Economics & Marginalia: November 13, 2023

Hi all,

Once again the links come to from a precariously balanced laptop, perched on cramped knees on a bus, packed cheek-by-jowl with other commuters. It's not a set-up that inspires great chattiness. Nor does Sri Lanka's ongoing attempt to plumb depths of cricketing incompetence never before plumbed, or—speaking of new lows—do the latest ill-considered pronouncements from the Home Secretary, or indeed the current front runner to be the next US President. These are dark days, both literally and metaphorically; but there are nevertheless small glimpses of light.

1. Chief among these glimpses of light is the wonderful annual tradition of the Development Impact Job Market Papers series. Every year the DI crew provide a platform to brilliant young economists with something new and interesting to say; every year I find someone who I wind up keeping tabs on and learning from for many more years. It's early days, but this piece by Justine Knebelman makes her an early contender for that role: the looks at how bureaucrat discretion in decision-making can lead to not only lower tax revenues, but more regressive taxation, in this case driven, it seems, by their lower ability to accurately assess the value of high-value houses (specifically, they tend to underestimate values). An algorithm, whether incorporating some bureaucrat inputs or not, performs better. But all of the series is worth reading: the effect of climate change on non-agricultural firms in one, the effect of typhoons on outmigration from the Philippines in another.  

2. Public services in the UK are another thing finding new nadirs every day. Tim Harford draws on anecdote and data to paint a grim picture of the 'doom loop' our services have entered, and can find no neat solutions. You cannot solve 40 years of systematic underinvestment quickly, but each year we put off doing something about it makes solutions even harder to find. You may need to register or subscribe for this one; it's the FT and worth it.

3. Andrew Gelman's rule of thumb is that you need a sample size 16 times larger to estimate an interaction than to estimate a main effect. This is the stuff of nightmares, but it's worth reading his full explanation to see the exceptions, and cases for which it is not true: it may save you much angst.

4. Dani Rodrik has a typically nuanced take on economic nationalism, arguing that it does not necessarily need to be the disaster many project of it. Recently, I had a conversation with a friend in which I said I lean towards optimism, and am drawn to optimists; but this is too optimistic even for me. It's true that protectionism and inward-looking policy can spur improvements to welfare and productivity, and it's true that many countries have used such methods to that end. But where we are now, with the policies and politics we're entering, I see much more potential for disaster than for success (again, registration may be required).

5. I learn something every time Ken Opalo puts pen to paper. This piece on Ethiopia's need for a navy and seaport is no exception; the former, at least, is something I had never previously considered.

6. I've been thinking about soft power a lot recently; a nebulous but oft-invoked concept, This VoxDev write-up of research by Lukas Wellner and others investigates the soft power returns of Chinese aid and find that completed development projects are associated with better perceptions of China in the countries it has most strategic interest in. Ironically, then, just as China is successfully using aid as a soft power instrument, the UK is undermining its own approach (and for no good reason).

7. Con artists usually make for good, upbeat thriller movies: think Ocean's 11 or The Sting; I think the Sam Bankman-Fried saga will eventually be something rather different. A black comedy, perhaps, or just a straight up tragedy. It's hard to listen to NPR's account of his recent trial—the guilty verdict from which has him facing a life in prison—without a sense of pathos (transcript here). I normally end on a pop culture note, but this time I'm ending on the anticipation of one. This will be a movie by David Fincher (or perhaps the Coen Brothers) one day. 

Have a great week, everyone!

R