Links round-up

Hi all,

I’ve only ever cancelled the links just because I was in a foul mood once, on Friday 24 June, 2016 (if you live in the UK, you won’t need google to remember the headlines that day). Today, almost exactly four years later, I was considering doing the same but I’ve thought better of it. I’m not going to go through the five stages of grief here: rather, I’m just going to hold on to two for the next few months: anger and acceptance, simultaneously. The anger is self-explanatory – DFID is a brilliant organisation, one which exemplifies the qualities of a mission-driven bureaucracy where people live the work they do and care deeply about getting it right. It became so because its mission was clearly articulated, inspiring and genuine: working in the department meant arguing with people over what the right thing to do was. During my days at DFID I had many, many conversations with angry, frustrated colleagues, furious because someone else wanted to do what they saw as the wrong thing. These weren’t “someone disagrees with me” conversations, but “we’re going to get this wrong, and that’s completely unacceptable” conversations. An organisation that achieves a culture like that discards it at its peril. It’s the kind of culture that contains within itself a constant drive to improve and learn, to excel. It can’t simply be recreated by putting nice words on the new letterhead; it has to be lived. People self-select into organisations, choosing to work where they both get paid and appreciated and contribute to something that they believe in. That’s the anger. The acceptance, though, is that the change is now coming, and the future has to be something better than simply sewing together DFID and the FCO. What came before was not a political equilibrium, which led to many compromises and fudges. The chance to shape a new organisation into one that both does good and is a political equilibrium can’t be missed. I’ll write something about this next week, but the main point is that the mission driven bureaucrats, on both sides of the merger, still have agency and can use it for good.

  1. That was a long intro, and I am still pretty angry, so just to cool off: a man very reluctantly wading into a pond. Watch to see why, it is joy.  
  2. As long as I’ve studied or worked in development (which is now more than half my life), agricultural productivity has been something of a puzzle; it’s one of those areas where every advance seems to directly set the seeds for new ones. VoxDev have a nice write up of work that looks at the effect of using multiple survey methods to improve the accuracy of productivity estimates for smallholders; long story short, measurement error is substantial, and can be meaningfully narrowed, with real implications for policy.
  3. In a similar vein, Markus Goldstein summarises a new paper by Lucia Diaz-Martin and co-authors on what women’s groups actually achieve on a range of different metrics. What I love about these blogs Markus does is that they basically teach you how to read a paper well: you learn not just about the paper he’s read, but also from what details he picks up, and how he presents the information. There is still a lot we’re learning about how to do development well; and I hope the capacity to keep doing this (and funding this kind of thing) is one that we protect.
  4. In non-DFID news, the world out there is still pretty terrible; and on Juneteenth, it seems appropriate to highlight the failings of the economics discipline on race. Ben Casselman has been covering this for the NYT for a while and his latest piece makes for sobering reading, again. Econ is lousy with insider networks and petty exclusionary politics, which makes life doubly hard for those from outside the pre-existing networks, be they race, class, institution or gender-based. And they may be self-perpetuating: a new paper finds big effects from having a paper tweeted by an influential professor. Matt Collin’s twitter thread digs in (and yes, the effect size is only two citations, but for many academics that’s a really big effect!).
  5. I don’t think it’s controversial to say that 2020 has completely sucked: between the global pandemic, mass social unrest, imbeciles defending statues of slavers and the pause in the NBA season, you’d imagine all the major economic indicators are doing the statistical equivalent of the scared Bruce Lee gif. But nope: the stock market’s not exactly been chill, but it’s definitely got a lit cigar and a glass of scotch right now. Planet Money run through some of the theories, including the ‘perfect storm of stupid’ theory, which is pretty much my meta-theory for 2020.
  6. More from our ‘the whole world sucks’ department: another paper to add to the growing pile of Coronavirus-is-making-gender-inequality-worse research; Anne Case and Angus Deaton on how the Coronavirus is hitting those who most suffered from deaths of despair particularly hard; and a behavioural scientist on how she fell prey to the very biases she studied.
  7. One for the stats geeks: I know he’s a bit of a spanner sometimes, but I found this piece by Nassim Nicholas Taleb on the uselessness of single-point estimates in a pandemic very good.
  8. I’m struggling to find stuff that put me in a good mood this week. I published my first blog at CGD (on how to improve decision-making under extreme uncertainty), which was fun. And I’ve discovered that AI cannot distinguish between a lung infected with coronavirus and a cat – though I suppose this is technically bad news. So to cheer us all up, let’s instead turn to Baloji.

Have a great weekend, everyone!


This entry was posted in Links round-up. Bookmark the permalink.

Comments are closed.