Does research on economic sanctions suffer from publication bias? A meta-analysis

Binyam Afewerk Demena, Gabriela Benalcazar Jativa, Alemayehu S. Reta, Patrick B. Kimararungu, Peter van Bergeijk

Research output: Working paperAcademic

13 Downloads (Pure)

Abstract

We meta-analyse 36 primary studies on determinants of the effectiveness of economic sanctions published over the years 1985-2018, using the Protocol of the Meta-Analysis in Economics Research-network. We investigate the impact of trade linkage, sanction duration and prior relations on sanction success. While the descriptive analysis and weighted averages suggest that the impact of the three variables of interest is significant and conforms to a priori theoretical expectations, our econometric analysis uncovers significant publication bias in the results. Bias is significant and large for the three variables of interest and the genuine impact of these variables on success and failure of sanctions after correction for publication bias is insignificant. Moreover, we find that bias in this literature increases over time.
Original languageEnglish
Place of PublicationThe Hague
PublisherInternational Institute of Social Studies (ISS)
Number of pages26
Publication statusPublished - Mar 2021

Publication series

SeriesISS working papers. General series
Number674

Bibliographical note

JEL Classification: F14
link in repub: hdl.handle.net/1765/135342

This is the pre-peer review version of Chapter 6 of the Research Handbook on Economic Sanctions (Edward Elgar, forthcoming 2021). This research is based on a MA project that was awarded the ISS prize for the best Student Research Project in 2018.

A preliminary version of this paper was presented at the 19th Jan Tinbergen Peace Science Conference and summaries appear as van Bergeijk et al. (2019) and van Bergeijk (2019).

Research programs

  • ISS-DE

Series

  • ISS Working Paper-General Series

Fingerprint

Dive into the research topics of 'Does research on economic sanctions suffer from publication bias? A meta-analysis'. Together they form a unique fingerprint.

Cite this