· I gave a talk at UCSD last week, and they interviewed me afterwards for their backstory podcast. You get a little bit of the backstory for my paper with Jacobus Cilliers and Nour Elashmawy on PDS Lasso (blogged here), but then most of the discussion is on broader issues such as how I think about problem selection in research on applied methods, firms, and migration; how the life of a researcher in a policy institution differs from academia; creating public goods such as this blog and whether there are private returns as well as public returns, and where I see a gap for others to provide public goods; the problem of “why can we never do anything?”; advice for what a second-year student could do over their summer; and how I do and don’t use AI in my work.
· A quintet of researchers at Williams College, including development economists Pam Jakiela and Owen Ozier, have a new working paper summarizing trends in the use of statistical software in Economics, Political Science, and Statistics. “Web-scraped metadata and student data collection, together covering more than 10,000 papers, reveal clear disciplinary patterns: Stata remains dominant in economics, while R is increasingly popular in political science and is the standard in statistics. Within the social sciences, a growing number of articles also use multiple software platforms within a single manuscript.”….” In 2010, the proportion of replication packages that included two different types of statistical software was below 10% in both the Dataverse and AEA/ICPSR samples. By 2024, more than 40% of replication packages in both data sets included at least two different types of statistical software. Thus, while Stata remains dominant in economics, it is increasingly the case that economists use Stata in conjunction with other statistical computing tools such as MATLAB, R, and Python.” The paper also discusses how this was a teaching exercise, and has the student assignment as an appendix, along with ideas for other student activities.
· On VoxDev, Craig McIntosh and Andrew Zeitlin summarize their work on cash-benchmarking, comparing an integrated nutrition and WASH program in Rwanda to cash, as well as a second program that compared an entrepreneurship training program – to look at both overall effects as well as cost-effectiveness. They randomly vary different amounts of cash with the cash treatment arms. “One takeaway from the results of our benchmarking studies is the benefit of concentrating resources, both in the sense of focusing on programmes that deliver tangible benefits, and also in spending more per beneficiary so as to move the needle for a whole group in a meaningful way.” See their prior post on Development Impact for some more details and methods.
· Unexpected consequences edition: Also on VoxDev, Jorge Luis García on how the MGNREGA program in India that offers employment guarantees actually reduced female labor force participation: “By providing a stable source of earnings for households, the programme reduces the need for women to act as ‘insurance’ workers in times of economic uncertainty. In India, rural women participate in the labour force mainly as insurance workers. Indeed, their families prefer that they do not work at all, if economically viable, due to cultural norms.” In another article on VoxDev, Tamim et al summarize their work on the limited benefits of housing subsidies for Syrian refugees, and the social tensions it can create with local communities who don’t get the assistance. “Our findings indicate that the housing assistance programme offered limited short-run economic improvements that dissipated following the programme, while negative psychological and social cohesion effects proved more lasting.”
· Funding: The IDB Group’s Gender and Diversity Knowledge Initiative has launched a Call for Proposals on Fostering the Growth of Women-Led Businesses in Latin America and the Caribbean. “This call seeks to fund rigorous and innovative research that identifies effective strategies to support women-led MSMEs, close gender gaps in access to finance, training, and markets, and remove structural barriers that limit firm growth. The deadline for submissions is May 12, 2025.
Join the Conversation