The Federal Reserve Board eagle logo links to home page

The Growth of Chinese Exports: An Examination of the Detailed Trade Data

Brett Berger** and Robert F. Martin**

NOTE: International Finance Discussion Papers are preliminary materials circulated to stimulate discussion and critical comment. References in publications to International Finance Discussion Papers (other than an acknowledgment that the writer has had access to unpublished material) should be cleared with the author or authors. Recent IFDPs are available on the Web at http://www.federalreserve.gov/pubs/ifdp/. This paper can be downloaded without charge from the Social Science Research Network electronic library at http://www.ssrn.com/.


Abstract:

Over the past decade, Chinese exports have boomed, increasing far faster than GDP growth. What can account for this explosion? Our paper uses finely detailed Chinese export data (8-digit HS codes) combined with U.S. trade data to explore this question. Although exchange rate policy clearly boosted the trade surplus, and the structure of the economy, e.g. abundant cheap labor, encouraged investment, these alone cannot account for the changing composition and acceleration of exports. We find that the growth in exports is most likely a product of effective Chinese industrial policy and fortuitous timing. The detailed trade data reveal that key "new" technology goods, such as cell phones, LCD screens, and laptops played a critical role.
Finally, we use the data to examine the relationship between Chinese exports and global manufacturing, in particular U.S. manufacturing employment. We find that increased Chinese competition in both domestic and U.S. export markets likely lowered U.S. manufacturing employment between 2000 and 2007. Chinese policy is not, however, wholly responsible. Some job losses, such as in textile production, were no doubt the result of China's natural comparative advantages, while other U.S. job losses are attributable to relatively low investment and slow GDP growth in the United States following the 2001 recession.

Keywords: China, trade, manufacturing

JEL classification: F12, F40, E65


Introduction

Between 2000 and 2007, the value of Chinese exports more than quadrupled and rose from 20 percent to 35 percent of GDP (Figure 1). Imports failed to keep pace and the Chinese economy became more dependent on external demand; the current account surplus, mirroring the trade surplus, ballooned from less than 2 percent of GDP to a peak of 10 percent.

Figure 1: Chinese Export Boom

Data for Figure 1 immediately follows

Source: China National Bureau of Statistics.

Data for Figure 1

TimeExports Quarterly, Billion $
0:01 59804.86
0:02 63182.35
0:03 63823.05
0:04 62534.41
1:01 67710.71
1:02 65894.01
1:03 66202.22
1:04 66882.95
2:01 74111.75
2:02 77400.34
2:03 85403.47
2:04 87519.95
3:01 98337.11
3:02 103762.82
3:03 110953.79
3:04 122926.67
4:01 131206.11
4:02 142344.51
4:03 149407.97
4:04 166712.74
5:01 176630.41
5:02 186619.55
5:03 192136.8
5:04 203507.17
6:01 223304.19
6:02 232396.32
6:03 245202.44
6:04 263172.21
7:01 290363.26
7:02 297629.51
7:03 306694.46
7:04 323074.17
8:01 350119.29
8:02 365385.17
8:03 375388.34
8:04 338422.86
9:01 280475.5
9:02 279343.35
9:03 296751.15
9:04 338958.53
10:01 364255.76
10:02 393463.73
10:03 391769.17
10:04 423294.6
11:01 455630.17
11:02 480655.36

Data for Figure 1 cont'd

YearPercent of GDP
2000 20.8
2001 20.1
2002 22.4
2003 26.7
2004 30.7
2005 33.9
2006 35.9
2007 35.2
2008 32
2009 24.1
2010 25.2

Although other economies have had success utilizing an export-led growth strategy, such success in a country of China's size is unprecedented. In 2009, China became the largest global exporter and in 2010 passed Japan as the second largest economy in the world.1 How was China able to accomplish this goal? In this paper, we use finely detailed trade data to differentiate amongst the various reasons given for China's rapid growth of exports. The reasons for China's success also shed light on the effects China's export boom have had on the manufacturing sectors of other countries.

The data make clear that no one story, such as an undervalued exchange rate, can explain the surge in growth, as the growth was concentrated in specific sectors--textiles/apparel/furniture, metals (steel), and machinery--and within the machinery sector, growth was highly concentrated in specific products. Thus to answer the question of why exports boomed, one must answer why these particular sectors and products experienced outsized growth relative to others.

Textile and apparel producers benefited from the expiration of restrictive trade agreements, and these sectors along with furniture producers were also disproportionately aided by China's entry into the World Trade Organization, which allowed the country to take greater advantage of its vast pool of low-skill labor. Heavy industry, such as steel production, benefited from energy and capital subsidies, as well as from reform of state-owned-enterprises (SOEs), whose new profitability translated into a rapid expansion of capacity.

A key contribution of this paper is to shed some light on machinery exports. Aggregate trade data show that machinery exports accounted for almost half of Chinese export growth between 2000 and 2007. This has been cited by some as evidence of Chinese mercantilism, under the assumption that a developing country is unlikely to be an exporter of capital goods without official support. However, the highly disaggregated Chinese trade data, compiled by hand for this paper, show that the growth of Chinese machinery exports was concentrated in a few specific high-tech goods--cell phones, LCD displays, laptops, and integrated electronic circuits--not the prototypical examples of capital goods.

Although, we attribute some of the success of Chinese high-tech exports to industry-specific supports, such as the establishment of science parks, China's export boom likely would not have been achievable without a healthy dose of good luck in terms of timing. Global demand skyrocketed over this period for products based on these technologies, and Chinese producers were particularly well placed to take advantage: China's own domestic market became a key source of demand; labor costs in neighboring high-tech producers were on the rise; and, investment in the West, particularly in the high-tech sector, was severely impaired from the 2001 recession.

As Chinese exports surged, we find that China's gains in global manufacturing came primarily at the expense of advanced economies, with the U.S. share falling the most. As part of our examination of the effect on the United States, we look at manufacturing employment and conclude that the rise in Chinese exports led to some employment losses in certain sectors, but subpar GDP growth and investment were the principal factors behind the lack of a rebound in manufacturing employment following the 2001 recession.

The welfare implications for the United States of the increase in Chinese exports is beyond the scope of this paper, but it is important to note that it is not at all clear that the costs outweighed the benefits. The cost in loss of jobs is obvious, but the benefits included lower prices, more rapid adoption of new technology, and efficiency gains from the removal of trade barriers and through increased competition.

The paper is organized in 4 sections. Section 1 gives some background on the evolution of the Chinese economy and the factors that led to the boom in exports. This section examines Chinese exports at the readily-available 2-digit level of the Harmonized Commodity Description and Coding System (HS), which has been adopted by most countries. Section 2 uses 8-digit HS codes to delve further into the details of Chinese trade. Section 3 examines other countries' manufacturing sectors over the period, with a focus on the United States. Section 4 speculates on the implications of our research for the prospects of future Chinese export growth following the 2009 financial crisis. Section 5 concludes.


Section 1:  The Growth of Chinese Exports: An Initial Look

In the three decades since economic reforms were enacted in the late 1970s, China has experienced a remarkable period of consistently robust economic growth, with real GDP increasing 23-fold since 1977. The composition of growth, however, has evolved away from consumption over this period. Household consumption fell from an already low 50 percent of GDP in the early 1980s to around 35 percent by the mid-2000s as investment soared and the economy became more export-oriented. This process accelerated significantly at the start of the new millennium. In 2000, China's exports, measured in dollars, were a third of those of the United States and around half of those of Japan and Germany. By 2009, China had become the largest exporter in the world.

Figure 2 sets the stage by showing, even at a fairly aggregated level, that Chinese exports growth was relatively concentrated in a few sectors. This growth was dominated first and foremost by machinery exports (HS categories 84 and 85), which accounted for about 45 percent of export growth. Textiles/furniture and metals accounted for 15 percent and 11 percent, respectively. The rest of this section examines each of the major categories of Chinese export growth.

Figure 2: China's Export Growth is Concentrated in a Few Sectors

Data for Figure 2 immediately follows

Source: China Customs (from CEIC) and U.S. Census Bureau.

Data for Figure 2

Classes of GoodsShare of total export growth
Machinery Exports 0.466687063
Textiles, Apparel, and Furniture 0.159709922
Metals including Iron and Steel 0.102228759
Optical 0.032469496
Other 0.23890476

Textile, Apparel, and Furniture Exports2

These categories of exports grew 220 percent from 2001 through 2007. No doubt, trade policy was a primary cause. Prior to China's entry into the World Trade Organization (WTO) in December 2001, China faced prohibitive tariffs and constraining quotas in textile/apparel and furniture markets. The production of these goods is intensive in low-skilled labor, an area in which China has an obvious natural comparative advantage.

Until 2005, Chinese exports of apparel and textiles were limited by a series of gradually less constraining multilateral agreements--the Multifiber Arrangement (MFA, through 1995) and subsequently the Agreement on Textiles and Clothing (ATC). Without these agreements, China's enormous supply of cheap labor likely would have led to a much larger share of the global market earlier. Indeed, Brambilla et al. (2009) found that China was constrained more than any other nation by these agreements. Consequently, when the MFA expired, textile and apparel export growth from China rose rapidly, and surged further with the expiration of the ATC. As China gained market share, exports from most other regions declined.

Figure 3: U.S. Furniture Imports (HS 94)

Data for Figure 3 immediately follows

Source: U.S. International Trade Commission Dataweb.

Data for Figure 3

YearChinaTaiwanCanadaMexicoOther
19962.3947811.2596513.0935681.7785853.484692
19972.9943611.2052443.6999932.2379113.944428
19983.9469581.2335364.3335742.6982364.54293
19995.5471661.2770874.7276283.3358135.483141
20007.2019211.2777965.3130943.8207276.21266
20017.4920340.941684.9144053.9138415.955039
20029.9207310.9731964.9470114.5431676.319155
200311.818320.9240085.0854645.0580326.774231
200414.417450.9297745.6112965.1466217.601129
200517.044630.8715825.794025.2632438.219429
200619.350520.8258525.7725215.4859318.315633
200720.403980.8163325.306735.6698078.780581
200819.472370.8218884.5860115.0422928.413911
200916.04190.611362.9594853.8529346.598356
201019.935360.7763053.5029665.6414437.88546

Chinese furniture exports (HS 94), began to increase rapidly back in the early 1990s and accelerated in the 2000s. Some of this production was shifted from Taiwan, which accounted for a large share of global furniture exports in the early 1990s. By 2007, Taiwan's share of U.S. furniture imports had fallen to 2 percent and China was by far the largest source.

Metals3

By 2000, China was already the largest producer of steel, but output was only 25 percent higher than either Japan or the United States (Figure 4). But by 2009, China was producing 6½ times more steel than second place Japan and almost 10 times more than the United States, each of whom have experienced large declines in steel production since the beginning of decade.4 Overall, Chinese metals exports grew 630 percent from 2001 through 2007. Why would a developing country with an enormous supply of labor experience some of its greatest export growth in a capital and energy intensive industry?

Figure 4: Global Steel Production

Data for Figure 4 immediately follows

Source: Steel Statistics Yearbook 2010.

Data for Figure 4

YearUSChinaJapanRussiaGermanyOtherSouth KoreaIndiaBrazilUkraineItalyOther
2000101.803128.5106.44459.13646.376406.67543.10726.92427.86531.76726.759250.253
200190.104151.634102.86658.9744.803402.69643.85227.29126.71733.10826.545245.183
200291.587182.366107.74559.7774.515458.1845.3928.81429.6043.452.666348.256
200393.677222.336110.51161.4544.809437.13246.3131.77931.14736.9322.758288.206
200499.681282.911112.71865.58346.374464.24147.52132.62632.90938.73828.604283.843
200594.897353.24112.47166.14644.524472.81347.8245.7821.6138.64129.35289.612
200698.557419.149116.22670.8347.224495.19248.45549.4530.90140.89131.624293.871
200798.102489.288120.20372.38748.55517.651.51753.46833.78242.8331.553304.45
200891.35500.312118.73968.5145.833504.27753.62557.79133.71637.27930.59291.276
200958.196567.84287.53460.01132.67417.77148.57262.83826.50629.85519.848284.152

First, energy prices were heavily managed by the government and significantly subsidized. As the 2006 U.S. Manufacturing Energy Consumption survey confirms, steel and aluminum production are among the largest industrial energy consumers in terms of energy per dollar of value added.5 Second, the cost of capital for SOEs, which dominate China's heavy industry, was extremely low. SOEs had ready access to bank borrowing at low interest rates because of implicit government backing. Third, the SOEs made substantial strides in improving efficiency and lowering costs. Hsieh and Klenow (2009) estimate that improvements in resource allocation account for about 2 percentage points per year of Chinese total factor productivity growth between 1998 and 2005.

Reform of the SOEs began in the mid-1990s and dismantled the "iron rice bowl," the system of housing, pensions, and health care that accompanied SOE employment. As a result, the SOEs, which were money-losers in the decades prior to reform, began earning substantial profits for the first time. Since SOEs did not pay dividends to the government, they piled up retained earnings and generally had few options but to reinvest the earnings in expanding capacity. This fed a circle in which profits led to greater capacity and still greater profits. From 1995 to 2000, China's steel industry averaged $400 million in annual operating profits and crude steel production rose at an average pace of 6 percent per year. Beginning in 2000, profits and production began to rise rapidly in tandem--profits climbed from $2 billion in 2001 to more than $21 billion in 2007, and crude steel production rose at an average annual rate of 25 percent. Given the incentive structure, i.e. subsidized inputs and political approval of output growth, this reinvestment was optimal from the perspective of Chinese steel producers. Domestic demand could not absorb this massive production growth and China went from being a net importer of steel, as late as 2004, to the largest net exporter in the world.

Machinery6

This category of exports grew 520 percent from 2001 through 2007 and, as noted earlier, accounted for roughly 45 percent of China's total export growth. A wide range of economists have attributed much of these gains to China's most visible trade policy, its exchange rate regime. For much of the decade, the renminbi was pegged to the U.S. dollar, and, as evidenced by China's massive accumulation of foreign exchange reserves, authorities have intervened heavily to keep it from appreciating. But the detailed trade data, discussed in the next section, indicate that this was likely not the primary factor in the growth of China's machinery exports as growth was concentrated in a few select products, whereas the exchange rate would be expected to have a broader effect by lowering the prices of Chinese goods generally.


Section 2:  Machinery Exports: The Devil in the Details7

In order to understand the growth of Chinese machinery exports, we now turn to the detailed trade data. Figure 5 shows exports of the machinery categories and the optical category (HS 90) at the 4-digit level; on the left is data for 2002 and on the right is data for 2007. As the figure illustrates, growth in the machinery categories was highly concentrated in high-tech goods. Digging even deeper, the 8-digit categories, shown in Figure 6, reveal that the growth of machinery exports was dominated by four products--cell phones, liquid crystal displays (LCDs), integrated electronic circuits, and laptops--which together accounted for more than a third of the growth.

This concentration of export growth argues against China's exchange rate regime playing the major role. The exchange rate should have a more-or-less even handed influence across China's export industries, as all goods are made relatively cheaper, and therefore it is not a plausible explanation for the outsized growth of particular categories. Instead, the influence of the exchange rate can perhaps best be seen in the wide range of smaller bars in Figure 4. Most subcategories of machinery exports increased significantly over the period, but their growth was dwarfed by a few categories for which we have special stories. For this reason, we believe papers that explain Chinese trade using more aggregated data, such as Ahmed (2010), Marquez and Schindler (2007), and Thorbecke and Smith (2010) likely overestimate the importance of the exchange rate. But even using the relatively high elasticities common in the literature, the exchange rate would still account for only a minority of machinery export growth over the period. For example, assuming the renminbi became 30 percentage points more undervalued from 2001 to 2007 and the elasticity of Chinese exports was 1½, then the exchange rate would have accounted for less than 10 percent of machinery export growth.

For laptops, cell phones, LCDs, and integrated electronic circuits, China became the dominant global manufacturer and exporter because these products share characteristics that turned out to favor China. First, they are by-and-large "new" products. Though laptops and cell phones have been around for more than twenty years, each succeeding generation incorporates new technologies and production methods. Hence, there was an opportunity for new participants to enter the market and the need for established participants to open new facilities. Second, these products require large capital investments which were likely a barrier to entry for many firms. In particular, high-tech firms in the United States, reeling from the dot-com bust and a capital overhang, were not well positioned to invest. In contrast, Chinese firms and other Asian firms, several years removed from the Asian financial crisis, were better positioned. Third, existing production of related technologies was dominated by Japan and China's newly industrialized neighbors, such as Taiwan and Korea, which had historically dominated China's foreign direct investment and where labor costs were rising. Fourth, there was an explosion in global demand for these products which came to dominate their broader market categories--computers, phones, televisions and computer monitors. Fifth, the leader in the booming global demand for these products was China's own domestic market, where urban disposable income more than doubled between 2001 and 2007.

Figure 5: Within 4 and 8 Digit HS Categories Exports are even more Concentrated

Figure 5 has 6 panels of bar charts in 2 columns and 3 rows. The panels show Chinese exports at the HS-4 digit level in 2002 (the left column) and 2007 (the right column). The top panels shows Chinese exports for HS 84, non-electrical machinery. For all of the panels each bar represents the level of exports at each 4-digit level within the 2-digit larger category. The middle pnaels shows the exports for HS 85, electrical machinery. The bottom panels shows exports for HS 90, optical apparatus. In the top left panel (2002 HS 84), only two categories stand out Automatic Data Processing Machines (computers), about $30 billion, and parts for these machines, about $20 billion. The others categories look like they are practically zero because they are on the scale necessary to show the 2007 values in the right panel. In 2007, shown in the top right paanel, almost all of the categories show growth but most are still quite small with all except 3 with well less than $10 billion. Computers (8471) has climbed to $94 billion, computer parts is $33 billion , and copiers, scanners, and printers (8443) is $19 billion. The story is very similar for the middle panels which show HS 85. Nothing stands out in 2002 like it did for HS 84. In 2007 a few categories do stand out, telephones (8517) at $79 billion, monitors (8528) at $37 billion, and integrated circuits (8542) at $24 billion. The bottom panels for HS 90 show smaller bars and are on a smaller scale. In 2002, none of the bars reach even $2 billion. In 2007, the chart is dominated by liquid crystal devices (9013) at $21 billion, as none of the other categories reach $2 billion.

Source: China Customs Statistical Yearbooks, 2002 and 2007.

Figure 6: A Glance at the 8-Digit Data

Figure 6 contains 4 panels of bar charts similar to those in figure 5. In this case each panel shows Chinese exports in 2007 of HS 8-digit categories for the key 4-digit HS codes identified in Figure 5. The bars shows the exports in billions of U.S. dollars. The top left panel is for computers (HS 8471). It shows that almost all of the growth of Chinese exports of computers was in the portable computer category, or laptops with $53 billion in exports in 2007.The top right panel shows a similar story for computer parts (HS 8473) reaching $28 billion. The bottom left panel is for telephones (HS 8517) and shows that nearly all of the growth in exports was in cell phones at $36 billion and associated parts at $16 billion. The bottom right panel is for monitors and projectors (HS 8528) and the two largest categories are for laptop screens at $18 billion and LCD computer monitors at $5 billion.

Source: China Customs Statistical Yearbooks, 2002 and 2007.

Reactors, Boilers, Machinery and Appliances (HS 84)

Returning to Figure 5, the top two panels show the 4-digit subcategories of HS 84 in 2002 and 2007, respectively. The top-left panel looks empty using the scale needed to show the 2007 data, shown to the right. Growth in computers (HS 8471) and associated parts (HS 8473) dominate, accounting for 52 percent of the total growth in HS 84.

These two categories were the largest in 2002 as well, but the composition of the computers making up the categories changed drastically over the period. Portable computers (i.e. laptops) made rapid gains in market share, relative to the traditional desktop computer, over the past decade. In 2002, laptops accounted for 7 percent of HS 8471 exports. By 2007, its share had surged to 42 percent. As shown in the upper left panel of Figure 6, portable computers (HS 84713000) were by far the largest subcategory of HS 8471. Similarly, parts and accessories for HS 8471 (HS 8473090) were the largest subcategory of HS 8473 (top right panel).

Figure 7: U.S. Imports of Laptops

Data for Figure 7 immediately follows

Source: U.S. International Trade Commission Dataweb.

Data for Figure 7

Country2000200120022003200420052006200720082009
China0.0110620.0225080.6321624.1582537.71579610.6703312.8277117.0018319.2354122.90866
Malaysia0.0493541.089382.9117013.8513585.1484996.5807717.6317197.7676816.4089132.46151
Taiwan3.7960623.0684833.407822.1656751.2432890.3719990.2302430.1866170.0967890.102735
Other3.2623.2583.493.011.9271.7982.0622.0941.8621.452

From 2000 to 2007, during China's export boom, global shipments of personal computers grew 105 percent, with laptops accounting for 47 percentage points of the growth.8 In 2009, portable computers were estimated to have accounted for the majority of computer shipments for the first time.9 At the same time, China became the largest producer of laptops, surpassing Taiwan, which had accounted for 64 percent of global laptop production in 2002.10

The surge in Chinese exports of laptops was not primarily a case of mainland production displacing other producers, but this was part of the story. Figure 7 shows U.S. imports of laptops by country of origin and reflects the move away from Taiwanese production of laptops to the mainland. In 2000, Taiwan accounted for more than 50 percent of U.S. laptop imports, whereas China accounted for less than ¼ percent. By 2007, China accounted for 63 percent of U.S. imports of laptops, and by 2009 it accounted for 85 percent. Indeed, since 2007, Malaysia and China have accounted for more than 90 percent of U.S. imports of laptops. Nevertheless, through 2007, the value of non-Chinese imports actually increased as the overall market for laptops expanded more rapidly than Chinese imports. Given the price declines in portable computers over this period, real shipments grew much faster still. According to China's National Bureau of Statistics (NBS), Chinese production of microcomputers increased from 7 million units in 2000 to 143 million units in 2007.

Government polices helped nurture domestic capabilities in consumer electronics and other advanced areas that would most likely not have developed in their absence (Rodrick (2006)). China's laptop industry, for example, was aided by the creation of "science parks," the development of which have been instrumental in the surge of Chinese high-tech exports generally. Although these parks were originally formed to promote indigenous innovation, they evolved to depend more on foreign investment and technology transfer. As noted in Sutherland (2005), in 2000, the Ministry of Science and Technology and Ministry of Foreign Trade approved a trial for about a third of the parks to become "high-technology export bases." Now, all of the parks have a bias toward this type of production. Foreign investment is drawn to these parks in particular by preferential tax policies. In addition, cheaper labor was a draw for high-tech investment, as in the case for Taiwanese laptop producers. As shown in Figure 8, between 1995 and 2000, Taiwanese labor costs in computer, electronic, and optical manufacturing rose by more than a third. As high-tech production moved to the mainland, wage growth in Taiwan slowed. Foreign firms were also drawn to China because of the potentially huge domestic market. The Chinese market soared from 21 million personal computers in use at the end of 2000 to 86 million in 2007, second to only the United States.11

Chinese laptop exports illustrate one of the primary points of Amiti and Freund (2010), the only other paper we are aware of that used detailed Chinese Customs statistics. Using 6 and 8-digit HS codes, they concluded that Chinese export growth was mainly accounted for by growth in trade of existing products. Laptops were an existing product, with the first laptop computers manufactured in the early 1980s. However, new technology for producing LCD screens, new integrated electronic circuits, along with the widespread adoption of wireless technology, moved the laptop from a business-only luxury to a household consumable. In this sense, laptops were a new good in the 2000s.

Figure 8: Rising Wages in Taiwan Led Production to Shift to China

Data for Figure 8 immediately follows

Source: CEIC.

Data for Figure 8

YearEarning Index
199574.20448
199675.58524
199779.43853
199884.43927
199989.42567
200092.83054
200191.29733
200293.16378
200395.44944
200494.09004
200597.04502
2006100
2007101.8078

Electrical Machinery (HS 85)

Electrical machinery exports (HS 85, the middle panels of Figure 5) also grew as the result of new goods. The growth is not quite as concentrated as in HS 84, but it is still dominated by three product categories, phones (HS 8517), monitors and televisions (HS 8528), and integrated electronic circuits (HS 8542). Moreover, the 8-digit data (bottom panels of Figure 6) show that the growth in these categories was also almost wholly accounted for by cell phones, liquid crystal displays (LCDs), and microprocessors and memory. LCD technology also accounted for the majority of the surge in the optical apparatus (HS 90) category (the bottom panels of Figure 5). As was the case with laptops transforming the computer market, cell phones and LCDs revolutionized the telephone, television, and computer monitor markets.

Cell Phones

Like for laptops, soaring global and domestic demand fueled China's rise in cell phone production and exports. Global cell phone users rose from about 750 million in 2000 to roughly 4 billion in 2008.12 Chinese mobile phone subscriptions rose from 85 million in 2000 to around 550 million in 2007 and 750 million in 2009, the largest market in the world. Led by Motorola, all of the major multinational cell phone manufacturers transferred at least some, and in some cases all, of their handset production to China.13 Although initially dominated by these firms, Chinese companies are gaining market share both at home and abroad. U.S. trade data reflects the growth of the global market and China's market share. U.S. imports of telephone sets rose from $12 billion in 2001 to $55 billion in 2007, with China's share rising from 12 percent to 38 percent (Figure 9). Of this 2007 total, $28 billion were imports of cell phones with China accounting for half of these.

Figure 9: U.S. Imports of Phones14

Data for Figure 9 immediately follows

Source: U.S. International Trade Commission Dataweb.

Data for Figure 9

Country2000200120022003200420052006200720082009
China2.7261982.9568914.2861955.4634978.92083713.5264716.9993622.530123.1627121.88311
Japan3.292922.0234411.41081.1212821.2681141.4609831.1806021.2576971.238250.978805
Malaysia1.8182441.8506312.228843.4061743.6870257.3847367.6355944.699154.4242953.453447
Taiwan0.8917950.806410.6507280.5938520.6223081.0894852.1076012.0344312.30772.723109
Other20.3739316.954116.7208517.6183621.1914621.995721.2378524.5114429.5592728.50757

Liquid Crystal Displays (LCDs)

The market for televisions and computer monitors experienced a technical revolution over the past decade. In 2000, cathode ray tube (CRT) technology dominated the two markets. By 2004, LCDs and CRTs each had about half of the global market for computer monitors, but CRTs still accounted for around 90 percent of the television market. However, by 2008, LCDs accounted for more than 50 percent of the television market and, in 2009, around 70 percent.15 China's NBS estimates that production of color televisions in China doubled from 42 million units in 2001 to 84 million in 2007. China accounted for only 3 percent of the $8 billion in U.S. imports of televisions in 2001. By 2007, 39 percent of the $39 billion in U.S. imports came from China (Figure 10).

Figure 10: U.S. Imports of Televisions and Monitors (HS 8528)

Data for Figure 10 immediately follows

Source: U .S. International Trade Commission Dataweb.

Data for Figure 10

Country20002001200220032004200520062007200820092010
China0.170.250.811.452.284.887.6115.0517.114.515.06
Japan0.661.231.582.112.722.281.61.160.920.610.84
Korea0.130.210.320.780.970.720.40.540.440.440.51
Taiwan0.130.130.210.511.121.681.761.971.510.911.09
Mexico4.554.714.795.297.49.6314.1218.6718.3215.0615.68
Other1.471.492.21.971.972.282.21.471.611.642.24

Integrated Electronic Circuits

Although microprocessors and memory have been a part of personal computers since their introduction in the late 1970s, the rapid turnover of the technology in these products, with new manufacturing processes required for each successive generation, makes them similar to "new" products every few years. For example, between 2000 and 2008, there were 4 different production processes, characterized by ever smaller etching technology, utilized by the main semiconductor manufacturers. Because of the ever-changing technology and highly specialized processes, the capital requirements can be enormous. For example, a new semiconductor fabrication plant can cost as much as $5 billion dollars, with the equipment and necessary inventory holdings costing billions more.16 To be profitable, the plants need to run at high volumes. According to the NBS, Chinese production of semiconductor integrated circuits increased from less than 6 billion pieces in 2000 to 42 billion pieces in 2007.

Unlike for the other categories of exports discussed, the United States is not one of the primary markets for Chinese exports of integrated circuits. U.S. imports of integrated circuits have remained relatively flat since 2002 at around $22 billion, with China accounting for only $1.4 billion in 2007. This is likely because integrated circuits are an intermediate good, with China primarily exporting them to other countries as a step in the production process; the United States, however, may well be the primary market for the final product.


Section 3:  The Effect of the Rise in Chinese Exports on Global Manufacturing

From 2000 to 2007, the value added of Chinese manufacturing output more than doubled in real terms, and its share of world manufacturing rose from less than 9 percent to 14½ percent (Figure 11).17 This nearly 6 percentage point increase in China's share of global manufacuring came mainly at the expense of advanced economies. The U.S. share of world manufacturing fell 2½ percentage points (from 23 percent to 20½ percent), Japan's fell 1¼, percentage points, Italy's and the United Kingdom's fell about 1 percentage point, and the shares of Germany, Canada, and France each fell a little more than ½ percentage point. Besides China, the only countries to gain at least ½ percentage point were South Korea and India.

Figure 11: Change in Global Manufacturing Share, 2000-2007

Data for Figure 11 immediately follows

Source: United Nations Statistics Division.

Data for Figure 11

China Korea India Brazil France Germany UK Italy Japan USA
5.908436 0.541749 0.491593 -0.07989 -0.56033 -0.66761 -0.91035 -0.96687 -1.24782 -2.55347

The loss of manufacturing shares by the advanced economies was not because manufacturing output fell, but because it did not grow as fast as in emerging market economies. Nevertheless, increases in productivity, owing in part to heightened competition from China, led to declines in manufacturing employment in most advanced economies. The United States and the United Kingdom led the declines with manufacturing employment falling by more than 20 percent in the United States and 27 percent in the United Kingdom, despite the fact that manufacturing value added grew 14 percent in the United States and was flat in the United Kingdom. To gain a bettter understanding of the effect of the rise in Chinese exports on the manufacturing sectors of the advanced economies, we examine more closely the impact on manufacturing employment in the United States.

United States

The manufacturing sector has been shrinking as a share of the U.S. economy since the early 1970s, but through 2000 real manufacturing production continued to climb and employment in the manufacturing sector remained stable except for fluctuations correlated with the business cycle (Figure 12). However, between January 2000 and December 2007, manufacturing employment in the United States fell by 3.6 million, nearly 21 percent, to 13.7 million.

Figure 12: U.S. Manufacturing Employment

Data for Figure 12 immediately follows

Source: Bureau of Labor Statistics.

Data for Figure 12

Year-MonthMillions of Jobs
1970 - Jan18424
1970 - Feb18361
1970 - Mar18360
1970 - Apr18207
1970 - May18029
1970 - Jun17930
1970 - Jul17877
1970 - Aug17779
1970 - Sep17692
1970 - Oct17173
1970 - Nov17024
1970 - Dec17309
1971 - Jan17280
1971 - Feb17216
1971 - Mar17154
1971 - Apr17149
1971 - May17225
1971 - Jun17139
1971 - Jul17126
1971 - Aug17115
1971 - Sep17154
1971 - Oct17126
1971 - Nov17166
1971 - Dec17202
1972 - Jan17283
1972 - Feb17361
1972 - Mar17447
1972 - Apr17508
1972 - May17602
1972 - Jun17641
1972 - Jul17556
1972 - Aug17741
1972 - Sep17774
1972 - Oct17893
1972 - Nov18005
1972 - Dec18158
1973 - Jan18276
1973 - Feb18410
1973 - Mar18493
1973 - Apr18530
1973 - May18564
1973 - Jun18606
1973 - Jul18598
1973 - Aug18629
1973 - Sep18609
1973 - Oct18702
1973 - Nov18773
1973 - Dec18820
1974 - Jan18788
1974 - Feb18727
1974 - Mar18700
1974 - Apr18702
1974 - May18688
1974 - Jun18690
1974 - Jul18656
1974 - Aug18570
1974 - Sep18492
1974 - Oct18364
1974 - Nov18077
1974 - Dec17693
1975 - Jan17344
1975 - Feb17004
1975 - Mar16853
1975 - Apr16759
1975 - May16746
1975 - Jun16690
1975 - Jul16678
1975 - Aug16824
1975 - Sep16904
1975 - Oct16984
1975 - Nov17025
1975 - Dec17140
1976 - Jan17287
1976 - Feb17384
1976 - Mar17470
1976 - Apr17541
1976 - May17513
1976 - Jun17521
1976 - Jul17524
1976 - Aug17596
1976 - Sep17665
1976 - Oct17548
1976 - Nov17682
1976 - Dec17719
1977 - Jan17803
1977 - Feb17843
1977 - Mar17941
1977 - Apr18024
1977 - May18107
1977 - Jun18192
1977 - Jul18259
1977 - Aug18276
1977 - Sep18334
1977 - Oct18356
1977 - Nov18419
1977 - Dec18531
1978 - Jan18593
1978 - Feb18639
1978 - Mar18699
1978 - Apr18772
1978 - May18848
1978 - Jun18919
1978 - Jul18951
1978 - Aug19006
1978 - Sep19068
1978 - Oct19142
1978 - Nov19257
1978 - Dec19334
1979 - Jan19388
1979 - Feb19409
1979 - Mar19453
1979 - Apr19450
1979 - May19509
1979 - Jun19553
1979 - Jul19531
1979 - Aug19406
1979 - Sep19442
1979 - Oct19390
1979 - Nov19299
1979 - Dec19301
1980 - Jan19282
1980 - Feb19219
1980 - Mar19217
1980 - Apr18973
1980 - May18726
1980 - Jun18490
1980 - Jul18276
1980 - Aug18414
1980 - Sep18445
1980 - Oct18506
1980 - Nov18601
1980 - Dec18640
1981 - Jan18639
1981 - Feb18613
1981 - Mar18647
1981 - Apr18711
1981 - May18766
1981 - Jun18789
1981 - Jul18785
1981 - Aug18748
1981 - Sep18712
1981 - Oct18566
1981 - Nov18409
1981 - Dec18223
1982 - Jan18047
1982 - Feb17981
1982 - Mar17857
1982 - Apr17683
1982 - May17588
1982 - Jun17430
1982 - Jul17278
1982 - Aug17160
1982 - Sep17074
1982 - Oct16853
1982 - Nov16722
1982 - Dec16690
1983 - Jan16705
1983 - Feb16706
1983 - Mar16711
1983 - Apr16794
1983 - May16885
1983 - Jun16960
1983 - Jul17059
1983 - Aug17118
1983 - Sep17255
1983 - Oct17367
1983 - Nov17479
1983 - Dec17551
1984 - Jan17630
1984 - Feb17728
1984 - Mar17806
1984 - Apr17872
1984 - May17916
1984 - Jun17967
1984 - Jul18013
1984 - Aug18034
1984 - Sep18019
1984 - Oct18024
1984 - Nov18016
1984 - Dec18023
1985 - Jan18009
1985 - Feb17966
1985 - Mar17939
1985 - Apr17886
1985 - May17855
1985 - Jun17819
1985 - Jul17776
1985 - Aug17756
1985 - Sep17718
1985 - Oct17708
1985 - Nov17697
1985 - Dec17693
1986 - Jan17686
1986 - Feb17663
1986 - Mar17624
1986 - Apr17616
1986 - May17593
1986 - Jun17530
1986 - Jul17497
1986 - Aug17489
1986 - Sep17498
1986 - Oct17477
1986 - Nov17472
1986 - Dec17478
1987 - Jan17465
1987 - Feb17499
1987 - Mar17507
1987 - Apr17525
1987 - May17542
1987 - Jun17537
1987 - Jul17593
1987 - Aug17630
1987 - Sep17691
1987 - Oct17729
1987 - Nov17775
1987 - Dec17809
1988 - Jan17790
1988 - Feb17823
1988 - Mar17844
1988 - Apr17874
1988 - May17892
1988 - Jun17916
1988 - Jul17926
1988 - Aug17891
1988 - Sep17914
1988 - Oct17966
1988 - Nov18003
1988 - Dec18025
1989 - Jan18057
1989 - Feb18055
1989 - Mar18060
1989 - Apr18055
1989 - May18040
1989 - Jun18013
1989 - Jul17980
1989 - Aug17964
1989 - Sep17922
1989 - Oct17895
1989 - Nov17886
1989 - Dec17881
1990 - Jan17799
1990 - Feb17896
1990 - Mar17870
1990 - Apr17847
1990 - May17796
1990 - Jun17775
1990 - Jul17703
1990 - Aug17648
1990 - Sep17610
1990 - Oct17575
1990 - Nov17428
1990 - Dec17394
1991 - Jan17331
1991 - Feb17214
1991 - Mar17141
1991 - Apr17095
1991 - May17069
1991 - Jun17042
1991 - Jul17016
1991 - Aug17025
1991 - Sep17011
1991 - Oct16998
1991 - Nov16960
1991 - Dec16916
1992 - Jan16840
1992 - Feb16831
1992 - Mar16805
1992 - Apr16830
1992 - May16834
1992 - Jun16825
1992 - Jul16820
1992 - Aug16783
1992 - Sep16761
1992 - Oct16750
1992 - Nov16758
1992 - Dec16767
1993 - Jan16791
1993 - Feb16806
1993 - Mar16795
1993 - Apr16771
1993 - May16766
1993 - Jun16742
1993 - Jul16740
1993 - Aug16741
1993 - Sep16769
1993 - Oct16777
1993 - Nov16800
1993 - Dec16815
1994 - Jan16854
1994 - Feb16863
1994 - Mar16896
1994 - Apr16932
1994 - May16961
1994 - Jun17011
1994 - Jul17026
1994 - Aug17082
1994 - Sep17113
1994 - Oct17143
1994 - Nov17187
1994 - Dec17218
1995 - Jan17261
1995 - Feb17265
1995 - Mar17262
1995 - Apr17278
1995 - May17259
1995 - Jun17249
1995 - Jul17218
1995 - Aug17239
1995 - Sep17246
1995 - Oct17215
1995 - Nov17207
1995 - Dec17229
1996 - Jan17208
1996 - Feb17230
1996 - Mar17192
1996 - Apr17204
1996 - May17222
1996 - Jun17227
1996 - Jul17222
1996 - Aug17255
1996 - Sep17252
1996 - Oct17268
1996 - Nov17276
1996 - Dec17283
1997 - Jan17299
1997 - Feb17317
1997 - Mar17339
1997 - Apr17351
1997 - May17363
1997 - Jun17388
1997 - Jul17388
1997 - Aug17451
1997 - Sep17465
1997 - Oct17513
1997 - Nov17556
1997 - Dec17587
1998 - Jan17623
1998 - Feb17627
1998 - Mar17637
1998 - Apr17635
1998 - May17623
1998 - Jun17609
1998 - Jul17421
1998 - Aug17563
1998 - Sep17557
1998 - Oct17511
1998 - Nov17465
1998 - Dec17447
1999 - Jan17432
1999 - Feb17395
1999 - Mar17368
1999 - Apr17343
1999 - May17333
1999 - Jun17296
1999 - Jul17308
1999 - Aug17286
1999 - Sep17279
1999 - Oct17273
1999 - Nov17281
1999 - Dec17277
2000 - Jan17292
2000 - Feb17284
2000 - Mar17302
2000 - Apr17298
2000 - May17279
2000 - Jun17298
2000 - Jul17321
2000 - Aug17286
2000 - Sep17226
2000 - Oct17215
2000 - Nov17202
2000 - Dec17178
2001 - Jan17114
2001 - Feb17029
2001 - Mar16939
2001 - Apr16803
2001 - May16662
2001 - Jun16516
2001 - Jul16378
2001 - Aug16225
2001 - Sep16113
2001 - Oct15971
2001 - Nov15825
2001 - Dec15710
2002 - Jan15598
2002 - Feb15518
2002 - Mar15446
2002 - Apr15394
2002 - May15338
2002 - Jun15297
2002 - Jul15250
2002 - Aug15164
2002 - Sep15115
2002 - Oct15059
2002 - Nov14992
2002 - Dec14910
2003 - Jan14867
2003 - Feb14780
2003 - Mar14722
2003 - Apr14608
2003 - May14556
2003 - Jun14493
2003 - Jul14401
2003 - Aug14378
2003 - Sep14347
2003 - Oct14334
2003 - Nov14315
2003 - Dec14300
2004 - Jan14292
2004 - Feb14277
2004 - Mar14288
2004 - Apr14316
2004 - May14342
2004 - Jun14331
2004 - Jul14333
2004 - Aug14343
2004 - Sep14330
2004 - Oct14334
2004 - Nov14305
2004 - Dec14285
2005 - Jan14261
2005 - Feb14274
2005 - Mar14271
2005 - Apr14253
2005 - May14258
2005 - Jun14227
2005 - Jul14223
2005 - Aug14201
2005 - Sep14174
2005 - Oct14191
2005 - Nov14181
2005 - Dec14189
2006 - Jan14210
2006 - Feb14215
2006 - Mar14220
2006 - Apr14232
2006 - May14209
2006 - Jun14217
2006 - Jul14190
2006 - Aug14159
2006 - Sep14125
2006 - Oct14070
2006 - Nov14030
2006 - Dec14001
2007 - Jan14007
2007 - Feb13998
2007 - Mar13973
2007 - Apr13949
2007 - May13939
2007 - Jun13916
2007 - Jul13898
2007 - Aug13833
2007 - Sep13788
2007 - Oct13753
2007 - Nov13745
2007 - Dec13726
2008 - Jan13721
2008 - Feb13691
2008 - Mar13654
2008 - Apr13604
2008 - May13577
2008 - Jun13528
2008 - Jul13456
2008 - Aug13363
2008 - Sep13270
2008 - Oct13129
2008 - Nov12999
2008 - Dec12822
2009 - Jan12543
2009 - Feb12377
2009 - Mar12212
2009 - Apr12063
2009 - May11911
2009 - Jun11782
2009 - Jul11739
2009 - Aug11682
2009 - Sep11634
2009 - Oct11577
2009 - Nov11552
2009 - Dec11534
2010 - Jan11556
2010 - Feb11572
2010 - Mar11591
2010 - Apr11629
2010 - May11668
2010 - Jun11681
2010 - Jul11717

No doubt some of these losses were attributable to the 2001 recession. But as shown in Figure 13, overall manufacturing job losses during the 2001 recession were not unusual by historical U.S. standards. The panel plots manufacturing employment in the months centered around recession troughs (as designated by the NBER), indicated by the vertical line. The blue line corresponds to the average experience over the post-war recessions prior to the 2001 recession, which is plotted as the red line. The pattern of manufacturing employment was very typical prior to the recession and through its trough, but atypical from that point onward. In the past, manufacturing employment began to rebound within 6 months of the trough, but following the 2001 recession, manufacturing employment contined to fall for 2 years and then only flattened out rather than recover.

Figure 13: U.S. Recessions and Manufacturing Employment

Data for Figure 13 immediately follows

Source: Bureau of Labor Statistics, Bureau of Economic Analysis.

Data for Figure 13

Months around recession trough (0)Median Manufacturing Employment Pre-2000 RecessionsManufacturing Employment 2000 Recession
-36104.8731247110.3633
-35105.0781154110.2496
-34105.2061788110.1548
-33105.3630037109.921
-32105.3531538109.7504
-31105.8858381109.5924
-30106.1570283109.5292
-29106.6216883109.2954
-28107.0626434109.3712
-27107.4842612109.2322
-26107.8284732109.188
-25108.2872732109.1501
-24108.7186837109.2006
-23108.8515041109.1754
-22108.7162014109.2701
-21107.6567997109.2196
-20109.2861845109.3333
-19109.1710309109.3081
-18109.5259568109.188
-17109.6062985109.3081
-16109.7534934109.4534
-15110.1654312109.2322
-14109.8301996108.8531
-13109.795853108.7836
-12109.6088317108.7014
-11109.0593564108.5498
-10108.7164902108.1453
-9107.7123961107.6082
-8107.317605107.0395
-7106.2359551106.1801
-6105.3416466105.2891
-5104.6968578104.3665
-4104.1677672103.4945
-3103.2165244102.5276
-2102.2836285101.8199
-1100.829317100.9226
0100100
199.8694095999.2733
2100.030244498.56556
3100.394696598.06003
4100.029152597.60506
5100.599370397.27646
6101.077724996.92259
7101.633611596.66351
8102.022646996.36651
9102.374154295.82306
10102.934264895.51343
11103.335403495.15956
12103.891206294.73618
13104.236977294.21801
14104.579440193.94629
15104.759323393.39652
16104.716357493.03002
17104.99341392.30964
18104.827822191.98104
19105.327839791.58294
20105.626341391.00158
21105.624571390.85624
22105.673437990.66035
23106.059240890.5782
24106.296564390.45814
25106.80475190.36335
26106.986137390.3128
27106.845943390.21801
28106.607397590.28752
29106.696018690.46445
30106.533204890.62875
31106.562466190.55924
32106.477301690.57188
33106.510995590.63507
34106.483398390.55292
35106.413393990.5782
36106.374378290.39494

To better quantify the different forces affecting U.S. employment, we regress the log difference of quarterly manufacturing employment on a constant, and the log differences of U.S. GDP, foreign GDP (weighted by U.S. exports), and investment in equipment and software (E&S). The right-hand side variables were estimated using third degree polynomials of four lags. We also include a first-order autoregressive error term, AR(1). The results are reported in Table 1, with the statistics for the sum of the lags reported for the economic variables. We estimated the equation through 1999:Q4 in order to allow for out-of-sample forecasts from 2000 forward, our period of interest.

Table 1. Dependent Variable: Log-difference Manufacturing Employment

VariableCoefficientStd. Errort-Statistic
C -0.015 0.002 -7.747
US GDP 0.604 0.277 2.182
For GDP 0.948 0.256 3.712
E&S Inv. 0.112 0.074 1.517
AR(1) 0.319 0.096 3.319

R2 =0.816

Adj. R2 =0.791

*Independent variables are log differences.

The coefficients on the GDP variables are of the expected sign and statistically significant at the 5 percent confidence level. The coefficient on investment falls just short of signficance at the 10 percent level. The inclusion of the autoregressive term has little impact on the coefficients and significance of the other variables as its relatively low value would suggest. In Figure 14, we show dynamic forecasts of the model through the 1980 and 1990 recessions. The model does a very good job of capturing both the contours and levels of employment, despite the very different characteristics of the two recessions.

Figure 14: Model Fit of the 1980 and 1990 Recessions

Data for Figure 14 immediately follows

Data for Figure 14

Year/QuarterManufacturing Employment Fundamentals ModelFundamentals - 1  S.E.Fundamentals + 1  S.E.
1980Q119.261448119.23919.1360077219.34199228
1980Q218.7694341718.7318.5880674818.87193252
1980Q318.4798098118.37818.2068511918.54914881
1980Q418.4733523318.58218.3844434418.77955656
1981Q118.6144775418.63318.410437718.8555623
1981Q218.5867304218.75518.5115586318.99844137
1981Q318.5402957618.74818.4857101819.01028982
1981Q418.2767405418.39918.1225863518.67541365
1982Q117.8481947217.96217.6756934418.24830656
1982Q217.4326392617.56717.2722329817.86176702
1982Q317.0095833417.17116.8693483117.47265169
1982Q416.6705581616.75516.4462147717.06378523
1983Q116.5342340916.70716.3882343617.02576564
1983Q216.6638164616.8816.5466087217.21339128
1983Q316.9027728117.14416.7939586617.49404134
1983Q417.2360489317.46617.0973506617.83464934
1984Q117.4792828217.72117.335642518.1063575
1984Q217.6330336117.91817.5179824618.31801754
1984Q317.7086685918.02217.6092581918.43474181
1984Q417.7292132218.02117.5970445618.44495544
1985Q117.7162193717.97117.5368933418.40510666
1985Q217.6655585217.85317.4099482418.29605176
1985Q317.6217131917.7517.2981151318.20188487
1985Q417.6428185617.69917.2368432118.16115679
1986Q117.5753776117.65817.1881162618.12788374
1986Q217.5444801417.5817.1016531418.05834686
1986Q317.4957575417.49517.0088946717.98110533
1986Q417.3927216717.47616.9838898617.96811014
1987Q117.2390324817.4916.9936047317.98639527
1987Q217.3182682617.53517.0277981418.04220186
1987Q317.4332983917.63817.1189894818.15701052
1987Q417.4870793117.77117.2420580418.29994196
1990Q117.8315342417.85517.7596535617.95034644
1990Q217.6661723117.80617.6724102317.93958977
1990Q317.5568748917.65417.4913988617.81660114
1990Q417.3652235317.46617.2802939217.65170608
1991Q117.1099846217.22917.0244260317.43357397
1991Q216.9657029517.06916.8467901917.29120981
1991Q316.8603113517.01716.7784769517.25552305
1991Q416.8045354216.95816.7038516717.21214833
1992Q116.7531348616.82516.5562595217.09374048
1992Q216.8011317516.8316.5459110917.11408891
1992Q316.8425567616.78816.4893103917.08668961
1992Q416.8800535316.75816.4453343317.07066567
1993Q116.820334316.79716.4727185917.12128141
1993Q216.777277816.7616.4243387117.09566129
1993Q316.7157413816.7516.4038319217.09616808
1993Q416.7479225516.79716.4387908417.15520916
1994Q116.8286958916.87116.4999856817.24201432
1994Q216.8759826616.96816.5851566817.35084332
1994Q316.9175803117.07416.6796963517.46830365
1994Q417.0389110317.18316.7755516317.59044837
1995Q117.1112638517.26316.8437167817.68228322
1995Q217.0039766817.26216.835540717.6884593
1995Q316.9004624417.23416.8006106317.66738937
1995Q416.8321182717.21716.7760796517.65792035
1996Q116.8251698917.2116.7601733217.65982668
1996Q216.8950688217.21816.7573592117.67864079
1996Q316.9818738817.24316.7711725117.71482749
1996Q417.0458669217.27616.7937037817.75829622
1997Q117.0996261817.31816.8256189117.81038109
1997Q217.2121317917.36716.8629065717.87109343
1997Q317.3693955217.43516.9178919517.95210805
1997Q417.3988130417.55217.0257278818.07827212

Figure 15 shows our out-of-sample forecast for manufacturing employment from 2000 through the fourth quarter of 2007. The model matches the general shape of employment through this period but significantly underestimates the job losses from the trough of the recession in the fourth quarter of 2001 through the end of 2003.18 It is important to note that the model shows that even after the recession's trough, U.S. GDP and investment growth were not sufficiently robust to prevent a further decline in employment over the next two years or a substantive rebound thereafter. We could not find a domestic macro variable capable of explaining the gap between the model fit and the data.

Figure 15: Out-of-Sample Forecast of 2001 Recession

Data for Figure 15 immediately follows

Data for Figure 15

Year/Quarter Manufacturing Employment  Fundamentals Model Fundamentals - 1 S.E. Fundamentals + 1 S.E.
2000Q1 17.293 17.39649539 17.30347514 17.48951565
2000Q2 17.292 17.55943281 17.42665019 17.69221543
2000Q3 17.278 17.55885735 17.39623784 17.72147685
2000Q4 17.198 17.46566694 17.27888671 17.65244718
2001Q1 17.027 17.28473321 17.07806988 17.49139655
2001Q2 16.66 16.98127494 16.75886117 17.20368871
2001Q3 16.239 16.63635828 16.40100349 16.87171306
2001Q4 15.835 16.39036985 16.14248528 16.63825442
2002Q1 15.521 16.26188358 16.00102336 16.5227438
2002Q2 15.343 16.16594524 15.89259664 16.43929384
2002Q3 15.176 16.08446335 15.79921794 16.36970876
2002Q4 14.987 15.91455781 15.61977581 16.2093398
2003Q1 14.79 15.75338625 15.44967467 16.05709783
2003Q2 14.552 15.67989532 15.36618924 15.99360139
2003Q3 14.375 15.74661045 15.4205122 16.07270869
2003Q4 14.316 15.82013696 15.48177154 16.15850239
2004Q1 14.286 15.77743656 15.42959886 16.12527426
2004Q2 14.33 15.76867906 15.41095569 16.12640243
2004Q3 14.335 15.8002426 15.43198111 16.16850409
2004Q4 14.308 15.86620098 15.48679542 16.24560655
2005Q1 14.269 15.85426767 15.46578506 16.24275028
2005Q2 14.246 15.81439366 15.41776907 16.21101825
2005Q3 14.199 15.83084617 15.42488566 16.23680669
2005Q4 14.187 15.81767846 15.40333154 16.23202538
2006Q1 14.215 15.9101301 15.48476732 16.33549288
2006Q2 14.219 15.89712703 15.46369493 16.33055913
2006Q3 14.158 15.80684565 15.36766535 16.24602594
2006Q4 14.034 15.76837759 15.32222669 16.2145285
2007Q1 13.993 15.73291253 15.27988574 16.18593931
2007Q2 13.935 15.72428551 15.26376678 16.18480424
2007Q3 13.84 15.68184387 15.21497628 16.14871145
2007Q4 13.741 15.66495358 15.19112653 16.13878063

Chinese exports, competing with U.S.-produced goods in the domestic and foreign markets, likely account for part of this unusual pattern. U.S. job losses were concentrated in those sectors where Chinese exports grew most rapidly--more than half of the losses were in the apparel and textiles, furniture, metals, and computer industries (Figure 16).

Figure 16: U.S. Employment Losses by Industry, 2000-2007

Data for Figure 16 immediately follows

Source: Bureau of Labor Statistics.

Data for Figure 16

Textiles and ApparelComputersMetals and MineralsTransportationPrint & PaperPlas. & Chem.FurnitureMachineryElectricalMisc.Food & Bev.
633 524.5 419.7 410.5 348.9 335 291.8 274 163 85.9 79.2

An Examination Industry-by-Industry

To quantify the effect of Chinese competition on U.S. manufacturing employment, we regress the log change in employment on the log change in Chinese import penetration, the log change in import penetration excluding China, and the log change in U.S. industrial production. Import penetration is defined as the ratio of imports to domestic demand, and demand is calculated as industrial production minus the change in inventories and net exports. The first two columns of the table quantifies the estimated impact of import pentration on jobs. The number takes the coeficient on the relevant variable and multiplies it by the change in that variable between 2001 and 2007. The next four columns give the results of the regressions for the import pentration variables. Numbers highlighted in bold are statistically significant. The first line of the table shows the results of a panel regression with industry fixed effects (at the 3-digit NAICS level). A one percent increase in Chinese import penetration reduces employment across these manufacturing sectors by almost 3 percent. In contrast, the same increase in imports from other countries boosts employment by over 8 percent. Both coeficients are highly statistically significant. Moving down the table, regressions are conducted one industry at a time. The estimated job losses are concentrated in three industries. Fabricated metal, machinery, and computers. These three industries have relatively large coefficients and were large employers in 2001.

The panel regression coefficient implies that domestic competition from Chinese exports can account for approximately ¾ million of the 3½ million manufacturing jobs lost. This regression may underestimate the effect on U.S. employment of Chinese exports by not fully capturing the effect of Chinese competition with U.S. exports in foreign markets.

Table 2. Cummulative Impact of Import Penetration on Jobs*

 Chinese Jobs Impact World Jobs
Impact
Chinese
Imports
T-StatWorld Imports
ex China
T-Stat
Panel Estimate -788464-2.89-2.68.394.3
Industry Estimates: 311 Food 12-261.140.27.03-0.9
Industry Estimates: 312 Bev. and Tobacco 0-64.151.7-8.99-0.6
Industry Estimates: 313 Textiles 843.450.77.820.5
Industry Estimates: 314 Txtl. Mill Products -469-34.85-2.615.371.0
Industry Estimates: 315 Apparel -251-8.30-1.30.700.1
Industry Estimates: 316 Leather -134-50.31-1.719.900.7
Industry Estimates: 321 Wood 32610.851.419.547.6
Industry Estimates: 322 Paper -293-8.07-1.45.340.7
Industry Estimates: 323 Printed Matter -80-2.14-0.30.020.0
Industry Estimates: 324 Petroleum 000.220.2-2.08-0.4
Industry Estimates: 325 Chemicals -88-1.66-0.94.620.9
Industry Estimates: 326 Plastics 38-138.840.6-7.03-0.3
Industry Estimates: 327 NonMetal Mineral -1619-13.60-1.842.876.7
Industry Estimates: 331 Primary Metal -118-0.28-0.18.281.3
Industry Estimates: 332 Fabricated Metal -242155-29.04-3.541.983.9
Industry Estimates: 333 Machinery -12371-16.16-2.530.023.6
Industry Estimates: 334 Computers -14939-13.41-0.516.830.8
Industry Estimates: 335 Electrical Equip. -2647-11.25-0.630.061.5
Industry Estimates: 336 Transportation 24162.281.017.171.5
Industry Estimates: 337 Furniture -613-2.24-0.217.731.0
Industry Estimates: 339 Misc. -1819-9.23-2.99.832.0

* Percent change in employment to percent change in import penetration

We attempt to illustrate this competition in Figure 17. Each point of the scatter plot represents the value of U.S. exports in 2000 (horizontal axis) and 2007 (vertical axis). The points are drawn from the 100 largest 4-digit HS categories in 2007. We shaded a point red if that category was also one of the top 50 categories of Chinese export growth measured in dollars. The red dots primarily fall below the trend line (dashed-red). Moreover, the red-shaded points that fall below the 45 degree line, indicating that those categories fell in dollar value between 2000 and 2007, are all high-tech goods.

As these high-tech goods evolved rapidly in the 2000s, the United States did not make the capital investment necessary to maintain its competitive position because of the bursting of the tech bubble and ensuing recession. In 2002 and 2003, the stock of private fixed assets in the computer and electronics industry fell for the first time in the postwar era. In contrast, Chinese investment in high-tech was soaring. Chinese fixed asset investment in communications, computer, and other electronics industries doubled between 2004, the first year of available data, and 2007. Of course, it is difficult to determine whether U.S. investment was lacking because of Chinese competition or whether the lack of investment opened the door to said competition. But regardless of the cause, with U.S. manufacturing capacity stagnant, there was an opportunity for Chinese producers to gain market share, both of the U.S. import market and the global market.

Figure 17: Chinese Competition with U.S. Exports

Data for Figure 17 immediately follows

Source: U.S. International Trade Commission Dataweb.

Data for Figure 17

HS Code20002007
854254.0980841.58776
870830.9946134.57804
847130.9292924.30857
880225.325253.50371
847324.6487715.65774
870317.2336944.79258
988015.7068526.91253
841115.6075327.15228
880315.2391421.95535
851712.8910418.3539
847912.278596.314796
90189.10066617.25416
85257.7115694.626867
84317.35382815.87453
90307.1652296.257367
27107.10675226.73326
85366.8205828.091127
87046.60810113.70651
84076.513296.870782
30046.50008916.48099
85415.9108178.170474
85445.8748387.04294
12015.31270510.01622
71085.29262811.79211
85295.1607155.055722
10054.71394410.0999
84094.5915435.094469
39264.3855034.530331
71024.23837711.91136
85044.2339164.369901
84143.766835.323414
84213.7552586.094264
84813.5028236.281801
90273.421976.082227
10013.3875018.344748
90323.3434553.286943
90313.3254724.916007
84133.2700535.54887
39073.0504325.152346
39202.9438713.804439
94012.9391053.297619
90012.7987142.76355
39012.6983026.443451
47032.6958683.621009
97012.6799055.581577
39232.6713714.073722
85432.6705143.708941
73262.6392443.071137
84082.5951144.820822
87012.5925734.860597
90212.5216676.502778
31002.4854443.741922
38222.4752014.137121
40112.4134623.573826
98012.3526214.074527
84292.2637885.923723
84182.2554672.755958
30022.2154888.752614
29332.1921014.169998
76062.1577623.210091
49012.148142.634647
94032.0921482.556966
85382.0453932.918966
27012.0444774.161278
85182.0166512.375267
90132.0132332.101809
48041.9512952.77302
52011.914894.580338
29021.8995064.826869
84191.8640422.993625
84831.7876033.092921
2071.7840643.259545
21061.6843612.974067
90221.6716153.416548
85231.5880855.969436
84241.5528542.574898
71101.5283833.900049
85011.5163612.442179
84331.5120942.697578
38241.5028973.186087
85281.4746213.643175
85371.4605212.907322
40021.3714692.981997
71121.3602254.987
39021.3595033.931721
71131.3335824.470029
27111.2996784.61399
33041.2764072.666729
90261.2599522.534374
2031.2519272.487563
84431.2456416.120288
47071.242992.775561
87161.1698792.955991
95041.1697624.20705
28441.1504342.38605
72041.0057636.964051
8020.9987782.775726
85020.7810872.533465
74040.5448122.927775
76020.496832.994806
848609.180212

The welfare implications for the United States of the increase in Chinese exports is beyond the scope of this paper, but it is important to note that it is not at all clear that the costs outweighed the benefits. Although the loss of manufacturing jobs is obviously costly to those individuals, the benefits to society included lower prices, more rapid adoption of new technology, and efficiency gains from the removal of trade barriers and through increased competition. Kamin et al (2006) estimated that for every percentage point increase in China's import share, U.S. import price inflation was reduced by about 1 percentage point. According to this estimate, the increase in China's import share between 2000 and 2007 reduced the 2007 level of import prices by about 8 percent. The United States also appears to have efficiently reallocated its labor resources, with economy-wide job growth sufficient to push the unemployment rate in 2007 down to well below its average of the 1990s and within ½ percentage point of its all-time low in early 2000, even as the manufacturing sector shed workers.


Section 4:  Implications for the Future

In the aftermath of the global financial crisis, U.S. investment has been lackluster. Like in the early part of the 2000s, the stock of high-tech fixed capital fell in 2009 and 2010.19 This may once again open the door for new technologies to be dominated by foreign producers. The IMF and others project China's trade surplus will again surge. For example, in its 2011 Article IV for China, the IMF projected that Chinese exports will more than double from $1.6 trillion in 2010 to $4.2 trillion in 2016. But which sectors will generate this growth and to whom will these goods go?

For apparel, furniture, and steel, it is unlikely that China will be able to repeat the massive surge in exports experienced earlier this decade. China now has a high market share in these categories and the categories themselves are unlikely to experience tremendous global growth.For high-tech products, China may be able to repeat its success, but it will be difficult. For example, China already produces most of the world's laptops. For some of the products discussed in this paper, there may be some room for China to increase market share further, but China's ability to gain additional market share in these now more mature markets, with more established players, is far from certain. Although China retains the draw of its huge domestic market, recent increases in wages have likely eroded, at least to a degree, its cost advantage relative to its neighboring emerging market economies. When the next product comes along that is adopted quickly and globally, such as the cell phone, it is not clear that China will be the country best positioned to dominate the market.

A final point is that, no matter the sector, there must be an importer on the other side of China's exports. If Chinese exports increase by $2.6 trillion, then the rest of the world's imports must also increase by that sum. This notion often gets swept under the carpet. For example, the IMF's April 2011 WEO projects the current account surplus of the emerging and developing economies will increase by more than $520 billion between 2010 and 2016 but that the deficit in the advanced economies will increase by only $200 billion--the majority of the burgeoning surplus in the emerging markets is added to the statistical discrepancy.


Section 5:  Conclusion

This paper examines the underlying causes of China's rapid growth of exports over the past decade. An undervalued exchange rate likely contributed to a degree, but export growth was relatively concentrated in select industries when one would expect the exchange rate to have a broader impact. Labor-intensive industries such as apparel, textiles, and furniture, benefited from China's WTO ascension. The apparel and textile industries saw massive gains following the expiration of multilateral agreements which had limited China's exports. Capital and energy-intensive industries, particularly iron and steel, benefited from government subsidies. These industries also benefited from Chinese reforms that led to state-owned enterprises becoming increasingly profitable. These profits, retained by the SOEs, resulted in an expansion of capacity and production beyond the domestic market's ability to absorb it.

These explanations, however, can take one only so far, as nearly half of Chinese export growth occurred in the "machinery" categories. It is only by examining more detailed Chinese trade data that one can see that this growth was heavily concentrated in a few specific high-tech products--cell phones, laptops, liquid crystal displays, and integrated electronic circuits. China was able to rapidly increase its exports of these products because of: industrial policy, such as science parks, that specifically encouraged these types of exports; an explosion in global demand for these products, with Chinese domestic demand leading the way; and, finally, a sharp fall in U.S. high-tech fixed investment, which contributed to China's ability to dominate these new technologies.

Further, this expansion of Chinese exports led to a decline in manufacturing share of the advanced economies, and in many cases a decline in manufacturing employment. We show that Chinese exports can help account for a significant portion of U.S. job losses at the industry level, but that macroeconomic fundamentals explain the majority of the fall in U.S. manufacturing employment between 2000 and 2007. Also, as we have noted earlier, quite a number of the jobs lost over this period, particularly in the textile, apparel, and furniture industries, would have moved to China earlier, owing to its natural comparative advantage in labor-intensive production, if not for protective tariffs and trade agreements. Chinese industrial policy aided the country's exports of high-tech goods, but China was also able to gain a competitive position in key products because it invested heavily at a time of burgeoning global demand for these products and diminished U.S. investment.


References

Ahmed, Shaghil (2009), "Are Chinese Exports Sensitive to Changes in the Exchange Rate?" International Finance Discusion Papers, 987.

Amiti, Mary and Caroline Freund (2010), "The Anatomy of China's Export Growth," in China's Growing Role in World Trade, editors Robert C. Feenstra and Shang-Jin Wei, University of Chicago Press.

Brambilla, Irene, Amit K. Khandelwal, and Peter K. Schott (2010), "China's Experience under the Multi-Fiber Arrangement (MFA) and the Agreement on Textiles and Clothing (ATC)," in China's Growing Role in World Trade, editors Robert C. Feenstra and Shang-Jin Wei, University of Chicago Press.

Cheung, Yin-Wong, Menzie Chin, and Eiji Fujii (2010), "China's Current Account and Exchange Rate," in China's Growing Role in World Trade, editors Robert C. Feenstra and Shang-Jin Wei, University of Chicago Press.

Congressional Budget Office, "How Changes in the Value of the Chinese Currency Affect U.S. Imports," Congressional Budget Office, July 2008.

Hsieh, Chang-Tai and Peter J. Klenow (2009), "Misallocation and Manufacturing TFP in China and India," The Quarterly Journal of Economics, Vol. 124, Iss. 4.

Kamin, Steven B., Mario Marazzi, and John W. Schindler (2006). "The Impact of Chinese Exports on Global Import Prices," Review of International Economics, Vol 14, No. 2, pp. 179-201.

Marquez, Jaime and John Schindler (2007), "Exchange-rate Effects on China's Trade," Review of International Economics, Vol. 15, Iss. 5.

Rodrik, Dani (2006), "What's So Special about China's Exports?" China & World Economy, Vol. 14, No. 5.

Sutherland, Dylan (2005). "China's Science Parks: Prodcution Bases or a Tool for Instituional Reform?" Asia Pacific Business Review, Vol 11, No. 1, pp 83-104.

Thorbecke, Willem and Gordon Smith (2010), "How Would an Appreciation of the Renminbi and Other East Asian Currencies Affect China's Exports," Review of International Economics, Vol. 18, Iss. 1.



Footnotes

**   The authors are economists in the International Finance Division of the Federal Reserve Board. The views in this paper are solely the responsibility of the authors and should not be interpreted as reflecting the views of the Board of Governors of the Federal Reserve System or of any other person associated with the Federal Reserve System. Return to text

1.   Exports and GDP measured in U.S. dollars at market exchange rates. Return to text

2.   This section includes HS categories 42, 51, 52, 58, 60, 61, 62, 63, 64, 94 Return to text

3.   The metals section includes HS categories 28, 72, 73, 74, 76, 79, 83. Return to text

4.   Based on data from the World Steel Association (www.worldsteel.org), Steel Statistical Yearbook 2009. Return to text

5.   http://www.eia.doe.gov/emeu/mecs/mecs2006/2006tables.html (Tables 1.1 and 6.3) Return to text

6.   The machinery section includes HS categories 84 and 85. Return to text

7.   In this section we use data from the China Customs Statistical Yearbooks for 2002 and 2007. Return to text

8.   Laptop sales data from Computer Industry Almanac (www.c-i-a.com/pr0707) and Gartner (http://www.gartner.com/it/products/newsroom/index.jsp)  Return to text

9.   International Data Corporation (IDC). http://www.idc.com/getdoc.jsp?sessionId=&containerId=prUS22383910  Return to text

10.   http://english.peoopledaily.com.cn/200401/08/print20040108_132140.html  Return to text

11.   Computer Industry Almanac. www.c-i-a.com/pr0608.htm. www.c-i-a.com/pr0701.htm. Return to text

12.   International Telecommunications Union. http://www.itu.int/en/pages/default.aspx Return to text

13.   "China's Burgeoning Mobile Phone Industry," China Daily, September 2003. http://www.chinatoday.com.cn/English/e2003/e20039/9p12.htm  Return to text

14.   Includes HS 8517 for all years and 852520 for 2000-2006. Return to text

15.   Display Search. http://www.displaysearch.com/cps/rde/xchg/displaysearch/hs.xsl/index.asp  Return to text

16.   http://spectrum.ieee.org/semiconductors/design/the-new-economics-of-semiconductor-manufacturing  Return to text

17.   U.N. Statistics. Value-added in U.S. dollars by economic activity at constant 2005 prices. Return to text

18.   The same exercise done in-sample produces essentially the same result. Return to text

19.   Source U.S. Department of Commerce, Bureau of Economic Analysis. Current-cost net stock of private fixed assets in the computer and electronic products industry. Return to text


This version is optimized for use by screen readers. Descriptions for all mathematical expressions are provided in LaTex format. A printable pdf version is available. Return to text

Home | Economic research and data | Publications and education resources
Accessibility | Contact us
Last update: May 4, 2012