Since 1999, the World Bank has merged its private sector policy teams with the investment staff of the International Finance Corporation (IFC) no fewer than five times, with the most recent union now less than a year old. Each new arrangement was billed as a fundamental rethink of how the bank supports firms and markets, yet none lasted a decade.
The reorganizations are a reflex—an institutional response to a sense that the World Bank’s private sector work is spinning its wheels. The bank has never clearly articulated how its private sector work supports its primary mission of poverty alleviation, and a structural remedy cannot resolve this analytical problem.
World Bank President Ajay Banga introduced the most recent restructuring in late 2023, organizing the bank’s work into five thematic “verticals”: people, prosperity, planet, infrastructure, and digital. In October 2025, he announced that the knowledge teams of the World Bank’s public lending arms would merge with those of the IFC, its private investment arm, to shift from a model that rewards production to one that rewards impact. But without a clear standard for what work private sector development should do, the latest reorganization of the bank’s private sector development (PSD) may also falter.
In a new paper, I proposed a three-part test for the World Bank’s PSD operations: they should diagnose a binding market failure, address the economic distortion causing that failure, and change an investment or market outcome that would otherwise occur. Together, these elements provide a fixed standard for judging what PSD operations should do.
Going around in circles
In 1988, World Bank President Barber Conable commissioned a review of the World Bank Group’s private sector work. A report, an action program, and a standing committee followed. The bank then began to support private sector development, working through governments on the rules firms operate under. This put its new mandate on a collision course with the IFC’s.
Collisions were settled with mergers; new joint structures arrived in 1999, 2003, 2007, and 2014, predating President Banga’s verticals. The cycle of fusing IFC and PSD reflects recurrent institutional pressures to align PSD policy work with the IFC’s investment work and creates tension between World Bank policy advisers and IFC dealmakers. Success for an IFC unit is defined by deal return, disbursements, and commitment volumes. For a World Bank country director, it is defined by sovereign disbursements and completed operations and reforms. This clash in work cultures ultimately undermines any hope of alignment and accounts for the dissolution of each merger (though it does not account for the decision to attempt the next one).
Why the structures never hold
Reorganizations are justified by the need to address organizational silos, duplication, weak coordination, or slow client response. When reorganizations fail, people blame poor change management, weak incentives, and cultural resistance. Yet if coordination were the real problem, it seems one of these reorganizations would have produced a stable structure by now.
The absence of an analytical standard explains both why PSD is the practice that gets repeatedly redesigned and why each new leader redesigns it differently. If the World Bank cannot explain how private sector work contributes to its overall mission, then it cannot claim that a reorganization has failed in any meaningful way—it can only say that the reorganization was poorly executed.
Thinking on how private sector work should contribute to growth and poverty reduction is constantly shifting: privatization (the 1991 Developing the Private Sector: The World Bank’s Experience and Approach), investment climate (the 2005 World Development Report: A Better Investment Climate for Everyone), financial inclusion (the 2008 Finance for All?: Policies and Pitfalls in Expanding Access), private capital mobilization (the 2017 Maximizing Finance for Development), institutional evolution (the 2022 Evolution Roadmap), and now industrial policy. Six orthodoxies in four decades. Each new idea implies a different portfolio and a different approach to collaboration between the World Bank and the IFC. And with each shift, the temptation to restructure the practice becomes alluring once again.
Every World Bank practice goes through intellectual fads—orthodoxies. However, in agriculture, infrastructure, and macroeconomics, each has a single, unchanging objective. PSD is different. Each changing PSD orthodoxy above concerns what blocks firms. An orthodoxy can be replaced by a better orthodoxy, but an objective cannot be replaced; it can only be pursued differently. With PSD, the practice’s orthodoxies appear as competing ideas on how the private sector can reduce poverty. When a new idea emerges, there is nothing against which to test it; instead, the new idea replaces the objective itself. Once the objective changes, the units built around the old objective no longer fit, leading to a push to reorganize.
What would break the cycle?
In my new paper, I propose a three-hurdle test for every private sector operation (the hurdles would serve as admissibility conditions for a PSD operation—an appraisal is still necessary to establish positive, expected net social benefit and a link to the poverty and jobs objective):
1. Justification: The operation must identify a binding market failure.
2. Effectiveness: The instrument must address the distortion identified by the diagnosis and change the resulting behavior.
3. Additionality: There must be evidence that without World Bank Group participation, the investment, behavior, or market outcome would not have occurred on comparable terms and scale, or within a comparable timeframe.
The justification hurdle determines whether World Bank support removes a binding constraint or merely compensates for it. A misdiagnosed constraint results in an instrument selected for the wrong constraint, and such an instrument may compensate for the binding constraint without removing it. A failure of diagnosis is therefore a failure of justification, and it may become observable only when the instrument is examined at the effectiveness hurdle.
A review by the Independent Evaluation Group of 579 evaluated IFC investment projects in middle-income countries found that only 60 percent realized both financial and nonfinancial additionality. The same review found that 82 percent of projects anticipated both financial and nonfinancial additionality. While 96 percent realized some additionality, the gap between the 82 percent anticipated and the 60 percent realized suggests that operations teams habitually overestimate additionality. Operations teams deliver less than what they expected on additionality and would be surprised when the opposite happens.
Some may view the framework as simply the latest orthodoxy in a succession of doctrines, each replacing its predecessor. That interpretation, however, misunderstands its purpose. Earlier orthodoxies sought to identify the binding constraint on private sector development. Each identified a different binding constraint, whether state ownership, regulatory barriers, access to finance, or the absence of private capital. One cannot wrench an answer to these questions using welfare economics. That very old and well-studied branch of economics identifies no constraint and predicts nothing about what most constrains the private sector. Instead, it provides an anchor: a test a proposed operation must satisfy before public money is committed and it can be labeled a PSD operation. Earlier doctrines were expected to perform this function, but none succeeded. An orthodoxy about how private sector development works cannot also serve as the standard for evaluating claims about private sector development.
Agriculture, infrastructure, and macroeconomic work could be reorganized without losing the standard against which their work was judged. PSD has no standing objective, and changing the structure was the lever management could affect. A welfare economics anchor provides the fixed criterion PSD needs. It justifies using public funds to affect private activity when three hurdles are cleared. Instituting a practice around a test presents its own challenge, since a test cannot be housed in a structure in the same way a subject area can.
The World Bank could demonstrate, operation by operation, that its capital was deployed in areas where private capital would not have stepped in. And it cannot do so too soon: in 2025, official development assistance fell by 23 percent in real terms—the steepest annual contraction ever recorded by the OECD. As development finance becomes scarcer, institutions face growing pressure to demonstrate that public funding changed outcomes rather than simply displaced private investment.
