Fund Diversion in UN-Implemented, MDB-Financed Projects: Extending MDBs’ Contractual Reach and Fiduciary Instruments

Multilateral Development Banks (MDBs) channel funds to the United Nations (UN), either directly, when a UN agency is the recipient of an MDB grant (e.g., the World Bank’s grant  disbursed directly to the Food and Agriculture Organization of the United Nations under the Afghanistan Emergency Food and Security Project), or indirectly, when a borrower country engages a UN agency to implement an MDB-financed project (e.g., the Somali government engaged UNICEF with a World Bank grant under the Somalia Crisis Recovery Project).  As the UN is facing a financial crisis while MDBs are increasing lending headroom and considering deepening collaboration with the UN, UN agencies may receive more MDB financing. Allegations of fund diversion, a form of financial misconduct, in UNOPS’ S3i and UNDP’s Iraq operations have raised concerns about possible systemic failures in some UN agencies’ oversight mechanisms. Given the UN’s restrictions on external audits, MDBs face challenges in ensuring that financing for UN-implemented projects is used for its intended purposes.

Fraud and corruption are common forms of fund diversion. Ensuring that funds are used for their intended purposes is a mandate under MDBs' charters (e.g., EBRD, IBRD, AIIB, ADB and AfDB all have such charter mandate).  Pursuant to this mandate, MDBs investigate fraud and corruption and impose sanctions on the wrongdoers. When implementing MDB-financed projects, UN agencies may, with MDB approval, use their own procurement rules to source goods or services (see here and here). If MDBs seek access to information concerning alleged misconduct by contractors engaged by UN agencies, the UN may assert its privileges and immunities (P&I) or invoke its  single audit principle, potentially restricting MDBs' investigative and oversight authority.

MDB investigations involve interviewing concerned parties and inspecting records related to an allegation.  UN officials and experts on mission enjoy P&I that include immunity from legal process for “words spoken or written and all acts performed by them in their official capacity.” According to the same convention, the UN’s archives are also subject to P&I. In practice, the UN has routinely denied donors access to its accounting records and other financing documents, citing its single audit principle. It is debatable whether fraud and corruption by UN personnel constitute “acts performed by them in their official capacity,” whether project-specific procurement documents fall within the UN’s “archives,” and whether the single audit principle bars donors verifications undertaken solely to substantiate allegations of fraud and corruption, rather than to conduct a standard financial audit. Applying its policies conservatively, the UN may cite these grounds to restrict MDBs’ investigative and sanctions authority, as reflected in earlier World Bank–UN negotiations, limiting MDBs’ ability to ensure funds are used as intended.

As a workaround, MDBs could require UN agencies to include clauses in their goods and services contracts that expressly subject external contractors to MDBs’ investigative and sanctions authority and negotiate broader preventive fiduciary requirements with the UN. While UN personnel may qualify as officials or experts on mission entitled to P&I, external contractors may not be covered by the Convention on the Privileges and Immunities of the United Nations (1946). External contractors and service providers may include private companies and non-governmental organizations engaged through the UN’s procurement system under commercial contracts. The UN’s template contract explicitly excludes interpretations of employment and principal-agent relationships with these suppliers and service providers, suggesting that the UN’s personnel P&I would not automatically extend to them. Even where their books and records under a UN-implemented project may arguably be subject to the UN’s archive P&I, MDBs could still investigate these parties through interviews and site visits. External contractors may play a substantial role in MDB-financed, UN-implemented projects (e.g., under the Afghanistan Health Emergency Response Project, $270 million of the $330 million World Bank grant was budgeted for UNICEF’s performance-based contracts with service providers).  Securing contractual acknowledgment of MDBs’ investigative and sanctions authority over these parties contractually would likely encounter fewer UN policy obstacles. Express MDB oversight and remedies directly against the UN’s external contractors could strengthen safeguards over a significant part of MDB financing.

MDBs could also resort to a broad spectrum of proactive fiduciary instruments that are less likely to contravene the UN’s P&I and single audit principle. Before engaging a UN agency as a project implementer, MDBs may evaluate its oversight system, using the UN’s self-assessment as a baseline and testing it against the risk environment of a particular project. Depending on the results, MDBs may require written assurances from the UN, typically in the form of contractual covenants, to mitigate the project-specific weaknesses identified.

The World Bank’s standard-form agreement preserves its right to investigate third parties but does not expressly require those parties to acknowledge that right directly, potentially complicating enforcement. The World Bank’s earlier fiduciary principles framework focused on emergency projects, involved a limited number of UN agencies, and relied heavily on UN self-certification (see historical framework). The African Development Bank also adopted a fiduciary principles arrangement with the UN, which similarly covered certain UN agencies. Public data on other MDBs’ proactive integrity risk management measures for UN-implemented projects is sparse. Given the gaps, MDBs could apply the contractual instrument and proactive measures described above in proportion to each UN-implemented project’s risk profile, regardless of whether such UN agency is covered by an existing framework fiduciary agreement. MDBs could also form a coordinated negotiation task force to secure more transparent and coherent terms with UN agencies, particularly where different MDBs cofinance a project, including on a parallel basis, and where risks fall outside existing fiduciary principles framework or standard-form agreement. With the UN’s growing and urgent need for funding, MDBs may have more leverage in negotiations. Terms that were historically hard to close  may now be possible to agree upon.

Failures in the UN’s oversight system can erode donors’ trust. The same reputational risk may spill over to MDBs absent enhanced oversight of UN-implemented, MDB-financed projects. Requiring external contractors engaged by implementing UN agencies to contractually acknowledge MDBs’ investigative and sanctions authority and implementing more proactive fiduciary measures could help MDBs not only address fund diversion in UN-implemented projects but also maintain the confidence of MDBs’ shareholders  and funding partners.

Min (Loftus) Zhang is an LL.M. candidate in International Legal Studies at New York University School of Law and a Graduate Editor at the NYU Journal of Law & Business. Prior to his LL.M., Loftus spent more than six years at a multilateral development bank, advising on anticorruption legal and policy matters. Earlier, he interned at the Massachusetts Superior Court, the Office of the United Nations High Commissioner for Refugees, and a leading Malaysian law firm. During his master’s studies at Peking University School of Transnational Law and his undergraduate studies at Sichuan University Law School, he won Best Oralist awards at the Jessup (China National Round) and the CIETAC Cup International Commercial Arbitration Moot. His academic interests center on development and project finance, sovereign debt, and fraud and corruption risks in MDB financing.

The views expressed in this post are solely those of the author and do not necessarily reflect the views or policies of any institution with which the author was previously affiliated. This post draws solely on publicly available information.