Do Economic Sanctions Work?
Analysts have interpreted the recent openings in Myanmar and North Korea as the finally successful result of years of international pressure and economic sanctions. At the same time, debate is hot on the scope for similar measures in Iran, Syria and, closer to us, Belarus and Hungary. Does economics have anything to say on this? What can we learn from the analysis of past experiences?
On February 29th, after decades of frustrating attempts by the outside world with sticks and carrots, but mostly economic and diplomatic isolation, North Korea announced that it would suspend its enrichment of uranium and its tests of weapons and long range missiles. It would even allow an inspection by the International Atomic Energy Agency, the first one since the country walked out of the Nuclear Non-Proliferation Treaty in 2003. The recently inaugurated leader, young Kim Jong Un, asked, in exchange, some tons of food aid and the promise of talks. Some believe this was inspired by another recent unexpected “opening”: the turn-of-the-year developments in Myanmar, where a cease-fire and the release of many of the political prisoners prompted a slow but sure thawing in the country’s diplomatic relations with the rest of the world. Some months on, the government’s intentions to move from a military dictatorship to greater pluralism still seem sincere enough. Many have interpreted these events as the finally successful result of years of international pressure and economic sanctions on the two countries. Is the tide turning for sanctions enthusiasts?
At the same time, though, concerns are rising that EU member Hungary is moving in quite the opposite direction, after a change in the constitution that endangers the independence of the media, the judiciary and the central bank. Hungarians protesting in the streets are openly talking about authoritarian evolution drawing parallels with the behavior of the government in Belarus, which only months ago attracted harsh criticism – and stringent sanctions. Hungary might follow suit in this respect as well: its credit line with the IMF is still hanging from a thread, and the EU threatened law suit over the constitutional changes, while a potential limitation of the country’s voting rights in Brussels is whispered as the “nuclear option”.
Although the situation looks increasingly, explosive both in Syria and Iran, even in these cases the hopes of the international community rest exclusively on economic coercion. Syria’s economy is now under severe pressure, after even the Arab League imposed sanctions. This is first time such a decision is taken against a fellow member. Near all trade and financial relations have been cut off, with the exception of some banks in Lebanon and perhaps a few business friends in China and Russia that might still offer assistance to Bashar Assad’s regime. But the country’s foreign reserves, already low one year ago at the offset of the crisis, should be running out by now, and inflation is rising as many consumption goods become scarce. At the same time, although Saudi Arabia is arming the rebel groups, a military intervention sanctioned by the international community seems unlikely, given the recent Libyan precedent.
The sanctions faced by Iran over its nuclear program are also growing to unprecedented severity, and also in this case military action does not seem to be considered an option – except by (understandably) jumpy Israel. Given the stage that the nuclear program has reached, and the level of protection built around it, bombing is not likely to stop it. Experts say that a successful US-lead operation could at most delay it some ten years. Arguably, this would only result in an even angrier Iran equipped with nuclear weapons, in ten years from now. Hence it would appear much more fruitful to try to change the population’s attitude, so that Iranians themselves can in turn affect their political leaders’ attitude, even if this needs replacing the regime altogether. This way the prospect of a nuclear Iran would not look as scary.
As the international community considers over and over its stance in all these thorny situations, a legitimate question in everybody’s mind is: What is the likelihood that the sanctions will work? Does the economic literature have anything to say on this matter?
Achieving the goal
According to Richard Baldwin, Professor of International Economics at the Graduate Institute of Geneva, “[i]t would be difficult to find any proposition in the international relations literature more widely accepted than those belittling the utility of economic techniques of statecraft.” In other words, a prominent scholar’s synthesis of the literature is that economic sanctions do not work. The anecdote most widely cited by advocates of sanctions is of course South Africa. The economic pressure imposed on the country in the mid-1980s certainly contributed to the strain that the inefficient and costly apartheid regime was increasingly suffering, finally leading to its dismissal. At the opposite end of the spectrum stands Iraq, where neither the comprehensive sanctions nor the oil-for-food program, in principle a quite clever combination of sanctions and aid, could achieve anything. The success of the following military intervention is also a subject of debate, though not one I will address here. Some have drawn the conclusion that the discriminating factor lies in how important for the target regime is the recognition of and identification with the sanctioning part. Others argue the probability that the sanctions succeed is linked to the cost born by the target, or by the sanctioning part (also called the sender), or other observable factors. If truth be told, these are both quite special cases, hard to generalize. But then again, one could argue that every episode involving international disputes is a special case. It follows that the systematic study of economic sanctions with the evaluation of their effects is not a straightforward task at all.
The first step to evaluate the success of imposed economic sanctions is to establish what the goal is. In the most basic terms, there are two types of explicit goals. In some cases, the imposition of an economic sanction is purely punitive towards a policy or act of a regime, or towards the regime itself, and aims at expressing disapproval from the initiating part, when inaction can signal complicity. Hoffman [8] was one of the first to suggest that “sanctions are mostly adopted to alleviate cross pressure situations, resulting when a (foreign) government faces demands for action but war is undesirable”. In this case, it makes little sense to talk about success or failure, as the imposition of sanctions is a goal in itself.
In the extreme case, this type of sanctions aims at destabilizing the target regime, inducing political change. This seems to be part of the aim of actions taken against Syria, although an end to the Iranian theocracy, and Lukashenko’s regime in Belarus, for that matter, would certainly be welcome as well. An analysis of the historical records from 1914 to 1989 [4] reveals that the probability of success with this goal has been 38% when the regime was very stable to start with and up to 80% in “distressed” countries. The single most important factor of success is hence, not surprisingly, the pre-sanctions stability of the political system in the target country. In some cases, paradoxically the imposition of sanctions stimulated political cohesion in the target country – the so called rally-round-the-flag effect. This is what seems to be happening, at least at this stage, in Hungary. The evidence suggests that there is a threshold of political cohesion above which external intervention strengthens the target government. According to Lindsay [13], three factors make it more likely that sanctions produce political integration rather than regime collapse:
- If they are seen as an attack on the whole country rather than on a specific faction
- If identification with the sanctioning part is weak or even negative
- If no alternative to the sanctioned course of action is available or perceived as better
In this light, measures that can be manipulated to punish only or prevalently the regime’s domestic supporters and political base are to be considered as superior. Travel bans and freezes of assets, foreign bank accounts and property of functionaries are examples of this type of measures. Financial restrictions, in addition to be perceived as comparatively fairer, have also been more effective in the past. Moreover, also to the point that the sanctions should not, if possible, hurt everyone indiscriminately, they are preferable to measures that hurt the productive sector, like trade restrictions.
Alternatively, sanctions are designed to compel a specific policy change in the target country. This is the case of Hungary and its new constitution, and formally of Iran, which is only required to drop its quest for nuclear weapons. The emerging consensus in the sanctions literature is that concessions are most likely at the threat stage [11]. Nevertheless, there are cases where the threat of sanctions fails and sanctions are then actually imposed. And, although the success rate becomes lower at this stage, there are examples where the target yields only after the sanctions are imposed. It might seem tempting then to investigate whether observable variables can predict the likelihood of success in these cases, because this would teach us something about the current crises around the world. However, trying to understand when and why sanctions have success based on the analysis of empirical data is complicated by a number of challenges.
First of all, there are at least two sources of censoring in the sample of imposed sanctions: because it is only a specific type of disputes that reach this stage, the evaluation based on them will be biased. The first reason why these are special cases is due to the fact that imposed sanctions have already failed at the threat stage. Hovi et al. [9] look at this situation from a game-theoretic perspective and argue that, if sender and target are rational, a threat of sanctions could fail because of one of three reasons: 1) it is not credible, so no actual sanctions will follow the threat; 2) it is not sufficiently potent, meaning that the target considers sanctions to be a lesser evil than yielding; 3) it is noncontingent, i.e. the target expects sanctions to be imposed regardless of whether it yields or not. If any one of these is true, then the target that did not yield at the threat stage will not yield after sanctions are imposed either (or no sanctions will be imposed if alternative 1 is true). Imposed sanctions will work only if at least one of these factors is initially not known with certainty, or wrongly perceived by the target: if the target believes the threat non credible, but then sanctions are actually imposed; if the target was wrong in judging the cost of the sanctions and realizes it only after sanctions are actually imposed; or if the target thought that sanctions would be imposed regardless of its behavior, but is subsequently persuaded that, in fact, the sanctions will cease if it yields. Otherwise, with perfect knowledge and rational decision-making, sanctions that are actually imposed are bound to fail precisely because they were imposed, i.e. because they failed at the threat stage.
Further selection occurs even earlier than the threat stage. The literature has examined thoroughly how strategic interaction during the sanction episode affects sanctions outcomes and duration (for example, [15], [7], [14], [5], [6], [12]). Much fewer studies have undertaken the possibility that states also act strategically before episodes, when choosing whether to challenge the status quo and how much to demand of the target. Theories around this stage of the “game” are referred to as endogenous demand theories. Krustev [11] proposes the idea that perhaps “strategic demands can account for the widely cited discrepancy between the frequent use of sanctions and the modest success rate of these instruments”. His game-theoretic model has the implication that oftentimes sender governments strategically choose hard cases, because “the uncertain prospects that the target agrees to a large demand might outweigh the certain prospects of receiving minor concessions”. This also results in a low observed success rate.
Beyond the difficulties related to selection, another challenge that the analyst faces is to isolate the effect of sanctions. Usually, sanctions are not adopted in a vacuum, but rather complement other types of actions (e.g. diplomatic pressure, military action), which interact with the success of the measures. Similarly, there is the issue of unintended consequences, that also affect the costs on both parts, and hence the likelihood of success. Most importantly, some of these unintended effects might change the situation so drastically that talking about success or failure does not make sense anymore.
Unintended consequences
Besides the success or failure with the specific goals they are intended to obtain, economic sanctions bring about a host of more or less foreseeable unintended consequences as well. One especially undesirable outcome of trade sanctions has recently been brought to attention from the analysis of former Yugoslavia [2]. Under a regime of import restrictions, private and public actors might be pushed towards the use of unlawful methods in order to avoid the sanctions and reach the international market through unofficial ways. An unhealthy cooperation between politicians, organized crime and smuggling networks might then establish itself and persist even beyond the duration of the sanctions.
This consideration speaks against isolating the target country from trade flows. A case in itself concerns, though, trades which already lie on the boundary of lawfulness and little contribute to the productive sector, such as arms traffic. These can and should be decisively stopped. Aside from the security benefits to such a move, this also has the potential to dry up a significant source of revenue for the contested leadership.
Be it on credit or on trade, it goes without saying that any restriction will hurt the economy. The political consequences of an economic downturn caused by the sanctions are not easy to foresee. Recent research on fragile states [3] studies the relationship between national incomes and two types of political violence: repression, i.e. unilateral violence by the incumbent government, and civil conflict, two-sided use of violence on the part of the state as well as insurgent groups. The link with the national income prospects is given by the consideration that both parts, deciding whether to resort to violence, evaluate the cost and benefits of violent action. The incumbent government has a cost-advantage, being able to dispose of the state resources. The costs for potential insurgent factions go down with deteriorating economic conditions, for example in presence of high unemployment, because then those involved have less to lose. Insurgence then becomes more likely. This theory is consistent with the last century’s worth of evidence, including the recent wave of revolutions in the Arab world, suggesting that countries seeing a decline in incomes move towards democracy considerably faster. The evidence is anecdotal, though, and more rigorous empirical analysis [1] revealed no significant pattern.
Moreover, the step between opposition insurgence and the establishment of a new, possibly democratic, regime might not be rapid at all, as the Syrian tragedy is reminding us of every day. The question is then whether the leverage of economic measures from outside is likely to make any difference during this phase. As analysts push for the political and logistical backing of the international community to the revolt in Syria, and as Saudi Arabia is arming the rebels, we must consider that also measures aimed at supporting eventual opposition factions, or the democratic system in general, might have undesirable consequences. Comparative statics in the context of the same theoretical framework referred to above show that, for example, the promise of financial assistance conditional on free multi-party elections may raise the incumbent’s perception of instability and hence raise the risk of repression and increased looting, unless combined with reforms to strengthen executive constraints. Even pressure for the release of political prisoners might set out a ransom system, with perverse incentives to taking more and more prisoners to be exchanged with economic assistance – this might still be a risk in Myanmar, given the abundance of political prisoners still held by the government.
Another important difference between trade and financial restrictions is that the former are likely to result in accumulation of debt. The burden of this debt, that the sanctioned regime is responsible for, will weigh on the future growth of the country, hence on future generations of taxpayers and potentially on a future government, which ideally should not be held accountable for the course of action chosen today by a contested leadership. Alternatively, in the case of a collapse of the economy, the debt could be defaulted. This risk is on the countries or financial institutions that today lend money to the sanctioned regime. In other words, interrupting trade without at the same time closing the lines of credit would put the sanctioning part or third part lenders in the least desirable situation.
In some cases, the target has the possibility to resort to alternative lenders in third countries. Although this is preferable to a situation where the sanctioning part itself bears the risk on the debt, it is not ideal because it frustrates the sanctioning effort. An innovative proposal has been put forward by Jayachandran and Kremer [10], related to the legal doctrine of odious debt. They propose that any debt incurred by a particular regime, that could be argued to be contracted without the consent of the people and not for their benefit, is declared by some supranational institution illegitimate and nontransferable to successor regimes. This would create disincentive for lenders in third countries, and potentially eliminate equilibria with illegitimate lending. Even this type of loan sanctions hurt the economy and hence ultimately the population; however they create a long-run benefit for the population by preventing the accumulation of an unjust debt that today finances mismanagement, looting or repression and tomorrow has to be repaid by someone who never agreed to incur it. It would be very interesting to see this solution implemented in practice!
Conclusion
In short, sanctions are difficult to implement so as to reach the intended goal and minimize the unintended effects, but are maybe even more difficult to study systematically. International disputes are often complicated matters, situations that evolve over long time horizons. The traditional research question of when sanctions work might not be the most relevant one. Including in the analysis the strategic behavior occurring at the threat stage, and even before that, is a first step, although basing policy on the prediction that threats work better than sanctions does not strike me as a very useful conclusion.
The fact that evaluation is problematic and generalization almost impossible does not mean, however, that the study of sanctions is useless altogether. Economic analysis may still be informative for decision-making, and produce innovative ideas on the design of supranational institutions for conflict management, like the proposal on odious debt illustrates.
▪
Bibliography
[1] D. Acemoglu, S. Johnson, J.A. Robinson, and P. Yared. Reevaluating the modernization hypothesis. Journal of Monetary Economics, 56(8):1043–1058, 2009.
[2] P. Andreas. Criminalizing consequences of sanctions: Embargo busting and its legacy. International Studies Quarterly, 49(2):335–360, 2005.
[3] T. Besley and T. Persson. Repression or civil war? American Economic Review, 99(2):292–297, 2009.
[4] J. Dashti-Gibson, P. Davis, and B. Radcliff. On the determinants of the success of economic sanctions: An empirical analysis. American Journal of Political Science, pages 608–618, 1997.
[5] H. Dorussen and J. Mo. Ending economic sanctions. Journal of Conflict Resolution, 45(4):395, 2001.
[6] D.W. Drezner. The hidden hand of economic coercion. International Organization, 57(03):643–659, 2003.
[7] J. Eaton and M. Engers. Sanctions: some simple analytics. The American Economic Review, 89(2):409–414, 1999.
[8] F. Hoffmann. The functions of economic sanctions: A comparative analysis. Journal of Peace Research, 4(2):140–159, 1967.
[9] J. Hovi, D.F. Sprinz, and R. Huseby. When do (imposed) economic sanctions work? World Politics, 57(4):479–499, 2006.
[10] S. Jayachandran and M. Kremer. Odious debt. The American economic review, 96(1):82–92, 2006.
[11] V.L. Krustev. Strategic demands, credible threats, and economic coercion outcomes. International Studies Quarterly, 54(1):147–174, 2010.
[12] D. Lacy and E. Niou. A theory of economic sanctions and issue linkage: The roles of preferences, information, and threats. Journal of Politics, 66(1):25–42, 2004.
[13] J.M. Lindsay. Trade sanctions as policy instruments: A re-examination. International Studies Quarterly, pages 153–173, 1986.
[14] T.C. Morgan and A.C. Miers. When threats succeed: A formal model of the threat and use of economic sanctions. In Annual Meeting of the American Political Science Association, Atlanta, GA, 1999.
[15] G. Tsebelis. Nested games: Rational choice in comparative politics, volume 18. Univ of California Press, 1990.
