Commentary | Trade and Globalization

China trade and jobs: Responding to myths and critics

In Unfair China Trade Costs Local Jobs, I showed that between 2001 and 2008, 2.4 million jobs were lost or displaced by growing trade deficits with China alone. Since then, many critics have come forward to dispute those findings and the methodology used to obtain them. Their critiques and my responses are documented here.

Daniel Ikenson of the Cato Institute has criticized the methodology used in this analysis. The basic logic of our model was explained by my colleague Josh Bivens in Trade Jobs and Wages:

Job loss is by far the most visible and easily understood way that international trade can affect American living standards. The effect of trade flows on American jobs is actually pretty complicated and so requires a bit of untangling. First, trade creates new jobs in exporting industries and destroys jobs when imports replace the output of domestic firms. Because trade deficits have risen over the past decade, more jobs have been displaced by imports than created by exports.

Our research employed a widely used methodology that has been used by proponents and opponents of trade liberalization for decades—the Federal Reserve Bank of New York published a piece with all-but-identical methodology in 2005, finding that between 1997 and 2003 alone rising trade deficits displaced roughly 3 million American jobs.

The U.S. China Business Council (USCBC), an association of multinational companies that have earned massive profits by outsourcing American jobs to China, has also critiqued my estimates of jobs lost or displaced due to growing China trade deficits. This critique springs from the gate with an incorrect assertion—that “U.S. manufacturing jobs…have been in a long decline over the past 4 decades.” A simple look at the Bureau of Labor Statistics Web page would have shown them that the level of manufacturing employment fluctuated between 16.6 and 19.3 million between 1979 and 2000. Then, starting in 2001, employment in this sector cratered and never recovered and it now stands at 11.6 million—fully 30% below its minimum value reached between 1979 and 2000. In short, the level of manufacturing employment was actually quite stable before 2001.

The only quasi-substantive critique of our model forwarded by the USCBC is that it is “built on the faulty assumption that every product imported from China would have been made in the U.S. otherwise.” This is a puzzling interpretation. For one, the model is focused on net imports— imports minus exports. Nowhere does it make the claim that “all imports” from China could or should be replaced with U.S. production. Rather, it makes a ceteris parabis (all else equal) assumption about what the employment effects would be of closing the bilateral trade deficit with China—that is, either exporting more to China or importing less from China.

If the USCBC critique is that bilateral trade deficits don’t matter and that only global trade flows should be examined for their employment impacts on the U.S. economy, then we have done that study. However, we would argue that once one determines that the overall trade deficit is too large for comfort in the United States (and there is widespread agreement on this), one must take the next step and try to ascertain which trading relationships are contributing the most to the overall deficit and hence which of those must undergo a significant adjustment. Given that the bilateral deficit with China now accounts for 80% of the total non-oil deficit, it seems odd to act as if this relationship is trivial to the larger question of external imbalances.

Further, a revaluation of the Chinese currency would not be undone by growing trade deficits with other countries, as the USCBC suggests; rather it would probably have strong multiplier effects in allowing the currencies of many other low-wage exporters to the United States to gain strength without losing market share to China. In short, the U.S./China trade relationship is the linchpin to the overall trade deficit in the United States. While one understands that it’s in the USCBC’s interest to deny this, this doesn’t change the fact that it’s true.

Additionally, while the USCBC extols the virtues of the Chinese market as a destination for U.S. exports, our model includes the effects of both imports and exports, and the net impact of changes in the trade balance, including all component trade with China. If Intel ships a microprocessor to China, and that chip goes into a computer that is shipped back to the United States, our model accounts for both the jobs gained from exports and the jobs displaced by imports.

The effects of growing U.S. trade with China are outlined in Table 1 in my report, which shows changes in trade and jobs supported or displaced between 2001 and 2008. The growth of U.S. exports to China in this period supported about 353,000 U.S. jobs. However, the growth of U.S. imports from China displaced 2,771,000 jobs. The net effect of the growth in imports and exports thus displaced 2.4 million U.S. jobs—the difference between jobs supported by the growth in exports and imports.

The growth of U.S. imports from China after 2001 could have been offset by growing exports. President Clinton, who approved the agreement for China’s entry into the World Trade Organization (WTO), claimed that exports would grow rapidly because of that deal. Kenneth Lieberthal, Clinton’s special advisor for Asia, predicted that the U.S. trade deficit with China would shrink after that agreement was implemented. Instead, the U.S.-China trade deficit more than tripled between 2001 and 2008. 1

U.S. access to China’s market did not increase after it entered the WTO, and our trade deficit with China increased dramatically because China failed to buy enough goods to offset their exports to the United States. Currency manipulation was one of several tools China used to maximize its trade surplus with the United States and other countries. It also suppresses labor rights, maintains extensive tariff and non-tariff barriers to imports, and engages in extensive, illegal subsidization and dumping of exports.

Leading international economists agree that eliminating currency manipulation by China could support the creation of hundreds of thousands or millions of job. Nobel Laureate Economist Paul Krugman has described China’s currency policies as “predatory” and “mercantilist.” He estimates that “for the next couple of years Chinese mercantilism may end up reducing U.S. employment by around 1.4 million jobs.” In Testimony before the House Ways and Means Committee, Fred Bergsten, director of the Peterson Institute of International Economics, estimated that eliminating Chinese currency manipulation “would generate an additional 600,000 to 1,200,000 U.S. jobs.”

Productivity growth—The scoundrel’s last refuge
The USCBC and others have tried to blame manufacturing job loss on productivity growth. But this canard simply does not stand up to close scrutiny. Employment growth in any economic sector is essentially the difference between growth in output and productivity (output per hour). Output growth, all else equal, spurs employment
while productivity growth dampens it. The Figure below illustrates why manufacturing employment has fallen so rapidly over the last three years.

[Figure: Output, productivity growth, and employment in U.S. manufacturing]

Between 1989 and 2000, manufacturing output and productivity growth averaged, respectively, 3.5% and 3.8% per year. As a result, the two largely offset one another and manufacturing employment was relatively stable, as shown in the figure. Between 2000 and 2007 (the last business cycle peak), manufacturing productivity growth actually declined slightly, relative to the previous decade, falling 3.7% per year. Output growth, however, cratered, and has averaged only 0.5% per year in this period. Employment fell 3.1% per year as a result. In short, it is slow growth in manufacturing output—not an acceleration in productivity—that makes 2000-07 different from the previous decade and explains the steep fall in manufacturing employment in that period. Overall, 3.5 million manufacturing jobs were lost between December 2000 and December 2007. An additional 2.5 million manufacturing jobs were lost through December 2009 due to the collapse in output associated with the great recession

Slow growth in U.S. output between 2000 and 2007 can be almost entirely attributed to this rise in imports and the loss of export sales due to currency manipulation and other unfair trade practices. Apparent consumption of manufactures in the United States grew 4.2% per year between 1989 and 2000, and 3.2% per year between 2000 and 2007.2 Furthermore, the growth of world export markets accelerated from 6.4% per year between 1989 and 2000 to 9.4% per year between 2000 and 2007.

The U.S. share of world manufacturing trade increased slightly between 1989 and 2000 and collapsed thereafter, rising from 12.0% in 1989 to 12.3% in 2000 and then falling to 8.1% in 2007, a decline of 4.2 percentage points (35%) (International Monetary Fund 2010). If the U.S. share of world export markets had recovered to 12.3% in 2008, U.S. exports would have been $670 billion higher than they actually were. U.S. exports of manufactured goods were $1.2 trillion in 2008, and would have been $1.8 trillion but for the decline in U.S. shares of world export markets. If the U.S. share of world export markets had recovered in 2008, we could have run a trade surplus in manufactured goods of up to $200 billion.

But for currency manipulation and other trade practices by China and a few other Asian countries, U.S. manufacturing exports and output in the 2000-07 period would have grown at least as rapidly as they did in the 1989-2000 era, and manufacturing employment would have remained stable or even increased, despite high levels of productivity growth in manufacturing. It should be noted that the Congressional Budget Office shares our view that rising trade competition is a primary determinant of U.S. manufacturing employment levels.

Currency revaluation will reduce the U.S. trade deficit
Mr. Ikenson of the Cato Institute claims that “appreciation [of the renminbi] will not reduce the bilateral trade deficit.” The USCBC hedges a bit, stating that “China’s exchange rate is probably not as significant a factor in the U.S. trade deficit that some make it out to be.” There are several problems with these arguments, which are based on analysis of short-run changes in trade flows in the wake of a 20% revaluation in the renminbi (RNB) that took place between 2005 and 2008.

First, despite the earlier revaluation, the RNB remains substantially undervalued by approximately 40% relative to the U.S. dollar according to William Cline and John Williamson of the Peterson Institute. It is clear that a 40% increase in the price of imports from China and a comparable decrease in the cost of U.S. exports will, over time, improve the trade balance, as suggested by the trade and jobs estimates of Krugman and Bergsten, above.

Second, China is just one of a group of Asian countries that are manipulating their currencies, including Hong Kong, Malaysia, Taiwan, and Singapore, all undervalued by 25% to 32%. The Japanese yen is also undervalued by approximately 15%, relative to the dollar. Undervaluation of these currencies reduces U.S. exports not only to these countries, but also to third-country markets where goods made in the U.S. must compete with these unfairly traded exports.

It is important to note that although the real, trade weighted value of the dollar fell 13% between early 2002 and the end of 2005, the trade deficit did not begin to improve until 2007, some two years after China allowed the RNB to appreciate beginning in July 2005. Although the U.S.-China trade deficit did continue to grow in the 2005-07 period, the U.S. global trade balance improved due to the rapid growth of exports to the rest of the world.3

Global currency realignment, along the lines suggested by Cline and Williamson, would reduce the real, trade-weighted value of the U.S. dollar by an additional 5% to 10%, resulting in a cumulative 26% to 31% depreciation in the dollar since February 2002. There have been two prior periods when trade policy interventions resulted in sustained depreciation of the dollar on a similar scale, in 1971 (following the Nixon import surcharge) and between 1985 and 1987 (the Plaza Accord era). In each case, substantial improvement in the U.S. current account balance was achieved within two to four years.

Mr. Ikenson twisted himself into knots claiming that a 40% increase in the price of imports from China and a comparable decrease in the cost of U.S. exports would not, over time, improve the trade balance. It seems that in Cato’s universe, international trade is the only market where the price mechanism doesn’t work.

While a 40% increase in the yuan may hurt the profits of multinational companies who make up the USCBC, which includes many Fortune 500 companies such as Amway, Apple Inc., Bank of America, Boeing, BP, Exxon, Fed-Ex, Home Depot, IBM, Intel, Target, and Wal-Mart, it would be beneficial for the larger U.S. and Chinese economies. It’s time we put national interests before corporate interests.

—The author thanks Ross Eisenbrey and Josh Bivens for comments.


1. See Unfair China Trade Costs Local Jobs, p.4, for full quotations and source notes.

2. Sources: BEA (2010) data on gross output (shipments net of intra-industry trade) in manufacturing and U.S. International Trade Commission (2010) data on U.S. trade in manufactured products. Apparent consumption equals output plus imports less exports.

3. It typically takes 18 to 24 months for the trade balance to begin to improve after the initial effects of depreciation take effect, due to well-known “j-curve” effects. Mr. Ikenson correctly notes that the initial response to a revaluation is primarily observed in higher prices for imports drives up the value of imports. He failed to note that trade volumes (both imports and exports) usually respond with a lag.

Bergsten., C. Fred. 2010. Testimony before the Committee on Ways and Means, U.S. House of Representatives. “Correcting the Chinese Exchange Rate: An Action Plan.” March 24.

Bivens, Josh. 2008. “Trade, Jobs, and Wages.” Issue Brief #244. Washington, D.C.: Ec
onomic Policy Institute. May 6.

Bureau of Economic Analysis, U.S. Department of Commerce. 2010. Gross-Domestic-Product-by-Industry Accounts: Gross Output by Industry. Washington, D.C.: Bureau of Economic Analysis.

Cline, William R. and John Williamson. 2010. “Notes on Equilibrium Exchange Rates: January 2010.” Policy Brief 10-2. Washington, D.C.: Institute for International Economics.

Congressional Budget Office. 2008. “Factors Underlying the Decline in Manufacturing Employment Since 2000.”

Groshen, Erica L., Bart Hobijn, and Margaret M. McConnell. 2005. U.S. jobs gained and lost through trade: A net measure.” Federal Reserve Bank of New York, Current Issues in Economics and Finance, 11(8).

Ikenson, Daniel. 2010. “China Trade and American Jobs: Studies Suggest that One-Half to Two-Thirds of the Value Of ‘Chinese’ Imports is Added in Other Countries, Including the U.S.” International Monetary Fund. 2010. International Financial Statistics. CD-ROM. January.

Krugman, Paul. 2009. Chinese New Year. The New York Times. December 31.

Scott, Robert E. 2009. “Re-balancing U.S. Trade and Capital Accounts.” EPI Working Paper #286. Washington, D.C.: Economic Policy Institute. December 31.

Scott, Robert E. 2010. “Unfair China Trade Costs Local Jobs: 2.4 Million Jobs Lost, Thousands Displaced in Every U.S. Congressional District.” Briefing Paper #260. Washington, D.C.: Economic Policy Institute. March 23.

U.S. China Business Council. 2010. “Flawed Study on China Jobs, Focus on Currency Distract from Real Issues, USCBC Says.” Washington, D.C.: U.S.-China Business Council. Press Release. March 24.

U.S. International Trade Commission (USITC). 2010. “USITC Interactive Tariff and Trade DataWeb.” Washington D.C.: USITC.

See related work on China trade | Trade and Globalization

See more work by Robert E. Scott