Reassessing world bank conditionality : beyond count measures

Tools

Winter, Jacob and Cormier, Ben and Kramarz, Teres and Manger, Mark (2025) Reassessing world bank conditionality : beyond count measures. Political Science Research and Methods. ISSN 2049-8489 (In Press)

[thumbnail of Winter-etal-PSRM-2025-Reassessing-world-bank-conditionaltiy] Text. Filename: Winter-etal-PSRM-2025-Reassessing-world-bank-conditionaltiy.pdf
Accepted Author Manuscript
Restricted to Repository staff only until 1 January 2099.
Download (5MB) | Request a copy

Abstract

Many studies argue the World Bank grants favorable loan conditions to allies of its powerful principals. These studies typically use the count of loan conditions as a proxy for how demanding loans are on borrowers. But not all conditions are equal: the text of conditions varies, making some conditions more difficult to comply with than others. We propose to use the stringency of conditionality text as the unit of analysis in studies of World Bank conditionality. Using the Bank’s internal reviews to define stringency, we construct a measure of conditionality stringency in Bank loans using Latent Semantic Scaling quantitative text methods. Using this new measure, we find that powerful principals have no generalizable effect on World Bank conditionality text. Instead, the stringency of loan conditions is associated with factors that inform bureaucratic assessments of risk. This highlights that evidence of dominant principals or bureaucratic behaviors influencing Bank lending depend on the specific aspect of loan conditionality being used as the dependent variable. To facilitate future research on the substance of loan conditions, our replication materials provide a new dataset of World Bank loan condition text and a corresponding measure of text stringency for all loans in the dataset.

ORCID iDs

Winter, Jacob, Cormier, Ben ORCID logoORCID: https://orcid.org/0000-0001-9278-5308, Kramarz, Teres and Manger, Mark;
CORE (COnnecting REpositories)