The world is so odd right now. In the last week, we’ve seen the US President tweet out an advert set to the Batman music (thanks to Huw, I was able to watch it in all of its glorious insanity before it was taken down), featuring the line “they call you a racist”; the Daily Star deciding – in 2019 – that it might be a bit weird to have topless women right at the front of the newspaper (well done, guys, you’re only around 2 centuries behind); Chinese scientists adding human genes relating to brain function to a monkey (Pogo?); and LeBron James not being in the NBA Playoffs. Eventful, too: Assange being dragged out of the Ecuadorian embassy looking like Trevor Slattery, Omar al-Bashir being removed in Sudan (and replaced by the army – can anyone suggest what I should read to understand all of this better? Does this mean Sudan get a version of this t-shirt?), Joaquin Phoenix gearing up to be the third iconic Joker (seriously: this trailer has me more excited than the Endgame ones).
- Let’s keep with momentous events with the first link, shall we? A team of 200 researchers have collaborated to produce the first ever image of a black hole, which looks exactly like what any six year old would have drawn when tasked with designing one. But what’s incredibly cool about all of this is not actually the image, but the technology they’ve used to assemble it. The researchers have basically used a series of techniques to put together multiple highly incomplete data sources to produce a near complete composite picture. As 538 point out, Black Holes might be one of the least valuable applications of this work, which might one day revolutionise medical diagnosis and imaging, for example. Among the brilliant researchers working on this project was an absurdly young woman in charge of creating the algorithm underlying much of this, building on work she started as a 23 year old. She’s 29 now. Another shovelful of dirt with which to bury stereotypes about gender and science. Related: CEPR’s new Women in Economics initiative.
- While we aren’t at all certain what the final implications of that particular technological innovation will be, Owen Barder and colleagues at CGD make the case for the central role of technology in development in this blog, using the relationship between child mortality and income to illustrate their points. While at any point in time, there is a general negative relationship between income and child mortality (that is, if you have more money, your children are less likely to die young), Owen and co point out that this curve has itself been shifting downwards. Essentially, improvements in how we deal with child mortality means that over time, your ‘child health purchasing power’ per dollar has been increasing. While a lot of the writing on technology and development has been so poor as to be almost absurd, this is worth reading. Always keep in mind, though, what they mean by technology: not typically ultrasounds or fancy new machines but things like vaccines, plastic tubs to transport clean water and so on.
- Another couple of great pieces on the CGD blog this week: first, as befits the man who wrote Getting Better, Charles Kenny takes down Robert Kaplan’s recent assertion that his famous article ‘The Coming Anarchy’ has held up well over time. Kenny points out that while a couple of his doomy predictions look good, West Africa has outperformed virtually all of his other dire predictions. And Michael Clemens says what Donald Trump could have discovered with five minutes of research or critical thought, beyond him as they may be: cutting aid to the Northern Triangle is not going to help him reduce migration.
- Guo Xu, whose feat as a DPhil student in coding an entire colonial civil service database still leaves me slightly breathless in admiration, has written a new paper with Marianne Bertrand and Robin Burgess, arguing that allocating civil servants to the areas they’re from reduces their performance, and in particular leaves them less able to withstand political pressure.
- Speaking of data, I liked this by Tim Harford: what we measure, how we measure it and what we miss out have serious implications for how the world changes and can be changed. Sometimes we can be obsessed by ‘big data’, and I’m struck by how often the people ask me ‘how many observations do you have’ when I talk about potential research, but the most important first question is ‘how good are the data’? Pollsters have known this for years: it’s far better to have a small(ish) sample that is well balanced across a population than a big sample with a defined blind spot, which is much more likely to bias your conclusions. As Tim points out, the blind spot often have isn’t so much a ‘spot’ as a ‘half’, given the paucity of data about women.
- Related: sometimes it’s super expensive to collect good data, so checking if readily available alternatives are usable is always helpful. Sometimes local experts are as reliable as price surveys.
- April means two things: the NBA Playoffs start, and County Cricket kicks off again. The playoffs are a funny event. Everyone already knows the ending: Golden State win, deploying a line up featuring four hall of fame locks, three of the four greatest shooters of all time and two of the five best players of the century. But what happens before that might still be interesting, even if it will never involve Dirk Nowitzki again, standard bearer for unathletic, unlikely superstars. Not supporting a county myself, the main interest is keeping tabs on who’s up for England this year, and the early signs are VERY encouraging: Haseeb Hameed has scored a century in a warm up and is currently unbeaten with a fifty in his first match. With any luck, this is the future of English batting, right here.
Have a great weekend, everyone!