How TUFF Is It To Track Chinese Aid? Insights from Beijing’s Latest White Paper on Foreign Aid
Despite China’s growing reputation for investing in the development of other countries, they have been historically reluctant to provide details regarding what they are funding and where. On July 10, China’s State Council published a white paper providing an overview of its funding for overseas development activities and its 2010-2012 aid delivery framework.
The white paper does not provide a full accounting for Chinese official financing to other countries, but it does reference several examples of China’s development finance activities. We took this as an opportunity to test how extensively AidData has been able to capture Chinese funding to Africa with our Tracking Underreported Financial Flows (TUFF) methodology.
Since 2013, AidData has been working to shed light on these financial flows, publishing the most comprehensive database on Chinese development finance activities in Africa to date. We collect and synthesize data from media reports, government documents and databases, Chinese embassy websites, NGO reports and scholarly articles to capture a more complete picture of official Chinese financing to Africa.
Projects Identified by AidData Compared to Chinese White Paper
Comparing AidData’s estimates of Chinese development finance to Africa against the new white paper, we were eager to explore whether and to what extent our methodology might suffer from systematic biases. The chart below shows the results of our analysis, displaying all Chinese development projects explicitly mentioned in the white paper alongside the corresponding project entries on china.aiddata.org.
So, how well did we do? In short, it looks like AidData’s database of Chinese development finance activities to Africa and projects closely matches those reported in the white paper, at least for certain types of development financing. Using the TUFF methodology, AidData previously identified ninety-three percent of economic infrastructure projects and eighty percent of medical supply projects mentioned in the white paper. However, our approach appears to underestimate Chinese activities in smaller-scale development activities, such as technical support in the agriculture sector or drilling wells. This divergence may be due to a tendency among media outlets and official information sources to place emphasis on more visible and tangible projects.
As more information about China’s development activities becomes available, we’ll continue to test our approach and refine the TUFF methodology. This summer, our team has also begun enriching our existing data by adding specific subnational locations for communities benefiting from these projects. For example, we have identified Chinese financing for the construction of the Stade d'Angondje in Gabon and the Grand National Theater in Senegal, as well as the Tappita Hospital in Liberia. So far, we have been able to assign over 3000 sub-national locations for Chinese development finance activities and corroborated most of them through multiple media reports, independent third-party citations, and official government reports.
In the coming months, we will release these new geocoded data on Chinese development finance activities to Africa along with an interactive geospatial dashboard that allows users to view these investments in broader context – for example, by visualizing China’s development finance activities vis-à-vis subnational data on poverty, conflict, population density, etc. We look forward to sharing these new data and tools with you and hope that it will fuel more research and analysis around the effects of Chinese development finance to Africa.
Charles Perla is a Research Associate and Harsh Desai is a Research Assistant with AidData focusing on Chinese development finance in Africa. The above table was revised on August 5, 2014 to reflect the following changes: in "Agriculture," AidData actually captured 5 out of 5 rather than 5 of 7 Demo. Centers, which places the percentage for Demo. Centers at 100% and the total percentage for Agriculture at 57%. Also under "Agriculture," AidData captured 3 of 8 "Experts (Sent)," putting the category at 38% and the overall "Agriculture" category at now 64%.