· A nice 20 minute discussion where Isaiah Andrews interviews Guido Imbens and Josh Angrist about their approach to co-authoring and advice on how to choose co-authors. Angrist says picking a co-author is like picking a spouse and students are sometimes too promiscuous in co-authoring.
· Scott Cunningham interviews Peter Hull. They talk about how he got into doing applied econometrics, his work on shift-share designs, and his new paper on potential issues that can arise in regressions with multiple treatments and a set of control variables – and the implications of this for judge IV designs. A couple of points that stood out for me. Peter noted that at MIT they would take a field class in the first year alongside the macro, micro, and econometrics core, and he just happened to take the labor class that Josh Angrist was teaching because people said it was interesting, rather than any specific field interest in labor per se – and this introduced him to a set of applied problems and got him working with Josh. I think most PhD programs just hammer their first years with theory, which makes for a tough year and can make ideas seem problem-set driven rather than real-world application-driven – so I like the idea of a mandatory applied field class in the first year. Second, Peter and Scott discuss whether some basics of causal inference should be taught in high schools, and the idea of having science fair experiments also include using IV or DiD. It made me remember that I actually did this in high school – my swim team got a new pool that allowed us to train every day through the winter, and so without knowing what DiD was, I wanted to see if this led to improved times, and so compared changes in times for swimmers in our team to those in other teams who didn’t have a new pool - so totally emphasize with the idea that causal questions are coming naturally to kids at that age – if only synthetic control had been around back then!
· On VoxEU, Gary Charness and co-authors summarize results from a survey of 1,400 economists about the peer review system. “As referees, our respondents wrote a mean (median) of 10.2 (8) reports per year. This distribution is similarly skewed, with around 15% of referees responsible for 40% of all reports written… editors request reports from some researchers disproportionately often. Tenured professors tend to write more reports than they think is reasonable, while PhD candidates and postdocs write fewer. In fact, only about 25% of respondents write the number of reports they think is reasonable”. Berk has previously discussed how he signs his reports. Others remain uncertain about this. “Respondents were quite sceptical about a policy where “referees sign their reports and the entire review history […] is disclosed” (Figure 10). They were more open to making reports public in an anonymized way, with junior respondents being the most favourable (Figure 11). However, respondents appeared uncertain about the benefits of this policy, suggesting journals might want to experiment further in this area.”
· John List notes that the new JPE Micro will include Registered Reports, with guidelines coming soon.
· On the CGD blog, Lee Crawfurd, Susannah Hares and Justin Sandefur ask what scales in global education – with school meals and extending the length of the school day being things that seem to be relatively easy to work in large-scale government run systems.
· On the Data Colada blog, a discussion of a paper on making police relatable and crime – and the illusion of robustness that can come from doing lots of specifications that don’t differ that much from one another. See also the authors’ response discussion. I do like the robustness heat map the authors provide, since it is clear in showing not just where the results remain significant, but also what it takes for them to start breaking down – something not all robustness checks do.
Join the Conversation