It is a week for rare occurrences. At the moment, it’s taking every ounce of self-restraint I have not to jump in the car and drive straight to the Peak District, where for only the second time ever recorded in the UK, a Lammergeier has been spotted. In birding terms, this isn’t a million miles from having a tiger stroll down Regent Street – the only place I’ve seen a Lammergeier is in Ethiopia, while hiking the Simiens and trying to avoid huge herds of Geladas (who have teeth like this, despite being vegetarian, essentially making them the animal world’s equivalent of Count Duckula). It’s not just the drive that’s stopping me though, but the equally rare prospect of England managing to crack the 400 barrier in the first innings of a test. Apparently, I should have kept my mouth shut as just as I wrote that, Stokes departs for 176, with England still 5 runs away. Lest I further jinx the team, let’s go straight into the links.
- How much of the world do you need to explain before your explanation matters? In his interview with Melissa Dell, Tyler Cowen presses her on the limits to what her work can say, and I really like the way she frames her answers. Melissa’s work looks at how institutions and institutional capacities from deep in the past can explain patterns of economic development today, but as she points out, these explanations are necessarily partial because the world is a really complicated place. What’s more, persistence is only visible (or rather studied) in places where there hasn’t been a massive interceding disruption, which isn’t surprising, but is important for understanding and contextualising the results. The common criticism of this kind of research is that the policy implications amount to ‘have a different history’; but perhaps this points to a different response: have a revolution and break the institutions that hold you back.
- That tees me up to go into my first proper rant of the week. Planet Money ran two really good pieces on representation this week, the first about the absolutely pitiful racial diversity of the economics profession (transcript), and the other about Hollywood (transcript). I really liked the second one, because it gets into one of the mechanisms by which diversity gets squashed, and it’s consequences. One of the reasons good film scripts were not being made was that many of the gatekeepers who decide what gets sent to the big bosses decided that they were ‘too out there’ or of niche interest to move up the chain, despite thinking they were really good scripts. A simple mechanism helped (partially) break this bad equilibrium: someone asked a bunch of junior executives to list their favourite unmade scripts, with no qualifiers about who would make it, or star in it, or fund it. And it turned out that many of the most popular scripts were not just universally highly rated, they were from Hollywood outsiders. Within three years of the list going public one of them, Juno, had won an Oscar. Diversity doesn’t just matter intrinsically, it helps find things we otherwise might not. This isn’t a new insight, it just turns out it’s really hard to act on.
- Post-World Bank Penny Goldberg continues to be on a roll. Her latest Project Syndicate piece makes the case for removing global restrictions to trade, immigration and capital flows. If the aim is a better world, it probably isn’t complete without this.
- I really liked this piece by David McKenzie on how to think about information interventions, and in particular his quick thought experiment near the end, which demonstrates why you should rarely expect really large effect sizes. The one quibble I have is that this is not generally a good justification for juicing up the sample size: the size of the sample should depend instead on the minimal economically important effect size, not what effect size will get you a significant result for a small effect. If an effect size isn’t big enough to matter, then does that second asterisk in your table of results really make a difference?
- A nice study reported in VoxDev looking at how cultural similarity affects production performance. There have been a few of these studies in the Indian context, using caste; I’d be really keen to see if the results replicate in other settings, where other dimensions of identity matter much more.
- Prospect’s list of the 50 most important thinkers in 2020 is out. It missed out Kendrick, so it’s obviously trash.
- Lastly, if I had to make a guess as to which article this week would make me have a ‘something in my eye’ moment, I probably would have guessed it would be something about the abuse of statistics in political discourse (I know, I’m a big softy) or bad news about prospects for a coronavirus vaccine. I definitely would not have guessed that it might be a statistical analysis of sporting mascots, but then there’s something about grown-ups dressed as animals that hits a nerve sometimes. (For the avoidance of doubt, though, the best mascot is definitely the Raptors’s). And just in case Benny the Bull isn’t your jam, your regular reminder that Clueless is the best, and much much much better than Emma.
Have a great weekend, everyone!