Bapat, Heinrich, Kobayashi and Morgan (2013) used the Threat and Imposition of Economic Sanctions (TIES) data and an extreme-bounds-style robustness analysis to establish which factors correlate with sanctions success, finding that involvement of international institutions and severe costs on the target are robustly associated with success at both the threat and the imposition stage, while several factors change sign or significance between stages. They explicitly declined to model the selection process that connects the two stages and called for theory. The open problem is to construct and empirically validate a causal theory of sanctions effectiveness in which selection across the threat and imposition stages is explicit: a formal model whose equilibria reproduce the stage-general and stage-specific patterns, identifying the mechanisms behind the robust correlations, and tested with empirical designs that separate selection from behavioral mechanisms on TIES-type data. Later contributions treat single variables or single stages rather than the joint threat-to-imposition selection process, and the canonical sender-target bargaining model has so far failed its empirical tests (Morgan and Kobayashi 2021).

Background

Economic sanctions are restrictions one state (or a coalition of states) imposes on economic relations with another state - the target - to coerce a change in its policies; the state applying them is the sender. Whether and when sanctions "work" (induce the target to concede, fully or partially, or reach a negotiated settlement) is a long-standing empirical question in international relations, and decades of studies produced conflicting findings: some attributed success to high costs on the target, others to "smart" design concentrating costs on elites, to the target's regime type, or to multilateral rather than unilateral sender coalitions.

Two diagnoses of this inconsistency motivated the source study. First, model dependency: published findings typically rest on a single regression specification, so results may not survive alternative control variables or model choices. Second, selection bias: most earlier datasets recorded only sanctions that were actually imposed. But sanctions are moves in a bargaining game (Eaton and Engers 1992): a sender first threatens, and a target that expects to concede may do so before anything is imposed. Imposed sanctions are therefore a non-random, self-selected subsample of "hard" cases, which can make sanctions look less effective than they are and distort every estimated determinant (Nooruddin 2002 modeled this selection bias explicitly; Drezner 2003 argued the threat stage systematically hides successes). The Threat and Imposition of Economic Sanctions (TIES) dataset (Morgan, Bapat and Krustev 2009) was built to repair this: it records 888 cases from 1971-2000 in which sanctions were threatened and/or imposed, allowing analysis of the threat stage (cases ending in target concessions before imposition) and the imposition stage separately.

Bapat, Heinrich, Kobayashi and Morgan (2013) combined TIES with a systematic robustness analysis: they took eighteen factors the literature had proposed as determinants of success, estimated every logistic regression obtainable from combinations of those covariates - 262,143 specifications - and asked which coefficients keep sign and significance across them (an approach of the kind introduced by Leamer and Sala-i-Martin in economics). Running this separately for pooled cases, threat-stage successes, and imposition-stage successes, they found that exactly two factors are robust at every stage: involvement of international institutions (participation of bodies such as the United Nations or regional organizations in the sanctioning effort) and severe economic costs imposed on the target. Other factors are fragile, and several reverse between stages: financial sanctions are associated with failure at the threat stage but success at imposition; involvement of multiple issues in the dispute is negative at the threat stage but positive at imposition; offers of positive inducements ("carrots", e.g. aid or concessions offered alongside threats) help at the threat stage, while imposition-stage carrot cases appear to be selected "hard" cases where the carrot had already failed. The criterion is distributional: the results are distributions of coefficient estimates and t-statistics over those specifications, a factor counts as robust when essentially all of them keep sign and significance, and as weakly robust when the estimates are consistently signed but about half of the t-statistics fall below the significance threshold. On that criterion high issue saliency is robust at the threat stage but ambiguous at imposition; target trade dependence is positively and fairly robustly related to success for imposed sanctions but ambiguous for threats, which the authors explicitly name as an interesting puzzle; and democratic target, export restrictions and United States sender are ambiguous at imposition yet clearer and more robust at the threat stage.

The paper's own suggested interpretation is that these stage differences may be fingerprints of selection - the authors write that "one reason that accounts for these results may be a selection effect," and the abstract reports stage differences that "suggest[] specific selection effects that should be explored in future work" - since the cases that reach imposition are those in which the target already refused to concede to the threat, so stage-by-stage regressions on non-random subsamples conflate selection with behavior. Crucially, the authors state explicitly that they do not model the selection process between threats and impositions and that modeling the entire process is beyond their scope; at the close of their stage comparison they ask future theoretical work to "explicitly consider possible selection mechanisms and address why these factors but not others may matter at different stages," while their conclusion calls for further theoretical as well as empirical investigation of the mechanisms by which these factors contribute to success and flags measurement uncertainty, choice of statistical model, and unexplored interactions (for example between multiple issues, institution involvement, and multilateral sanctions) as sources of uncertainty for future work. The paper itself estimates only unconditional additive effects. Selection-corrected designs, which model jointly which cases are observed at each stage and what outcome follows, are an established method class in this literature: Nooruddin (2002) modeled the selection bias explicitly, and the later selection-aware studies of sanctions outcomes are surveyed in the introduction of Morgan and Kobayashi (2021). A decade of subsequent work has extended the data (Morgan, Bapat and Kobayashi 2014), tested mechanisms for individual variables (Early and Spice 2015; Walentek et al. 2021), and applied alternative model classes (Caetano, Galego and Caleiro 2023), but the joint theory of the threat-to-imposition selection process and its selection-corrected empirical validation remain missing.

Problem Statement

The problem, as posed by Bapat, Heinrich, Kobayashi and Morgan (2013) at the close of their stage comparison and in their conclusion, is to move from their robust correlations to a causal and theoretical understanding of when and why economic sanctions succeed, by making the selection dynamics that connect the threat and imposition stages explicit. Concretely, the open program is to: (1) construct a formal theory of the sanctioning process - a game-theoretic or otherwise explicit model in which the sender's threat, the target's response, and the sender's decision to impose are strategic, so that the cases observed at the imposition stage are endogenously selected rather than taken as given; (2) show that the theory accounts for the documented stage patterns and addresses why these factors but not others matter at different stages: why international-institution involvement and severe target costs are associated with success at both stages, while financial sanctions, involvement of multiple issues, and carrots change sign or strength between stages (factors the source reports as weakly robust - consistently signed but with specification-dependent significance - rather than robust in the strict sense), why high issue saliency is robust at the threat stage but ambiguous at imposition, why target trade dependence goes with success at the imposition stage but not at the threat stage, and how conditional (interaction) structure across institutional and issue contexts modifies these effects; (3) identify the causal mechanisms behind the robust correlations - for example whether institutional involvement works through enhanced sender credibility, legitimacy, cost generation, or coercion of the target; and (4) test the theory with empirical designs that separate selection from behavioral mechanisms in the joint threat-to-imposition process, using TIES-type data that cover both stages. The four components are interdependent: accounting for the stage patterns is what connects a formal theory to the source's evidence, and the empirical validation is what distinguishes a selection explanation from a behavioral one. The program is resolved by a formal selection model that accounts for the documented stage patterns and identifies the mechanisms behind the robust correlates, together with an empirical test that separates selection from behavioral mechanisms; the patterns listed in (2) define the scope the theory must account for.

References

  1. Navin A. Bapat, Tobias Heinrich, Yoshiharu Kobayashi, T. Clifton Morgan (2013). Determinants of Sanctions Effectiveness: Sensitivity Analysis Using New Data. International Interactions 39(1):79-98. DOI: 10.1080/03050629.2013.751298. URL: https://doi.org/10.1080/03050629.2013.751298
  2. T. Clifton Morgan, Navin Bapat, Valentin Krustev (2009). The Threat and Imposition of Economic Sanctions, 1971-2000. Conflict Management and Peace Science 26(1):92-110. DOI: 10.1177/0738894208097668. URL: https://doi.org/10.1177/0738894208097668
  3. T. Clifton Morgan, Navin Bapat, Yoshiharu Kobayashi (2014). Threat and imposition of economic sanctions 1945-2005: Updating the TIES dataset. Conflict Management and Peace Science 31(5):541-558. DOI: 10.1177/0738894213520379. URL: https://doi.org/10.1177/0738894213520379
  4. Peter A. G. van Bergeijk, Muhammad S. H. Siddiquee (2017). Biased Sanctions? Methodological Change in Economic Sanctions Reconsidered and Its Implications. International Interactions 43(5):879-893. DOI: 10.1080/03050629.2016.1242584. URL: https://doi.org/10.1080/03050629.2016.1242584
  5. Bryan R. Early, Robert Spice (2015). Economic Sanctions, International Institutions, and Sanctions Busters: When Does Institutionalized Cooperation Help Sanctioning Efforts? Foreign Policy Analysis 11(3):339-360. DOI: 10.1111/fpa.12038. URL: https://doi.org/10.1111/fpa.12038
  6. Dawid Walentek, Joris Broere, Matteo Cinelli, Mark M. Dekker, Jonas M. B. Haslbeck (2021). Success of economic sanctions threats: coercion, information and commitment. International Interactions 47(3):417-448. DOI: 10.1080/03050629.2021.1860034. URL: https://doi.org/10.1080/03050629.2021.1860034
  7. T. Clifton Morgan, Yoshiharu Kobayashi (2021). Talking to the hand: Bargaining, strategic interaction, and economic sanctions. European Economic Review 134:103685. DOI: 10.1016/j.euroecorev.2021.103685. URL: https://doi.org/10.1016/j.euroecorev.2021.103685
  8. Gabriel Felbermayr, T. Clifton Morgan, Constantinos Syropoulos, Yoto V. Yotov (2021). Understanding economic sanctions: Interdisciplinary perspectives on theory and evidence. European Economic Review 135:103720. DOI: 10.1016/j.euroecorev.2021.103720. URL: https://doi.org/10.1016/j.euroecorev.2021.103720
  9. Bryan R. Early, Menevis Cilizoglu (2020). Economic Sanctions in Flux: Enduring Challenges, New Policies, and Defining the Future Research Agenda. International Studies Perspectives 21(4):438-477. DOI: 10.1093/isp/ekaa002. URL: https://doi.org/10.1093/isp/ekaa002
  10. Yewon Kwon (2024). The Double-Edged Sword of Carrots: Exploring Their Simultaneous Effects on Threat and Imposition Stages of Economic Sanctions. International Interactions 50(5):723-749. DOI: 10.1080/03050629.2024.2374363. URL: https://doi.org/10.1080/03050629.2024.2374363
  11. Jose Caetano, Aurora Galego, Antonio Caleiro (2023). On the Determinants of Sanctions Effectiveness: An Empirical Analysis by Using Duration Models. Economies 11(5):136. DOI: 10.3390/economies11050136. URL: https://doi.org/10.3390/economies11050136
  12. Gabriel Felbermayr, Aleksandra Kirilakha, Constantinos Syropoulos, Erdal Yalcin, Yoto V. Yotov (2020). The global sanctions data base. European Economic Review 129:103561. DOI: 10.1016/j.euroecorev.2020.103561. URL: https://doi.org/10.1016/j.euroecorev.2020.103561
  13. Jiakun Jack Zhang, Spencer Shanks (2024). Measuring Chinese economic sanctions 1949-2020: Introducing the China TIES dataset. Conflict Management and Peace Science 42(3):328-352. DOI: 10.1177/07388942241248274. URL: https://doi.org/10.1177/07388942241248274
  14. Patrick M. Weber, Gerald Schneider (2020). Post-Cold War sanctioning by the EU, the UN, and the US: Introducing the EUSANCT Dataset. Conflict Management and Peace Science 39(1):97-114. DOI: 10.1177/0738894220948729. URL: https://doi.org/10.1177/0738894220948729
  15. Binyam A. Demena, Alemayehu S. Reta, Gabriela Benalcazar Jativa, Patrick B. Kimararungu, Peter A. G. van Bergeijk (2021). Publication bias of economic sanctions research: a meta-analysis of the impact of trade linkage, duration and prior relations on sanctions success. Research Handbook on Economic Sanctions, Edward Elgar Publishing. DOI: 10.4337/9781839102721.00012. URL: https://doi.org/10.4337/9781839102721.00012
  16. Omer Zarpli (2022). When Do Imposed Sanctions Work? The Role of Target Regime Type. Journal of Conflict Resolution 67(7-8):1482-1509. DOI: 10.1177/00220027221139809. URL: https://doi.org/10.1177/00220027221139809
  17. Irfan Nooruddin (2002). Modeling Selection Bias in Studies of Sanctions Efficacy. International Interactions 28(1):59-75. DOI: 10.1080/03050620210394. URL: https://doi.org/10.1080/03050620210394
  18. Daniel W. Drezner (2003). The Hidden Hand of Economic Coercion. International Organization 57(3):643-659. DOI: 10.1017/s0020818303573052. URL: https://doi.org/10.1017/s0020818303573052
  19. Jonathan Eaton, Maxim Engers (1992). Sanctions. Journal of Political Economy 100(5):899-928. DOI: 10.1086/261845. URL: https://doi.org/10.1086/261845