Microdata for Research in Macroeconomics and Finance

This paper studies the role of micro-level data for research in macroeconomics and finance, with a focus on low-income countries and the engagement of international institutions, particularly the IMF. Combining bibliometric data, LLM-processing, and human verification, I construct a novel dataset of more than 1,900 academic publications using nine types of microdata (credit registry, balance sheet, mobile money and digital payments transactions, and others) drawn from the top 50 journals in economics and finance between 2000 and 2025. I document a sharp rise in the use of microdata across the literature: from under 1\% of all papers published per year in the early 2000s to nearly 4\% annually by 2025. Despite this growth, low-income countries remain underrepresented. Economic development, proxied by GDP per capita, is strongly associated with more frequent and more highly cited research, as well as more available datasets and better replication practices. The availability of microdata and the structure of research networks play a central role. When new microdata become accessible in a country for the first time, the volume of publications rises steadily in the long term. Research networks in this field tend to be fragmented, with the exception of the literature using credit registry data, with this interconnectivity emerging from collaborations between academics and central bank officials. Within the IMF, microdata have gained traction in working papers, while its presence in policy documents, though meaningful, remains more modest. I conclude with four recommendations to promote the use of microdata research in macroeconomics and finance, especially in LICs: (i) create structured data hubs modeled after global best practices to improve access, standardization and the integrity and independence of research, especially for less conventional data sources; (ii) support central banks as hubs of talent and institutional development; (iii) promote open data and replicability across settings; and (iv) strengthen the integration of research insights into policy work.