I’m not sure what the best bit of the week has been so far – whether the unexpectedly stirring poetry (more in the final link), the unexpectedly stirring century by Angelo Matthews, or the fact that Seth Rogen and Ted Cruz are having a very high profile spat right now (mentally, I read that tweet in an *extremely* Jean-Ralphio voice). But actually my favourite bit of the week has been putting the final touches on a podcast I’ve recorded with Matt Collin (eccentric development economist by day; eccentric development economist by night) called Paper Round – keep an eye on twitter for its first post next week. The idea was seeded during lockdown 1: stuck at home with no new Marvel movies to make jokes about, Matt and I committed to reading a paper roughly every month and talking about it on Zoom, a partial replacement for all the casual conversations about economics we would normally be having in our respective offices. So, for those of you feeling a gap in your life which would normally be filled by people trying to explain economic concepts (in episode 1, it’s poverty traps) using Batman and Trading Places, this is for you. And with that audience-limiting pitch, it’s on to the links.
- Well, this one doesn’t half make me look silly. It turns out that the Dunning-Kruger effect – punchline of around 1/3rd of my jokes, and the only theory that ever seemed to satisfactorily explain selection into political careers – may simply be a statistical artefact and, ironically, I was not smart enough to notice it myself. (This is a real galaxy-brain one; was my failure to identify the flaw in the Dunning-Kruger effect … an example of the Dunning-Kruger effect?). I say ‘may’ here because after reading this McGill explainer of why it’s not real, I’m still not wholly satisfied as to why it appears to be a statistical anomaly rather than a real effect with a causal explanation. I think it has something to do with the fact that the original effect arises from people guessing their percentage score on a test, creating upper and lower bounds to their guesses (0 and 100). This means those nearer the bottom of the scoring charts have more ‘room’ to make mistakes that are higher scores than their real ones, and those near the top have more ‘room’ to guess lower scores than their actual one… But I would like a statistician to confirm that intuition. Now, if someone conclusively proves that The Peter Principle is a lie, I’m going to lose all my joke material, so I’m issuing a cease-and-desist on all further inquiry into the issue.
- On to more important – and depressing – issues. A new VoxEU piece looks at the impact of Covid-19 on childcare services, and by extension on female employment and finds predictably grim effects. It’s not surprising: if discrimination at some point in the chain leads to women earning less or selecting into lower-earning work, this kind of effect is predictable. It made me think about how a lot of great economic research compartmentalises problems into small, solvable pieces, but sometimes misses the stage where all these pieces are stitched together (to be clear, this is not a criticism of this, excellent, piece). A bit like a translator working word by word, creating a barely coherent sentence. The benefit of having specialists (or even generalists) then looking across all the research to put together a broader picture is immense, and under-incentivised in economics.
- And if you need more downbeat news on gender and the workplace, there’s this: evidence from the Swedish army suggests that mixed working environments have positive effects on attitudes about gender… but they don’t last long. This is not the kind of thing we’re going to nudge our way to a solution for.
- One thing I absolutely love about Andrew Gelman’s blog is the amount of time he and his co-bloggers spend talking about how best to communicate statistics. Two great examples: the comments on this blog are great (quite possibly the first time anyone has ever written that sentence), in response to a question about how medical doctors should communicate probabilities and risks. There are a lot of people thinking about how we compute and act on probabilities, and sometimes its good just to go down the rabbit hold and learn a bit from their work. And second: this piece looking at a different way of visualising distributions. Possibly just for the completists, but reading this blog has made me much more aware of how I try and communicate data, and how people might be responding to it.
- I’ve spoken to a surprising number of people who are hesitant about taking vaccines recently (even in my little bubble, extremely skewed towards overeducated geeks as it is), and I remain completely befuddled as to why anti-vaccine sentiment arises. I cannot for the life of me understand how the vast majority of people who share anti-vax sentiments get any benefit from this movement (perhaps I’m being naïve?), which makes me think there must be some returns to just talking about this much more and answering questions and responding to doubts people express much more openly and regularly. FiveThirtyEight has the scoop on the small army of people doing just that on Facebook now.
- There has been far too little coverage of the practicalities of how East Asia has tackled the Covid-19 pandemic so much more successfully than the rest of the world (talking to family and friends in Hong Kong is always a little painful in that respect), so I was very glad for this piece on Duncan Green’s blog going into it in a little more depth. I had always thought that if SARS had happened in the UK we’d have dealt with it just as well – I now have no doubt we wouldn’t have.
- Everyone has been talking about Amanda Gorman this week – and with good reason. Not only was her poem amazing, and uplifting, it was also a relatively rare event: few Presidents have poets read at their inauguration. LitHub have them all here, starting with Robert Frost in 1961 – a poem with a great story to it. He arrived with a new poem in his pocket, but discovered that he couldn’t read it in the light of the podium and instead recited one he knew by heart. For what it’s worth, my favourite is Elizabeth Alexander’s Praise Song for the Day, Obama’s choice the first time around: “We cross dirt roads and highways that mark/ the will of some one and then others, who said/ I need to see what’s on the other side. I know there’s something better down the road.” And on that note…
Have a great weekend, everyone!