You know, every once in a while my mood just plummets after spending five minutes reading the news and not just for the obvious reasons. I’m very much a big-picture optimist, in the Max Roser camp of taking the long view to realise how much better things are getting. Every once in a while, though, people behave like such complete s**ts that even a gregarious optimist like me has no response but a massive face palm and an expletive filled rant at the world. Two examples: the NRA campaigning against a domestic violence bill because they think it’s overkill to deny convicted abusers of their guns (not The Onion); and Danny Rose having had all the joy sucked out of being a professional footballer by the sheer level and frequency of the racist abuse he has to endure. Seriously, even if things are getting better, there’s still a lot of selfish, mean-spirited idiocy in the world. It’s incumbent upon all of us to do something against it.
- At the CSAE conference, Nick Lea said (to spontaneous applause) that the requirement of every paper to be built around a piece of causally robust econometrics was hampering the discipline. Many of the problems that we should care most about may have no prospect of being ‘solved’ by such approaches, he said . He’s not alone. A number of economists I’ve spoken to have expressed similar concerns; one, a recently-retired (and very brilliant) professor, said he thought what young economists are taught now is orthogonal to what they should care about. David Evans takes up the (measured) defence of regression-based economics, arguing that it is our best way of falsifying hypotheses and therefore being scientific about the accumulation of knowledge; he also talks about modelling as an alternative way of structuring econ papers. I rather suspect that the key part of Nick’s complaint was that we start with the constraint to use econometrics as our falsifying procedure. Instead we should start with the selection of the most important question and then pick the best available falsification, even imperfect.
- Back in the intro, I mentioned the racist abuse Danny Rose is subject to as a football player. In not-unrelated news, research shows that media outlets are far more likely to report a crime if it’s performed by an ‘outsider’ (in this case an immigrant) than a native-born person. We have a moral responsibility to argue for and promote a more thoughtful, evidenced and compassionate approach to migration, as Michael Clemens and Kate Gough do here, even when the messages aren’t the easiest ones to sell.
- I’ve often said that one of the best things about 538 is how scrupulously it examines its own performance in forecasting and making predictions. This piece by Nate Silver, looking at thousands of forecasts the site have made is fantastic, demonstrating just how important making probabilistic forecasts is.
- More economic policy should start with David Bowie. When Bowie was in the midst of his career, he wanted to raise a substantial lump sum to invest in future musical ventures. What he did was sell the rights to a fraction of his future royalties for a defined period of time. The idea has now spread to the education sector, where in some countries, you can be fully funded for university education in exchange for a small percentage of your future earnings, subject to a ceiling. Planet Money has the details (transcript). And since it all started with Bowie, Starman.
- I really liked this VoxEU piece on zombie firms, something that should be a much bigger topic in developing countries than it is. The idea is that these are firms that can’t actually compete on productivity or cost grounds but do not fail and die either, due to some other advantage they have (in the VoxEU case, it comes from the banking system, but it could easily be political connections). The existence of these firms has knock on effects for economy wide productivity which are not all obvious.
- Branko Milanovic summarises all the fights Francis Fukuyama tried to pick with mainstream economics in his new book, The Origins of Political Order.
- Speaking of slapdash economics, I found this via MR: almost all studies about the effects of social media use / screen time are based on fundamentally unsound measurement, so very few of those ridiculous ‘ten extra minutes on facebook makes you 20