Download the dataset here

How do donors fund multilaterals?

In the face of strategic shifts and funding changes, annual figures are important when considering flows of multilateral funding. High-profile pledges to individual organizations are often made at replenishment events and conferences without binding agreements, meaning that a simple overview of pledges made may obscure facts critical to advocates, such as pledge fulfillment rates, timespans, and conditions that enhance or detract from funding effectiveness. This page brings clarity by showing donors' annual disbursements and pledges across multiple data sources, revealing not only what donors have pledged but also what they have allocated in their budgets and what has been transferred on an annual basis and on what conditions.

The dropdown menus in the top right corner allow you to select the donor and multilateral of interest. The 'Status' and 'Channel' buttons allow switching between two relevant dimensions: pledged vs disbursed funding and core vs earmarked funding.

Status: pledges vs contributions

Pledges are political commitments announced at replenishment conferences, often non-binding and spread over multiple years. Contributions are funds actually disbursed or firmly committed through formal agreements and budget allocations. This distinction is relevant as pledges do not guarantee funding will materialize. For example, across the Global Fund's replenishment history, approximately 8.5% of pledges have not been converted to contributions due to payment delays, economic challenges, or political shifts. In addition, disbursement patterns also vary. Australia shows irregular payment schedules to the Global Fund with significant annual variation, whilst Denmark tends to disburse more evenly. Understanding where donors are in the pledge-to-contribution pipeline helps advocates identify funding gaps early.

Channel: Core vs earmarked funding

As donors move to more closely align development funding with national interest, they are increasingly funding through earmarked contributions, which has been referred to as a "bilateralization" of multilateral organizations. This shift is critical when considering the relative "quality" of ODA, which offers nuance beyond comparisons of absolute figures.

Core funding goes directly into organizational budgets and can be allocated as deemed appropriate within their mandates. Earmarked funds are restricted to specific projects, regions, or countries. Core funding is typically considered more effective for project-level outcomes and preferred by multilaterals because it offers greater flexibility. Earmarking, on the other hand, allows donors greater control and allows political priorities to exert a greater influence over multilateral funding. However, research shows that this can undermine recipient-country ownership by bypassing national priorities and fragmenting development efforts. From official OECD data, we can see that donors like the US and Denmark typically earmark about half of multilateral funding, whereas donors like France and Belgium reliably earmark only a minimal portion of contributions

The figures in the Multilateral Funding Tracker represent our best estimate of current and near-term multilateral funding flows. They are based on the sources and hierarchy described in the Purpose and Approach section. Beyond 2024, no official data exists, and therefore estimates are likely incomplete. From 2026 onwards, donors may still make commitments, for example, to CEPI's next replenishment period.

Donors sometimes elect to not report funding to the OECD, which can cause gaps in OECD data that do not accurately reflect contributions to multilateral organizations. For the purposes of this dataset, if there is no funding reported to the OECD by a particular donor for a particular year that otherwise has OECD data, data is pulled as available from other sources according to the prioritization methodology, excluding calculated values (see section Purpose and Approach). If no other data source is available, funding is reported as zero.