Estimating the effect of renminbi appreciation on US jobs: A comment on Francois' China result

William Cline 23 April 2010



In a recent study, Francois (2010) estimates that if China appreciated the renminbi by 10%, the US trade balance would rise by $100 billion but the number of US jobs would decline by 430,000. He uses a computable general equilibrium (CGE) model to make this calculation. He allows for below-full capacity and sticky wages so that it is possible for a change in the external balance to affect the level of employment. The paradoxical negative sign on employment as a consequence of the currency correction stems from the model specification that emphasises induced losses of jobs throughout the economy that result as a consequence of the increase in costs of intermediate inputs imported from China and used in the US economy. Francois argues that the gain of employment in exports and import substitutes would be too small to offset the loss of jobs in the general economy; hence the net loss of 430,000 jobs. This column examines whether these results make sense.

Direct and indirect US employment effects of a renminbi appreciation

Francois’ Table 2 suggests that in 2009 the United States imported approximately $100 billion in intermediate inputs from China (proxied by “office machines, electronics, and parts thereof”). So the postulated 10% appreciation of the currency would have a direct cost impact of $10 billion on US manufacturing intermediate input costs. In 2009, value added in manufacturing was about $1.2 trillion (BEA, 2010a). The ratio of gross output to value added in manufacturing industry in the input-output tables is 3.1 to 1 (BEA, 2010b). So gross output in manufacturing in 2009 amounted to $3.7 trillion. The extra cost of intermediate inputs into manufacturing from a 10% appreciation of the RMB would thus amount to $10 billion/ $3.7 trillion = 0.27 percent of gross output value. Applying a price elasticity of unity, US manufacturing output would decline by 0.27 percent. US manufacturing employment in 2009 stood at 11.9 million (BLS, 2010). Applying the 0.27 percent decline, the loss of jobs from the impact on intermediate input costs in manufacturing would amount to 32,000 jobs.

In comparison, consider the direct job effect of the trade balance change. In 2009 gross output per worker in manufacturing amounted to $311,000 per worker (i.e. $3.7 trillion / 11.9 million workers). Dividing the Francois estimate of $100 billion trade balance gain by this parameter, the direct job effect from the trade balance gain would amount to 320,000 jobs.

This exercise suggests that something appears to have gone wrong in the Francois calculations. A reasonable approximation of his two opposing effects suggests that the 10% RMB appreciation would create 320,000 jobs from the US trade balance improvement and eliminate only 32,000 jobs from the induced effect of higher intermediate input costs to US manufacturing. These estimates suggest that the standard result of job gain from trade balance improvement (under unemployment conditions) would hold even for appreciation of China’s currency, because the induced intermediate input costs have an adverse impact that is an order of magnitude smaller than the direct trade effect.


BEA (2010a). "Table 6.1D. National Income Without Capital Consumption Adjustment by Industry", National Income and Product Accounts.
BEA (2010b). "The Use of Commodities by Industries before Redefinitions (1997 to 2007)", Industry Economic Accounts.
BLS (2010). "Table B-1. Employees on nonfarm payrolls by industry sector and selected industry detail", Establishment Data, Historical Employment.
Francois, Joseph (2010). "Deconstructing Sino-US codependence: revaluation, tariffs, exports and jobs " in S. Evenett (ed), The US-Sino Currency Dispute: New Insights from Economics, Politics and Law, CEPR and VoxEU: London.



Topics:  Exchange rates

Tags:  competitiveness, renminbi, Francois

Senior Fellow, Peterson Institute for International Economics


CEPR Policy Research