Coercive Diplomacy and Economic Sanctions Reciprocity: Explaining Targets’ Counter-Sanctions

ABSTRACT

Though reciprocity is an important aspect of coercive diplomacy, little is known about whether and when sanctioned countries (i.e., targets) respond to foreign pressure with their own counter-sanctions. The purpose of this article is to offer a comprehensive analysis of the conditions under which targets are more likely to employ economic counter-measures against their senders. Analyzing data for sanctions reciprocity episodes in the Threats and Imposition of Economic Sanctions (TIES) dataset, we find that targets with wealthier economies, less democratic regimes, or higher trade dependence on their senders are more likely to initiate reciprocal sanctions. Our findings also denote that sanctions reciprocity is more likely when targets are subject to sanctions by senders with poor economies or when the issue that instigates the initial sanctions is less salient. As the first cross-national, quantitative analysis of sanctions reciprocity, our analysis provides a more complete picture of how strategic ties between senders and targets unfold, and why some sanctions are more likely to fail or result in stalemate due to counter-sanctions employed by targets.

KEYWORDS:

• Economic sanctions
• coercive diplomacy
• counter-sanctions

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Hufbauer, Jeffrey Schott, and Oegg (2007) contend that targets facing political turmoil are more prone to capitulate to sanctions. However, putting their data to systematic multivariate analyses, existing studies find no consistent evidence for this argument (see Drury 1998; Hart 2000; Ang and Dursun 2007).

2. It is possible that hostility instigated by initial sanctions fades away as time goes by, which implies that follow-up sanctions by targets long after the onset of initial sanctions may have nothing to do with reciprocity. We estimated how long initial sanctions incentivize targets to reciprocate and found that such an effect diminishes after 10 years since the onset of initial sanctions and disappears approximately after 16 years. In our sample, very few reciprocal sanctions are used 10 years after the onset of initial sanctions. More specifically, five reciprocal sanctions are initiated 10 years after the onset of initial sanctions and only two after 16 years. Further, our key findings do not change significantly in the models estimated using only the counter-sanctions cases occurred within 10-years following the initial sanctions (see Tables A4 through A6 in the online appendix).

3. Although the result in Figure A1 denotes such an effect begins to attenuate after approximately 10 years, there is only one overlapping initial sanction case that remains 10 years after its onset.

4. The TIES dataset contains 1,412 different sanctions episodes, including the episodes involving international organizations (IOs). Fortunately, the dataset designates one of the sender countries in most IO-led sanctions as the primary sender. In the IO sanctions cases where a primary sender country exists, we considered that country as the sender and excluded the cases without a designated primary sender from the analysis. We also dropped the cases where the primary target is an international institution in the TIES.

5. We also considered estimating separate models for threats and imposition of reciprocal sanctions. Though we expect that similar factors are likely to drive targets’ decision to issue threats of sanctions and/or impose them, it is still an empirical question worth exploring. However, threats and actual sanctions are implemented in the same year in a significant portion of reciprocal sanctions cases in our data. Once such cases are excluded, only few threat and imposed reciprocations remain in the full model, which renders it difficult to make valid interpretations of the results for reciprocal sanctions threats and imposition.

6. We use the rivalry data from 2002 for the years after 2002 (i.e., 2003-2005) because the data for this variable from Klein, Goertz, and Diehl (2006) are available until 2002.

7. The mixed findings for imposed sanctions might be attributed to the binary nature of the imposed sanctions variable as it does not capture the association between the cost of initial sanctions and targets’ reciprocity. Hence, we estimate additional models that control for the sanctions cost and found no significant evidence that the severity of initial sanctions increases the probability of reciprocal sanctions (See Table A7 in the appendix).

8. The substantive effects are predicted based on the model where the democratic dyad variable is dropped as it is statistically insignificant. In doing so, we confirm that the substantive impact of target regime type does not pick up the impact of insignificant democratic dyad variable. Even after excluding the democratic dyad variable, the estimates for all the other variables remain unaltered.

9. The substantive effects do not change significantly when we calculate them based on the full model in Table 2 (See Figure A2 in the appendix).