Weekly links March 1: causal ML, female labor supply, role models and STEM, and more…

·       A new (free online) book on Causal inference and machine learning by Chernozhukov and co-authors.

·       A new VoxDevLit is out, on female labor force participation, with Rachel Heath as the senior editor. “We identify several policies that consistently increase women’s labour supply: increasing childcare availability, empowering women within households, psychological interventions, prompting businesses to offer amenities like flexibility that female employees value, and increased global exposure among export industries that disproportionately hire women. Other policies we discuss, such as skills training or depression treatment, yield more mixed results and are probably insufficient by themselves to increase women’s labour supply in environments where they face other large barriers, such as social norms or childcare and other household obligations. We also identify several important barriers to women’s labour supply – namely, discrimination and a lack of safety and harassment in public spaces”

·       On the IGL blog, James Phipps and Rob Fuller summarize some of the lessons and highlights from a program of experimentation on SMEs in the UK called the Business Basics Programme – which was intended to involve 17 RCTs plus a bunch of proof of concept pilots. One key takeaway is also how hard it is to successfully conduct trials in this space – many of their planned RCTs failed to get sufficient uptake, or had high attrition or were only able to measure short-term and intermediate outcomes.

·       Also via the IGL newsletter, a JPAL summary of research on advancing women’s representation and opportunities in STEM fields through exposure to role models – based on 7 RCTs in 4 high-income countries (France, Norway, the U.S. and South Korea).  

·       On the Econ that Matters blog, Ignacio Rodriguez Hurtado summarizes his job market paper on the causal effect of Mexican drug trafficking organizations (DTOs) on school dropout. “in summary, I find more DTOs increase school dropout and also teen crimes with offenders that dropped out of school.”