of the North American Free Trade Agreement
|
|
of the North American Free Trade Agreement |
U.S. Gross Domestic Product
Discussion of NAFTA's effect on U.S. GDP is divided into two parts. Under short-run effects, we discuss those occurring during NAFTA's first three years that, with the United States at or near full employment of resources, will not necessarily persist over time. Under the long-run effects, we discuss those effects of NAFTA on
U.S. GDP that can be expected to endure. The longer-term effects are partially felt in
NAFTA's first three years; but, their full effect will be fully realized only over time. Several studies have concluded that short and long-run effects of the NAFTA for the
U.S. GDP are positive.
Short-Run Effects
DRI estimated that U.S. GDP in 1996 was $13 billion higher than it would have been absent NAFTA, controlling for the peso crisis. The methodology used in those calculations was to apply the results from DRI's disaggregated model of NAFTA's impact on bilateral trade to DRI's large macroeconomic model of the U.S.
economy.
The ITC reported in its study that it was unable to obtain consistent results for U.S. GDP from its econometric analysis
16
because the data it used were insufficient to draw reliable statistical inferences.
Long-Run Effects
Short-run effects are by their nature transitory; however, the long-run effects of trade
agreements like NAFTA are not. It is to those long-run effects that we now turn.
Because export industries tend to be among the most productive and highest wage sectors of the economy, increased exports translate into increased average U.S. labor productivity, real labor compensation, and increased real income in the United States. Reduced restrictions on imports also benefit the economy by increasing
purchasing power of American consumers, providing them with greater choice in products and services at lower prices, helping to restrain inflation, and offering
alternative sources of inputs for America's own production in industries where we are less productive -- thus permitting U.S. companies to enhance their competitiveness and efficiency. In addition, the reduction or elimination of barriers to trade can act to raise the rate of return on productive investment, thus encouraging additional investment and stronger growth.
For these reasons, the long-run benefits of NAFTA are best measured in terms of its effect on overall trade. The impact of increased trade on the U.S. economy can be estimated using long-run trade models. These models can be simulated both with and without regard to barriers, thereby enabling one to isolate the impact of trade
barriers on trade, GDP, and other factors.
A number of studies have projected the long-run effects of NAFTA. A University of Michigan study is representative.
17
Based on a model it proposed, the study concluded that the total long-run effect of NAFTA will be to raise GDP by 0.1 percent annually. Using the trade estimates in the DRI study,
described above, the Michigan models would suggest that U.S. real income in 1996 would have captured somewhere between $3 to $6 billion of the long-term, permanent gain expected from NAFTA. This gain is additional to the short-term, but more transitory gain of the $13 billion reported above.
18
Some recent studies have predicted even larger effects. For example, a recent Federal Reserve Board of Chicago study estimated that over the long-run the NAFTA will increase annual GDP by 0.24 percent (or $18 billion relative to the size of the economy in 1996), long-run real wages by 0.25 percent, and total trade between
the United States and Mexico by 19 percent.
19
In May 1992, the ITC published a report on the proceedings of a symposium held by the ITC on the modeling of the long-run economic effects of NAFTA.
20
Twelve economy-wide models were presented, discussed, and assessed at the symposium. The ITC summarized the various studies stating that most of, "the models estimate a NAFTA would cause U.S. real GDP to expand by 0.5 percent or less." This is a per year gain, reached with full implementation and full adjustment to the NAFTA's rule changes. Applied
to current GDP, the upper end of this gain reaches nearly $40 billion a year.
It is important to note that the "jobs per billion dollar" ratio derived from the Commerce Department's work on exports cannot be used to calculate job changes due to imports. In fact, imports do not necessarily displace U.S. production. The mainstream economic community has not developed any broadly agreed methodology to sort out from the nearly $1 trillion in U.S. annual imports those imports that might displace U.S. production, as well as the degree to which such production is displaced. 22
Clearly, some imports may have a job-displacement effect. Just as clearly, not all imports displace U.S. domestic production or jobs. Indeed, some imports provide competitive inputs for U.S. production thereby supporting, rather than displacing, U.S. jobs. Other imports are goods that are in short supply or are not produced in the United States, and do not have a one-for-one job displacement effect. Had there been no U.S. imports from Mexico of petroleum ($6.4 billion in 1996) or coffee ($540 million in 1996), the result would either have been reduced supply and higher prices in the United States, or higher U.S. imports from third countries, not increased U.S. production.
To take another example, had the United States not imported simple apparel items from Mexico over the past several years, much or all of the substitute supply could be expected from third country suppliers with lower U.S. content, such as China, rather than from increased production in the United States itself. U.S. content in apparel imports from Mexico is 64 percent higher than for apparel imports from many other countries.
As indicated, it is incorrect to apply the "jobs per billion dollars" ratio, derived from the Commerce Department's work on exports, to the dollar value of U.S. imports to obtain an import job displacement effect. The calculation of the U.S. labor content of exports is completely unrelated to any measurement of the job displacement effects of imports, and such import effect calculations are not based on any recognized conceptual underpinnings.
Moreover, this is tantamount to concluding that trade deficits always have a net job displacement effect and trade surpluses a net job creation effect. This invites the question whether the shift in the trade balance with Mexico from a surplus of $1.7 billion in 1993 to a deficit of $17.5 billion in 1996 had a net job displacing effect in the United States. Generally, the facts do not support the view that trade surpluses increase employment levels or that trade deficits necessarily result in reduced employment. Over the last 40 years, U.S. employment generally grew more and the unemployment rate was lower during periods when the U.S. trade balance was deteriorating.
| This should not be surprising; trade deficits and employment both tend to rise when the economy is growing rapidly and consumers and businesses increase their spending. For example, the U.S. current account was nearly in balance in 1991, while the unemployment rate peaked at 7.0 percent in December of that year. (Figure 4). Since 1992, the U.S. economy has recovered more quickly and has grown faster than those of many of our major trading partners. The relative strength of U.S. growth contributed to the expansion of the current account deficit to $148.2 billion in 1996. By the end of 1996, the U.S. unemployment rate was down to 5.3 percent and from the recession year of 1991 to mid-year 1997 13.8 million net new jobs had been created in the U.S. economy. (Mexico's trade surplus with the U.S. in 1995 was accompanied by a loss of over 1.0 million Mexican jobs.) |
|
In addition, the DFR and DRI studies confirm that NAFTA has worked to raise exports to Mexico more than imports from Mexico. Thus, NAFTA, on its own, if anything, improved the trade balance; any deterioration is associated with factors other than NAFTA. It is extremely difficult to estimate the impact of a trade agreement such as NAFTA on net employment except indirectly by estimating the net job growth associated with net income gains. The DRI estimates of modest positive income gains cited above belie any possibility of net job losses.
NAFTA-TAA Program
The NAFTA Transitional Adjustment Assistance program (NAFTA-TAA) was enacted to ensure that workers would be provided support to the extent adjustment to new trade patterns with Canada or Mexico caused dislocations. Both the Administration and Congress recognized that, while expanded trade provided real opportunities, there would be inevitable dislocations as jobs shift to capitalize on those opportunities. The NAFTA-TAA program is operated by the Department of Labor and provides expanded worker retraining, and income support benefits, as well as other adjustment assistance, for those workers who may lose employment due to trade or production shifts to Mexico or Canada.
The number of workers certified under the NAFTA-TAA program is often cited as those who have been displaced because of NAFTA. However, a close examination of how the program actually operates reveals quite clearly that certification numbers overstate the number of workers displaced because of trade with Canada and Mexico, and, in any event, do not provide estimates of job losses due to NAFTA. There are two reasons for this overstatement. First, being displaced is not a prerequisite to being certified by the Department of Labor. Second, those certifications, in any event, are not dependent on identifying NAFTA as the cause of the possible dislocation.
Under the program, workers must be certified to apply for benefits. Workers may be certified if the worker is at risk of losing his or her job due to trade or production shifts in North America. Certification, however, does not mean that all workers certified have actually lost jobs. In fact, actual job lay offs will only total some fraction of certifications.
Moreover, certification does not require that NAFTA be the cause of the dislocation or risk. For example, employees at the Mattel Corporation were certified under NAFTA-TAA, even though imports from Mexico and Canada of products produced by these workers were free of U.S. tariff duties prior to the NAFTA. Certification is provided even if the increased imports (or shift in production) were only one among other possible causes for layoffs or the risk of lay offs.
Appreciating the distinction between certification and displacement helps explain the difference between the 99,497 U.S. workers who were certified under NAFTA-TAA in its first three years, 23 and the 12,193 who actually applied for benefits during that time. 24 (In addition to these 12,193, another almost 20,000 workers were dual certified for both NAFTA-TAA and the regular Trade Adjustment Assistance, but received services and benefits under the latter. (Under the latter 25 program, income benefits are available without accompanying training obligations if a waiver is obtained. NAFTA-TAA does not provide for such waivers.)
These facts indicate that the number of workers actually displaced because of trade with Canada and Mexico -- not necessarily as a result of NAFTA -- is below the 99,497 certified and above the approximately 32,000 who applied for benefits. The Department of Labor does not have information to estimate the precise number of displaced workers within this range. As previously noted, some of the workers covered under a certification are not laid off, or may remain certified even though recalled to work.
It is also important to put this displacement in the context of the overall dynamism of the U.S. labor market. The U.S. economy, on average, generated more than 100,000 net new jobs (excess of job gains over losses) every two weeks during the first three years of the NAFTA.
The NAFTA and Investment
The following discussion addresses the extent to which investment in new or existing production or other operations in the United States has been redirected to Mexico as a result of the NAFTA and the effect, if any, of such redirection on the United States employment.
U.S. Direct Investment in Mexico
Some early critics alleged that NAFTA would result in a one-for-one reduction in productive investment in the United States as investment was relocated to Mexico. For a variety of reasons, if anything, NAFTA has had a positive, although moderate, effect on investment in both countries.
The NAFTA eliminates both impediments to direct investment in Mexico, and certain
provisions that made foreign direct investment (FDI) in Mexico the only way to sell into the domestic Mexican market. The local manufacturing requirement and other provisions of Mexico's earlier auto decrees are prime examples. U.S. producers
who otherwise would have exported to Mexico were required to invest in Mexico if they wished to serve the Mexican market. In this respect, the NAFTA should
tend to reduce certain types of investment in Mexico. The net result should be to increase the scope for rational, market-based investment decisions, enhancing higher paying job opportunities and global competitiveness on both sides of
the U.S.-Mexican border.
The types of productive activities in which foreign direct investors tend to invest in Mexico are the types of activities in which Mexico's main competitors would typically be other low and middle income countries, not high income
countries like the United States or Canada. The competition for that investment would be from other Latin American countries and Asia.
Foreign direct investment in Mexico, and other countries, also helps support U.S. exports. When U.S. firms build plants in other countries, they tend to rely on U.S. inputs, both in the form of capital goods when constructing the plant, and later, for materials and other inputs. Finally, contributing to a stronger Mexican economy through investment also helps build markets for U.S. exports.
When considering U.S. FDI in Mexico, it bears observing that in 1996 alone, total private business fixed investment in the United States totaled nearly $800 billion (over two times the size of Mexico's GDP). Investment in the United States in producers' durable equipment as a share of U.S. GDP is at its highest level since the 1950s. The United States is the world's largest destination for FDI, with $77 billion invested from abroad in 1996.
26
It is, in fact, the large net inflow of foreign capital -- not just FDI -- that constitutes the U.S. capital account surplus, in turn, offsetting the U.S.
current account deficits.
By contrast, U.S. FDI flows to Mexico in 1994, 1995, and 1996 were very small relative to the $1.1 trillion in U.S. gross private domestic fixed investment in 1996, amounting to approximately 0.4, 0.3, and 0.2 percent of total U.S. investment. It was also quite small when compared to world FDI flows into the United States, which
reached a level more than 28 times greater than what the United States invested directly in Mexico in 1996. Not surprisingly, the ITC concluded in its FDI analysis
that any fluctuations in U.S. FDI in Mexico (whether or not attributable to NAFTA) have had only a minimal impact on aggregate U.S. investment.
27
There is also no particular evidence in the aggregate since 1994 that U.S. FDI in Mexico was withdrawn from other markets and redirected to Mexico because of superior Mexican investment opportunities. During 1989-94, U.S. FDI in Canada and in the Asia/Pacific region grew at significantly higher rates than U.S. FDI in Mexico. In 1995, a large-scale surge in total U.S. direct investment abroad ended up
primarily in Europe, the traditional host region for most U.S. FDI.
U.S. FDI flows to Mexico rose from $2.5 billion in 1993 to $3.7 billion in 1994, but fell back to $3.0 billion in 1995 and $2.7 billion in 1996.
28 Mexico accounted for 2 percent of the stock of U.S. FDI
abroad in 1995. Mexico's share of U.S. outflows of FDI shifted from 3.3 percent in 1993 to 5.3 percent in 1994, and dropped to 3.4 percent in 1995, the year of the peso crisis, and to just 3.1 percent in 1996.
Moreover, while NAFTA contains important investment-related provisions that broaden the degree of protection for U.S. investments in Mexico, the Mexican government undertook unilateral initiatives toward foreign direct investment
in the pre-NAFTA years, which contrasted markedly with Mexico's earlier inward-looking policies. Mexico broadened the list of sectors of the Mexican economy
eligible for FDI and loosened government oversight of FDI in other sectors. These reforms are of at least equal importance to NAFTA reforms. They make assessment of investment trends in the NAFTA period particularly problematic.
FDI in the United States
Because NAFTA is expected to boost U.S. GDP, it will also likely increase U.S. investment over time, and empirical estimates suggest that NAFTA has already had a modest positive effect. DRI estimated that NAFTA had raised U.S. fixed investment by $5 billion by 1996. The Chicago Federal Reserve's long-run model
predicts that U.S. capital stock will eventually be about 0.37 percent higher
on account of NAFTA. Since the United States has likely achieved about one-fourth of the long-run effect of NAFTA thus far, that would imply that NAFTA has generated a boost of about 0.1 percent to capital stock. As the value of non-residential fixed private capital in the United States in 1995 was $8.0 trillion, a 0.1 percent
increase translates to an $8 billion increase.
During the period 1994, 1995, and 1996, inflows of FDI to the United States totaled $45.7 billion, $67.5 billion, and $77 billion, respectively. Outflows of FDI from the United States in those three years totaled $69.3 billion, $86.7 billion, and $87.8 billion.
29
Direct investment of Canadian firms in the United States continued to grow after NAFTA's implementation, following a longstanding trend. In 1996, the FDI position of Canadian firms in the United States was $53.8 billion on a historical-cost basis, nearly 59 percent of the U.S. direct investment position in Canada. The FDI position of Canadian firms in the United States accounted for 8.6 percent of the direct
investment position of all foreign firms in the United States in 1996.
Canada currently is the fifth largest direct investor in the United States, behind the United Kingdom, Japan, the Netherlands, and Germany, in that order. Just over 40 percent of Canadian FDI stock in United States is invested in manufacturing.
Canadian direct investment flows into the United States during the first three years of NAFTA were $5.0 billion in 1994, $7.1 billion in 1995, and $5.7 billion in 1996. These were the three largest such annual flows in nominal terms. By comparison, Canadian direct investment flows to the United States were $1.3 billion in 1992 and $3.8 billion in 1993. Manufacturing accounted for 49.5 percent of the Canadian FDI inflows into the United States during 1994-95, insurance for 18.9 percent, and
retail trade for 11.7 percent.
The stock of Mexico's foreign direct investment in the United States is small -- less than six-tenths of one percent of total FDI in the United States. Although it more than doubled from $1.0 billion in 1993 to $2.3 billion in 1994, it declined to slightly under $2.0 billion in 1995 and declined further to just over $1.0 billion in 1996.
Capital inflow during this time period was also low, and was positive only in 1994 at $1.2 billion, or 2.7 percent of total capital inflow into the United States that
year.
The NAFTA'S Effect on Mexico's Economy, Employment and Wages
| The Mexican authorities responded to the financial crisis in 1995 by firmly implementing a strong economic adjustment program -- backed by U.S. and other international support -- and fully respecting its NAFTA obligations to liberalize trade with the U.S. and Canada. The Mexican economy began to recover by late 1995, and over the course of 1996, Mexican GDP grew over 5 percent in real terms. (Figure 5.) Quarterly data show that Mexico had regained its pre-crisis level of GDP by the fourth quarter of last year, two years after the crisis began. Foreign investor confidence also returned quickly, as foreign lenders returned to Mexico only seven months after the crisis, compared to seven years after the 1982 debt crisis. |
|
The comparison may become even more stark. Private forecasters (such as the WEFA Group, a major U.S. macroeconomic forecasting organization) today expect Mexico's recent growth to continue at a pace of 4.5 percent to 5.0 percent over the next two years. Assuming a 4.5 percent average growth rate over the next four years, Mexico's GDP would be nearly 20 percent above its pre-crisis level. In contrast, for the six years following the 1982 crisis, the Mexican economy registered zero real growth.
Mexican overall imports from the U.S. also recovered much faster after the 1995 crisis than in 1982. In the initial months after the two crises, imports fell by similar proportions (i.e., they dropped 20 percent within six months of each crisis). In the 1995-96 period, Mexican imports -- including imports from the United States -- rebounded, surpassing their pre-crisis level by the third quarter of 1996. By comparison, it took seven years after the 1982 crisis for U.S. exports to reach their previous peak (Figures 6 and 7).
|
|
It is likely that NAFTA contributed to this recovery both directly and indirectly. First, unlike in 1995, Mexico in the early 1980s imposed heavy restrictions on imports, such as tariff increases to 100 percent and strict import licensing requirements. Those measures likely hurt subsequent output growth by shielding inefficiencies in Mexican industry and by limiting access to imported materials and capital goods. Second, looked at more broadly, NAFTA helped the Mexican recovery and long-term growth prospects in several ways. It supported the resolve of the Mexican Government and its people to stay the course of market-based economic reforms. The NAFTA gave investors confidence that Mexico was, in fact, committed to market-oriented economic reforms.
Real wages of Mexican workers fell sharply after both the 1982 and the 1995 economic crises, although the decline was somewhat more severe after the 1982 crisis. Real manufacturing wages were more than 30 percent below their pre-crisis level 27 months after the 1982 crisis began, compared to a drop of 23 percent in the same period after the 1995 crisis. Looking forward, however, most private analysts expect real wages to begin to recover this year, while real wages were still falling more than six years after the 1982 crisis.
General Employment and Wage Trends in Mexico
The Mexican labor force grew 4 percent per year during the 1970s and 1980s and grew 2 to 3 percent in the 1990s (compared to a U.S. rate of 1 percent). This trend reflects, in part, the rapid growth in the working age population, but is also a result of a steep rise in labor force participation rates beginning in the early 1980s.
Participation rates rose by 0.7 percentage points per year in the 1980s, and by 0.5 percentage point per year between 1991 and 1995, with increases for both men and
women.
Of the nearly 40 million people in the economically-active population (itself about 42 percent of the Mexican population), approximately 35 percent are employed in the formal sector and are covered by social security and other related programs. Another 25 percent work in small enterprises in the semi-formal sector where few are
covered by social security, and most of the remaining 40 percent are marginally under- or self-employed in the informal sector where they receive few, if any, of the
benefits and protections afforded workers in the formal sector.
Formal sector employment contracted sharply during the 1995 recession, as workers
registered with Mexico's Social Security Institute (IMSS) declined by almost one million, or 8 percent, between November 1994 and September 1995.
However, Mexico's economic recovery brought a rapid rebound in employment. IMSS
registrants recovered their pre-crisis peak by November 1996, and by April 1997, were 4 percent above the pre-crisis level.
Real wages, in peso terms, also fell sharply in Mexico during the 1994-95 financial and economic crisis. Real manufacturing wages fell nearly 20 percent between November 1994 and November 1995, and declined further during 1996. Even in March 1997, they were still 23 percent below their March 1994 level, before the
crisis hit.
Although there has been a decline in real wages of 23 percent, manufacturing jobs that are associated with exports pay more than other jobs in Mexico
30
following the general pattern that prevails in the U.S. regarding export supported
jobs. Manufacturing firms with a significant percentage (i.e., 40-80 percent) of their total sales going to exports during the 1994-96 period paid wages that at a
minimum were 11 percent higher than non-export oriented manufacturing firms in Mexico. As the percentage of export sales increased (i.e., above 80 percent) during the 1994-96 period, manufacturing firms paid between 67 percent (1994) and 58
percent (1996) more.
The higher pay levels of Mexican manufacturing workers in the export sector have not spared them 20-plus percent reductions in real wages experienced by other Mexican workers in the wake of the peso crisis. The decline in the real wages of manufacturing workers in the maquiladora sector between 1994 and 1996 has been
considerably less at 12 percent. The maquila wage has risen from rough parity with non-export workers in 1993 to 16 percent above non-export workers in 1996.
Footnotes:
17. Drusilla Brown, Alan Deardorff, and Robert Stern, "A North American Free Trade Agreement: Analytical Issues and Computational Assessment," The World Economy 15 (1), January 1992, pp. 11-29.
18. Estimates from the aggregate DRI study suggested that, to date, NAFTA has increased total U.S.-Mexico trade by 5.5 percent, while the disaggregated DRI study found a 15 percent increase. The Michigan model simulations imply that between 35 and 100 percent of NAFTA's long-term effect on trade has already been realized. Based on that estimate, the DRI trade estimates translate into a $3 to 6 billion increase in two-way trade through 1996. Applying the DRI study to the Chicago Federal Reserve results noted below, suggests that NAFTA raised annual GDP by $5 billion to $14 billion as of 1996.
19. Michael A. Kouparitsas, "A Dynamic Macroeconomic Analysis of NAFTA," Economic Perspectives, Federal Reserve Bank of Chicago, 1996, pp. 14-35.
20. U.S. International Trade Commission, "Economy-Wide Modeling of the Economic Implications of a FTA with Mexico and a NAFTA with Canada and Mexico," Inv. No. 332-317, May 1992.
21. The Commerce Department periodically publishes estimates on the number of U.S jobs supported by exports. These figures include both workers employed in exporting sectors as well as U.S. workers employed producing inputs or components for export production, or whose employment is otherwise required to produce exports and move them to market. Commerce's estimate of jobs supported by exports are based on the examination of U.S. sectoral input-output relations and U.S. employment data. These estimates take into account actual annual changes in
the volume of output for exports and domestic demand, composition of that output, productivity, and technology change, among other factors effecting the labor
requirements for the production of U.S. exports. Because of the lag in Commerce's reporting of its formally derived estimates, this Study uses an extrapolation for recent years, currently for 1995 and 1996, which reflects the fact that
productivity gains and price inflation reduce the average number of U.S. jobs supported each year for a given dollar value of U.S. exports.
22. Hinojosa-Ojeda has started preliminary work using Armington elasticities to take into account that estimated job displacement effect of imports are smaller than the job creating effect of exports.
23. Of the 99,497 workers certified, 50 percent were covered due to a shift in production to either Mexico or Canada and about 50 percent were
covered due to increased imports from those countries. Mexico was
identified as the location of production shift or the source of imports for about 60 percent of workers covered (just under 60,000), Canada was identified as the
location or source for about 23 percent of the workers covered, and no single source was identified for about 17 percent of the covered workers.
24. Out of this number, 5,886 U.S. workers actually received job training paid for by the Department of Labor, 24 and 3,854 have received extended income support (beyond the 26 weeks provided by regular unemployment insurance.)
25. The Department of Labor does not identify how many of the 12,193 and 20,000 benefits applicant groups were certified with respect to Mexico, because applications for benefits are made at the state level and do not require the applicant to identify the basis for the original certifications.
26. In 1995, the latest year available, China was second largest recipient of FDI with $37 billion, 40 percent less than the U.S. total in 1995.
27. ITC study, June 1997 pp. 3-36.
28. The stock of U.S. FDI in Mexico, at historical cost basis, was $15.2 billion in 1993, $16.1 billion in 1994, $16.0 billion in 1995, and $18.7 in 1996.
Continue on to Chapter 2: Sectoral Trade
16. The ITC, however, reported on the basis of all information available to it (not simply the formal econometric analysis) that it "found positive, although modest, effects on the U.S. economy after three years of the NAFTA." (p. XVIII). The Congressional Research Service in a recent paper also indicated that "[t]he data suggest that the NAFTA has had a positive, but small,
effect on U.S. trade with Mexico... ." NAFTA: Economic Effects on the United States After Three Years, CRS Report to the Congress, Arlene Wilson, June 13, 1997.
