Read on Twitter
One read of development economics is the following: what works for one country won't necessarily work for a different country, in part because there are a host of institutional variables that are difficult to capture, both formal and informal.
For example, the importation of the US constitution to other countries seems to have failed in many cases. Why this is the case is a difficult question, but the lesson is clear: saying "well this seems to work here, let's try it somewhere else" is a fraught approach.
To me, this is why cross-country regressions and comparisons are so damn hard, and why they may have limited utility for our understanding of why some places got rich and others didn't.
I can't stop thinking about the parallels to the pandemic debate. It seems like some countries have gotten it right, but I wonder how generalizable those lessons are for othe countries. How would trying a South Korea-style approach work out in countries with bad govts, etc?
For that matter, how would a Swedish-style approach work out in the US? It's hard to tell. My sense would be "maybe not so great", it's questionable how well it has worked there, but even in the best of cases we should probably be much more cautious in pointing to "successes"
and claiming that their approach is the golden ticket and will be effective always and everywhere, etc.
Just like we shouldn't make naive cross-country comparisons w/ development policy, we shouldn't make naive comparisons w/ pandemic policy. Social science is, alas, hard.
Just like we shouldn't make naive cross-country comparisons w/ development policy, we shouldn't make naive comparisons w/ pandemic policy. Social science is, alas, hard.
I'm willing to say that some things seem obviously good. Like, test and trace if you can, wear masks, etc. But we all need to be thinking about if our preferred approaches rely on underlying variables that are tough to observe, and if we think those things are transferrable.
There are no shining cities on the hill-there are only tradeoffs and constraints. And those things are tough to get at even in the best of circumstances.
