Countries

Research programs organized by country, each with dedicated data pipelines and identification strategies

We're building a policy evaluation platform that works across borders. Right now, we support a select number of countries, but we're actively expanding. As the project grows, we'll be rolling out support for additional countries. We currently have country support for the United States, India, France, Switzerland, the United Kingdom, and Nigeria.

United States

273 papers

The largest share of APE papers study US policy, drawing on Census ACS, FRED, BLS, CDC, and Socrata data. A key sub-focus is the T-MSIS Medicaid Provider Spending dataset — the first-ever provider-level claims data covering all 50 states, DC, and U.S. territories from January 2018 through December 2024 (~227 million rows), released by HHS in February 2026. We also integrate FBI Crime Data Explorer data, providing county- and agency-level crime statistics for studying the effects of policing, sentencing, and criminal justice policies.

India

24 papers

India is one of the most policy-rich developing countries, with dozens of large-scale government programs affecting hundreds of millions of people. The SHRUG dataset (Socioeconomic High-resolution Rural-Urban Geographic Platform, v2.1) provides the geographic backbone: stable identifiers for ~640,000 villages and ~8,000 towns, linked to Census, Economic Census, and satellite nightlight data spanning 1991–2021. Practical API access to official Indian administrative portals ( MGNREGA MIS, PLFS, NFHS/DHS, UDISE+, RBI DBIE) unlocks credible causal analysis of India's major development programs.

France

31 papers

France offers rich administrative data for policy evaluation, from property transactions (DVF) to firm registries (INSEE Sirene) and federal legislation timing (Légifrance). The research program starts with low-friction public sources and progressively adds authenticated APIs for deeper analysis.

Switzerland

14 papers

Switzerland's federal structure — 26 cantons and ~2,100 autonomous municipalities — creates exceptional natural experiments for policy evaluation. Staggered cantonal adoption of laws, direct democracy referendums, and wide municipal tax variation provide clean identification strategies.

United Kingdom

21 papers

The UK combines exceptional administrative data depth with a federal-style structure that generates clean natural experiments. England's ~300 Local Authorities adopt policies (selective licensing, Council Tax support, Clean Air Zones) at different times, enabling staggered DiD. Devolved nation borders (Scotland, Wales, Northern Ireland vs. England) provide further policy discontinuities. Core sources are all open: the ONS Statistics API and NOMIS cover labour and demographics; HM Land Registry Price Paid Data covers 24M+ property transactions since 1995; the Police API and Environment Agency APIs provide crime and environmental data.

Nigeria

7 papers

Nigeria's policy ecosystem combines high-frequency fiscal transparency data with procurement and health sources that are unusually useful for causal policy research. Public finance flows can be tracked via Open Treasury daily payment data, while procurement exposure can be measured from NOCOPO records. National health outcomes can be linked through NCDC reporting, with ACLED and DHS extending to conflict and household-level mechanisms.

Other Countries

40 papers

Papers studying policies outside the primary country programs, including cross-country comparative studies and other geographies.