Africa’s net capital flows have shifted sharply as repayments to China and other bilateral lenders now exceed new funding, highlighting a structural change in the continent’s development finance landscape.

Analysis from ONE Data, published alongside the launch of the Development Finance Observatory, shows that over the past decade Chinese lenders moved from being a net provider of finance to a net extractor. Between 2010 and 2014, Africa received roughly 30 billion US dollars in net Chinese financing. Between 2020 and 2024, the continent repaid 22 billion US dollars, representing a net outflow of 22 billion US dollars, a 52 billion US dollar swing.

The Observatory, backed by four million US dollars from Google.org and The Rockefeller Foundation, is an interactive platform integrating financial inflows, including official development assistance and new sovereign lending, with outflows such as debt servicing. It aims to address long-standing gaps in development finance data, which are often fragmented and siloed.

David McNair, executive director of ONE Data, said accurate and timely information was critical for decision-making in a period of tightening finance. “This platform allows policymakers, investors and citizens to follow the money and understand the impact it is having,” he said.

The Observatory also shows a broader retreat in private finance. Public and publicly guaranteed long-term external debt from private sources fell from 19 percent of net flows in 2010–2014 to just one percent in 2020–2024. Net new private financing dropped from 115 billion US dollars to 7.3 billion US dollars over the same period.

At the same time, multilateral lenders have expanded their role. Net financing from institutions such as the World Bank grew 124 percent, now accounting for 56 percent of total net flows to developing countries, equivalent to 378.7 billion US dollars between 2020 and 2024.

The Observatory’s platform uses Google’s Data Commons infrastructure and AI-enabled tools to allow users to query complex datasets using natural-language questions. The Rockefeller Foundation is contributing 700,000 US dollars to support country-level analysis and integration of fragmented datasets.

Mike Muldoon, chief of staff at The Rockefeller Foundation, said the platform would provide policymakers a comprehensive view of aid, debt and repayments. “It ensures that every dollar, public and private, is spent efficiently,” he said.