Late-Stage Displacement and the Additionality of Multilateral Development Banks

Sammanfattning

The development finance architecture has increasingly relied on multilateral development banks (MDBs) to mobilise private capital in support of the Sustainable Development Goals. Practitioners, however, describe a recurring frustration: MDBs entering transactions after the private sector has already established contact and shown appetite to invest. This thesis examines whether this pattern on the project-level matches the aggregate data on the country-level and uses a mixed-methods design combining thirteen semi-structured interviews with cross-sector practitioners across eight regions with a two-way fixed effects panel of 105 developing countries over 2000-2024. Disaggregated OECD DAC data isolates IFC-specific flows from general MDB lending, enabling a more precise test than prior aggregate specifications. The qualitative results document late-stage displacement as a consistent pattern grounded in practitioner perceptions. The quantitative analysis documents a robust positive within-country correlation between general MDB lending and gross fixed capital formation, alongside a null result for IFC-specific flows. The absence of a detectable IFC effect at the country level is consistent with the mechanism operating beneath the aggregate level. If the IFC displaces private capital in certain transactions whilst catalysing it in others, the net-country effect could be indistinguishable from zero. The thesis contributes a theoretical account of late-stage displacement as a form of crowding-out in the development finance context, the first recurrent practitioner-based documentation of the late-stage displacement mechanism, and a quantitative decomposition isolating IFC activity from aggregate MDB flows.