As noted in class, deindustrialization is a long-standing trend in the developed world, but the elasticity story is different than in agriculture: it’s that productivity rises faster than demand, so that the need for workers falls. In agriculture, it’s not so much that productivity rises faster than demand, but that demand doesn’t rise. So in order to gain income parity with the rest of a growing economy, the number of workers must fall.
The dynamic across countries has been that demand for manufactured goods rises very quickly with income, as consumers want (to take the Japanese examples) electric fans, electric rice cookers, and TVs. (Before that it was of course electric lights, and let’s not forget bicycles and — in the days before ready-made clothing — sewing machines.) Thereafter we switch into upgrading mode, and demand rises less quickly, and productivity overtakes demand. So we see a humped-shaped shift in the share of the labor force in manufacturing.
Is this the case for China? What of the future? Just in time for our shift to a new topic, Tyler Cowen (in the blog Marginal Revolution) cites a paper by Dani Rodrik on “premature deindustrialization”, to wit:
Somehow this paper has not been receiving adequate notice (Ryan Avent is one meritorious exception), so here it is:
I document a significant deindustrialization trend in recent decades, that goes considerably beyond the advanced, post-industrial economies. The hump-shaped relationship between industrialization (measured by employment or output shares) and incomes has shifted downwards and moved closer to the origin. This means countries are running out of industrialization opportunities sooner and at much lower levels of income compared to the experience of early industrializers. Asian countries and manufactures exporters have been largely insulated from those trends, while Latin American countries have been especially hard hit. Advanced economies have lost considerable employment (especially of the low-skill type), but they have done surprisingly well in terms of manufacturing output shares at constant prices. While these trends are not very recent, the evidence suggests both globalization and labor-saving technological progress in manufacturing have been behind these developments. Premature deindustrialization has potentially significant economic and political ramifications, including lower economic growth and democratic failure.
And while I’m at it, a new working paper has elasticities for cereals and meat, noting that they fall as incomes rise. They use meta-analysis, comparing results from 36 papers. (This approach is widely used in medicine, where clinical trials collect data on many things but have different sample populations, drug regimines and so on, often modest in size. Meta-analytic techniques use formal methods to cull information from disparate studies in the hopes that this will serve to increase the effective sample size and increase the ability to infer relationships for which none of the individual studies provide enough data for reliable statistical analysis.)
Their best estimate is that rice demand will rise by 50% by 2030. Given what I know of Japan’s case, and the trends we already see in urban China, I think that’s too high. My own impetus would be to scan the 5 or so most recent studies, but that would still leave me with a wide range (0.36-0.12 for urban) and mean I’d have to read the papers and judge which might be best. Given variations in datasets and the details of how they set up their regressions, I might not have a basis to do so, and I’d have to mimic them, use an average from the papers.
Our results indicate that national-level income elasticities for general cereals and general meat were 0.40 and 0.48, respectively, in 2000 and that they are projected to decline to 0.12 and 0.36, respectively, by 2030.