Looking at foreign policy differences between the past two administrations, one gets the impression that while the spirit and tools have changed a bit, the challenges and goals have stayed mostly the same. Persistent, hollow threats from North Korea, chemical warfare in Syria, and bickering with Russia have held on to their positions as the top news stories of the day, just as they were for most of the past decade.

What has changed, then? Most obviously, it seems that the present administration has lost its faith in the tools of its predecessor, namely diplomacy and sanctions. It is difficult to quantify the efficacy of the former, given its personal, and usually secret, nature, but the latter permits detailed analysis into its use and results. Thus, one might naturally ask, should the Trump administration continue to move away from sanctions, as exemplified by its pursuit of more muscular acts in Syria and elsewhere? What are the factors that lead to successful sanctions, and how might we devise sanctions regimes that accomplish their objectives quickly and without harming innocents?

How effective are sanctions?

Many noted economists, as well as governmental agencies, have conducted studies into such questions. For example, a 1992 GAO report to Congress offers mixed reviews of historical sanctions efforts. The Pearson Institute for International Economics, a leading DC think tank, maintains a repository of case studies offering perhaps the most detailed analysis of US and international sanctions efforts, ranging from the famous Cuba and Iran sanctions to obscure cases like Guinea-Bissau and Fiji. These they rate using a simple scale, assessing each sanctions regime on a scale of 1-4 on policy result (to what extent did the regime succeed in changing target country policy) and sanctions contribution (how much of the policy change was due to the effects of the sanctions). These scores are then multiplied to arrive at a total score.

So for example, the PIIE rates 2004 international sanctions against Ivory Coast as 4 for policy results, as the rogue president of Ivory Coast, Laurent Gbagbo, was indeed arrested and is currently awaiting trial by the ICC. But it only gave a 2 for sanctions contribution, as this outcome was really only brought about by French military intervention seven years later, giving a total score of 8, just below their threshold for “success” of 9.

While this rating scheme is somewhat subjective, it does align fairly well with both intuition about current programs and mainstream scholarship about historical ones. For example, 1948 international sanctions against the Netherlands for imperialism in Indonesia, a resounding success in reversing Dutch policy, received a rating of 16, while recent sanctions against Syria, largely ineffective, received a 2.

Why is success so unpredictable?

The creation of such quantitative measures allows for a more rigorous statistical analysis of the results. A 2014 blog post for the Cato Institute by Gary Hufbauer, an author of the PIIE’s flagship sanctions text, examines the then-immediate issues of changes in sanctions against Russia and Iran, and how historical data might lead to their improvement. Several conclusions stand out. First, friendly nations were more likely to respond to economic pressures than enemies. Additionally, democracies are far more easily convinced than autocracies. Indeed, unfriendly autocracies are only likely to respond to sanctions pressure if such sanctions are actually seen as a precursor to or threat of military force, as was the case with US sanctions against Libya in the 1970’s. Thus, while the authors found an overall success rate of 33%, only 18% of sanctions levied against unfriendly regimes had any success.

The reasons are intuitive: first, friendly nations are more likely to actually have economic ties which could be affected by sanctions. Second, autocracies are more able to simply shift the burden of such sanctions to the populace as a whole, as was the case with sanctions against Saddam Hussein, which resulted in the deaths of thousands of Iraqi citizens and no change in Saddam’s policies.

Policy Implications

This conclusion has important consequences for most contemporary issues. Few could regard our relations with Iran, Russia, or Syria as friendly, and most would consider their governments autocratic. Thus the imposition of sanctions with sweeping objectives, like halting military action or catalyzing regime change, is unlikely to succeed without the concurrent threat of military action. It may be possible to modify our sanctions template for the next Syria to include such threats, thereby increasing the likelihood of their success, but our inability to do so against a great power like China or Russia indicates a need to find alternate tools.

One such possible tool was first introduced by the US Treasury in 2005, by taking advantage of the US Dollar’s status as sine qua non of large-scale international transactions. Both the debts and foreign exchange reserves of most governments are predominately denominated in dollars, as well as most oil sales. This grants the United States exceptional leverage over other countries, which one might expect to see used as part of a sanctions package.

In a response to Hufbauer’s article, sanctions expert Eric Lorber describes two such cases related to sanctions against Iran in Russia. In Iran’s case, the United States promulgated regulations forbidding any company or government doing business in Iran from doing business in the United States. While many foreign nations and firms don’t seem to have significant business dealings in the US anyway, the restriction also has the important consequence of preventing access to US financial markets. Given the prevalence of dollar-denominated debt, this is a serious threat. Any country that chose to permit business in Iran would be hampered in servicing its existing debt or issuing new debt. Similarly, in response to Russian aggression in Ukraine, the Treasury applied the same restriction directly to several large, politically-connected or state-owned firms. This led to a significant recession in Russia, with the Russian ruble losing nearly half of its value against the dollar (see figure 1).

Figure 1: Rouble per Dollar, 2010-2017. Higher numbers correspond to weaker rouble.

From Recession to Regime Change

Even with sanctions better able to affect economic conditions, the question remains as to whether such hardship will translate into political changes. According to the framework developed by Hufbauer, this is so, but only when relations are close, politics are democratic, and the target is smaller than the nations imposing sanctions.

In a 1997 response to Hufbauer’s research, however, A. Cooper Drury conducted a logistic regression analyzing Hufbauer’s data with the addition of more control variables and a slightly more nuanced view of the relationships between variables. He found that many of the variables cited by Hufbauer as significant were not so, and vice versa. For example, Hufbauer’s original analysis included the “sanctions contribution” component describing the degree to which sanctions, rather than other measures, contributed to success, as well as a separate variable controlling for the presence of other activities, such as covert military action, in the case in question. This presents an obvious endogeneity and multicollinearity problem as by definition, a case solved entirely by sanctions (a 4 on the “sanctions contribution” scale) will necessarily have no such supplementary policy actions, while a case ultimately resolved by military force will definitely have a positive “other measures” value, as well as a low “sanctions contribution” score (and likely, a high success rate, since military action tends to be definitive).

With these issues in mind, Drury jettisons the old dual-tracked scoring system, replacing it with a single 1-4 scale of whether objectives were achieved. He also takes Hufbauer’s hypotheses of significant factors and adds several other from more recent studies. First, taking into account only the policy suggestions of Hufbauer, he finds very few significant variables, with only the damage caused to the foreign economy (p = .063) and the year sanctions were imposed (p = .075) achieving statistical significance at even the 10% confidence level. Next, when including new variables, such as threats posed by the target to international security, the cooperation of international institutions in sanctions enforcement, and the presence of saboteurs like China (happy to continue trade with the likes of Iran and North Korea over international objections), many of the original conclusions are discredited.

Crucially for our study, it is found that the damage done to the target’s GNP is significant (p = .081), and the strongest predictor, besides international institutional cooperation, of success. The latter result, meanwhile, is important, in that it demonstrates the coordinating ability of bodies like the United Nations, and suggests that multilateral sanctions often fail because their scattered implementations and aims diverge in the absence of some kind of coordination. Indeed, the coefficient on international cooperation without institutions is negative, and significant (p=.012), underscoring the need for coordination.

The Way Forward

Taken all together, the above papers and studies provide several concrete reasons why the effectiveness of sanctions varies, and thus offer guidance on the considerations for which policymakers should account when deciding to levy sanctions. First, sanctions work best when they cause significant damage to national economies, and when imposed by an international body rather than unilaterally or multilaterally without coordination. Second, simply being larger than the target country is no guarantee of success, no matter whether that country is friend or foe. This informs our actions against North Korea and Cuba, who have resisted international pressure for decades despite their small size. Finally, sanctions unable to cause significant economic damage must instead be backed up by the threat, and appear a signal for the imminent use, of military force. Under such constraints, it will be difficult to design sanctions regimes that avoid harming the target’s civilian populace. Further, it will be nearly impossible for sanctions to act as a deterrent to a great (or even regional) power, given its economic strength and the high costs of military conflict. These challenges all point to the eventual decline of sanctions as a panacea for violence-free conflict resolution, and the need to develop a new framework for dealing with our enemies, large and small.

Why Sanctions Do(n’t) Work
Tagged on: