World Bank Reform Analysis · April 2026
Institutional Power Architecture and Portfolio Distortion at the World Bank
Economist Dominance in Country Management, DPO Instrument Bias, and Development Outcome Costs
↓ Download the Full Paper (PDF)Lowest of any GP at scale
104 evaluated projects
below standard since 2015
across 8 Africa CMUs studied
The World Bank’s project portfolio is not primarily shaped by development need or comparative advantage. It is shaped by the internal power structure of the institution. The Macroeconomics, Trade and Investment (MTI) Global Practice records the worst development outcome rate of any practice at meaningful scale — 18.5 percent Satisfactory globally, 13.5 percent in fragile and conflict-affected states — while managing the largest committed volumes through the Development Policy Operation instrument.
This paper documents the mechanism. Across 60+ Country Director tenures in eight major Sub-Saharan African CMUs from 2000 to 2023, Country Directors with primary training in macroeconomics and careers rooted in the Bank’s PREM and MTI networks consistently allocate 1.5 to 2 times more of the IDA lending envelope to DPOs than Country Directors from engineering, public health, agriculture, or operations backgrounds. The differential averages 18–20 percentage points of total committed lending.
The paper is organised in three parts. Part I documents the institutional power architecture: the four-stage economist pipeline from Young Professional entry to Country Director; the MTI portfolio inversion (largest footprint, worst results); the DPO instrument as a mechanism of institutional power; and the CMU as the battleground on which GP competition is decided. Part II presents seven cross-cutting empirical findings, including the Marie-Nelly natural experiment — three countries, same instrument preference, near-experimental evidence that professional identity not country context determines DPO intensity. Part III sets out five structural reform recommendations and the priority CV analysis follow-on.
Annex A contains the full country-by-country evidence across all eight CMUs: Nigeria, South Africa, Ethiopia, Kenya, Tanzania, Senegal, Ghana, and DRC. Every Country Director tenure is documented with professional background, DPO profile, and estimated DPO share. The Ghana Kerali-Laporte natural experiment — 12 percent DPO share versus 43 percent in consecutive tenures in the same country, ending in debt distress — is the strongest individual-level evidence in the dataset.
All outcome data are from the IEG Master Database March 2026, covering 3,507 deduplicated projects rated FY2015–2026. These numbers supersede earlier figures from the February 2026 analytical series.