Facing
the Dragon: Prospects for
Andrew B. Bernard, J. Bradford Jensen, and Peter K. Schott*
May, 2004
Overview: Will everything
be “Made in
The large and growing trade deficit between the United
States and China, the significant wage differential between Chinese and American
manufacturing workers, and the loss of 2.7 million jobs in the U.S.
manufacturing sector over the past three years have led to the fear in some
quarters that the U.S. manufacturing sector is (increasingly) being relocated
to China. Pundits proclaim the imminent
end of
The purpose of this paper is to go beyond this uninformed hyperbole
and assess the long-run prospects for
We have three goals.
The first is to demonstrate that not all
To summarize our key findings:
·
Import penetration from low-wage countries
is accelerating: The share of
·
The industries most at risk are low-skill,
low-wage and employ relatively few workers: Imports from low-wage
countries have been, and will continue to be, concentrated in low-wage,
low-skill, labor-intensive sectors like Apparel and Footwear. It is important to note that these sectors
employ relatively few workers compared to industries where the
·
The industries least at risk are
high-skill, high-wage: Industries consistent with
·
Reallocation, reallocation,
reallocation: Industries with relatively little competition
from low-wage countries saw employment increase
an average of 2.3% per decade over the past thirty years. By contrast, industries facing the highest
levels of low-wage country competition experienced employment declines
averaging 12% per decade. The net result of
these trends is a reallocation of
The Past: Which
industries did low-wage countries enter over the last twenty years?
Even as total imports have increased faster than GDP,
imports originating in low-wage countries have grown more rapidly than overall
imports. As illustrated in Figure 1, the
share of
Figure 1:
Notes: Figure displays actual and forecast share of
the value of
There is little chance that this re-orientation of
The surge in imports from low-wage countries starting in the
1980s is driven largely by
Figure 2: A
Breakdown of
Notes: Figure
displays share of
The aggregate increase in low-wage country import shares shown in Figure 1 has not been uniform across industries. Some sectors, such as Transportation and Chemicals, have faced relatively little low-wage country competition, while others, for example Apparel and Leather goods, have attracted much more. A breakdown of manufacturing exposure, by industry at ten-year intervals beginning in 1972, is reported in Table 1.
The key message of Table 1 is that stark differences in industry exposure to low-wage competition persist over time. In 1981, several years before the surge in aggregate exposure noted in Figure 1, three industries faced exposure greater than 10% while exposure was less than 2% in nine others. The high-to-low range of exposure was 21 percentage points. Twenty years later, in 2001, similarly sharp differences remained: seven sectors have low-wage import shares greater than 20%, while six others have exposure less than 10%. Moreover, the high-to-low range of exposure increased to 60 percentage points.
Table 1:
Competition by Industry, 1972 to 2001
Percent of Imports from Low-Wage Countries by
|
||||
|
Industry |
1972 |
1981 |
1991 |
2001 |
|
20 Food |
11 |
11 |
8 |
8 |
|
22 Textile |
25 |
21 |
19 |
22 |
|
23 Apparel |
3 |
15 |
30 |
41 |
|
24 Lumber & Wood
Products |
4 |
8 |
12 |
10 |
|
25 Furniture |
1 |
7 |
7 |
33 |
|
26 Paper |
0 |
0 |
1 |
7 |
|
27 Printing |
0 |
1 |
4 |
19 |
|
28 Chemicals |
2 |
7 |
3 |
4 |
|
29 Petroleum |
1 |
8 |
5 |
7 |
|
30 Plastic & Rubber
Products |
0 |
1 |
19 |
30 |
|
31 Leather Goods |
2 |
5 |
28 |
61 |
|
32 Stone & Concrete
Products |
1 |
2 |
7 |
22 |
|
33 Primary Metal |
1 |
4 |
3 |
6 |
|
34 Fabricated Metal |
1 |
2 |
6 |
17 |
|
35 Industrial Machinery |
0 |
1 |
1 |
12 |
|
36 Electronics |
0 |
2 |
7 |
18 |
|
37 Transportation Equipment |
0 |
0 |
0 |
1 |
|
38 Instruments and Controls |
0 |
1 |
3 |
9 |
|
39 Miscellaneous (E.g.
Toys, Jewelry) |
3 |
7 |
25 |
43 |
|
All Manufacturing |
2 |
4 |
7 |
15 |
|
Notes: Industry identifiers are preceded by their
two-digit Standard Industrial Classification (SIC) code. Each cell reports the percent of industry
imports originating in countries with less than 5% of |
||||
The growing exposure gap within manufacturing indicates that not every industry is heading toward the same fate. Between 1981 and 2001, for example, Leather Goods experienced the largest increase in exposure to low-wage competition, as the low-wage import share grew from 5% to 61%. Other industries with large rises include Apparel, Plastic & Rubber Products (e.g. gaskets, hoses, and pipes), and Miscellaneous Products (which includes toys). In contrast, Transportation, Chemicals, and Instruments and Controls experienced much more muted increases in exposure, such that it was still in the single digits by 2001. As we discuss in further detail below, a critical factor in determining the level of exposure to low-wage competition that an industry will face is the level of technological sophistication. High-wage, capital-using, and skill-intensive industries will continue to attract fewer imports from low-wage countries than low-wage, labor-intensive industries.
The Future: How will competition from low-wage countries
change from 2001 to 2011?
Knowing which industries are likely to experience the largest gains in low-wage country imports in the coming decade is of crucial importance for both managers and policymakers. As we discuss below, both employment and production growth are substantially lower in industries facing the highest levels of low-wage country competition. We are able to forecast changes in industry exposure to low-wage country imports over the coming decade using detailed data on product market penetration and industry characteristics. Our research reveals a lag between low-wage countries initially entering a product market and their subsequent increase in market share. By looking at the product markets that low-wage countries are entering in 2001, we can make a reliable forecast of where volumes will likely grow between 2001 and 2011.
Figure 3: Low-Wage Countries First Establish a
Beachhead and Later Gain Market Share
Notes: Product penetration is the number of products
imported from at least one low wage country divided by the total number of
products imported each year. Value
penetration is the total value of low-wage country imports divided by the total
value of imports.
Figure 3 captures the lag between initial product-market
penetration and subsequent market share increases for
The empirical model we develop suggests that current market share, current product penetration, and industry capital and skill intensity are useful predictors of future import share levels. Using this empirical model, we forecast future levels of low-wage imports.
Table 2: Forecasted Change in
Low-Wage Country Imports, 2001 to 2011
|
|
Percent of Imports from Low-Wage Countries |
|
|
||
|
Manufacturing
Industry |
2001 |
2011 |
Change |
2001 |
2001 |
|
31 Leather
Goods |
61 |
87 |
26 |
0.3 |
$10 |
|
23 Apparel |
41 |
67 |
25 |
3.2 |
$9 |
|
25 Furniture |
33 |
57 |
24 |
2.9 |
$12 |
|
39 Misc (E.g Toys, Jewelry) |
43 |
65 |
22 |
2.1 |
$12 |
|
32 Stone &
Concrete Products |
22 |
36 |
14 |
3.2 |
$15 |
|
34 Fabricated
Metal |
17 |
30 |
13 |
8.4 |
$14 |
|
27 Printing |
19 |
31 |
13 |
8.4 |
$15 |
|
30 Plastic
& Rubber Products |
30 |
42 |
12 |
5.4 |
$13 |
|
22 Textile |
22 |
32 |
10 |
2.7 |
$11 |
|
36 Electronics |
18 |
28 |
10 |
9.2 |
$14 |
|
24 Lumber
& Wood Products |
10 |
19 |
8 |
4.4 |
$12 |
|
26 Paper |
7 |
14 |
7 |
3.6 |
$16 |
|
35 Industrial
Machinery |
12 |
19 |
6 |
11.4 |
$16 |
|
38 Instruments
and Controls |
9 |
15 |
6 |
4.7 |
$15 |
|
37
Transportation Equipment |
1 |
4 |
3 |
9.9 |
$19 |
|
20 Food |
8 |
11 |
3 |
9.6 |
$13 |
|
28 Chemicals |
4 |
7 |
2 |
5.8 |
$18 |
|
33 Primary
Metal |
6 |
7 |
2 |
3.7 |
$16 |
|
29 Petroleum |
7 |
5 |
-2 |
0.7 |
$21 |
|
All
Manufacturing |
15 |
24 |
9 |
100.0 |
$14 |
|
Notes: Industry identifiers are preceded by their two-digit
Standard Industrial Classification (SIC) code. Rows are sorted by forecast change in
low-wage country import share between 2001 and 2011 (column 4). The employment share is the fraction of |
|||||
Our forecast predicts a 9 percentage point rise in the share of imports from low-wage countries between 2001 and 2011. This forecasted change is higher than that achieved by low-wage countries in any ten-year interval over the last 30 years (compare to the bottom row of Table 1). While this aggregate gain is large, it will be distributed unevenly across industries. Indeed, the industries most at risk from future low-wage country import competition employ relatively few workers and are both low-wage and labor-intensive. Table 2 reports our forecast, as well as other characteristics, by industry. In the table, industries are sorted according to their predicted change in low-wage country exposure between 2001 and 2011.
Four sectors – Leather Goods, Apparel, Furniture and
Miscellaneous – are forecast to experience increases in low-wage country import
shares of more than 20 percentage points by 2011. These industries pay below-average wages and
have a small share of
Figure 4
Notes: Size of marker proportional to industry
employment. Average hourly wage data by
two-digit SIC industry is from the U.S. Bureau of Labor Statistics
(www.bls.gov). Forecast low-wage country
import shares by two-digit SIC industry are from Table 2.
In Figure 4, we plot the forecasted low-wage import shares in 2011 against the average industry wage in 2001. The relationship is clearly negative: industries with high wages today are forecast to have relatively small increases in imports from low-wage countries. Conversely, the highest expected increases in low-wage country import share are found in industries with the lowest average wages. The sizes of the circular industry markers in Figure 4 are proportional to their manufacturing employment share in 2001. These markers reinforce the message that manufacturing industries with relatively large employment bases are expected to see the smallest increases in low-wage country imports.
Nine industries are forecast to experience only single-digit
increases in low-wage country import competition. These industries, which represent more than
half of
Real output growth exhibits a similar pattern. Industries with the lowest exposure increased their real output by 15% in each ten year period (roughly 1.5% per year). Output from industries experiencing the least competition from low-wage countries, however, grew more than twice as fast: 38.7% in each ten-year interval or roughly 3.9% per year.
The Consequences: What are the effects of low-wage country
competition?
Over the past 30 years, industries that have faced the highest levels of low-wage import competition have experienced an average net employment loss of over 12% per decade (see Table 3). In contrast, industries that faced the lowest levels of low-wage country imports actually saw their employment rise by 2.3% in each ten year interval even as employment in all of manufacturing was declining.
Table 3: Consequences of Low-Wage Country Competition
|
Initial Exposure to
Low-Wage Country Imports |
Average Decade-Long |
Average Decade-Long |
|
Low |
2.3% |
38.7% |
|
Middle |
-4.4% |
32.4% |
|
High |
-12.8% |
15.0% |
|
Notes: Industry exposure to
low-wage country imports is classified as low, middle or high according to
the level of exposure to low-wage imports at the beginning of each
decade. Industry classification as
well as average employment, real output and real export growth are across
two-digit SIC industries. Employment
data are from the U.S. Bureau of Labor Statistics (www.bls.gov) and output
data are from the NBER Productivity Database (www.nber.org). Decades are 1972
to 1982, 1982 to 1992 and 1992 to 2001.
Output is deflated by shipment price indexes from the NBER
dataset. Data on low-wage country
import exposure are unavailable after 2001 and data on real output are
unavailable after 1996. As a result,
averages for the final decade for each series are scaled up appropriately. |
||
These divergent trends are driven largely by comparative advantage. In 1981, Apparel, a relatively low-skill, low-wage industry, employed 1.2 million workers and low-wage countries accounted for 15% of total imports. By 2001, Apparel had shrunk to 566,000 employees while low-wage imports had surged to 41% of the total.
The Printing industry, relatively high-wage and high-skill, also employed 1.2 million workers in 1981. Printing, however, faced relatively little low-wage import competition in 1981, and its total employment actually grew to 1.5 million between 1981 and 2001.
Real output growth exhibits a similar pattern. Industries with the lowest exposure increased their real output by 15% in each ten year period (roughly 1.5% per year). Output from industries experiencing the least competition from low-wage countries, however, grew more than twice as fast: 38.7% in each ten-year interval or roughly 3.9% per year.
Similar outcomes are also evident across firms within industries. Plants in industries with high levels of low-wage import competition are more likely to close, have lower employment growth, and have lower output growth. Recent research in Bernard et al. (2003) indicates that a ten percentage point increase in the share of imports coming from low-wage countries increases the probability of a plant closing by 3.3 percentage points (or 13 percent). Further, plants that face higher import shares from low-wage countries experience lower employment growth and lower output growth. Each 10-percentage-point increase in import shares from low-wage countries is associated with a 1.3 percent lower annual employment growth rate for an individual plant and also lower annual output growth.
While plants facing increased low-wage country import shares are more likely to fail and have lower employment growth, certain plant characteristics appear to insulate plants from the effects of low-wage competition. Results in Bernard et al. (2003) show that, within an industry, capital usage by plants reduces the probability of closure and is especially beneficial when low-wage competition is high. As competition from low-wage imports increases, the benefits of higher capital usage increase for plant survival, job creation and output growth.
Even in exposed industries, firms can position themselves to avoid the harmful consequences of low-wage country competition. Some upgrade the products they produce within their industry while others switch to new lines of business. Bernard et al. (2003) find that plants facing higher levels of low-wage imports are more likely to switch industries than plants in industries facing lower levels of low-wage imports. In addition, the plants that switch are more likely to move to more capital-intensive and more skill-intensive industries and the higher the low-wage import share, the bigger the jump. The plant switching results confirm that plants that have higher capital intensity are less exposed to low-wage imports – capital-intensive plants within an industry are less likely to switch products than other plants in the industry.
Increasing low-wage country imports over the next decade will
provide a powerful force for the reallocation of resources in the manufacturing
sector. This reallocation will be driven
by differential outcomes across industries, within-industry growth of skill and
capital-intensive firms, and the changes in firm product mix towards high-end
goods. These forces will continue to
change the composition of
The Forgotten: What is the role for exports?
The lesson from the previous sections is that most
industries and most manufacturing workers are not threatened by the looming
increase in low-wage country imports.
However, while the focus of this brief is on exploring the future path
of imports from low-wage countries, the export side of international trade
cannot be completely overlooked.
International trade – including imports from low-wage countries – drives
change in the
Table 4 illustrates that
Table 4:
|
Initial
Exposure to Low-Wage Country Imports |
Average Decade-Long 59.2% 56.7% 52.4% |
|
Low |
|
|
Middle |
|
|
High |
|
|
Notes: Industry exposure to
low-wage country imports is determined as noted in Table 3. Export data are from Feenstra
et. al (2002) and are deflated by industry shipment price indexes available
from the NBER (www.nber.org). Decades are 1972 to 1982, 1982 to 1992 and 1992
to 2001. Data on real exports are unavailable after 1997. As a result, the average for the final
decade is scaled up appropriately. |
|
Exporters are different, and better, than non-exporting
plants. As Bernard and Jensen (1995, 1999) report, exporters pay 12% higher
wages, are 20% more capital intensive, and are 19% more productive. Across all industries, exporters have higher
annual employment and output growth, between 2% and 5% faster on average, and
these firms are less likely to fail. The
combination of superior productivity, higher wages, faster growth and higher
survival chances mean that increased opportunities to export have a range of
desirable outcomes. Both sides of the
trade equation, imports and exports, are working to realign
A Challenge to Public Policy
Competition from low-wage countries induces a long-run reallocation
of
Surveys demonstrate that fear of job loss and wage
deterioration are important factors in public sentiment against trade
liberalization.[5] Thus, our forecast presents an important
challenge to policy makers. The prospect
of accelerating change presents a natural temptation to try to protect
industries that face competitive pressures.
To give in to this temptation would inhibit one of the great strengths
of the
In summary:
·
· Sectors within manufacturing will have dramatically different experiences with low-wage imports over the next decade.
· Low-wage, labor-intensive industries will face the largest increases in low-wage import competition and are likely to experience the largest decreases in employment and the largest number of plant closings.
· Industries that are capital-intensive and skill-intensive will face smaller increases in low-wage import competition and may add jobs even as they greatly increase output.
· Inside industries facing competition from
low-wage countries, firms that are capital-intensive and skill-intensive will
have better prospects than their industry rivals.
· Both within and across industries,
References:
Bernard,
Andrew B. and J. Bradford Jensen. (1995). “Exporters, Jobs, and Wages in
Bernard, Andrew B. and J. Bradford Jensen. (1999). “Exceptional Exporter Performance: Cause, Effect, or Both?” Journal of International Economics, Vol. 47, pp. 1-25.
Bernard,
Andrew B., J. Bradford Jensen and Peter K. Schott. (2003). “Survival of the
Best Fit: Competition from Low-Wage Countries and the (Uneven) Growth of
Bernard,
Andrew B., J. Bradford Jensen and Peter K. Schott. (2004). “Forecasting the
Timing of Market Entry by Low-Wage Countries.”
Tuck
Feenstra, Robert, John Romalis and Peter K. Schott. (2002).
“
Kletzer, Lori. “Job Loss From Imports: Measuring the Costs.” Institute for International Economics, September 2001.
Kletzer, Lori and Robert E. Litan. “A Prescription to Relieve Worker Anxiety.” Institute for International Economics, February 2001.
Scheve, Kenneth F. and Matthew J. Slaughter. “Globalization and the Perceptions of American Workers.” Institute for International Economics, March 2001.
Table 5: Low Wage Countries
|
Country |
Average Percent of |
|
Average Percent of |
|
|
0.03 |
|
0.01 |
|
|
0.02 |
|
0.01 |
|
|
0.01 |
|
0.01 |
|
|
0.02 |
|
0.01 |
|
|
0.04 |
|
0.02 |
|
Burkina |
0.01 |
|
0.02 |
|
|
0.01 |
|
0.05 |
|
|
0.01 |
|
0.01 |
|
|
0.03 |
|
0.01 |
|
|
0.02 |
|
0.03 |
|
|
0.01 |
|
0.01 |
|
|
0.01 |
|
0.01 |
|
|
0.02 |
|
0.02 |
|
|
0.01 |
|
0.04 |
|
|
0.04 |
|
0.05 |
|
|
0.04 |
|
0.01 |
|
|
0.04 |
|
0.02 |
|
|
0.02 |
|
0.01 |
|
|
0.00 |
|
0.02 |
|
|
0.02 |
|
0.01 |
|
|
0.02 |
|
0.04 |
|
|
0.02 |
|
0.03 |
|
|
0.01 |
|
0.01 |
|
|
0.03 |
|
0.02 |
|
|
0.02 |
|
0.01 |
|
|
0.03 |
|
0.01 |
|
|
0.01 |
|
0.01 |
|
|
0.03 |
|
0.02 |
|
|
0.04 |
|
0.03 |