A Preliminary Evaluation of NAFTA
by C. Fred Bergsten, Peterson Institute for International Economics
and Jeffrey J. Schott, Peterson Institute for International Economics
Testimony before the Subcommittee on Trade Ways and Means Committee
United States House of Representatives
September 11, 1997
There has been much talk about structural change in the current United States economy, and our session chair has presented some of the hard facts underlying it. In keeping with the mandate of the panel—and recognizing the comparative advantage in forecasting of the chair, Monday's speakers, and many in the audience—I will stick to the general question of what monetary policy's response to structural change should be. Then I want to go a bit further and suggest what the apparently successful U.S. monetary framework at present can do to lock-in and build upon the achievements of the last few
Any evaluation of NAFTA to date must analyze the detailed trade and investment flows among the three member countries since the agreement was implemented at the outset of 1994. We will do so in this statement. We will also offer our evaluation of the Administration's official report on the agreement.
Before proceeding, however, we wish to emphasize several broad strategic considerations that must be accorded substantial weight in assessing the results to date. NAFTA must be judged against a series of fundamental US policy goals rather than simply on a narrow assessment of changes in trade and investment, and their effects on the American economy. Indeed, we believe that NAFTA shall be evaluated primarily on these broader considerations. We feel compelled to emphasize these factors in our own evaluation because the Administration's report, like its overall trade policy, gives short shrift to these strategic perspectives.
Even before addressing the strategic impact of NAFTA, however, three caveats must be stressed. First, it is far too early to reach a considered judgment on the success or failure of the arrangement. Like any trade agreement, NAFTA aims to improve the economic structures of the participating countries. It is not a cyclical or short-term tool for creating jobs or anything else. It will be years, if not decades, before anyone can objectively reach a comprehensive judgment on the impact of NAFTA.
Our assessment covers a longer period than the 3 1/2 years since the agreement was formally launched. The reason is that much of Mexico's liberalization and deregulation in the late 1980s and early 1990s was undertaken at least partly to enable it to enter into NAFTA negotiations. From the time that Mexico proposed the idea in 1990, the United States made clear that Mexico would have to greatly improve its trade, investment and related policies before an agreement could be concluded. Hence NAFTA deserves credit for a substantial part of the Mexican economic reforms implemented since 1990 as well as the substantial further reduction of Mexican tariffs, and other trade and investment barriers, since NAFTA formally took effect.
The second caveat is that NAFTA necessarily has a very small impact on the American economy. The addition of Mexico to the Canada-United States Free Trade Agreement essentially expanded our free trade area by 4 percent—the ratio of Mexico's GDP to our own. Any impact is therefore inherently modest. For example, the US economy has created 7.5 million jobs since NAFTA went into effect, swamping any conceivable “NAFTA effect”.1 Neither judgments about the American economy nor about future American trade policy can be driven very far by NAFTA.
The third caveat is that much of the NAFTA debate, certainly during the Congressional approval process and even today, addresses the wrong questions. Neither NAFTA nor any trade agreement should be judged on its contribution to achieving full employment in the United States. The total level of jobs in our economy is basically determined by macroeconomic and monetary policy. The current period presents dramatic evidence of this conclusion: we are at “full employment,” and the unemployment rate has dropped far below the level that most economists had felt was safe from the standpoint of price stability. This positive employment situation has developed despite a large and growing trade deficit, including an “adverse” shift in our trade balance with our NAFTA partners.
Trade rather helps to determine the quality of jobs in the economy. It shifts output from sectors where we are least productive into those where we are most productive. Hence it increases wage levels and standards of living; export jobs pay about 15 percent more than the national average. Indeed, the export boom of the past decade has stopped the decline of high-wage manufacturing jobs in our economy and, if it continues at the recent pace, could even restore the level of manufacturing employment to its previous high over the next decade or so. Since our major national economic problem has been a long-term stagnation of per capita incomes and wage levels, increased trade—including via agreements like NAFTA—clearly contributes positively to our national economic interest.
NAFTA's Strategic Objectives: An Early Appraisal
With those very important caveats in mind, let us begin the evaluation itself by analyzing the record of NAFTA to date in achieving seven of its central strategic objectives:
This key purpose of NAFTA was unfortunately put to a very early test with the peso crisis less than one year into the agreement. But NAFTA and the Mexican reforms clearly held: unlike virtually all previous cases, such as the debt crisis of 1982, the Mexican government responded with an appropriate package of macroeconomic and further structural reforms rather than by rolling back its past liberalization. Open access to the US market, reinforced by NAFTA, helped prevent an even more drastic recession and thus still greater pressure to reverse the reform program.
At the same time, Mexico should be faulted for paying too little attention to the macroeconomic and monetary implications of its trade liberalization. To be sure, NAFTA-related liberalization was only a minor factor in bringing on the peso crisis, and the United States responded properly by helping finance a constructive Mexican policy response, but preemptive action would have been far better and should have resulted from ongoing consultations between the US Treasury and its Mexican counterparts.
The results have been notable, although more progress needs to be made in restoring the real income levels of the poorer segments of Mexican society. Mexico already achieved 5 percent growth already in 1996, in stark contrast to the five-year recession that followed its 1982 crisis. Since mid-1996, the Mexican recovery has been led by a revival of domestic demand, primarily in the labor intensive construction sector. United States exports to Mexico exceeded their 1994 level by 11.8 percent in 1996. This stands in stark contrast to the 50 percent cut in US exports from their 1981 level in the aftermath of the Mexican debt crisis of 1982-3 when our sales did not recover to their pre-crisis levels until 1988. NAFTA thus passed its first major test with flying colors.
Here too, NAFTA has worked well. Mexico will eliminate tariffs that averaged about 10 percent on US goods compared with US tariffs that averaged about 2 percent on Mexican products. The NAFTA ratio is thus about 5 to 1 in our favor and, at least to date, full implementation (7 percentage points of Mexican tariff cuts, 1.4 percentage points for the United States) is proceeding on schedule. Even more important, NAFTA has provided a model for the proposed Western Hemisphere and APEC free trade arrangements where the ratios are even higher and where free trade is thus so clearly in the US interest.
During its first 3 1/2 years of existence, NAFTA has thus already fulfilled its most fundamental strategic goals to a considerable extent. On these criteria, it must be viewed as a major success for the United States (and for Mexico and Canada). We now turn to a more detailed examination of the trade and investment flows that have occurred, framed in terms of an appraisal of the official evaluation offered to the Committee today by Ambassador Barshevsky and Secretary Daley.
Evaluating the Administration's Evaluation
NAFTA assessments often confuse what has happened since NAFTA entered into force with what has happened because of the implementation of NAFTA-mandated reforms. Doing so attributes to the trade pact many developments that essentially are unrelated to the pact and would have occurred anyway. To its credit, the Administration report on NAFTA tries to parse out the effects of the Mexican peso depreciation from the effects of the NAFTA trade reforms to estimate what difference the NAFTA has made for the North American economies.
Overall, the Administration reports that NAFTA has had a very modest impact on the US economy. This conclusion is markedly different from its rhetoric during the NAFTA implementation debate in 1993 but unremarkably similar to most economic projections that forecast modest trade gains and insignificant employment effects. The following subsections review the key findings on trade and employment effects, as well as the operation of the side pacts. We also include an evaluation of the dispute settlement provisions of the NAFTA, which the Administration report ignored despite their surprisingly strong track record.
The report accurately records the impressive growth in US bilateral trade with Mexico and Canada since NAFTA entered into force. Total trade with our NAFTA partners increased by 43.3 percent since 1993 (the year before NAFTA took effect), significantly faster than US trade with the rest of the world (32.4 percent) over the period 1993-1996. US exports to Mexico and Canada increased by 37 percent and 33 percent, and US imports from our NAFTA partners grew by 83 percent and 41 percent, respectively. This growth continues a trend that preceded the NAFTA trade reforms. During the three-year period prior to NAFTA, US trade with Mexico and Canada also grew much faster than our trade with other countries (25.5 percent versus 14.9 percent).2
What this means is that the economic integration of the North American economies had been advancing long before the NAFTA, spurred by the new trade and investment opportunities created by the domestic economic reforms in Mexico since the mid-1980s and the inflation and budget-cutting initiatives in the United States and Canada. The NAFTA reinforced this trend, but regional trade and investment would have continued to expand even if NAFTA had never been broached.
How much of this trade growth is due to trade liberalization mandated by NAFTA obligations? The report acknowledges the difficulty of isolating the NAFTA effect, particularly in light of the 1995 peso crisis and sharp Mexican recession, the concurrent implementation of tariff cuts negotiated in the Uruguay Round, and the robust growth of the US economy over the past few years. It states that the effects of the peso depreciation and the relatively faster growth of the US economy were much more important than NAFTA trade reforms in explaining the increase in post-NAFTA trade, and references other studies that suggest that NAFTA reforms by themselves actually increased US net exports to Mexico since 1994.3 Those results are then used to calculate net US job gains generated by NAFTA-related reforms (see below).
What the report also could have noted is that, even before the NAFTA entered into force, the overvalued peso further complicated the analysis by dampening Mexican export growth to the United States and promoting larger US exports to Mexico than otherwise would have occurred. Hence, the shift in the US bilateral trade balance with Mexico since NAFTA took effect would have been substantially smaller, absent the peso misalignment that emerged well before the peso crisis of December 1994.4
The report provides a straightforward and accurate assessment of the impact of NAFTA on regional investment and the minimal effect US foreign direct investment (FDI) in Mexico has had on the US economy. It demonstrates persuasively that the alarmist fears of Ross Perot and others have not materialized, i.e. there has been no “giant sucking sound” of US plants and jobs heading south of the border.
To be sure, NAFTA has made it easier to invest in Mexico by cutting red-tape and removing key ownership restrictions, particularly in the financial services sector. In fact, Mexico accelerated the implementation of its NAFTA obligations to liberalize investment in the banking sector as part of its response to the peso crisis. However, these reforms will only attract foreign investors if those companies believe that the climate for economic growth is favorable, i.e. that domestic economic policy is sound and commands political support.
As noted earlier, Mexico's reforms have contributed to an increasingly strong inflow of FDI in Mexico. Since NAFTA entered into force, the US share of new FDI in Mexico5 (new investment and reinvested earnings) has been slightly more than 50 percent (down from 60 percent historically); these funds represent about one-half of one percent of the total investment in plant and equipment in the United States in 1996.
The report provides a half-step back from the Administration's persistent and exaggerated claims that the NAFTA would be an engine of job creation. Nonetheless, the authors seem obliged to link NAFTA reforms to net gains in US employment and thus rely on commissioned studies from DRI (that calculated that US net exports were $7 billion higher in 1996 than otherwise would have occurred due to NAFTA) to estimate export-related US job gains of 90,000 or more. This is small potatoes in an economy that counts a labor force of 136 million people.
Scant attention is given to job displacements due to import growth: the report correctly states that “imports do not necessarily displace U.S. production” and that the workers certified under the NAFTA-TAA program “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.” This conclusion is a bit cavalier. Clearly, the program does not require a clear link to NAFTA reforms and reportedly has certified workers displaced for reasons other than increased trade and investment with our NAFTA partners. Equally obvious, however, is that not all workers eligible for the program have taken advantage of it, so the number of certifications likely understates the number of affected workers.
However, any way one slants the numbers, the problem of job displacement due to NAFTA is very small relative to the total number of jobs displaced each year in the US economy (about 1.5 million), which as noted above have been more than offset by the substantial US job creation in recent years. Labor adjustment is an important issue for US policy, but it is not primarily a NAFTA or trade-related problem.
Inexplicably, the Administration report makes virtually no mention of the results to date of the NAFTA dispute settlement provisions, particularly chapter 19 reviews of final antidumping and countervailing duty decisions.6 Through April 1997, NAFTA panels have been convened in 26 cases, of which the United States was the plaintiff in 13 and the defendant in 10 cases. In general, the disputes have been handled expeditiously and objectively. Panelists have not displayed national bias nor have they “rubber stamped” decisions by national agencies.
The NAFTA system has worked well for the United States. Five of the eight cases that resulted in a panel finding (5 others were withdrawn during the panel process) led to the reappraisal of the Mexican or Canadian antidumping action against US firms by the national agency.
Labor and Environment Side Agreements to NAFTA
The Administration report provides a factual account of the activities that have been undertaken pursuant to the North American Agreement on Environmental Cooperation (NAAEC) and the North American Agreement on Labor Cooperation (NAALC). The side agreements have three specific objectives. First, the pacts monitor implementation of national laws and regulations in each country pertaining to labor and the environment, performing a watch-dog role to alert countries about abuses of labor and environmental practices within each country. Second, the pacts provide resources for joint initiatives to promote stronger labor and environmental practices. Third, the pacts establish a forum for consultations and dispute resolution in cases where domestic enforcement is inadequate.
Despite a slow and cumbersome start, the pacts have recently begun to show some results. Both side agreements have focused their efforts primarily on oversight of national laws and practices, sponsoring comparative studies, training seminars, and regional initiatives to promote cooperative labor and environmental policies. These efforts seem small in relation to the magnitude of the labor and environmental problems confronting the three countries, but they have directed additional attention and resources to these problems that would have been lacking in the absence of the side pacts.
To be sure, the dispute settlement provisions were the driving force behind the US initiative to secure the agreements, which sought additional trade provisions to address perceived labor and environmental problems in Mexico. In this area, the record to date has been mixed.
Disputes concerning unfair labor practices (primarily denial of right of association) have benefited from the glare of publicity afforded by the pacts. Eight cases have been filed with the NAALC secretariat — seven by the United States and one by Mexico. Two of the eight resulted in changes in the contested labor practices, two are the subject of ministerial consultations, three were terminated, and one recent case is under review. Trade sanctions have not been a factor in any of the cases.
In the environmental area, none of the eleven cases filed as of September 1997 has prompted changes in national practices (although some of the charges were determined to be unfounded). Five cases were terminated for various reasons, five are under review by the NAAEC secretariat, and one case has advanced to the stage of compiling a factual record of the dispute (which presages the convening of a dispute panel). The process is deliberately convoluted. Mexico and Canada staunchly resisted the incorporation of dispute provisions in the side pacts and only accepted a compromise process that was long on consultation and short on adjudication. Interestingly, more cases have been filed regarding US and Canadian practices than Mexican practices.
In sum, have these pacts been worth the effort? The answer is clearly “yes”: cooperation among the three countries on labor and environmental matters is greater than would have been likely in the absence of the joint initiatives and dispute panels established by the pact.
Could the United States have negotiated more detailed obligations regarding labor and environmental practices and enforcement procedures? and should we try to do more in these areas in future trade negotiations? These questions lie at the heart of the current debate over fast track authority, and they deserve a direct answer.
Labor and environmental issues should be an integral part of our bilateral relations, and should be pursued in a variety of fora. WTO agreements already cover several important problems in these areas but WTO members explicitly rejected the further negotiation of labor issues at the Singapore Ministerial meeting of December 1996. Few of our trading partners supported US proposals to initiate formal WTO consultations on labor standards, preferring instead to handle those matters in the International Labor Organization. In August 1997, our Latin American neighbors reiterated their opposition to the inclusion of labor issues in trade pacts, including the prospective Free Trade Area of the Americas. The reason for the almost universal reluctance to address labor issues in trade talks is straightforward: other countries recognize that the main reason US labor and environmental groups want international obligations in their areas blended into trade pacts is to take advantage of the trade sanctions available under the WTO's dispute settlement process, and they thus regard the US policy as a transparent threat of new US protectionism.
As a practical matter, the United States has little leverage to convince our trading partners to incorporate enforceable obligations on labor and environmental issues in new trade pacts. We basically want other countries to accept US norms or standards, so are not suggesting changes in US policies that would benefit our trading partners (unless we abstained from using trade measures to coerce foreign compliance with what we regard as appropriate policies and practices). But, as we noted earlier, our market is already generally open to their products, so the only threat we can make is to reduce that access — that is, impose new protectionist measures.
Whether they realize it or not, when US policymakers insist that new trade pacts include enforceable obligations on labor and environmental issues, they are effectively advocating either the introduction of new US protectionism or the withdrawal from new trade negotiations. For the myriad reasons cited at the start of this statement, that would be a bad result for the United States.
The NAFTA today is still a work in progress, but it has already produced tangible gains for the United States. Most importantly, it has helped support the crucial economic and political reforms in Mexico that have strengthened democracy and stability in our southern neighbor. The United States benefits when our neighbors prosper and deepen their democratic institutions. Second, the NAFTA has lowered barriers to an important and growing market of 90 million people, creating new trade opportunities for US exporting firms that on average pay wages about 15 percent higher than paid by non-exporting US manufacturing firms. In that regard, NAFTA clearly contributes to the improvement of our long-term problem of stagnant per capita incomes and wage levels. To be sure, US imports from Mexico have increased sharply, but this growth derived far more from the strong performance of the US economy and the peso depreciation than from the minimal US trade reforms required by NAFTA.
In sum, NAFTA has been a good deal for the United States, promoting crucial US economic and geopolitical objectives. We should build on that success and reassert our leadership role in the world trading system by launching new trade negotiations both regionally and multilaterally. To do so requires new fast track authority, and we urge the Congress and the Administration to work closely together to provide new negotiating authority as soon as possible.
Summary: NAFTA Results
1. Data cover the period January 1994 through June 1997.
2. This earlier period (1990-1993) covered the initial implementation of the US-Canada Free Trade Agreement, which led to faster growth in US-Canada trade despite the sharp recession in Canada in the early 1990s.
3 Of course, if one focused primarily on the impact of the NAFTA tariff cuts, the more sizable cuts by Mexico would lead to a positive trade balance effect for the United States.
4. Our colleague John Williamson raised concerns about the peso overvaluation in testimony before the House Committee on Small Business on May 20, 1993. Hufbauer and Schott also suggested that Mexico would have to take steps to avoid the further real appreciation of the peso in their widely-cited study, NAFTA: An Assessment, published by the Institute for International Economics in 1993.
5. New FDI comprises new investment plus reinvested earnings of US companies in Mexico.
6. The general dispute settlement provisions of NAFTA chapter 20 and the innovative procedures of the chapter ll regarding investment disputes have been used much more sparingly than chapter 19.