MDB Reform · Independent Analysis | mdbreform.com | March 2026
By Parminder Brar | Former Country Manager, Sierra Leone | Lead Governance Specialist, World Bank | mdbreform.com | March 2026
The Board of Executive Directors approves every World Bank loan. It meets twice a week. It reviews every project appraisal document. It formally authorises each operation. In eighty-two years it has never formally held management accountable for whether those operations achieved their development objectives. This is not an oversight. It is the architecture.
| Metric | Number |
| World Bank operations rated by IEG since 1973 | 13,273 |
| Did not achieve Satisfactory outcomes (S+) | 53% |
| Committed to below-Satisfactory operations | $577bn of $1,023bn total |
| MS+ (Moderately Satisfactory or above — Bank’s headline metric) | 72% |
| S+ (Satisfactory or above — the honest bar) | 47% |
| Times the Board has formally held management accountable for these results | 0 |
A note on the numbers: The World Bank typically reports its portfolio performance using MS+ — Moderately Satisfactory or above. By that measure, approximately 72 percent of projects pass. This series uses the stricter S+ threshold — Satisfactory or above — which reflects genuine achievement of development objectives. The gap between these two measures (72% vs 47%) is itself a governance question. Moderately Satisfactory means the project mostly worked, with significant shortfalls. It is not a result worth celebrating.
I. The Approval Architecture
The Board of Executive Directors is the World Bank’s governing body. It consists of 25 Executive Directors, who are resident in Washington DC and meet twice a week. It approves every single loan the Bank makes — approximately 300 operations per year. Before any money is lent, a Project Appraisal Document is prepared, reviewed, and voted on by the Board.
This arrangement is presented as rigorous oversight. It is the opposite. When the Board approves a project, it becomes a co-author of that project. It has formally decided the project is sound, well-designed, and worth lending money for. When that project subsequently fails — as more than half do — the Board cannot scrutinise the failure without simultaneously acknowledging that it approved the project in the first place.
A Board that approves everything cannot be held accountable for anything. The approval function and the oversight function are structurally incompatible when held by the same body.
II. The Performance Record
The IEG has rated 13,273 World Bank operations since 1973. 47 percent achieved Satisfactory outcomes or above. The trend over fifty years is the most important context for any governance discussion.
| Period | Projects | S+ Rate | MS+ Rate |
|---|---|---|---|
| 1970–74 | 674 | 81.8% | ~92% |
| 1980–84 | 1,206 | 62.1% | ~79% |
| 1990–94 — Wappenhans Report | 1,363 | 50.4% | ~72% |
| 2000–04 | 1,452 | 34.4% | ~62% |
| 2005–09 — lowest point | 1,684 | 26.4% | ~56% |
| 2020–24 | 221 | 38.9% | ~65% |
In 1992 a Vice President of the Bank — Wappenhans — wrote a report identifying an approval culture: the institution cared more about getting projects approved than about whether they would work. The S+ rate continued falling for fifteen years after that report and has never recovered to pre-1985 levels.
III. The Zedillo Finding: The Impossible Trinity
In 2009 a commission chaired by Ernesto Zedillo — former President of Mexico — examined World Bank governance. It identified what it called an impossible trinity: the Board was attempting to do three things simultaneously that cannot all be done by the same body.
| Board Function | The Problem |
|---|---|
| Represent member countries politically | Creates constituency interests that shape approval decisions |
| Share responsibility for Bank policy | Makes the Board a co-author of management decisions |
| Oversee management and hold it accountable | Cannot oversee what you have co-approved |
The Zedillo Commission’s structural recommendations — delegate project approval, establish formal performance review, create genuine separation between oversight and approval — were discussed, endorsed in principle, and not implemented. That was seventeen years ago.
“The Board’s involvement in operational work is so extensive that it compromises the Board’s ability to provide effective oversight.” — Zedillo Commission, 2009
IV. What the Board Costs — And What It Costs To Have No Oversight
The World Bank’s Board of Executive Directors costs approximately $84.4 million per year — the highest Board cost of any multilateral development bank. The AIIB, founded in 2016 with a non-resident Board, spends approximately $5 million annually.
| Institution | Board Model | Annual Board Cost | S+ Rate (recent) |
|---|---|---|---|
| World Bank | Resident — meets twice weekly, approves all loans | $84.4M | 38–42% |
| Asian Infrastructure Investment Bank | Non-resident — meets 6 times/year, strategic oversight only | ~$5M | Too early to assess |
| New Development Bank | Non-resident — delegated lending authority | Significantly lower | Too early to assess |
| European Investment Bank | Delegated lending authority, Board focuses on strategy | Lower | AAA rated |
The cost comparison is the least important argument. The more important argument is functional: a Board that spends its time approving 300 individual operations per year has no time to evaluate whether the previous 300 achieved their objectives. The approval workload structurally crowds out the oversight function.
V. Three Reforms
1. Delegate project approval to management
The Board should delegate approval authority to the President for operations within policy-compliant parameters — following AIIB and NDB precedent. The Board retains authority for operations above specified financial thresholds, new or high-risk categories, and operations raising novel policy questions. A Board that does not spend its time approving 185 individual operations per year has time to evaluate portfolio performance.
2. Establish a formal annual performance review of the President
The Zedillo Commission recommended this in 2009. It has never been implemented. Such a review would include IEG S+ rates by region, instrument, and practice; quality-at-entry trends; and management’s response to IEG findings. It would be public. This would be the first formal mechanism in the Bank’s history by which management is held accountable for development outcomes.
3. Have IEG report directly to the Board before management
Currently, management receives and responds to IEG’s annual findings before the Board encounters them. The Board’s first exposure to the performance data is through a management-framed response. IEG should brief a Board oversight committee before management receives the findings. Management’s response comes second — as a reaction to Board questions, not as the frame through which the Board first reads the evidence.
Delegating project approval to management is not a weakening of accountability. It is its precondition. A Board that approves everything can be held accountable for nothing. A Board that delegates operations and holds management accountable for portfolio results creates, for the first time in the World Bank’s history, a mechanism with actual consequences.
VI. The IDA21 Context
IDA21’s $100 billion replenishment — the largest in IDA’s history — has introduced strengthened borrower accountability through the Sustainable Development Finance Policy. It has introduced no strengthened Board oversight of management. Shareholders made commitments to their publics about development results. Those commitments cannot be enforced through a governance structure in which the Board that monitors results is the same Board that co-approved every operation through which those results were to be achieved.
Tomorrow: The IMF. The ship directly across 19th Street. It has no outcome ratings at all. Not one. In 80 years.
Data Disclaimer
The IEG Master Database underlying this analysis has been compiled from publicly available World Bank and IEG sources and is published on mdbreform.com in good faith as a research resource. Every reasonable effort has been made to ensure accuracy. This database is an independent compilation and has not been verified or endorsed by the World Bank Group or the Independent Evaluation Group. Users should verify individual project figures against primary sources before relying on them for official or formal purposes. mdbreform.com accepts no responsibility for errors or omissions. The analytical conclusions represent the independent views of the author and do not reflect the positions of the World Bank Group or any of its member institutions.
Analysis based on the IEG evaluation database (13,273 rated operations, FY1973–2026) matched to World Bank commitment data. S+ = Satisfactory or Highly Satisfactory. MS+ = Moderately Satisfactory or above. All data drawn exclusively from publicly available IEG and World Bank sources.
Parminder Brar | mdbreform.com | Former Country Manager Sierra Leone and Lead Governance Specialist, World Bank | 20 years. 500 missions.