones enfatizan la importancia de combinar zanahorias con amenazas de sanciones. No obstante, se recomienda precaución al usar zanahorias como sustituto de estrategias punitivas ineficaces, ya que estas pueden provocar efectos adversos.

Résumé

Cet article étudie l’efficacité des politiques incitatives, ou « carottes », pendant les phases de menace et de l’application de sanctions. Il met l’accent sur la prise en compte de l’endogénéité potentielle qui découle des processus d’autosélection dans la conception de la stratégie de sanctions. Pour que les sanctions soient efficaces, il faut que l’émetteur s’engage durablement jusqu’à ce que la cible cède, en faisant preuve d’une position politique cohérente et d’une volonté de supporter les coûts associés. Les « carottes » utilisées au cours de la phase de menace représentent une approche « œil pour œil, dent pour dent », complétant la position politique cohérente de l’émetteur. Cette approche combine conditionnellement la technique de la carotte et celle du bâton ; elle réduit potentiellement la perception de bluff et augmente ainsi la probabilité que la cible se plie aux sanctions. Toutefois, l’introduction de carottes au cours de la phase d’application des sanctions peut être interprétée à tort comme un changement soudain en faveur de l’apaisement, ce qui nuit à la cohérence de la politique et indique un manque de capacité ou de volonté d’appliquer les sanctions. Cette perception pourrait conduire la cible à attendre la levée des sanctions, ce qui signifierait un échec. Après l’atténuation de l’autosélection potentielle de l’émetteur en faveur de la carotte, les résultats soulignent l’importance de combiner la technique de la carotte avec des menaces de sanctions. Néanmoins, la prudence est de mise lorsqu’il s’agit d’utiliser les carottes comme substitut à des stratégies punitives inefficaces, car des effets négatifs peuvent survenir.

Keywords:

• Carrot-and-stick
• policy consistency
• self-selection based endogeneity
• threat and imposition stages of sanctions

Palabras clave:

• Estrategia de zanahoria y palo
• endogeneidad basada en la autoselección
• etapas de amenaza e imposición de sanciones
• coherencia de las políticas

Mots-clés :

• La carotte et le bâton
• endogénéité basée sur l’autosélection
• phases de menace et d’application de sanctions
• cohérence politique

Notes

1 In this paper, “carrots” are used conditionally in exchange for the target’s concessions.

2 In cases where explicit threats were not made and sanctions are directly imposed, the incurrence of actual and cumulative sanctions costs reveals the sender’s incentive to quickly end the sanctions. Understanding these incentives, the target might interpret the sender’s carrots as a sign of weakness. Further explanation of cases without explicit sanctions is provided in the theory section.

3 The sudden change in policy direction could be interpreted as undermining the sender’s firm resolve to enforce demands, such as weakening the domestic support base for the sanctions policy itself, thus potentially undermining credibility in the sender’s implementation of sanctions.

4 The theory explores the potentially different nature of incurred costs between the threat and imposition stages, specifically focusing on the existence of actual costs. In the threat stage, costs are not yet incurred but represent future virtual expenses, whereas in the imposition stage, costs are both real and accumulate over time. Consequently, the theory posits that the sender’s motivation to promptly terminate sanctions is more pronounced in the imposition stage due to the presence of actual and cumulative costs, contrasting with the threat stage where actual costs have yet to materialize.

5 Note that the principle of policy consistency may not apply in cases where sanctions are imposed without explicit threats. However, the argument regarding the negative impact of carrots during the imposition stage can still hold in such cases. This is because carrots during the imposition stage may undermine the perceived commitment of the sender, due to the actual cumulative costs incurred. In this case, the target may perceive that the sender prefers to swiftly end the sanctions by offering compensation, rather than continuing the punitive measures and increasing the cumulative cost of imposition. Therefore, if the target continues to resist while the sender in- creases the cumulative cost of sanctions, the target may expect the sender to lift the sanctions and withdraw first. Consequently, the sender’s use of carrots may showcase its weak commitment to continuing the sanctions. The robustness check, excluding cases without explicit threats, is reported in Supplementary Appendix. The results consistently show a negative and statistically significant coefficient, along with a marginal effect of carrots in the imposition stage.

6 If the sender’s expected payoff from bluffing is higher than the size of the audience cost for not living up to the sanctions threat, the sender has an incentive to issue bluffing. Therefore, if the expected payoff from bluffing is sufficiently high, the likelihood of bluffing can also be increased. Let’s consider two senders operating under conditions where the exact costs of sanctions are not fully observable to the target: Sender 1, employing a threat- only strategy, and Sender 2, utilizing a combined strategy of carrots and threats. Assuming both senders face identical audience costs and probabilities of threat success, the payoff for Sender 1’s threat resembles a pure pie with minimal costs. Therefore, even in scenarios where the sender faces a fifty-fifty probability of success, there exists an incentive to pursue a nearly cost-free victory by relying on bluffing, given the significant potential benefits of sanction success. In contrast, Sender 2 incurs costs in the form of compensation for threat success. Relative to Sender 1, the potential benefit of sanction success diminishes due to the costs associated with conditional rewards for the target’s compliance. This implies that the expected payoff for bluffing is comparatively lower under identical success probabilities, resulting in a reduced incentive for bluffing in the latter scenario.

7 Some studies indicate that sender states design sanctioning policies to incur only moderate costs on themselves and their domestic groups (e.g. McLean and Whang (2014)). However, as the duration of sanctions imposition prolongs, the cumulative costs of sanctions are expected to escalate, even if these costs are relatively minor. This extended period can impose a burden on the sender. Moreover, as sanctions continue, domestic political opposition within the sender may intensify. Even if the actual cost of sanctions is relatively low, the domestic public or opposition party may blame the prolonged sanctions for deadlock fatigue. Considering these cumulative costs and domestic political ramifications, it is acknowledged that the costs incurred during the imposition stage will mount as the duration of sanctions extends. Based on this rationale, it is argued that the utilization of carrots during the imposition stage can be interpreted as indicative of the sender’s intent to alleviate cumulative costs by expediting the lifting of sanctions.

8 When sanctions are unilaterally lifted by the sender, it can result in significant costs, such as domestic political criticism and a decline in international political reputation. In such cases, the sender may attempt to mitigate the negative consequences associated with the termination of sanctions by offering incentives (carrots), thereby creating the perception of extracting concessions from the target, even if only superficially.

9 For example, despite significant loans being withheld by both the World Bank and Japan for China’s economic development during the period of sanctions, the United States proposed reinstating these loan programs, contingent upon China allowing the dissident Fang Lizhi to leave the country. Additionally, inducements were extended to the Chinese government to enhance its Most Favored Nation (MFN) status within the critical U.S.-China trade relationship during this period of imposition (Hoffman and Devroy 1989; Mann 1991). Consequently, China could interpret these actions as signals of a potential relaxation of sanctions by the United States.

10 In the situation of policy design, it could be realistic that policymakers opt for particular policy—in this situation, carrots—based on a few prominent variables.

11 In assessing the potential adoption of carrot policies, additional factors influencing the sender were accounted for through both matching and statistical analysis. This involved conducting matching based on variables such as the US-sender, the target’s level of trade dependence on the sender, and the duration of sanctions. The analysis consistently yielded effects aligned with the results reported in the main body of the paper. After considering the model fit (log-likelihood) after probit regression with suggested variables and the degree of balance before and after matching, it was concluded that the model presented in the main text should be reported as the main result. These supplementary robustness test results are available in the Supplementary Appendix.

12 For cases where sanctions transitioned directly to the imposition stage, treating the threat as a missing variable results in 556 cases of failure and 266 cases of success, totaling 822 cases.

13 In the Robustness Check section, I exclusively considered explicit threats as the threat stage and conducted identical statistical tests. The results are broadly consistent with the primary findings of this paper.

14 The coding rule was primarily established because if carrots are already employed at the threat stage, the target may understand that the sender utilizes both incentives and penalties as part of the sanctions strategy. Consequently, in such instances, the utilization of carrots at the imposition stage may not compromise the policy consistency of the sender.

15 Given the rarity of the variables of interest, I also conducted Penalized Maximum Likelihood Logistic Regression (Firthlogit) to check for robustness. The results from the Firthlogit analysis align consistently with those of the main analysis. Further discussions are provided in the Robustness Check section.

16 The distribution of cases for the start and end years of sanctions does not exhibit significant unevenness. The most frequently targeted country is Yugoslavia in both the threat and imposition stages of sanctions (4 times), and the distribution of targeted countries is relatively balanced. However, in terms of the sender, the United States appears to issue threats 16 times and impose sanctions 23 times, which represents a substantial difference compared to the second-place senders, Germany for threats and Russia for imposition, each occurring 4 times. To address such skewed distribution in the sender variable, I include a United States dummy in the regression model for the statistical analysis. Further discussion on the model specification follows in the explanation of control variables.

17 For more details about each category, see “https://sanctions.web.unc.edu/wp-content/uploads/sites/18834/2021/04/tiesusersmanualv4.pdf” (Morgan, Bapat, and Kobayashi 2013).

18 Additional robustness checks were conducted by including militarized action and weapons of mass destruction in the original Issue Salience variable. The statistical results remained consistent with the main findings. For more details, refer to Supplementary Appendix: Section 5.

19 For detailed information on the coding criteria, please refer to Maoz et al. (2019).

20 https://data.worldbank.org/. Accessed 08-10-2022.

21 Additional robustness checks with variables such as the target’s trade dependence and sanctions duration yield consistent results, reported in Supplementary Appendix.

22 Additional robustness tests involving other factors are detailed in Supplementary Appendix.