Reconsidering economic sanctions reconsidered. A detailed analysis of the Peterson Institute sanction database

This paper analyses two vintages of the key resource for research on economic sanctions: the Peterson Institute database reported in Hufbauer et al. (2nd edition in 1990 and 3rd edition in 2007). The Peterson Institute has not reported transparently on these changes. We provide detailed tables in order to facilitate comparison between descriptive statistics and the findings of the two editions. One way to interpret our results is as are porting of the 2nd edition results corrected for changes in methodology and case selection. Using descriptive statistics, ratio analysis, first-difference method and probit we investigate how case selection, (re)coding and new observations impacted on sanction characteristics and assumed effectiveness of economic sanctions. About 17% of the common cases of the 2nd and 3rd edition is modified and changed to some extent. The number of goals assigned to these cases increased from 146 to 155. The average success score increases from 6.6 to 7.0 for the common cases. Indeed, the mean values for all categories of core variables for the common cases in the 3rd edition exceed those reported in the 2nd edition. A redefined index value of the ‘sanction contribution’ underlies these changes. The lowest value index is defined as zero or negative contribution in the in 2nd edition whereas is limited to negative contribution in the 3rd edition (upgrading all zero contributions by definition) Likewise ‘modest and significant contribution’ is used in the 3rd edition instead of ‘substantial and decisive contribution’, making it easier to get a high score. We provide a probit analysis that shows that the 3rd edition’s methodology in comparison to the methodology used in the 2nd edition is biased in favour of finding positive results for modest policy change, regime change and the use of sanctions to disrupt military adventures and to achieve military impairment.