,
8 tweets,
2 min read
Read on Twitter
1/ Can governments in poor countries facilitate development by supporting certain industries? When does such "industrial policy" make sense? This is an old and contentious question. A lovely new paper by Ernest Liu provides sharp new insights. I will try to explain ..
2/ The starting point is that each industry suffers from a potential "weakness", such as poor access to credit markets (imperfection in econ lingo). Should the government help industries with the biggest weakness? Or the largest industries with most value-added?
3/ Liu shows that the answer is *none of the above*. Instead, the key is to focus on industry network structure - or input-output matrix - and identify industries that serve as "fundamental inputs", i.e. as inputs in the supply-chain of other industries.
4/ In general, it's difficult to figure out which industries are more "fundamental" than others: A can be an input into B, and B into C, but C may itself be in input into A. So the nice theoretical result may not be of much practical use. BUT Liu shows a neat empirical fact ...
5/ input-output matrices in general are "hierarchical", i.e. in example above C is almost never an input into A. This is great news, as it implies an easy way to rank which industries are more "fundamental": industries that are more upstream in the input-output matrix.
6/ The second half of Liu's paper is a brilliant empirical exercise where he measures "upstreamness" of industries in korea and china, and shows that these countries subsidized exactly the set of industries that theory suggests they should have.
7/ Liu's paper is a great contribution to macro-development, a field that is yearning for new ideas.
8/ I only scratched the surface. I suggest you order a cup of coffee and start reading scholar.princeton.edu/sites/default/…
