For nearly two decades, the scale and scope of Chinese state-led finance across the Global South have reshaped infrastructure, debt architecture, and geopolitical alignments. Yet, for much of this period, a comprehensive public ledger of these activities remained elusive. The (version 2.0, released in 2021) fundamentally changed this landscape. The simple act of downloading this dataset —a CSV file of over 20,000 projects across 165 countries—represents more than just a data acquisition process. It is a critical step toward empirical rigor, policy accountability, and a balanced understanding of 21st-century development finance.
Before discussing the download process, one must understand what the dataset contains. Compiled by the College of William & Mary’s Aiddata lab in partnership with the German Development Institute, this dataset aggregates official Chinese-financed projects from 2000 to 2021. It records not only loans and grants from the two major policy banks (China Exim Bank and CDB) but also details often omitted from official Chinese or World Bank data: precise geolocations, interest rates, maturity periods, and sector classifications. Downloading the dataset (typically available in .xlsx or .csv format from the Aiddata website) grants researchers, journalists, and policymakers access to a unique, ground-verified source that triangulates Chinese government documents, host country budgets, and media reports.
Downloading the dataset is a straightforward, democratizing process. One navigates to the Aiddata research data portal (aiddata.wm.edu), locates the "Global Chinese Development Finance Dataset" under the "Datasets" tab, and selects the most recent version. The file is often compressed (ZIP format) and requires standard software like Excel, Python (Pandas), or R to open. A key consideration for users is version control: older versions (e.g., 1.0 from 2017) have been superseded by Version 2.0, which corrects double-counting errors and adds hundreds of new projects. Thus, downloading the current version is essential for accurate analysis.