Where to find international development data

International development data is one of the few categories in this directory where every serious option is free. These tools are built by universities, UN agencies, and government-funded programs as public goods, not commercial products, so the decision isn't about price — it's about matching the right source to the right question, and understanding that "free" doesn't always mean "unrestricted" or "guaranteed to still be there next year." This guide is for researchers, NGOs, journalists, and policy analysts who need data on aid flows, health, food security, or household welfare in developing countries. If you need data about a developed economy's population, look at census and demographic data instead.

Decide what kind of question you're asking

• "Who is funding what, where?" AidData tracks development finance flows, with particular depth on Chinese official lending and grants — a dataset of roughly 31,750 loans and grants across about 200 countries that doesn't have a close free equivalent elsewhere.
• "What are health and welfare conditions in a household or population?" The DHS Program runs nationally representative household surveys on population, maternal and child health, nutrition, and HIV across nearly 100 countries. The World Bank Microdata Library catalogs household-level survey and administrative microdata more broadly, including its own Living Standards Measurement Study collection.
• "Is a food crisis emerging, and where?" FEWS NET is purpose-built for this: monthly forecasts and IPC-compatible food-insecurity classifications across 30+ countries, aimed at humanitarian response planning rather than general research.
• "What indicators track child welfare specifically?" UNICEF Data compiles child mortality, nutrition, education, and protection indicators for nearly 200 countries, several produced jointly with WHO and the World Bank.
• "I don't know which dataset I need — show me what exists." the Humanitarian Data Exchange and the Global Health Data Exchange are catalogs: they index data produced by many other organizations rather than running their own surveys.

Catalogs vs. primary data producers

This is the distinction that trips people up most in this category. HDX and GHDx are data catalogs — HDX hosts over 20,000 datasets from more than 1,900 contributing sources on a CKAN-based platform, and GHDx indexes the surveys, censuses, and vital-registration systems that feed IHME's Global Burden of Disease study, with citation-level provenance rather than the underlying microdata itself in every case. AidData, the DHS Program, FEWS NET, UNICEF Data, and the World Bank Microdata Library are closer to primary sources: they produce, commission, or directly compile the underlying indicators and survey data, even when — as with the World Bank's library — some datasets are contributed by partner agencies. If you want to know what's available across the whole field, start with a catalog. If you already know the topic, go straight to the primary source.

Household microdata vs. aggregated indicators

A second real difference is the level of the data you get back. DHS and the World Bank Microdata Library give you de-identified household or individual-level records — you can build your own cross-tabulations, weighted by the survey's own weighting scheme, subject to registration and, for some World Bank-catalogued datasets, an access request because ownership sits with the contributing agency. UNICEF Data and FEWS NET publish already-aggregated country- or region-level indicators and classifications — faster to use if a country-level number is all you need, but you can't re-cut them by demographic group yourself.

Continuity is a real risk in this category, not a footnote

Because most of these are grant- and government-funded rather than commercial, funding disruption is a genuine operational risk you don't see with subscription software. The DHS Program's own record makes this concrete: after a 2025 US funding disruption affecting USAID, the program's continued website and archive access has depended on a Gates Foundation grant rather than its original funder. Before building a research pipeline or a monitoring dashboard around any one of these sources, it's worth asking who funds it and what happens to access if that funding changes — and keeping local copies of any dataset your work depends on.

How "pricing" works here

Every tool in this category is free at the point of use — funded by universities, federal research grants, UN member-state contributions, or (for AidData and DHS) USAID and philanthropic support. The catch is not cost but access terms: DHS microdata requires registration and an approved research use; a subset of World Bank Microdata Library datasets carries restrictions because a partner agency, not the World Bank, owns them. Budget time for registration and data-use agreements, not money.

A shortlist by situation

• You're tracking global aid and development finance, especially from China. AidData.
• You need household-level health survey microdata across developing countries. The DHS Program for health-specific surveys; the World Bank Microdata Library for a broader range including welfare and poverty microdata.
• You're monitoring or forecasting food insecurity. FEWS NET is the specialist tool for this.
• You need child-specific indicators across nearly 200 countries. UNICEF Data.
• You don't know where to start, or need to find data from many organizations at once. the Humanitarian Data Exchange for humanitarian and crisis data broadly, the Global Health Data Exchange if the search is specifically for health data provenance.

Questions to ask before you rely on a source

1. Is this a primary data producer or a catalog pointing to someone else's dataset — and if the latter, who actually owns and maintains it?
2. What geography and time-period resolution does the data provide, and does that match the decision I'm making?
3. Does accessing this data require registration, an approved research use, or a data-use agreement, and how long does that take?
4. Who funds this source, and what's the risk to continued access if that funding changes?
5. If I need this ongoing, should I keep a local archive rather than relying on the live site?

Common mistakes

Treating a catalog entry as if it were the primary dataset, when the underlying data may sit with — and require permission from — the contributing agency. Assuming free means unrestricted, when a meaningful share of development microdata carries real access conditions. And building a dependency on a single free source without a contingency plan, in a category where funding continuity is not guaranteed the way it is for a commercial subscription.

For head-to-head detail, see The DHS Program vs World Bank Microdata Library and FEWS NET vs Humanitarian Data Exchange, or browse every tool in this category.