Economics & Marginalia: May 24, 2024

Hi All,

This has been a week of extremes for me: I spent the first part of the week basking in sunshine in Barcelona for a (very good) conference, meeting people from around the world working on interesting problems, eating incredibly good food and enjoying one of the more walkable cities I’ve visited. And I’m closing the week on a train to the far North of England to see family, with the temperature dropping appreciably with each passing mile, rain chucking down near-constantly, and a hazy grey sky replacing the sunshine. Between the travel and late nights working (and, let’s be honest, following Luka Doncic’s advance through the NBA postseason), I am absolutely knackered. If the links trail off in the middle of a sentence and start snoring, I can only apologise.

  1. Something else that has kept me up at night recently: the threat of a full-scale trade war between the US and China. There were two very good pieces in Project Syndicate about this that are worth your time. First, Penny Goldberg came closes to articulating what my baseline view of the situation and the risks it carries in her very critical assessment of the recently announced US tariffs targeting a number of Chinese products, electric vehicles among them. She argues (and I agree), that the tariffs are an admission of US weakness, rather than strength, and that even if Chinese subsidies for electric vehicle production were unfair, that ‘[s]ince we cannot rewrite history, we should try to take advantage of the circumstances that history has created. From a climate perspective, availing ourselves of cheaply produced Chinese EVs would have been a step in the right direction.” The whole thing is very good. Arvind Subramanian’s piece took a very different perspective, one that I hadn’t considered fully: that there may, in the carnage, be opportunities for some developing countries. His take is, if not exactly optimistic, at least cognisant of possible benefits. What it is particularly good at is considering gradations of bad: it’s a piece very much in the mold of the classic Sir Humphrey advice: “if you’re going to do this damn silly thing, don’t do it this damn silly way”. There are ways that the US can act that will be more or less destructive to developing countries. We should hope they take the less destructive route.
  2. And while we’re on the subject of trade and globalisation, an interesting VoxDev piece describing the rise and distribution of cross-border patents.
  3. I have now read enough Judea Pearl to understand some of the very many ways that we can be led astray by thinking about even very reliable statistical associations without understanding the causal relationships that underpin them (which is, of course, not to say that there is always causality in correlation, but that there is a way to think about the causal bases of correlations—including the absence of any such causal basis). Tim Harford talks about Moore’s law, and argues that this famously robust statistical association has led us astray because the way it was formulated masked a rather important part of the causal relationship it reflects. Moore’s law is formulated in terms of time (we double the number of transistors in our integrated circuits every two years); but in fact, an important driving factor here is not time per se but the volume of production we undertake (what he calls Wright’s law). Harford points out that this shift in our understanding implies a shift in the lesson we should learn: “Moore’s Law suggests that good things come to those who wait. Wright’s Law says that good things come to those who act.” Misunderstanding this has important implications for public policy—and specifically that when this relationship applies for socially valuable products like solar panels, we should subsidise and incentivise additional production to drive prices down faster. Relating back to link 1, it strikes me that US policymakers should read Tim’s piece as soon as they can.
  4. I am not a technical whizz: I do not often think about integrated circuits or transistors, and could probably not pick one out of a line-up. And when I use the internet (which is roughly all of my waking hours), I do not often think about what makes it actually run and work. So this Planet Money piece on a hack that nearly brought the whole internet to its knees was a huge eye-opener for me. It’s an incredible story, a long con that required skill, creativity and genius: as one of the interviewees says, this should 100% be made into a movie one day (transcript).
  5. This is very good and thoughtful, about how AI and large language models can absorb and reproduce bias. Jessica Hullman discusses why it is not surprising that they do so, but also how the optimal policy response is not necessarily to try and manually ‘edit out’ all such biases. Some stereotyped beliefs of heuristics can be useful in generating accurate answers to certain questions.
  6. I really liked this, which I found via David McKenzie’s great links, on the process of policy implementation. It’s particularly good on contracting and why hiring external contractors can be worse than doing things in-house, even when you lack all of the necessary expertise. All ways of doing stuff are flawed. The choice in how to implement policy is very often about accepting what kinds of flaws and what kinds of risk you are willing to tolerate. A huge chunk of this is just understanding and applying principal-agent theory… which, luckily, is one of the things I teach to students doing the BSG Master of Public Policy course.
  7. I couldn’t choose whether to close the links this week with a list of great books about jazz or the news that Taylor Swift’s famous cats are apparently in a permanent state of agonizing pain, so I’m going to do both. First, LitHub have a list of the best books about jazz, and it’s a very good one; I particularly recommend Young Man with a Horn, Dorothy Baker’s barely disguised novelisation of the life of Bix Beiderbecke (though it’s not as good as her magnificent Cassandra at the Wedding). But there’s one horrendous omission: Eric Hobsbawm’s The Jazz Scene, a collection of his pseudonymous jazz criticism. And for those of you who don’t give a fig about jazz, here’s the news about Taylor’s cats, bred for cuteness at the expense of lifelong disability… which seems not great. I have a friend who thinks that all pet ownership is essentially a crime, and I dread to think what he’ll say about this.

Have a great weekend, everyone!




CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.