NAFTA: ESTIMATED U.S. JOB "GAINS" AND "LOSSES" BY STATE OVER 5-1/2 YEARS
CRS Report for Congress
NAFTA: Estimated U.S. Job “Gains” and
“Losses” by State Over 5½ Years
Updated February 2, 2000
Mary Jane Bolle
Specialist in International Trade
Foreign Affairs, Defense, and Trade Division
Congressional Research Service ˜ The Library of Congress
ABSTRACT
U.S. job "gains" and "losses" from trade with Mexico and Canada stem from changes in trade.
Job "gain" data cover the North American Free Trade Agreement (NAFTA)'s first five years
and job "loss" data cover a little more than NAFTA's first five and one-half years. Between
January 1, 1994 and September 28 1999, approximately 259,618 workers were certified as
potentially suffering NAFTA-related job "losses." These data are sorted by state. Earlier
versions of this report also included a table which sorted these data within states by business,
indicating for each business, its products, reason for certification, and industry. Because of
its increasingly unwieldy size, this table has been dropped from this report. However, these
data for any state are available from CRS by calling the author at 7-7753. In addition,
between January 1994 and the end of December 1998, nearly 710,000 net jobs were "created"
from new exports to Mexico and Canada. These figures are sorted by state. In all, six tables
present the data from varying perspectives, including major industries of job certifications and
major industries of increased exports to Mexico and Canada . This report is generally
updated once or twice a year.
NAFTA: Estimated U.S. Job “Gains” and “Losses”
by State Over 5½ Years
Summary
What has been the effect of the North American Free Trade Agreement
(NAFTA) on jobs in individual states? NAFTA — the trade agreement between the
United States, Mexico, and Canada — appears to have served primarily to accelerate
trade-related job trends that were already ongoing. Thus, any reference in this report
to effects “of NAFTA” is really a reference to effects “since NAFTA.” Since NAFTA
went into effect January 1, 1994, both imports from and exports to Mexico and
Canada have increased — boosted by reductions in tariff, nontariff, and investment
barriers, particularly in Mexico.
U.S. job “gains” and “losses” from trade with Mexico and Canada stem from
changes in trade. In 1998 the value of trade with Mexico was slightly more than half
the value of trade flows with Canada. Since NAFTA went into effect through the end
of 1998, exports to Mexico and Canada combined have increased 64%, while imports
have increased by 78%. Between 1993 and 1998, the annual merchandise trade
deficit with Canada increased from $13 billion to $24 billion, while a $1 billion surplus
with Mexico evolved into a $17 billion deficit. (The U.S. trade positions with both
Mexico and Canada worsened in 1998 over 1997.)
“Job gain” data cover NAFTA’s first five years, and “job loss” data cover
NAFTA’s first five-and-a-half years. Measuring “job gains” and “losses” from trade
with NAFTA partners is an inexact process. “Job loss” and “job gain” data were
derived from different methods and databases and, therefore, are not comparable. In
addition, job loss figures represent the outside limit of, rather than actual job losses.
That is, they represent total employment at plants where workers have been certified
to receive NAFTA-Transitional Adjustment Assistance (NAFTA-TAA) benefits.
Only 20-30% of these workers actually collect benefits. Trying to further apportion
these job gains and losses by state compounds the problems. Thus, the figures come
with strong caveats. Nevertheless, they provide some useful information.
Moreover, estimates of NAFTA-related job “gains” and “losses” are small
relative to total U.S. employment. Approximately 259,618 workers were certified
between January 1, 1994, and September 28, 1999, as potentially suffering NAFTA-
related job losses. This represents less than the number of jobs created in a single
month in 1998. On the other hand, an estimated 1,212,357 gross jobs and 709,989
net jobs were created from new exports to Mexico and Canada between 1994 and
1998. (The net figure factors in job losses from countervailing productivity growth
and inflation).
States suffering the greatest potential total job losses from trade with Mexico
and Canada since NAFTA went into effect are North Carolina (27,725), Texas
(23,386), Pennsylvania (18,663), New York (17,487), California (14,825), Georgia
(12,457), and Tennessee (12,191). States estimated to have experienced the greatest
gross job gains (not reduced by productivity and inflation growth) from trade with
Mexico and Canada since NAFTA went into effect are Texas (175,407), Michigan
(149,382), California (147,284), Illinois (60,181), New York (56,256), Ohio
(53,895), and Indiana (43,192).
Contents
Estimated Job “Losses” ..........................................2
Estimated Job “Gains”............................................5
Perspective on NAFTA-Related Job Effects............................8
Appendix:
Data and Explanation of Methodology...........................10
Algebraic Formula:
Relationship Between Gross and Net Jobs Created from New Exports since
NAFTA .................................................. 14
List of Figures
Figure 1. Quick NAFTA State Data Finder............................1
List of Tables
Table 1. NAFTA-TAA Certification by Reason, January 1, 1994-September 28,
1999 ..................................................... 3
Table 2. Potential Job Loss by State: Number of Cases and Workers Certified by the
NAFTA-TAA Program, January 1, 1994-September 28, 1999..........4
Table 3. Job Gain by State: Estimated Number of Gross and Net Jobs Supporting
New Merchandise Exports to Canada and Mexico Combined, 1993-98...7
Table 4. Major Industries of NAFTA-TAA Job Certification, Jan.1, 1994 -Sept. 28,
1999 ..................................................... 9
Table 5. Major Industries of Increased Exports to Mexico
and Canada, 1993-1998.......................................9
Table 6. Merchandise Exports to Canada, Mexico, and Combined,
1993-1998 ................................................ 11
Table 7. Gross and Net Jobs Supported by New and All Merchandise Exports to
Mexico and Canada Combined, by State, 1993-1998................12
Special thanks to Cathi Jones for assistance in compiling data for this report.
NAFTA: Estimated U.S. Job “Gains” and
“Losses” by State Over 5½ Years
What has been the effect of the North American Free Trade Agreement NAFTA
on jobs in individual states? NAFTA — the trade agreement between the United
States, Mexico, and Canada — appears to have served primarily to accelerate trade-
related job trends that were ongoing before NAFTA. Thus, any reference in this
report to effects “of NAFTA” is really a reference to effects “since NAFTA.” One
broad observation that can be made is that since NAFTA went into effect in January
boosted by reductions in tariff, nontariff, and investment barriers, particularly in
Mexico. 1
Figure 1. Quick NAFTA StateU.S. job “gains” and “losses”
Data* Finderfrom trade with Mexico and Canada
(the U.S.’s first and third largest
trading partners) stem from changes
Job Losses from New ImportsPagein trade flows, but are also affected
Potential Job Losses .............4by domestic increases in
productivity. Since NAFTA,
Gross and Net Job Gains from Newexports to and imports from Canada
Exportshave each increased by about 55-
Total, 1993-98 ..................760%. This has resulted in relatively
By Year......................12suggests few net job effects from
trade with Canada. Exports to and
Exportsimports from Mexico, however,
Total Exports to Mexico and have more or less doubled (exports
Canada.......................11less and imports more than
* State data may include data on the District ofdoubled). This suggests some
Columbia and U.S. territories.sectoral job “losses” from
production shifts to Mexico.
However, since about two-thirds of
the increase in imports is covered by an increase in exports, net job effects are2
estimated to be relatively small in relation to overall U.S. employment.
In fact, NAFTA’s overall effect on the U.S. economy has been relatively small.
Foreign trade itself accounts for about 14% of U.S. gross domestic product (GDP);
1For details on exports and imports traded with NAFTA partners, by industry, see NAFTA:
Estimates of Job Effects and Industry Trade Trends After Five and One-Half Years, by Mary
Jane Bolle. CRS Report 98-783E.
2See also, U.S. Library of Congress. Congressional Research Service. NAFTA, Mexican
Trade Policy, and U.S.-Mexico Trade: A Longer-Term Perspective, by J.F. Hornbeck. CRS
Report 97-811 E.
and trade with Mexico and Canada represents roughly one-fifth all U.S. trade. Larger
effects on the overall U.S. economy result from structural changes taking place as
companies shed employment due to trade and other economic factors. At the same
time, however, the U.S. economy is both dynamic and robust. As jobs are eliminated
in one industry, they are added in another. The United States, as a job-creating
nation, is the envy of many developed countries, particularly Canada and many
European nations, where employment growth has stagnated in recent years, and
where unemployment is in some cases one-and-a-half to two times the U.S. rate.
This report provides estimates on NAFTA-related job effects. Included, by state,
are estimates of job gains covering NAFTA’s first five years, and potential job-losses
covering roughly NAFTA’s first five-and-a-half years. (State export data, from
which job gains by state are derived, are available only annually.) Measuring job gains
and losses from trade with NAFTA partners is an inexact process. Trying to further
apportion these job gains and losses by state compounds the difficulties. Thus, these
figures come with strong caveats:
First, the terms “job gains” and “losses” are, to a certain extent, misnomers. In
an economy operating at full employment, trade results in neither net job gains nor
net job losses, only in relocations from less efficient to more efficient industries.
Economy-wide, job gains balance out job losses. However, at the industry and firm
level, job “gains” from trade will not likely equal job “losses” from trade.3
Second, it should be emphasized that job-effect estimates included in this report
were developed by different methods and are arguably incomplete and incompatible.
In particular, job “losses” captured by Department of Labor (DOL) certifications
arguably both underestimate and overestimate the actual number of job losses from
NAFTA, for reasons mentioned below. Job “gains,” while perhaps reasonably
reflective of actual job gains from exports at the national level, are not as accurate at
the state level, for reasons discussed below.
Estimated Job “Losses” 4
The DOL certifies the eligibility of workers to apply at the state level for benefits
under the NAFTA-Transitional Adjustment Assistance (NAFTA-TAA) Program. The
certification identifies potential dislocated workers who, if separated from their jobs,
would be eligible for training and income replacement benefits because, either
imports from Mexico and Canada “contributed importantly” to their job loss, or their
plant relocated to Mexico or Canada. Hence, the certified NAFTA job losses are but
a subset of total job losses from NAFTA; they include only those job losses for which
3Economists continue to debate what level the full-employment rate of unemployment should
be. Many believe unemployment rates below 5.5% suggest an excess demand for labor.
During the past several years, unemployment rates have dipped to under 4.5% — the lowest
they have been since the late 1960s.
4 The lengthy table included in previous versions of this report that listed each certified
business in each state, together with reasons for certification, has been omitted because of
size. However, these data for any state are available from CRS by calling the author at 7-
the displaced worker applied for benefits and a direct linkage to trade with Mexico
or Canada, or a shift in production to either of these countries, can be verified.
However, NAFTA-TAA certification figures may overestimate job losses among
certified workers. This is because not all workers certified actually lose their jobs.
In fact, recent data from the Department of Labor suggest that as few as 20- 30% of
the certified workers actually collect NAFTA-TAA benefits. (The others who were
precertified may either not have actually lost their jobs, may have found another job
in lieu of needing benefits, or for other reasons may not have collected their benefits.)
Table 1. NAFTA-TAA Certification by Reason,
January 1, 1994-September 28, 1999
% of all
certified
Reason for CertificationCasesWorkersworkers
C-1 Production shift to Mexico948120,88847
C-2 Production shift to Canada23323,0109
C-3 Increased customer imports from19622,8529
Mexico
C-4 Increased customer imports from 16814,9556
Canada
C-5 Increased customer imports not
identified by source country21427,49711
C-6 Increased company imports from19426,3529
Mexico
C-7 Increased company imports from668,8023
Canada
C-8 Increased company imports not193,4191
identified by source country
C-9 High and Rising aggregate imports from114111,8435
Mexico and/or Canada
TOTAL CERTIFIED2,179259,618100
Source of data: U.S. DOL Office of Trade Adjustment Assistance. Compiled by CRS.1
C-9 represents a new category in 1997.
Table 1 sorts certified workers by reason for certification. The growth in the
number of workers certified under the NAFTA-TAA program has leveled off in the
past year. The single most important reason for certification, and with it eligibility to
receive income and job training benefits under NAFTA-TAA, continues to be
production shifts to Mexico. This accounts for the largest and the largest growing
part of all worker certifications — 47%. Numbers of workers potentially dislocated
are sorted by state in table 2.5
5 The DOL Office of Trade Adjustment Assistance keeps several different lists of numbers
reflecting workers certified under the NAFTA-TAA program. Thus, totals in table 1 do not
(continued...)
Table 2. Potential Job Loss by State: Number of Cases and Workers
Certified by the NAFTA-TAA Program,
January 1, 1994-September 28, 1999
Total Total
Jan. 1, 1994-Jan.1, 1994-
Sept. 28, 1999 Sept. 28, 1999
NAFTA-TAA NAFTA-TAA
Certified Certified
STATE Cases Workers STATE Cases Workers
North Carolina17127,725Arizona302,054
Texas 252 23,386 Minnesota 20 1,921
Pennsylvania19318,663New Mexico121,771
New York12617,487Maine181,702
California 124 14,825 Kansas 13 1,364
Georgia11012,457West Virginia181,343
Tennessee 109 12,191 Connecticut 11 1,291
Indiana 59 9,406 Mississippi 4 1,144
Arkansas488,993Puerto Rico21,090
Michigan 74 8,334 Utah 13 1,047
Wisconsin 52 7,776 Montana 24 790
Washington 85 7,351 Alaska 5 780
New Jersey697,064Wyoming19620
Alabama406,627South Dakota5566
South Carolina466,551Iowa9454
Virginia 64 6,513 Vermont 4 429
Ohio536,074North Dakota4393
Missouri 67 5,984 Maryland 3 390
Florida 72 5,756 Oklahoma 12 331
Illinois 50 5,718 Nebraska 4 283
Oregon 90 4,907 Nevada 10 257
Louisiana184,688New Hampshire7224
Idaho 383,073Delaware00
Kentucky302,904Rhode Island00
Massachusetts 31 2,562 Hawaii 0 0
Colorado282,359Dist. of Columbia00
TOTAL 2,346 259,618
Source: U.S. Department of Labor, Office of Trade Adjustment Assistance. Database sorted
by CRS.
5 (...continued)
agree precisely with totals in table 2 because of errors in the data base. (Entries for which
certain data are missing are not picked up by various “sorts” of the data.)
While certification figures may overestimate some potential NAFTA job loses,
they may miss others. Other workers whose job losses may be related to NAFTA, but
who are not counted in the NAFTA-TAA figures, include the following major groups:
(1) primary job losers who for some reason either: (a) did not apply for NAFTA-TAA
benefits; or (b) applied and were rejected because they did not meet the criterion for
certification (e.g., trade with Mexico or Canada contributed “somewhat” rather than
“importantly” to their job loss); (2) some secondary job losers (who typically account
for more than half the total number of job losers) in supplier or distributor industries,
who for reasons similar to (1) above were not certified to receive NAFTA-TAA
benefits; and (3) other job losers whose job loss is less directly related to NAFTA.6
Some workers may be able to claim loss of jobs due to NAFTA even though
other factors may be involved. For example, some argue that labor-intensive jobs are
shifted out of the United States to Mexico because the U.S. job market is tight and
Mexico has a ready supply of workers willing to work for lower wages.
Estimated Job “Gains”
NAFTA-related job-gain estimates are based on Department of Commerce
(DOC) data. The DOC publishes data on the number of total jobs in the economy
supported by exports. This figure is derived through an input-output model which
incorporates output-per-worker ratios for each sub-industry. Thus, the model
estimates jobs added to the economy when output for any given sector increases.
DOC data on all jobs supported by total merchandise exports can be used to derive
the average number of jobs supported by $1 billion in merchandise exports each year.
The number of jobs supported by $1 billion in exports declines each year because of7
productivity gains and inflation.
Table 3 (p. 7) includes two measures of job gains from exports. One is a gross
number. It focuses only on jobs created by new exports each year. The other is a net
number. It takes into consideration jobs created by new exports and jobs lost by
productivity gains among workers producing for export both before and since
6U.S. Department of Commerce, Economics and Statistics Administration. U.S. Jobs
Supported by Goods and Services Exports, 1983-94, p. 27 suggests that approximately two
additional jobs support each manufacturing job by producing intermediate inputs, capital
goods, and transportation and other services to the goods going to market. The NAFTA-TAA
program covers some workers whose job loss is indirectly linked to trade with Mexico or
Canada — for example, workers in a business which supplies a company directly affected by
trade with Mexico or Canada. Others, however, whose supplier relationship is less direct,
may “slip between the cracks.”
7Data published by the Department of Commerce Economics and Statistics Administration
included in a November 1996 report: U.S. Jobs Supported by Goods and Services Exports,
1983-1994 and updated by a separate data release, show that the number of jobs supported
by $1 billion in merchandise exports was 15,123 in 1993,14,361 in 1994, 13,774 in 1995,
13,258 in 1996, and 12,755 in 1997. The average annual decline in the number of jobs
supported by a billion dollars worth of exports is about 4%; thus, the estimated figure for
NAFTA went into effect. Thus, the net figure measures the difference between all
workers needed to produce exports to Mexico and Canada in 1998 and all workers
needed to produce exports in 1993. Gross and net numbers would be identical if
productivity gains and inflation did not reduce the number of workers required to
produce a given dollar value of exports each year. (More detail on the mathematics
of this is included in the appendix, p. 14.)
Results of the gross and net measures of NAFTA-related job growth are shown
in table 3. Figures in table 3 are derived using data from two sources: 1) DOC state
export data (included in appendix table 6, p. 11); and 2) DOC averages of jobs
supported by exports (listed in footnote 6). From these numbers it can be estimated
that increased merchandise exports to Mexico and Canada combined during NAFTA’s
first five years (1994 through 1998) have created approximately 1,212,357 gross
export-related jobs in the United States (table 3, TOTAL). It can also be estimated
that when job losses from productivity gains among those producing for export before
and since NAFTA are additionally taken into consideration, new exports since
NAFTA have added to the total number of workers producing for export to Mexico
and Canada combined, only about 709,988 net jobs.
Estimated numbers of net jobs in each state which support total exports to
Mexico and Canada for 1993, and 1998 are included in columns (5) and (6) of table
For the country as a whole, the net jobs-from-new-NAFTA-exports figure
covering the period 1994-1998, is less than two-thirds the gross figure (59%).
However, net job gains may be a much higher percent of gross job gains in specific
states whose export levels started out small and increased since NAFTA went into
effect. See, for example, in table 3, Alaska, (right hand column) where the net figure
is 76% of the gross figure. For Alaska, exports, initially at a low level, grew by 172%
between 1993 and 1997 (as can be seen in appendix table 6, p. 11).
It should be emphasized that estimates of jobs supporting exports in each state
(in table 3) are only very rough estimates, and not an accurate reflection of actual jobs
supporting exports, for two reasons. First, state export data on which these figures
are based reflect total sales, not value added by each state. Second, the jobs-
supporting-each-billion-dollars-worth-of-exports figure includes jobs both directly
and indirectly involved in manufacturing the merchandise. Since more than half the
jobs are indirect — either “upstream” (primarily supplier) or “downstream” (primarily
distribution) — and indirect jobs can be carried out anywhere in the country, they are
not necessarily attached to the state which the export figures represent.
Table 3. Job Gain by State: Estimated Number of Gross and Net Jobs Supporting New
Merchandise Exports to Canada and Mexico Combined, 1993-98
Jobs added in export industries Jobs added in export industries
1993-1998 1993-1998
(3) (3)
NET jobs addedNET jobs added
(2)
GROSS (2)adjusted foradjusted for
(1) jobs (1)GROSSproductivityas a % ofproductivityas a %of
STATEaddedSTATEjobs addedchanges*gross jobschanges*gross jobs
Texas 175,407 116,816 67 Connecticut 8,990 3,237 36
Michigan 149,382 79,304 53 Colorado 8,402 4,312 51
California 147,284 101,406 69 Louisiana 6,594 4,858 74
Unallocated 1 79,859 27,713 35 Delaware 5,802 3,124 54
Illinois 60,181 37,389 62 Vermont 5,501 (996) (18)
New York56,25628,38750Mississippi5,2623,99176
Ohio 53,895 24,583 46 Arkansas 5,123 3,207 63
Indiana 43,192 24,591 57 Oklahoma 4,784 2,812 59
Pennsylvania 38,56523,26160Puerto Rico4,2962,18651
N. Carolina35,29824,55970N. Hamp.4,2292,62362
Minnesota 27,733 19,206 69 Nebraska 4,163 2,775 67
Wisconsin 23,119 14,798 64 Maryland 4,115 2,019 49
Kentucky20,63615,11273W. Virginia3,2972,21767
Tennessee 20,506 12,212 60 Utah 2,947 1,598 54
New Jersey20,0498,77544Nevada2,5261,90375
S. Carolina19,65214,15772N. Dakota2,3871,47962
Arizona 19,128 12,861 67 Maine 2,126 763 36
Georgia 19,036 12,701 67 Alaska 2,075 1,584 76
Missouri 14,860 8,791 59 Idaho. 1,973 1,195 61
Mass.14,4744,30430D. of Col.1,7991,48182
Washington13,6726,83350Rhode Island1,39242831
Virginia 13,437 8,692 65 Montana 1,387 877 63
Florida12,5255,36543S. Dakota1,00557657
Louisiana 11,938 8,190 69 Wyoming 544 366 67
Alabama11,7608,13269Hawaii 1282318
Oregon10,0846,99169New Mexico(7)(407) 5,758
Kansas9,6646,71970V. Islands(67)(106)158
TOTAL 1,212,357 709,988 59
1 Unall: unallocated among states.
* Net jobs added takes into account estimates of people already working to produce exports who would have lost their jobs to
productivity improvements between 1994 and 1998, and subtracts this from the number of jobs added from new exports.
Thus, a negative net jobs added number generally means either very slowly growing exports, or an actual decline in
exports to Mexico and Canada combined.
Source of data: calculated by CRS from DOC data. This table is a condensation of appendix table 7, p. 12. See footnotes to
that table. For explanation of difference between gross and net jobs, see discussion on p. 5, and discussion of the algebraic
formula beginning on p. 14.
Perspective on NAFTA-Related Job Effects
As mentioned, data on job losses and job gains due to trade are derived from
different methods and data bases. They are therefore incompatible and comparisons
between the two could be inaccurate.
Estimates presented above on NAFTA-related job losses are relatively small.
The more than 259,618 workers certified as of September 28, 1999 as potentially
suffering NAFTA-related job losses represent less than the number of U.S. jobs
created in a single month in 1998.
According to the Department of Commerce,8 roughly half of all jobs supporting
exports are included in the manufacturing sector. This suggests that of the
approximately 710,000 estimated net jobs gained from trade with Mexico and Canada
between 1993 and 1998, about 355,000 would be in manufacturing. After four
straight years of decline just prior to NAFTA, during NAFTA’s first four years,
manufacturing jobs increased from about 18 to almost 18.7 million. The NAFTA-
related estimated job gain in manufacturing represents about 50% of all manufacturing
jobs gained between 1993 and 1998. Therefore, it can be argued that NAFTA may
have made a significant contribution to the manufacturing employment turnaround.
To complete the larger picture, tables 4 and 5 identify major industries of
potential job loss and export gains since NAFTA. (Job gains from increased exports
have not been estimated for specific industries because the number of jobs supported
by each billion dollars worth of exports varies by industry.) Several industries
[identified by boldface type and by an asterisk (*)] are included on both lists. These
are: electronics, transportation equipment, nonelectrical machinery, apparel, paper
products, and scientific instruments. This suggests that certain less efficient parts of
these industries are being shifted to Mexico and Canada while more efficient parts are
expanding domestically. The apparel industry is the biggest potential job loser, with
As mentioned at the beginning of this report, in an economy operating at full
employment, trade results in neither net job gains nor net job losses, only in
relocations from less efficient to more efficient industries. Job gains from trade with
Mexico and Canada under NAFTA do not necessarily have to equal job losses from
such trade under NAFTA. Even before NAFTA went into effect there were some
estimates that job losses would be concentrated in early years after NAFTA was
adopted. Tables 4 and 5 document mid-term industry relocations.
8U.S. Jobs Supported by Goods and Services Exports, 1983-94, op. cit., p. 27.
Table 4. Major Industries of NAFTA-TAA Job Certification,
Jan.1, 1994 -Sept. 28, 1999
% of all
No. of jobsNAFTA-TAA
SIC Industrycertified certified jobs
23 *Apparel 73,568 28
36 *Electronics 33,684 13
37*Transportation equip.17,0907
34 Fabricated metals15,3726
22 Textiles 14,1505
35*Nonelec. Machinery11,7475
24 Lumber9,8264
38*Scientif. instruments9,4334
26*Paper products8,982 3
30 Rubber7,7223
31 Leather7,521 3
SUBTOTAL 209,095 81
Other Manufacturing35,17114
ALL MANUFACTURING244,26695
Non-Manufacturing 15,352 6
TOTAL259,618101
SIC: Standard Industrial Classification codes. Manufacturing includes 20 2-digit codes which span
numbers 20-39.
Source: NAFTA-TAA database, sorted by CRS. See also text footnote 5.
* indicates industries listed in both tables 4 and 5.
Table 5. Major Industries of Increased Exports to Mexico
and Canada, 1993-1998
Growth in Industry% of total
Export Value 1993-98
NAFTA
commodity% change
SICIndustryexport gainin $billions93-98
37*Transportation Equip165618
36 *Electronics 17 81 18
35*Nonelectric machinery167417
28 Chemicals 8729
33 Primary metals4765
30 Rubber4815
38*Scientific instruments3483
26*Paper products2693
23*Apparel21002
20 Food2453
SUBTOTAL 74 — 82
Other Manufacturing12—13
TOTAL MANUFACTURING876695
Nonmanufacturing 4 44 4
TOTAL9164100
Source: DOC Office of Trade and Economic Analysis.
* indicates industries listed in both tables 4 and 5.
Appendix:
Data and Explanation of Methodology
This appendix includes supplemental data and explanations. Table 6 includes
state exports to Canada, Mexico, and the two countries combined for 1993, 1997, and
Table 7 includes calculations supporting job gain figures in table 3. In table 7,
column (4) lists gross job gains, by state during NAFTA’s first five years. The same
figures also appear in table 3, p. 7, column (2). In table 7, column (7), lists net job
gains. The same figures also appear in table 3, p. 7, column (3). Net job gains are
gross job gains from new exports minus job losses from productivity growth and
inflation in manufacturing and services. Actual calculations are explained in the table
7 footnotes. In table 7, estimates of the total number of workers supporting all
exports to Mexico and Canada combined for 1993 and 1998, are included in columns
(5), and (6), respectively.
Pages 14 and 15 include an algebraic formula showing how the numbers were
calculated.
Table 6. Merchandise Exports to Canada, Mexico, and Combined,
1993-1998
U.S. Exports to
U.S. Exports toCanadaU.S. Exports toMexicoMexico and CanadaCombined
(in $millions)(in $millions)(in $millions)% ª
STATE1993 1997 19981993 1997 19981993 1997 199893-98
U.S.TOTAL 100,190 150,124 154,15141,635 71,378 79,010141,826221,503 233,16162%
Ala.622 1,2511,281185 814 380807 2,065 1,661106%
Ak.84 305 2291 2585 307 234175%
Ariz.533 1,0721,0591,087 1,963 1,9931,6213,035 3,05288%
Ark.421 820 68869 141179490 961 86777%
Calif.7,158 11,49212,6445,117 9,942 10,79812,274 21,434 23,44291%
Colo.595 672 729604 1,4181,1051,200 2,090 1,83453%
Conn.1,407 1,848 1,872336 530 5441,743 2,377 2,41739%
Del.628 788 938159 308 290788 1,0971,22856%
D.C.36 14215817 172953 159187253%
Fla.1,572 1,928 2,058770 1,2211,2722,341 3,149 3,33042%
Ga.1,469 2,000 2,116324 686 1,1361,793 2,685 3,25281%
Hi.14 33 170 1 214 34 1936%
Idaho157 27828036 44 56193 322 33674%
Ill.4,860 8,044 7,9431,364 2,190 2,7986,224 10,23410,74173%
Ind.4,265 5,060 5,6721,168 2,5733,0465,432 7,634 8,71860%
Iowa919 1,569 1,74278 168158997 1,737 1,90091%
Ks.473 834 879187 449 485660 1,284 1,364107%
Ky.1,058 2,369 2,325190 345 4511,248 2,714 2,776122%
La.372 663 73661 133196433 796 932115%
Maine362 557 52829 18 17391 576 54539%
Md.601 653 69096 199337698 752 1,02747%
Mass.2,541 3,677 3,388374 468 5642,915 4,145 3,952 36%
Mich.11,434 19,760 19,6665,630 6,458 7,88817,065 26,218 27,55461%
Minn.1,950 3,190 3,390229 823 8702,179 4,013 4,26096%
Miss.306 430 49325 127242331 558 735122%
Mo.1,113 1,490 1,570540 1,0421,1901,653 2,532 2,76067%
Mont.145 236 1931 2159146 257 25273%
Neb.296 529 52461 142143357 671 66787%
Nev.123 272 30113 6023137 332 324136%
N.H.377 672 64340 74 86417 746 72975%
N.J.2,539 3,837 3,873789 884 9543,328 4,721 4,82745%
N.M.47 56 68106 8787152 142 1552%
N.Y.6,581 10,6169,9571,171 1,8051,9367,752 10,42111,89353%
N.C.2,289 3,748 3,719365 1,3211,5652,654 5,069 5,28499%
N.D.227 428 3873 18 18230 445 40576%
Ohio7,672 10,47210,669927 1,584 1,9598,598 12,055 12,62847%
Okla.426 670 656158 240 295584 910 95163%
Ore.871 1,0821,329109 89 452980 1,1701,78182%
Penn.3,730 5,616 5,857627 1,1401,4254,358 6,756 7,28267%
R.I.286 330 37242 77 68328 407 44034%
S.C.1,009 1,621 1,711293 936 1,0541,303 2,557 2,765112%
S.D.108 167 166 4 1119112 179 18565%
Tenn.1,679 2,389 2,589650 1,1881,2882,329 3,577 3,87466%
Texas3,811 8,118 8,50612,861 18,86421,62716,672 26,98230,13381%
Utah343 514 50530 73 87374 58759258%
Vt.2,075 2,310 2,48612 9 112,087 2,3192,49720%
Va.1,052 1,536 1,836302 430 5471,355 1,966 2,38376%
Wash.1,723 2,457 2,360208 272 5831,931 2,730 2,94352%
W.Va.285 479 50321 34 56306 513 55983%
Wis.1,947 3,096 3,457288 427 5122,235 3,524 3,96978%
Wyo.38 88 764 5 642 93 8295%
P.R.387 690 657129 217 160517 906 81758%
V.I.10 4 30 4110 8 4-60%
Unall.15,163 17,16217,6583,744 9,288 7,958 18,907 26,451 25,61674%
Source: DOC, Office of Trade and Economic Analysis. Website: http://www.ita.doc.gov. Go to “Trade Statistics” and
then to “State Export Data.”
%ª = % change.
Table 7. Gross and Net Jobs Supported by New and All
Merchandise Exports to Mexico and Canada Combined, by State, 1993-1998
(See column explanations at end of table)
NET Jobs (after compensating forNET as
GROSS Jobs Added Each Year by New Exports to Mexico and CanadaProductivity Growth and inflation)from All Exports to Mexico and Canada:% of GROSS
(4) 1 (7) (8)
(1) (2)(3)TOTAL(5) (6)TOTALCol. (7)
STATE19941997 19981994-1998199319981993-1998/Col. (4)
ALL U.S.334,167410,175142,7771,212,3572,144,8342,854,823709,98959%
Alabama 2,939 7,175 (4,947) 11,760 12,205 20,337 8,132 69%
Alaska 534 1,589 (894) 2,075 1,281 2,865 1,584 76%
Arizona 3,304 5,651 208 19,128 24,508 37,369 12,861 67%
Arkansas 1,800 3,114 (1,151) 5,123 7,409 10,616 3,207 63%
California 24,953 39,157 24,586 147,284 185,618 287,024 101,406 69%
Colorado (556) 7,367 (3,134) 8,402 18,143 22,455 4,312 51%
Conn. 2,580 3,211 478 8,990 26,357 29,594 3,237 36%
Delaware 1,686 901 1,616 5,802 11,912 15,036 3,124 54%
D.C. 1,399 615 343 1,799 809 2,290 1,481 82%
Florida2,2569,9572,21612,52535,40840,7735,365 43%
Georgia 4,635 3,464 6,930 19,036 27,116 39,817 12,701 67%
Hawaii 78 (463) (184) 128 210 233 23 18%
Idaho 709 106 171 1,973 2,919 4,114 1,195 61%
Illinois 19,598 20,687 6,208 60,181 94,124 131,513 37,389 62%
Indiana 9,392 6,045 13,285 43,192 82,152 106,743 24,591 57%
Iowa 3,665 4,455 1,996 11,938 15,074 23,264 8,190 69%
Kansas 3,980 (707) 992) 9,644 9,982 16,701 6,719 70%
Kentucky 7,970 6,741 759 20,636 18,877 33,989 15,112 73%
Louisiana 1,790 1,369 1,665 6,594 6,553 11,411 4,858 74%
Maine 738 951 (367) 2,126 5,910 6,673 763 36%
Maryland 341 1,584 2,143 4,115 10,556 12,575 2,019 49%
Mass. 6,900 3,437 (2,363) 14,474 44,084 48,388 4,304 30%
Michigan 155,574 (5,334) 16,358 149,382 258,067 337,371 79,304 53%
Minnesota 3,724 3,904 3,024 27,733 32,953 52,159 19,206 69%
Miss. 1,243 1,004 2,179 5,262 5,008 8,999 3,991 76%
Missouri 6,722 1,981 2,792 14,860 25,002 33,793 8,791 59%
Montana (11) 642 (61) 1,387 2,208 3,085 877 63%
Nebraska 1,140 1,075 (49)5 4,163 5,392 8,167 2,775 67%
Nevada 386 729 (98) 2,526 2,064 3,967 1,903 75%
N.Hamp. 704 510 (208) 4,229 6,303 8,926 2,623 62%
N.Jersey 9,822 8,446 1,298 20,049 50,327 59,102 8,775 44%
N.Mexico (157) (114) 147 (7) 2,305 1,898 (407) (5758%)
New York15,21327,566(6,456)56,256117,231145,61828,38750%
N.C. 8,929 7,986 2,632 35,298 40,138 64,697 24,559 70%
N.Dakota 320 701 (502) 2,387 3,480 4,959 1,479 62%
Ohio 19,045 14,411 7,004 53,895 130,034 154,617 24,583 46%
Oklahoma 753 2,040 502 4,784 8,832 11,644 2,812 59%
Oregon 2,323 2,901 7,469 10,084 14,816 21,807 6,991 69%
Penn. 8,262 14,073 6,440 38,565 65,900 89,161 23,261 60%
R.Island (452) 778 404 1,392 4,959 5,387 428 31%
S.C. 5,748 5,314 2,547 19,652 19,698 33,855 14,157 72%
S.Dak. 352 87 86 1,005 1,689 2,265 576 57%
Tennessee 5,740 5,918 3,636 20,506 35,221 47,433 12,212 60%
Texas 36,236 60,951 38,581 175,407 252,132 368,948 116,816 67%
Utah 746 865 61 2,947 5,650 7,248 1,598 54%
Vermont (101) (857) 2,179 5,501 31,569 30,573 (996) (18%)
Virginia 3,329 3,409 5,106 13,437 20,485 29,177 8,692 65%
Wash. 4,836 479 2,620 13,672 29,201 36,034 6,833 50%
W.Va. 651 1,480 563 3,297 4,627 6,844 2,217 67%
Wisconsin 8,839 6,218 6,461 23,119 33,798 48,596 14,798 64%
Wyoming 151 260 (135) 544 638 1,004 366 67%
P.R.4,1851,321(1,102) 4,296 7,81710,0032,18661%
V.I. 16 (139) (49) (67) 155 49 (106) 158%
Unalloctd (70,790) 115,159 (10,211) 79,859 285,929 313,642 27,713 35%
For explanation of calculations, see next page. Numbers in parentheses are negative. Numbers may not total exactly because of
rounding.1
Figures for 1995 and 1996 are not shown [(between columns (1) and (2)] because of space constraints.
Explanation of columns in table 7: (Export data are taken from table 6).
Columns (1)!(4): GROSS jobs supported by new NAFTA-related exports (i.e.,
additional exports to Mexico or Canada since NAFTA).
Column (1): Total number of gross jobs supported by new exports to Mexico and
Canada in 1994: Value of export growth in 1994 (in $billions) times
14,361 (the estimated number of workers supported by each $billion
in exports in 1994). Not shown are figures for 1995 and 1996: the
value of export growth times 13,774 and 13,258, respectively.
Column (2): Same figure for 1997: Value of export growth in 1997 times 12,755.
(The number representing additional workers supported by each
$billion in exports each year takes into consideration both productivity
changes and inflation since the previous year.)
Column (3): Same figure for 1998: Value of export growth in 1998 times 12,245.
Column (4): Total number of gross jobs supported by new NAFTA-related exports
during NAFTA’s first five years (94 + 95 + 96 + 97 + 98).
Columns (4)!(8): NET jobs supported by NAFTA-related exports
Column (5): Total estimated number of net jobs supported by exports to Mexico
and Canada combined in 1993: Value of exports in 1993, in $billions,
times 15,123 (number of workers supported by $1 billion in exports).
Column (6): Same figure for 1998: Value of exports in 1998 times 12,245.
Column (7): Net job growth from new NAFTA-related exports, 1993-1998:
Columns (6)-(5).
Column (8): Net NAFTA-related job growth as a percent of gross NAFTA-related
job growth: Column (7)/column (4).
NOTE: See p. 14 for the algebraic formula by which these figures were calculated.
Algebraic Formula:
Relationship Between Gross and Net Jobs Created
from New Exports since NAFTA
This section sets forth the equation that quantifies the relationship between the
gross and net methods for calculating job changes from trade with Mexico and
Canada since NAFTA went into effect. (Gross and net job changes are included in
table 3, p. 7, and table 7, p. 12.)
The estimated gross number looks only at the increase in the dollar value of
exports each year, in billions, and multiplies it by the number of workers required to
produce a billion dollars worth of exports. It ignores any decline in employment of
those already working to produce exports, which occurs because productivity growth
renders them “redundant” or no longer necessary.
The net number looks at the total value of exports each year, in billions of
dollars, and multiplies that by the number of workers required to produce a billion
dollars worth of exports.
The estimated number of net workers added to the payrolls to produce exports
for each state calculated in this way reflects three things:
a. productivity gains that occurred during
that year; and
b. inflation (meaning that fewer items
produced would represent the same value
as the previous year.)
For any state, the difference between the two numbers (gross and net) is equal
to the sum, for each of the years (1994, 1995, etc.) of:
the value of exports (in $billions) to Mexico and Canada
combined in that year times the number of jobs supporting each
billion dollars worth of exports which are lost to productivity and
inflation in that year.
MATHEMATICS of CALCULATION:
Let symbols represent values as follows:
Lg = gross jobs supporting new exports to NAFTA partners, 1993-1998.
Ln = net number of jobs supporting new exports to NAFTA partners, 1993-1998.
Xi = value of exports in initial year “i” in billions of dollars (numbers listed in
table 6 divided by 1,000)
Ji = number of jobs estimated to produce a billion dollars worth of exports in
initial year; and
Jf = number of jobs estimated to produce a billion dollars worth of exports in
final year (“f”).
Where, for both X and J:
i = 1993;
i+1 = 1994;
i+2 = 1995;
i+3 = 1996;
i+4 = 1997; and
i+5 = f = 1998.
And in actual numbers,
J i = J93 = 15,123;
J i+1=J94 = 14,361;
J i+2=J95 = 13,774;
J i+3=J96 = 13,258;
J i+4=J97 = 12,755;
J i+5= Jf =J98 = 12,2459
The gross number of jobs supporting new exports to NAFTA partners can be
represented by multiplying the number of jobs supporting a billion dollars worth of
exports in a given year by the growth in exports for that year:
(1)Lg =Ji+1 (Xi+1—Xi)+Ji+2(Xi+2—Xi+1)+Ji+3(Xi+3—Xi+2) . . . J f (X f—X f-1).
The net number of jobs supporting new exports to NAFTA partners can be
represented by the following equation showing the difference between the number of
jobs supporting a billion dollars worth of exports and the value of exports for the
beginning and end years:
(2)Ln = JfXf—JiXi.
To find the difference between the gross and net numbers of new jobs created from
trade with Mexico and Canada since NAFTA went into effect equation (2) is
subtracted from (1):
Thus, for any state, for any year, the difference between the gross and net
estimates of jobs created from new exports to Mexico and Canada combined since
NAFTA went into effect represents the sum total of the number of jobs held by
previously employed workers producing exports which are subsequently lost to
productivity growth.
9 The figure for 1998 is a CRS estimate based on Department of Commerce estimates for the
previous 5 years.