VoxEU Column Politics and economics

Failure and success of economic sanctions

Can international economic pressure induce policy changes? The conventional wisdom, among economists at least, is that economic sanctions, for all their posturing, won’t achieve very much. For better or worse, this column shows that this is now changing.

The discussion about the effectiveness of economic sanctions as an instrument of foreign policy recently revived due to the recent oil embargo of Iran and the European embargo on equipment for the Syrian oil and gas industry. The consensus view seems to be that economic sanctions – despite the long history of and experience with this instrument – are still completely ineffective. This may have been the case, but over the last 15 years the situation has changed.

Table 1. Key characteristics of sanctions before and after 1990

 
1946–1989
After 1990
Ongoing sanctions
yes
no
yes
no
Annual average number of sanctions
2.6
2.3
4.5
3.8
Share of successes
32%
34%
39%
40%
Trade linkage
 
 
 
 
Share of target’s trade
22%
23%
45%
45%
Percentage of target’s GDP
6%
6%
11%
10%
Period (years)
8.9
7.8
4.2
4.1

Sources: Number of sanctions, success rate, and period calculated from Hufbauer et al (2008). Trade linkage based on many economic sources for trade and GDP as detailed in Van Bergeijk (2009, Appendix 6.A1, pp 138–46).

Table 1 reports key characteristics of 172 economic sanction episodes imposed since the World War II as reported in the Peterson Institute’s sanction database (Hufbauer et al 2008).1 The Hufbauer et al dataset is probably the best available data source on economic sanctions although its verdict on the success of sanctions has been contested, especially in individual cases, and because of changes in methodology between in the most recent edition (Shahadat Hossain Siddiquee and van Bergeijk 2012). Since 1990 the use of economic sanctions doubled and with better results: 40% is considered successful.2 The empirical lessons of the history of economic sanctions appear to have been taken to heart and this has led to a better design and selection of cases in which sanctions are deployed.

Why sanctions fail

The reasons sanctions fail are straightforward, as these three historical examples show.

  1. Probably the most adhered-to sanction of all time is the Canadian withdrawal of South African landing rights during apartheid. After the sanction was imposed, not a single South African aeroplane landed in Canada. The sobering fact, however, is that not a single South African aeroplane landed in Canada before the sanction either. It is clear that such a sanction will have no effect. This also true for a great many sanctions in which the pre-sanction economic relationships between the country (group) that imposes the sanction and the target nation are simply too small to have any (economic) impact. It is obvious for any economist that this manner of sanctions will be ineffective: the success rate is 50% if bilateral trade in per cent of the target’s GDP is 10% or more; if bilateral trade amounts to less than 2% of the target’s GDP then the failure rate is 80% (Figure 1).
  2. Reaching diplomatic agreement often requires a lot of time, as illustrated by the EU sanctions against the former Yugoslavia. The longer the negotiations, the better a sanction target can prepare, for example by building strategic reserves or changing its production structure. Again, the stylised facts are clear: two thirds of the successful sanctions reached their goals within three years; two thirds of the failed cases had a duration of (much) more than three years (van Bergeijk and van Marrewijk 1995).
  3. In the case of sanctions against Saddam Hussein, all the economic requirements for ‘a success’ were met. The sanctions were implemented quickly, on a global scale, and with a lot of economic damage. The decision and implementation was a matter of only four days, even traditionally neutral Switzerland participated, and the costs of economic sanctions in 1991 have been estimated at half the GDP (compare Smeets 1990). Still the economic sanctions against Iraq did not achieve their goal. Neither did the air strikes lead to the end of Saddam Hussein’s reign. The lesson is that some countries cannot be influenced because the decision-makers do not attribute decisive weight to the potential damage done to their country. Also this effect can be shown empirically: my logit analysis of the sanction cases reveals that the probability of a sanction succeeding against a democracy is two times as high as that against an autocracy.

Figure 1.Trade linkage, failures (left), successes (left) and success rate (right)

Sources: see Table 1.

Did policymakers learn?

As illustrated by these examples, much can go wrong with the organisation and implementation of sanctions. Obviously, however, policymakers have learned from their mistakes. As illustrated in Table 1, economic sanctions have a better track record as the success rate increased by about one fifth. Importantly, this improvement takes place in the context of a doubling of the trade linkage between sanction imposers and sanction targets, a halving of the duration of the sanctions, and a much larger share of democratically ruled target countries. One possible explanation is that the sanction instrument has been applied more selectively and more intelligently, but it is also possible that structural changes have increased the applicability.

  • First, globalisation has increased the importance of undisturbed international trade and investment for potential targets of economic sanctions: countries that three decades ago were more or less self-reliant can now be hit much harder by economic sanctions.
  • Second, the decision-making processes, especially in the United Nations, have been streamlined, increasing the speed of implementation.
  • Third, the world is much more democratic than during the Cold War.

Whatever the reason, it is clear that the idea that sanctions do not work needs reconsideration.

References

van Bergeijk, P A G (2009), Economic Diplomacy and the Geography of International Trade, Edward Elgar.
van Bergeijk, P A G and C van Marrewijk (1995), “Why Do Sanctions Need Time to Work?, Adjustment, Learning and Anticipation”, Economic Modelling, 12(2): 75–86.
Hufbauer, G C, J J Schott, K A Elliott, and B Oegg (2008), Economic Sanctions Reconsidered, 3rd edition, Peterson Institute for International Economics.
Shahadat Hossain Siddiquee, M and P A G van Bergeijk (2012), “Reconsidering Economic Sanctions Reconsidered: Detailed Analysis of the Peterson Institute Sanction Database”, ISS Working paper, Erasmus University: Rotterdam.

Smeets, M (1990), “Economic Sanctions Against Iraq: The Ideal Case?”, Journal of World Trade, 24(6):105–20.


1 Note that nine cases of the Hufbauer et al 2008 study had to be excluded since data for trade and or investment could not be collected from economic sources.

2 Table 1 distinguishes between sanctions that had ended when Hufbauer et al 2008 was published and cases that were still ongoing at that time.

6,404 Reads