Michael Pettis
Michael Pettis

@michaelxpettis

7 Tweets 1 reads Apr 18, 2023
1/7
"In a simulation exercise, the IMF said the long-term efficiency costs of the world shifting towards economic blocs with greater investment barriers at borders would be large. It estimated they could cut global economic output by 2 per cent."
ft.com
2/7
Because this study was widely reported, I recently finished reading the chapter in the IMF's World Economic Outlook on which it is based, and while this is indeed what it says, there wasn't a very convincing argument behind it.
imf.org
3/7
One of the major weaknesses of the way mainstream economics is practiced is in the tendency for economists to build models that implicitly embed their assumptions, in which case it is no surprise that the models nearly always confirm those assumptions.
4/7
But this isn't the way science works. It's not that I dispute their conclusion. "Friendshoring" policies may indeed reduce global output, and it is certainly possible to make legitimate and credible arguments suggesting just that.
5/7
My complaint is that while they propose a scientific-sounding 2% output loss, there is nothing scientific, objective, or meaningfully data-based about the exercise (even if it uses lots of data). Small changes in assumptions would have changed the conclusions.
6/7
China implemented disruptive "self-shoring" policies in the 1980s and 1990s, after all, and they were very successful. Why is it necessarily the case that today we shouldn't disrupt the pattern created by this previous disruption?
7/7
Perhaps I've read too much Albert Hirschman on the specificity of economic development, but I don't believe these simulations can usefully measure the costs and benefits of something so complex. All they do is confirm already-held assumptions and impress the unsophisticated.

Loading suggestions...