DOI: doi.org/10.71734/DP-2026‑6

This study shows that traditional methods underestimate the trade-reducing effects of sanctions by nearly 50 %. Using heterogeneity-robust techniques that take account of the imposition and lifting of sanctions, we demonstrate that complete trade sanctions erode approximately 58 % of bilateral trade. However, on average, trade relations recover surprisingly quickly once sanctions are lifted. We also analyze the factors associated with the severity of these disruptions and the speed of recovery.

Economic sanctions have become an increasingly prominent tool used in international policy. Between 1950 and 2020, the number of active sanctions and newly imposed measures rose sharply. Over this period, almost every country in the world was either a sender or a target of sanctions at some point. Despite this proliferation, existing research likely underestimates the trade impact of sanctions because of methodological limitations. Traditional estimation approaches typically treat sanctions as a single, homogeneous intervention, assuming that their effects are the same across cases and constant over time. This “one-size-fits-all” assumption disregards the fact that sanctions vary significantly in intensity, duration, and dynamics. We address this shortcoming by using modern econometric methods that explicitly take account of this heterogeneity during both the imposition and lifting phases. This approach reveals effects that differ substantially from those found in previous work.

A heterogeneity-robust approach to the effects of sanctions

We combine modern difference-in-differences estimators with gravity models of trade to evaluate the effects of sanctions over their full lifecycle. Our preferred specification is an extended two-way fixed effects model that is designed to handle heterogeneous effects across sanction periods and over time. We draw on the Global Sanctions Database, the most comprehensive dataset on economic sanctions, and merge it with bilateral trade data for 260 countries from 1950 to 2019. We focus on complete trade sanctions that were both imposed and later lifted, yielding a sample of 795 country pairs across 37 sanction groups. This framework allows us to isolate the specific impact of the imposition and lifting of sanctions without relying on the restrictive assumptions of homogeneity used in earlier studies.

Trade effects of complete trade sanctions: magnitude and heterogeneity

Our results show that complete trade sanctions have substantially larger effects on trade than previously thought: they reduce bilateral trade by around 58 % while they are in force. Traditional methods, by contrast, point to a decline of only 38 %. We attribute this discrepancy to the limitations of standard two-way fixed effects estimators in handling heterogeneity. These methods give arbitrary weights to different sanction cases, often overweighting those with relatively small effects. They also suffer from contamination bias, as estimates for periods of active sanctions are distorted by data from years after sanctions were lifted. We show that this contamination explains roughly one-third of the difference between our estimates and earlier findings.

We also study whether trade relations recover once sanctions end. On average, we find that there is no statistically significant long-term impact on trade after sanctions are lifted. This is in contrast to traditional estimates using the same data, which suggest positive post-sanction effects. However, this average masks important differences across sanction cases. When sanctions last six years or less, trade typically rebounds to, and sometimes exceeds, pre-sanction levels. By contrast, sanctions that persist for more than six years show little evidence of a full recovery, suggesting that prolonged restrictions may have lasting effects on commercial ties.

Finally, we document systematic heterogeneity in sanction effects. Partners that are geographically distant and those that share a land border experience larger trade contractions. Institutional ties, such as joint membership of the World Trade Organization or participation in a regional trade agreement, soften the blow during the sanction period and help trade to recover more quickly afterwards. Sanctions motivated by war-related objectives and those ultimately judged unsuccessful are associated with particularly severe and persistent trade effects.

Conclusion

This study shows that complete trade sanctions have both powerful and uneven effects on international trade. On average, they reduce bilateral trade by more than half, although the size and duration of these losses differ across sanction cases. Short-lived sanctions tend to trigger sharp yet largely reversible declines in trade, while long-lasting sanctions can permanently affect bilateral trade. Observable characteristics such as geography, institutional ties and stated objectives are closely linked to the magnitude of the trade decline and the extent of the subsequent recovery. The study highlights the importance of explicitly accounting for this heterogeneity and shows that systematic variation across sanction cases can provide valuable information for the design and assessment of sanction policies.

Kwon, O., A. J. Nagengast, J. Yoon, Y. V. Yotov (2026), From imposition to lifting: Estimating the effects of sanctions over their lifecycle, Bundesbank Discussion Paper, No 06/2026