Do sanctions change regime behavior?

The pharmacy line

The week the new sanctions took effect, nothing at home looked any different, and within a month everything cost more. Picture standing in a country that has just been cut off, watching the price of an imported medicine at the pharmacy climb from one visit to the next, watching the local money buy less of everything that comes from abroad, and watching savings that took years to build lose their meaning against a foreign currency that now costs several times what it did in the spring. On the television a foreign official explains that the measures are aimed at the government and not the people, and that the pressure will force those in charge to change course. In the pharmacy line the person actually doing the paying has two questions the official does not answer. The first is who is absorbing this. The second is whether it will change anything at all.

This chapter is written for that person, and it takes both questions as seriously as they deserve. The short version of the answer is that the money you send home and the medicine on the shelf tend to absorb the blow, while the people close to power turn the scarcity into a business; that a government rarely bends the sweeping way the official on the television promises; and that in the cases where sanctions did change a regime’s behavior, they were asking for something narrow and verifiable and were backed by more than one country at once. Getting from the promise to the record means looking first at whether sanctions work in general, and then at who pays when they do not.

What the record says about whether they work

The most-cited attempt to keep score is a catalogue assembled over three decades by Gary Hufbauer, Jeffrey Schott, and Kimberly Elliott, later with Barbara Oegg, who combed through more than a century of episodes in which one country tried to bend another with economic pressure. Their third edition runs to 174 case studies from the Allied blockade of Germany in the First World War through the year 2000, and their headline finding is the one every discussion of sanctions starts from: by their reckoning, sanctions were at least partly successful in about a third of the cases, roughly 34 in 100.

A third is the number that launched a thousand policies, and it is worth being exact about what stands behind it, because the counting is the whole argument. Hufbauer and his colleagues did not simply ask whether the target changed. They scored each episode twice: once for how far the sender’s stated goal was reached, on a scale from 1 to 4, and once for how much the sanctions themselves, rather than a war or a coup or a shift in domestic politics, contributed to that result, again from 1 to 4. The two scores are multiplied into a single number from 1 to 16, and a case counts as a success only if it reaches 9. Every one of those judgments is a reading of a messy history, which means the famous one-third is not a measurement in the way a temperature is. It is the sum of a great many verdicts about what caused what.

That is the opening a critic can drive through, and the political scientist Robert Pape did. Taking the earlier edition of the same catalogue, with its 115 cases and 40 that the authors marked as successes, he went back through the 40 one by one and accepted only 5. The rest, he argued, had been miscounted: 18 were really settled by force rather than by the economic pressure, 8 were simply failures in which the target never gave in, 6 were ordinary trade disputes rather than attempts to coerce a government, and 3 were too murky to call either way. By his count the true success rate was not 34 percent but under 5. His test for a genuine success was strict and worth stating, because it is the crux of the disagreement: the target had to concede a real part of what was demanded, the sanctions had to come before the target changed, and no more convincing explanation, above all the threat or use of military force, could account for the change.

The two numbers, a third and a twentieth, bracket the range of defensible readings of whether sanctions work, and the argument between them is not really about the facts of any single case. Both sides agree on what happened in Cuba and Iraq and Libya. They disagree about what to credit for it, and that disagreement cannot be settled by collecting more cases, because the thing in dispute is how to tell a sanction’s effect apart from a war’s or a revolution’s in a history where all three were happening at once. What survives from both counts is that outright success is the exception rather than the rule, and that a great many confident sanctions campaigns ended with the target exactly where it started.

When they do change behavior

Exceptions exist, and they are not random. The same catalogue that produced the one-third figure also sorted its cases by what the sender was trying to achieve, and the pattern in that sorting is more useful than the headline rate. When the goal was modest and specific, freeing a prisoner, reversing a single policy, extracting a defined concession, sanctions reached it about half the time. When the goal was grand, toppling a government, breaking a country’s military power, forcing a wholesale change of course, the success rate fell to roughly 30 percent, and for attempts to stop a military adventure already under way it fell to about a fifth. The lesson buried in those numbers is that sanctions can sometimes buy a small, well-defined thing and almost never buy a large, open-ended one.

Two other conditions run alongside the size of the demand. Sanctions did better when the target had been trading heavily with the sender beforehand, so that cutting the tie actually hurt, and when they were imposed quickly and decisively rather than tightened slowly while the target adjusted. Cooperation among several senders mattered too, though less evenly than the lone-embargo intuition suggests. For the ambitious goals it was present in far more of the successes than the failures; for modest demands it made little difference and was often not even sought, and it was never a strict requirement. A lone embargo does leak, since the target sells through whoever is still willing and the main effect is to hand that business to a rival, but cooperation is no guarantee: the failed campaigns against North Korea and Russia were among the most multilateral of all. The political scientist Lisa Martin, studying 99 post-war cases, found that the heaviest, most economically painful sanctions were the ones that came with broad international cooperation, because a sender bears a large cost of its own only when it has locked in partners who will bear theirs. That explains what makes cooperation form rather than what makes a sanction succeed, and cooperation is hard to sustain in any case, which is part of why the big cases so often fail.

Line up the exceptions against that template and they fit it. Libya was asked for something narrow and checkable, the surrender of two named men, under sanctions that much of the world enforced together, and in time it complied and went further. The financial pressure on apartheid South Africa, when foreign banks stopped rolling over the country’s loans in 1985 and capital fled, came from many directions at once and bore on a government already losing its footing at home. The nuclear agreement Iran signed in 2015 followed years of sanctions that, unusually, the major powers applied in concert around a single defined demand about enrichment, though the deal’s later collapse is a reminder of how contested even this one is. Set beside them, the long failures share the opposite features. The embargo on Cuba was mostly one country’s, aimed at nothing less than the end of the regime, and it has run for more than sixty years while the regime it meant to remove buried the presidents who imposed it. The demand was total, the coalition was not, and the result was the one the pattern predicts.

Who actually feels it

Return now to the pharmacy line, because the second question turns out to have a firmer answer than the first. Whether sanctions bend a government is genuinely uncertain; who absorbs them while the experiment runs is not. A broad economic squeeze lands first and hardest on the people with the least room to absorb it. The economist Dursun Peksen, working across a large sample of episodes, found that governments under sanctions became more repressive, not less, worsening their record on the most basic protections against disappearance, torture, and political imprisonment, and that the broader and more comprehensive the sanctions, the worse the effect. In a separate study he traced the same pressure into public health, where the measure was the death rate of children under five, and found that it rose with the economic pain the sanctions inflicted. The damage was not a side effect that careful design had failed to prevent. It scaled with the very thing sanctions are built to maximize, the economic hardship imposed on the target.

The reason the hardship travels downward rather than upward is that a government and its inner circle control the one thing that becomes precious when a country is cut off, which is access. Access to the remaining foreign currency, to the licences to import, to the official exchange rate that lets a well-connected buyer purchase dollars at a fraction of the street price. When everything from abroad grows scarce, whoever administers the scarcity is positioned to profit from it, and the same measures that empty an ordinary household’s savings can enrich the people standing closest to the state. This is the concentrated-benefit logic that governs so much of economic policy, that an organized few capture what a rule hands out while the costs scatter across everyone (Frameworks, on how an organized few capture what a policy hands out). Sanctions do not suspend that logic. They sharpen it, by making access the most valuable commodity in the country. The wider story of how a sanctioned economy’s money can collapse outright, and take a government down with it, belongs to a later volume (Volume V, on how a sanctioned economy’s money can collapse).

One case became the emblem of this whole debate, and it has to be handled with care, because it rested on numbers that did not hold up. Through the 1990s the claim that United Nations sanctions had killed around half a million Iraqi children circulated as established fact, repeated by officials and critics alike and grounded in a survey conducted in Iraq late in the decade. Years later, when researchers were able to examine the underlying data, the survey was found to have been manipulated, its child-mortality figures inflated by a government with every reason to inflate them, and the true rise in child deaths, while real, far smaller than the headline number that had done the arguing. The episode is a caution that cuts in an awkward direction. The suffering under the Iraqi embargo was genuine and severe, and it is exactly the kind of population-level harm the broader evidence would predict, yet the single most-quoted measure of it was a fabrication in the service of the regime the sanctions targeted. That is why the incidence panel above draws on Iran, where the mechanism is visible in an exchange rate rather than in a contested body count, and why the record on Iraq is that the harm was real, the arithmetic that made it famous was not, and a regime proved willing to manufacture its own people’s deaths as a weapon against the pressure.

Where the argument goes next

Sanctions are the outside world reaching into a national economy on purpose, but the same forces act on every country all the time, quietly, without anyone intending them as a weapon. The value of a country’s money against the world’s, the balance of what it buys from abroad against what it sells, and the flow of foreign capital in and out are the ordinary, permanent version of the pressure a sanction applies deliberately. Why the money you send home loses value, and whether a country that buys more than it sells is really losing, is the subject of the next chapter.