The Vote That Never Happens
You are a finance minister. Your country holds a modest shareholding in a major multilateral development bank, enough to feel like ownership, not enough to feel like power. One morning a project appears in the bank's published pipeline: a large hydropower dam, a highway corridor, a privatisation support programme in a neighbouring state. Nobody called you. No vote was scheduled, no formal objection invited, no record made of your country's position. The project moved through the institution like water finding its level, guided by channels you did not design and were never shown. You did not block it because the structure of the institution meant you were never asked.
This is not an oversight. It is, in most cases, the design.
Development banks, whether the World Bank Group, the African Development Bank, the Asian Infrastructure Investment Bank, or the dozens of smaller regional cousins, are governed by founding Articles of Agreement or charters that define exactly which decisions require shareholder approval and which ones belong entirely to a professional management class. Understanding that boundary, and how it was drawn, explains a great deal about which projects get funded and which grievances go formally unrecorded.
The Two-Tier Structure That Does Most of the Work
Every major multilateral development bank separates its governance into at least two bodies: a Board of Governors, which represents member countries at the ministerial or head-of-state level and meets infrequently, and a Board of Executive Directors (sometimes called the Board of Directors), which sits in permanent or near-permanent session and handles operational decisions. The Governors hold the supreme authority in theory. The Executive Directors exercise it in practice, almost continuously.
Then there is a third tier, less visible and more consequential for day-to-day lending: management itself, meaning the president and senior vice-presidents. The charter of almost every major development bank explicitly delegates a tranche of lending authority to management, without requiring any board vote at all. At the World Bank, this delegation is expressed in dollar ceilings and project-type categories. Loans under a certain size, or restructurings of existing loans below a defined threshold, can be approved by the president under what the institution calls "streamlined procedures." The board is informed after the fact, not consulted before.
The ceiling is not trivial. A project valued at, say, fifty million dollars in a low-income country can clear entirely within management's delegated authority, and for a small shareholder nation with a fractional voting share, that project will never appear on a ballot.
Consider how this plays out in practice. Two neighbouring countries both hold shares in a regional development bank: the first with a two percent voting share, the second with half a percent. A seventy-million-dollar urban water project in a third member state gets structured in two tranches of thirty-five million each, processed six months apart. Both fall under the management delegation ceiling. Neither shareholder is formally consulted. The first country might informally raise concerns through its executive director. The second, which does not hold a dedicated seat and instead shares a constituency chair with six other small states, has no practical channel at all. The project proceeds. This is standard operating procedure, not an aberration, and anyone who has spent time inside these institutions knows it.
Constituency Chairs and the Geometry of Representation
The board itself is not a parliament of equals. At the International Monetary Fund and the World Bank, the five or so largest shareholders each hold a permanent executive director seat. Most other countries are grouped into "constituencies," each of which elects a single executive director to represent them all. A constituency might contain twenty-five countries. The chair rotates, often on a two- or three-year cycle, among the more influential members of the group.
When a project comes before the board, that one executive director must represent the interests of every country in the constituency. If those interests diverge, and they very often do, the chair exercises discretion. A small island state in a constituency dominated by a middle-income regional power may find its objections absorbed into the larger bloc's consensus position, the way a minor key resolves into the chord the pianist intended all along. Its governor back home never voted, and its executive director spoke words that did not reflect its position. The formal record shows no dissent.
The African Development Bank's board structure illustrates this geometry well. African member states collectively hold a majority of shares, but they are distributed across many constituencies. Non-regional members, mostly wealthy industrialised nations, hold a significant minority block and have historically filled the senior management pipeline at rates disproportionate to their share count. Projects that attract non-regional capital can move faster through informal alignment between management and those shareholders' executive directors, before smaller regional shareholders fully organise a response. That is not a conspiracy. It is an incentive structure, and incentive structures produce predictable behaviour.
What People Get Wrong About "Shareholder Control"
The common assumption is that shareholding equals voting power, and voting power equals influence over projects. Wrong, in any operationally meaningful sense. The assumption holds only at the level of constitutional amendments, capital increases, and changes to the lending mandate itself. For the actual portfolio of projects, the more honest model runs like this: shareholding determines who sits on the board, board membership determines access to information and informal agenda-setting, and those two factors together shape what management chooses to bring forward and how it is packaged when it arrives.
Management learns, over years of institutional life, which projects will sail through and which will generate friction. The rational response is to structure borderline projects in ways that avoid friction, keeping them below delegation thresholds or pre-clearing them informally with the two or three executive directors whose constituencies hold enough votes to block. The shareholders who are never consulted are not being deceived. They are simply not in the room where the pre-clearance happens.
So ask yourself: if you discovered a project you oppose already listed in a bank's annual report, what recourse do you actually have? For a small-share member, the answer is almost certainly less than you assumed when your country signed the Articles of Agreement.
The Consequence of Invisible Approvals
None of this is secret. The delegation thresholds are published. The constituency arrangements are documented. The Articles of Agreement are public instruments. The design was deliberate, and its defenders make a reasonable argument: if every fifty-million-dollar project required a full shareholder vote, the institution would approve nothing on a useful timeline. Infrastructure lending would calcify. The operational logic of delegation is genuinely sound.
The political consequence, though, accumulates quietly. Countries that hold small shares in institutions funding projects within their own borders can find themselves in the paradoxical position of co-owning an institution whose specific decisions they had no formal opportunity to contest. Civil society groups in affected communities sometimes have more practical influence over outcomes, through the inspection panels and accountability mechanisms that several major banks have created, than the small-shareholder government sitting at the distant edge of a constituency bloc. That inversion should trouble anyone who believes that formal membership in an institution carries meaningful democratic weight.
The internal structure of a development bank is, at bottom, a theory of trust: a wager that professional management, supervised by the largest shareholders, will make decisions the others would have endorsed. History offers partial comfort on that score. What it does not offer is a guarantee, and the distance between those two things is where the smallest members of these institutions have always lived.