How international sanctions actually work — and why they usually don’t

Thirty years of research suggest sanctions rarely change the behaviour they target. Western governments keep imposing them anyway.

On February 19, 2022, Kamala Harris stood before the Munich Security Conference and delivered a sentence that carried the full weight of Atlantic alliance politics behind it. “If Russia further invades Ukraine,” she said, “the United States, together with our Allies and partners, will impose significant and unprecedented economic costs.” The phrasing was careful in its escalation — “further invades,” as though the existing occupation of Crimea and parts of the Donbas were a footnote — but the instrument itself was announced without qualification. Significant. Unprecedented. The coalition scope was implicit in the phrase “allies and partners”: this was not a unilateral threat but a promise of coordinated economic warfare on a scale not previously attempted against a G20 economy.

Five days later, Russia invaded. The sanctions followed. They were, by most technical measures, genuinely significant: freezing roughly $300 billion in Russian central bank reserves, excluding major Russian banks from SWIFT, a G7 oil price cap, extensive technology export controls, and the largest coordinated use of secondary sanctions since the Iran packages of the Obama era. The EU adopted thirteen sanctions packages over the following two years. By late 2022, over forty countries had imposed some form of economic measure on Russia.

Russia did not withdraw. It did not change its strategic calculus in any measurable way. By 2024, the war had settled into the grinding territorial attrition it has remained ever since.

None of this is surprising to researchers who study sanctions for a living, which raises a question more interesting than whether the Russia sanctions worked. Thirty years of academic literature on sanctions effectiveness has reached a contested but not flattering empirical consensus, and that consensus has been available to foreign policy establishments for most of those thirty years. Yet the gap between what the literature finds and how sanctions are publicly justified — as instruments of coercion, as mechanisms for changing the behaviour of sovereign states — has, if anything, widened. The Munich speech was not an anomaly. It was a template. And the question worth asking is not why sanctions failed on Russia, but why governments keep reaching for an instrument as if it works in the way they claim.

The Theory of the Instrument

The term “sanctions” covers three distinct mechanisms that share a political vocabulary but differ substantially in their causal logic.

Comprehensive sanctions are trade embargoes — measures affecting all or most commerce with a target state. The US embargo on Cuba, in place since 1962 and codified through the Helms-Burton Act of 1996, is the paradigm case. The theory of change is direct: deny the target economy sufficient trade that it contracts, producing hardship broad enough to force the government to alter its behaviour or collapse. The instrument has historical roots in the League of Nations’ attempt to discipline Italy after its invasion of Ethiopia in 1935 — an early and instructive failure, since oil was not included in the embargo and the measure was lifted within eight months.

Targeted sanctions — sometimes called “smart” sanctions — operate differently. Rather than restricting the target economy as a whole, they direct pressure at specific individuals and entities: travel bans, asset freezes, restrictions on named companies. The theory of change is surgical: sanction the decision-makers rather than their populations, avoiding the humanitarian damage of comprehensive measures while concentrating pressure on the people with actual authority to change policy. The turn toward targeted sanctions is usually traced to a series of government-sponsored expert reviews in the late 1990s and early 2000s — the Interlaken Process, the Bonn-Berlin Process, and the Stockholm Process — which emerged from a clear-eyed acknowledgment that comprehensive sanctions had produced widespread civilian harm in Iraq and Haiti without generating the political results their architects claimed. The “smart sanctions” turn was not a refinement of the instrument. It was a tacit admission that the original instrument had structural problems that better targeting might partially address.

Secondary sanctions occupy a different legal and political category altogether. They are not directed at the primary target but at third-country entities that continue to do business with the designated state. The mechanism that gives secondary sanctions their reach is the dollar-clearing system: because most international trade settles in US dollars and US dollar transactions must pass through the US correspondent banking system, the US Treasury can threaten to cut off any financial institution anywhere in the world from dollar-clearing access if it processes transactions involving sanctioned entities. This is not a theoretical leverage; it is an operational chokepoint. The legal basis — that entities choosing to use dollar infrastructure implicitly accept US jurisdiction over those transactions — is contested in international law but has proved practically unchallengeable.

The UN Security Council provides the formal multilateral framework for sanctions under Article 41 of the Charter, which authorises measures “not involving the use of armed force” to restore international peace and security. UN Security Council sanctions carry binding force under international law and cannot be legally circumvented by member states. US and EU unilateral measures carry no such binding force — they apply legal or financial pressure on third parties who face a choice between the US or EU financial system and whatever benefits the sanctioned relationship provides.

The dollar-clearing chokepoint
Secondary sanctions derive their extraterritorial power from a single architectural fact: the US dollar is the world's dominant reserve and settlement currency, and dollar transactions must clear through the US financial system. Any bank in any jurisdiction that processes a dollar payment on behalf of a sanctioned entity risks losing access to US correspondent banking — effectively losing the ability to clear any dollar transaction globally. For most institutions, losing dollar-clearing access is an existential threat, which means the US Treasury can apply binding financial pressure on foreign banks without any formal legal authority over them. The EU's objection to this architecture is codified in its Blocking Statute, first adopted in 1996 and updated in 2018 to cover Iran-related US secondary sanctions. The statute prohibits EU operators from complying with certain extraterritorial US measures and theoretically allows EU persons to recover in EU courts damages arising from non-compliance penalties. In practice, the statute has been ineffective: EU companies faced with a choice between the US financial system and trade with Iran have overwhelmingly chosen to exit Iran. No EU company can afford to lose dollar-clearing access over any single market. The statute records an objection to US extraterritoriality; it does not neutralise it.

The Record — What the Numbers Say and What They Don’t

The foundational dataset in sanctions research is Gary Clyde Hufbauer, Jeffrey Schott, Kimberly Ann Elliott, and Barbara Oegg’s Economic Sanctions Reconsidered, now in its third edition. Covering 174 case studies representing 204 sanctions episodes from 1914 to 2000, the study found a success rate of approximately 34 percent — meaning that in roughly one-third of cases, sanctions contributed to at least partial achievement of their stated objectives. For a political instrument routinely dismissed as ineffective, 34 percent might sound almost encouraging.

Robert Pape’s 1997 challenge, published in International Security, methodically disassembled this figure. Working through the Hufbauer cases, Pape argued that the dataset included a substantial number of episodes in which military force, not sanctions, was the operative coercive variable — cases where sanctions happened to be imposed alongside military operations or credible military threats that drove compliance. When Pape restricted the analysis to cases in which sanctions operated as the independent variable, without significant accompanying military pressure, his figure for genuine sanctions success was approximately 5 percent. The disagreement between 34 percent and 5 percent is not a minor academic dispute about coding. It is a dispute about whether the instrument works at all. The Hufbauer figure is frequently cited in policy justification; the Pape figure essentially never appears in a speech by a foreign minister.

The Global Sanctions Database assembled by Kirilakha, Felbermayr, Syropoulos, Yalcin, and Yotov — updated through 2019 and covering 1,101 cases — finds aggregate success rates in the 30 to 50 percent range depending on objective type and coding criteria. But the aggregate figure obscures the more important finding: effectiveness varies dramatically by what the sanctions are trying to achieve. Sanctions targeting trade practices, modest policy adjustments, or specific regulatory behaviour perform substantially better than the aggregate suggests. Sanctions targeting regime change, nuclear weapons program elimination, or territorial concession perform substantially worse. The cases that generate political demand for sanctions — the cases involving nuclear programs, invasions, genocide, and authoritarian consolidation — are precisely the cases where the record is worst.

This is the distinction the article’s central claim depends on. When we say sanctions usually don’t work, we mean they usually don’t achieve consequential political objectives under conditions of genuine state resistance. The 34 percent figure — even accepting Hufbauer’s methodology — does not survive disaggregation by objective type.

The Hufbauer/Pape methodological dispute
The core of the dispute is how Hufbauer et al. coded cases for success and which cases they included. Pape identified several categories of error: cases coded as sanctions successes where the target had already decided to comply before sanctions were imposed (anticipatory compliance); cases where military pressure was the primary driver and sanctions were a secondary element; and cases involving such modest objectives that "success" was nearly guaranteed by any sustained engagement. Beyond the success-rate calculation, Pape challenged the causal logic of several individual codings — most notably the second Hufbauer coding of the Korean War period and several Cold War episodes where US leverage derived primarily from security commitments rather than economic pressure. Hufbauer and colleagues responded in part by acknowledging some of Pape's corrections in subsequent editions while contesting his broad reinterpretation. The substantive implication is significant: if Pape is even partially right, the policy record is not "one-third effectiveness" but something considerably more dispiriting. The 5 percent figure may itself be too low — methodological critics have argued Pape's restriction to independent sanctions effect is itself too demanding — but the directional implication of his analysis has not been convincingly refuted.

Why the Theory Fails — Link by Link

The causal chain that sanctions theory requires has three links: economic pain reaches the target; that pain translates into domestic political pressure for change; and that political pressure translates into policy concession. Each link has structural problems. The second is the weakest. All three, operating simultaneously, are what the theory requires.

The first link — that sanctions produce real economic pain — is the one the empirical literature most readily supports, though the picture is more complicated than it appears. For comprehensive sanctions to generate coercive pain, the sanctioning coalition must represent a significant share of the target’s trade and the target must lack viable alternatives. This condition has become progressively harder to satisfy as global trade has diversified and as sanctioned states have learned, over decades of exposure to US sanctions policy, to cultivate relationships with non-participating economies. Russia after February 2022 is the genuine test case, and the record must be handled with precision. The claim is not that Russia was economically unaffected. The costs were real and documented: higher transaction costs from the exclusion of Russian banks from SWIFT, oil revenue impacts from the G7 price cap (Russia was selling oil to India and Turkey at discounts that reduced government revenue), meaningful technology denial in aerospace and military electronics affecting both the civilian economy and military-industrial production, and early severe ruble volatility that required emergency interest rate action by the Russian central bank. CREA tracking, IEA analysis, and Bruegel Institute research all document these costs. The claim is that these costs — significant by any measure — did not produce the coercive effect. Russia absorbed them, re-routed trade through Turkey, the UAE, China, and India, and restructured its economy around war production. Pain, in this case, landed. It did not translate.

That translation failure defines the second link, and it is worth understanding why it is so consistently where the chain breaks. Francisco Rodríguez’s 2023 systematic review of the empirical literature found “a remarkable level of consensus across studies that sanctions have strongly negative and often long-lasting effects on the living conditions of the majority of people in target countries.” This is the structural asymmetry at the heart of the problem. Governing elites — the people with actual authority to change policy — have foreign assets, parallel import channels, access to shadow financial networks, and control over the distribution mechanisms that determine who gets what inside a sanctions-constrained economy. Ordinary populations do not have these things. They absorb the costs that the instrument is theoretically designed to redirect upward toward decision-makers.

The political consequence runs in the opposite direction from what the theory predicts. Governments facing external economic pressure redirect that pain domestically as a narrative of foreign aggression — sanctions become evidence of hostile intent, not leverage for compliance. Sebastian Hellmeier’s analysis of authoritarian regimes using monthly data from 2003 to 2015 found that sanctions significantly increase pro-government mass mobilisation, an effect amplified by state media control. The people who might otherwise constitute a domestic political constituency for compliance are, in practice, rallied toward the government by the experience of being economically attacked by foreign powers. Julia Grauvogel and Christian von Soest’s research adds a further structural finding: sanctions fail to instigate democratisation when target regimes have strong legitimacy claims — and external economic pressure tends to strengthen precisely those claims. The instrument designed to produce internal pressure against the government produces internal solidarity with it.

The third link — that political pressure, even if it somehow reached decision-makers, would produce concession — has its own structural problem independent of the first two. For a government to concede to sanctions pressure, the cost of continued refusal must exceed the cost of giving in. In the cases that generate the most sustained sanctions campaigns — nuclear programs, territorial claims, regime survival — the thing demanded is the thing the government has determined is non-negotiable. Asking Russia to withdraw from Ukraine, or North Korea to dismantle its nuclear deterrent, or Iran to abandon uranium enrichment entirely, is asking governments to give up what they have identified as core security requirements. No level of economic pain that falls short of genuine regime collapse — a threshold that sanctions have essentially never independently reached — satisfies this calculus. The instrument asks governments to trade their strategic interests for economic relief, in cases specifically selected for how unconditionally those strategic interests are held.

Cuba: persistence without purpose
The US embargo on Cuba has survived nine presidencies, six decades, and continuous biannual votes in the UN General Assembly that reject it by margins consistently above 180-2 (the two being the US and Israel). By every coercive measure — Castro did not comply, his successors have not complied, and Cuban political structures have not democratised — the embargo has failed for longer and more thoroughly than any comparable measure in modern foreign policy history. What the Cuba case illustrates most clearly is the paradox of persistence. The embargo's durability has nothing to do with coercive hope and everything to do with domestic US politics: specifically, the electoral significance of Cuban-American communities concentrated in Florida, a perennial swing state. William LeoGrande's work on the political economy of Cuba policy documents in detail how domestic constituency pressure has repeatedly overridden foreign policy analysis — including within administrations that privately acknowledged the embargo's futility. The Obama administration's 2014 normalisation effort, which briefly tested what dropping the embargo might achieve, was reversed by the Trump administration within two years. The instrument serves functions entirely disconnected from its stated foreign policy objective, which is why its demonstrated ineffectiveness produces no political pressure to end it.

The Contested Cases — What the Successes Actually Show

The strongest version of the case for sanctions points to three cases: South Africa, Libya, and Iran. All three deserve engagement rather than dismissal. They are also, on examination, more complicated than the standard narrative allows.

South Africa is the most frequently cited success, and it has genuine substance. The US Comprehensive Anti-Apartheid Act of 1986 — passed over Ronald Reagan’s veto in a Senate vote of 78-21 and House vote of 313-83 — was part of a broader multilateral sanctions package that coincided with, and arguably accelerated, the end of apartheid. Nelson Mandela and F.W. de Klerk both acknowledged the role of economic pressure in the transition. The causal story is real but partial. What actually ended apartheid was a convergence: the sustained organisational capacity and moral authority of the ANC, the collapse of the Soviet Union (which removed the Cold War justification that had given South Africa’s apartheid government strategic cover with Western governments), sustained internal resistance across two decades, and economic pressure from sanctions and capital flight. Audie Klotz’s analysis of the South African case is sceptical of simple causal attribution, emphasising the role of norm diffusion — the delegitimisation of apartheid as a political system in international opinion — as a force structurally distinct from economic coercion.

More importantly: South Africa was not a consolidated authoritarian state in the sense that makes Link 2 of the sanctions causal chain fail so reliably. It had a functioning economy deeply integrated into Western capital markets. Its white business elite held substantial foreign assets and maintained relationships with international investors who were increasingly alarmed by the political trajectory. It had a political opposition — the ANC — with genuine organisational depth and external support. The structural conditions that normally prevent economic pain from translating into elite political pressure were, in South Africa, only partially present. This is what made it susceptible to a form of coercion that consistently fails elsewhere. South Africa is not evidence that comprehensive sanctions work. It is evidence that economic pressure can contribute when the conditions that normally short-circuit the causal chain are absent.

Libya’s 2003 agreement to dismantle its WMD programs and pay compensation for the Lockerbie bombing is the second canonical success. Richard Nephew’s analysis frames the Libyan case as a case where sanctions contributed meaningfully to an outcome, though as part of a broader coercive context. The broader context included the US invasion of Iraq in March 2003 — a fact that Muammar Gaddafi himself cited explicitly in explaining his decision. A leader who had watched Saddam Hussein’s regime collapse under US military force drew his own conclusions about the price of continuing to host WMD programs. Attributing compliance primarily to sanctions in 2003, without accounting for the demonstration effect of the Iraq invasion in the same calendar year, is a coding choice rather than an empirical finding.

Iran and the 2015 JCPOA is the most plausible sanctions success case in recent history, and it deserves the most careful treatment. The years of intensifying US and multilateral sanctions on Iran produced real and visible economic effects: Iran’s oil revenues contracted substantially, the rial depreciated sharply, inflation rose, and Iranian negotiators explicitly acknowledged the pressure they were operating under. What those sanctions achieved was an agreement to limit — not end — Iran’s nuclear program, subject to verified compliance and in exchange for sanctions relief. This is a constrained objective, but it is a genuine objective, and the causal connection between sanctions pressure and Iranian willingness to negotiate is credible. Iran is the case that most closely satisfies the structural conditions for sanctions effectiveness: the objective was limited and concrete; multilateral implementation reduced escape routes; and Iranian elites with foreign exposure were genuinely affected. The JCPOA outcome confirms the structural analysis rather than refuting it, by showing what conditions the instrument actually requires.

What happened next is equally instructive. The US withdrew from the JCPOA in May 2018, reimposing sanctions unilaterally and demanding comprehensive compliance with a maximalist set of demands extending well beyond the original nuclear agreement. Iran’s response was not renewed compliance; beginning in 2019, Iran exceeded JCPOA limits on uranium enrichment in systematic steps. The leverage that years of sustained pressure had built was destroyed by unilateral withdrawal — not because the sanctions became less painful, but because the reversibility condition that gave them coercive logic was eliminated.

The Political Economy of a Failed Instrument

If sanctions so consistently fail to achieve their stated objectives, why do governments keep imposing them? The answer requires disaggregating what sanctions actually do from what governments say they are doing. They are not the same.

Sanctions serve at least four distinct functions, and governments deploy all four simultaneously while foregrounding only one.

The coercive function is the one that gets the speeches — changing the target’s behaviour through economic pressure. The expressive function is different: recording formally that certain behaviour is unacceptable within the international order, creating a normative record regardless of whether it produces compliance. The alliance-management function is different again: demonstrating solidarity with partners, maintaining coalition cohesion, giving aligned governments a shared action to take that affirms commitment without requiring military engagement. And there is a degradation function — imposing material costs that reduce the target’s capacity over time, even without immediate compliance, by restricting access to technology, capital, and expertise.

None of these functions is fraudulent. The expressive function is real and, in a world where international norms require maintenance to remain operative, genuinely important. The degradation function may be the most honest account of what Russia sanctions are achieving: long-term technology denial affecting military-industrial capacity, constraints on defence-sector capital formation, and restrictions on the know-how transfer that Russian aerospace and electronics sectors need. Alliance management explains why the 2022 Russia sanctions package served EU cohesion regardless of what it achieved in Moscow — the package demonstrated that European states would absorb economic costs to maintain solidarity with Ukraine, which mattered enormously for the credibility of the Western response.

The problem is not that these functions exist. The problem is their conflation. Governments justify harms that only make sense under a coercive logic — severe civilian suffering in Iran and Iraq, long-term economic impoverishment in Venezuela, the compounding deprivations of comprehensive embargo in Cuba — by invoking coercive success claims the record does not support. A sanctions regime serving expressive, degradation, or alliance-management purposes could be designed to minimise civilian harm: targeted measures on decision-makers and their assets, technology restrictions on military sectors, coordinated financial measures on elites. The fact that existing sanctions architecture routinely imposes broad civilian costs suggests that “this is coercion” is doing political work that “this is an expression of norms” or “this is degradation of military capacity” cannot. Civilian suffering, under the coercive theory, is the mechanism — the pressure that will eventually force governments to change course. Under any of the other three theories, it is unnecessary collateral damage.

Daniel Drezner’s work on signalling and commitment functions identifies another layer: sanctions serve as a credibility-management device for sanctioning governments as much as a pressure device on targets. Imposing sanctions demonstrates resolve to domestic audiences and international partners, forecloses certain diplomatic options, and creates political path dependencies that make the instrument difficult to lift even when its failure is evident. This is the “act without acting” dimension — governments can impose costs on themselves and others through expressive and degradation channels while avoiding the more politically costly alternatives: military action, formal diplomatic accommodation, or the explicit acceptance of an adversary’s fait accompli. The political cost of admitting that the primary non-military coercive instrument of international order does not work — and therefore that the realistic alternatives are worse — is high enough that sustained overclaiming remains the rational strategy.

The Narrow Conditions of Genuine Effectiveness

Given what the structural analysis reveals about why the instrument fails, the question of when it works has a diagnostic answer. The conditions required describe an instrument substantially different from what is typically deployed.

Binyam Afewerk Demena and Peter van Bergeijk’s 2025 meta-analysis of determinants of sanctions success and failure, drawing on decades of quantitative research, and Hufbauer et al.’s own data, converge on a set of conditions that genuine effectiveness requires.

Limited, concrete objectives are the first condition. Regime change and nuclear disarmament have an essentially zero independent success rate under sanctions. Trade practice adjustments, specific regulatory compliance, and the release of detained nationals are achievable under the right circumstances. The mismatch between the maximalism of stated objectives and the instrument’s actual capacity is a persistent feature of sanctions policy, not an accident: domestic political audiences demand maximalist objectives, while the instrument can at best achieve modest ones.

Strong pre-existing economic integration is the second condition. The EU-Russia relationship before 2022 offered more genuine leverage than the US-North Korea relationship has ever offered, for the simple reason that there was vastly more economic value at stake on both sides. Russia was the EU’s largest energy supplier. The threat of disrupting that relationship was credible and costly. The US-Cuba relationship has been structured as an embargo for so long that there is almost no economic integration left to threaten; the leverage the US once had has been consumed by decades of embargo and long since replaced by Cuban adaptation.

Multilateral implementation that closes off escape routes is the third condition. UN Security Council sanctions are structurally more effective than unilateral US or EU measures because they cannot be legally circumvented through China, Russia, or other major powers. US secondary sanctions are the attempt to replicate this through dollar-system leverage, but they are imperfect substitutes: they can be worked around through local currency swaps, barter arrangements, and the gradual development of alternative financial infrastructure, as Russia has been actively developing since 2014.

Target elites who cannot insulate themselves from economic effects are the fourth condition — arguably the most demanding. This requires either that elites hold substantial assets within the sanctioning jurisdiction (making asset freezes and travel restrictions genuinely costly to them personally) or that the economic constraints cascade through the economy in ways that affect their political base. South Africa partially satisfied this condition; Russia in 2022 did not, because the most exposed Russian oligarchs had already been sanctioned after 2014 and had already adapted.

Short, explicit timelines with clearly stated reversibility conditions are the fifth condition. Permanent sanctions, or sanctions without defined benchmarks for relief, provide no inducement to comply: if the target cannot see a path to relief, the cost-benefit calculus runs entirely against concession. The JCPOA was designed with explicit reversibility — measurable compliance thresholds would trigger defined sanctions relief — which is precisely what made it the most functional sanctions-based agreement of recent decades.

The pattern that emerges from these conditions is uncomfortable for the foreign policy establishments that most frequently deploy sanctions. The cases where sanctions are most vigorously imposed — Russia, Iran, North Korea, Cuba, Venezuela — are almost all cases where these conditions are absent: objectives are maximal, pre-existing integration is limited or has been deliberately reduced, multilateral implementation is incomplete, target elites have proved capable of insulating themselves, and the reversibility condition is either absent or has been violated. Where these conditions are present, political demand for sanctions tends to be lower, precisely because the relationship is cooperative enough that the credible threat of disruption is itself sufficient pressure.

The JCPOA withdrawal and the reversibility paradox
The US withdrawal from the JCPOA in May 2018 is a precise case study in how violating the reversibility condition destroys leverage. The nuclear agreement had been designed with explicit conditionality: Iran would limit uranium enrichment, reduce centrifuge numbers, and accept intrusive IAEA verification in exchange for phased sanctions relief. IAEA verified Iranian compliance eleven times before US withdrawal. The Trump administration's reimposition of sanctions — without Iranian non-compliance triggering the snapback — sent an unambiguous signal: compliance does not guarantee relief. Iranian policymakers drew the rational conclusion. By 2019, Iran was systematically exceeding JCPOA enrichment limits, and by 2024 had enriched uranium to 60 percent purity — technically one step below weapons-grade. The sanctions were still in place. The leverage was gone. What US withdrawal demonstrated is that the reversibility condition is not optional architecture — it is the mechanism by which the coercive logic of sanctions functions. Remove it and you have economic pressure without a path to compliance, which is punishment rather than coercion.

Standing in Munich in February 2022, Kamala Harris was making a specific kind of promise: that the economic costs to be imposed on Russia would be significant enough, and coordinated enough, to function as a deterrent or, failing deterrence, as a coercive force for reversal. The promise made sense within a particular theory of the instrument. It is the same theory that has structured Western sanctions policy through the Iran nuclear negotiations, the North Korea standoff, the Cuban embargo, the successive rounds of measures on Venezuela, and dozens of smaller-scale episodes. The language has been consistent for decades: significant, unprecedented, coordinated, targeted at those responsible.

The empirical record — contested in its details but unambiguous in its direction — suggests this theory has been systematically overclaimed. Sanctions have contributed to outcomes in a minority of cases, under structural conditions that are usually absent precisely when political demand for sanctions is highest. They impose costs on civilian populations far more reliably than they impose costs on decision-makers. They strengthen the domestic political position of the governments they are designed to coerce. They persist not because their coercive record justifies confidence but because the alternative account of their purpose — that they are primarily expressive, degradative, or alliance-managing instruments — cannot publicly justify the civilian harm they routinely cause.

The paradox of persistence is not comfortable, but it is not nihilistic either. Decision-makers are not naive about the empirical record. They understand, at some level of institutional memory, that comprehensive sanctions on consolidated authoritarian states rarely change behaviour. The persistence is more rational than that: admitting openly that the primary non-military coercive tool of international order is largely theatrical in its most consequential applications would require answering a question that has no comfortable answer. If not sanctions, then what? The honest answer — that in most high-stakes cases the realistic alternatives are military action, diplomatic accommodation, or the explicit acceptance that an adversary has done something that cannot be reversed — is harder than anything the empirical analysis delivers. Military action has its own calculus of costs and risks. Diplomatic accommodation requires conceding legitimacy to the thing that prompted the sanctions. Acceptance is politically unsustainable when the stated objective has been maximalist.

What was significant and unprecedented about the 2022 Russia sanctions package was not its effectiveness. It was the size of the coalition willing to absorb economic costs to demonstrate that certain conduct would not pass unremarked in international order. That is a real and non-trivial function. It just is not the function the language claimed.

Gen AI Disclaimer

Some contents of this page were generated and/or edited with the help of a Generative AI.

Media

cottonbro studio – Pexels

Key Sources and References

Harris, Kamala, Vice President. Remarks at the Munich Security Conference, February 19, 2022. White House Archives. https://bidenwhitehouse.archives.gov/briefing-room/speeches-remarks/2022/02/19/remarks-by-vice-president-harris-at-the-munich-security-conference/

Hufbauer, Gary Clyde, Jeffrey J. Schott, Kimberly Ann Elliott, and Barbara Oegg. Economic Sanctions Reconsidered, 3rd ed. Peterson Institute for International Economics, 2007.

Pape, Robert A. “Why Economic Sanctions Do Not Work.” International Security, Vol. 22, No. 2 (Fall 1997), pp. 90–136.

Kirilakha, Aleksandra, Gabriel Felbermayr, Constantinos Syropoulos, Erdal Yalcin, and Yoto V. Yotov. “The Global Sanctions Data Base: An Update that Includes the Years of the Trump Presidency.” In Peter A. G. van Bergeijk, ed., Research Handbook on Economic Sanctions. Edward Elgar Publishing, 2021.

Rodríguez, Francisco. “The Human Consequences of Economic Sanctions.” Center for Economic and Policy Research, May 2023. https://cepr.net/report/the-human-consequences-of-economic-sanctions/

Hellmeier, Sebastian. “How Foreign Pressure Affects Mass Mobilization in Favor of Authoritarian Regimes.” European Journal of International Relations, Vol. 27, Issue 2 (2021), pp. 450–477.

Grauvogel, Julia, and Christian von Soest. “Claims to Legitimacy Count: Why Sanctions Fail to Instigate Democratisation in Authoritarian Regimes.” European Journal of Political Research, Vol. 53, Issue 4 (2014), pp. 635–653.

Demena, Binyam Afewerk, and Peter A. G. van Bergeijk. “A Meta-Analysis of Determinants of Success and Failure of Economic Sanctions.” Econometrics (MDPI), Vol. 13, Issue 2 (April 2025).

Comprehensive Anti-Apartheid Act of 1986. Pub. L. 99-440. Enacted October 2, 1986.

Klotz, Audie. Norms in International Relations: The Struggle Against Apartheid. Cornell University Press, 1995.

Nephew, Richard. “Libya: Sanctions Removal Done Right? A Review of the Libyan Sanctions Experience 1980–2006.” Columbia Center on Global Energy Policy, March 2018.

Drezner, Daniel W. The Sanctions Paradox: Economic Statecraft and International Relations. Cambridge University Press, 1999.

Drezner, Daniel W. “Sanctions Sometimes Smart: Targeted Sanctions in Theory and Practice.” International Studies Review, Vol. 13, Issue 1 (2011), pp. 96–108.

LeoGrande, William M., and Peter Kornbluh. Back Channel to Cuba: The Hidden History of Negotiations between Washington and Havana. University of North Carolina Press, 2014.

International Atomic Energy Agency. Verification and Monitoring in the Islamic Republic of Iran in light of United Nations Security Council Resolution 2231 (2015). Reports to the IAEA Board of Governors, 2015–2019. https://www.iaea.org/newscenter/focus/iran/iaea-and-iran-iaea-board-reports

Centre for Research on Energy and Clean Air (CREA). Russia Fossil Tracker and monthly analyses of Russian fossil fuel exports and sanctions, 2022–2024. https://energyandcleanair.org/

Bruegel Institute. Various analyses of Russia sanctions impact, 2022–2024. https://www.bruegel.org/

International Energy Agency. Oil Market Report (monthly editions), 2022–2024. https://www.iea.org/topics/oil-market-report

United Nations Charter, Article 41. 1945.