- DEC has a fantastic lecture series going on at the moment. This week we had Pascaline Dupas. Videos of the talks are online. Of particular interest to our readers, will be her discussion of the techniques used for how they managed to re-interview 95% of Ghanaian youth after 10 years; and of how they messed up asking about labor market outcomes the first time they tried due to the sporadic nature of work for many youth (and something I hadn’t thought about – people working for the government whose payments have been delayed, so are owed back wages, but didn’t actually get paid in the last month).
- In VoxEU, revealed vs reported preference – when asked if they saved or spent their stimulus payments, people’s answers were qualitatively informative of actual behavior seen from observed spending data; and when asked how much they spent, gives a reasonable measure of average spending propensity – but these questions aren’t so good at capturing which households respond more.
- Uri Simonsohn on 8 things to do to make your open research easier for others to use – e.g. “In tables and figure captions, I include links to code that reproduces them” and “Rule-of-thumb: At least one comment per every 3 lines of code.”
- The Upshot on how motherhood is strongly associated with the big gap in earnings by gender, even in Denmark. Although there is definite envy when reading “One new study, which used a data set including everyone in Denmark from 1980 to 2013, along with details about their jobs and families”...
- Cassidy and Ball on 9 things you do (but shouldn’t) in research communications on the from poverty to power blog: “Describe your research audience as ‘policy-makers, practitioners and the public’. You know you should be more specific, but no one seems to be questioning it. And, well we want to reach influential ‘people’. Next time, try asking: if you could get five people to read it, who would they be?”
Join the Conversation