Continental Drift: NAFTA and Its Aftershocks

The North American Free Trade Agreement (NAFTA) is a symbol of Mexico's incorporation into the U.S. economy as a low-wage manufacturing center. This economic integration will drive down wages, employment, and living standards, while rolling back environmental regulations in the United States as well as in Mexico. But NAFTA is only a symbol: the low-wage approach to economic integration continues apace with or without NAFTA. The treaty was mainly designed not to promote economic changes (which were happening anyway) but to improve the domestic political fortunes of Presidents Bush and Salinas. However, many Americans appropriately concerned about declining labor and environmetal standards that result from integration with Mexico have fallen into the trap of opposing NAFTA while giving less attention to the underlying economic strategies followed by both nations regardless of the treaty's formalization.

Consider the trading practices that have developed even without NAFTA. General Motors is now Mexico's largest employer. Ford, Chrysler, and GM already own a total of 64 plants in Mexico. Half a million Mexicans now work in some 2,000 border plants (maquiladoras) that export to the United States, paying minimal tariffs, low wages, and little attention to environmental standards. While not all these maquila jobs would otherwise move north of the border in the absence of these advantages (some might move to Asia or the Caribbean instead), clearly the maquila sector has displaced hundreds of thousands of American assembly manufacturing jobs. Defenders of maquila manufacturing assert that these plants create markets for American-made machine tools and unassembled components. But an American machine shop gains nothing when the assembly plant it supplies moves from Wisconsin to Mexico.

In 1989, American-owned transnational corporations added 48,000 new manufacturing jobs at home, while they added 46,000 in Mexico. This investment pattern cannot be attributed to the prospect of NAFTA--Mexico's President Salinas did not make his surprise announcement seeking a free trade agreement until 1990. American manufacturers already have sufficient incentives, without NAFTA, to create all the jobs they can in Mexico. In Mexico's export sector, labor productivity often equals that in the United States, while real wages have declined nearly 50 percent since 1982. Following this dramatic decline, Mexican wages are now substantially below Asian rates: by 1989, the average Mexican manufacturing wage had fallen to 16 percent of the U.S. wage, while the Singapore wage was 22 percent and the Korean and Taiwanese wages 25 percent of U.S. rates.

Mexico's unions, controlled by the ruling party, help attract export firms with solidarity pacts that, since 1988, have continued to hold wage increases below the inflation rate. A recently announced pact continues this policy through 1993. This summer, a Mexican court upheld Volkswagen's discharge of 14,000 employees who repudiated a contract negotiated by the union's government-sponsored leadership. In 1987, Ford Motor Company renounced its union contract, discharged 3,400 Mexican employees, and then rehired them with a new union contract at reduced wages and benefits. When workers protested in support of dissident union leadership, gunmen hired by the official union shot workers at random inside the Ford factory. Then 1,000 state police entered the plant to enforce the new contract.

Mexico's industrial environmental regulations, on the books, are comparable with those in the United States, but are rarely enforced. Despite President Salinas's pledge that Mexico would stop attracting investment with lax enforcement, the General Accounting Office last year examined six new pollution-prone American factories in the maquila sector and reported that none had obtained permits, required by Mexican law prior to operations, certifying that waste disposal and pollution control systems are in place. NAFTA won't change this. NAFTA contains bold provisions barring import to the United States of a product that is dangerous to consumers' health or safety, like a child's crib with sharp edges, but contains no provisions by which Americans can challenge the import of a crib that is unsafe for the Mexican workers who spray-paint it. Nor can we challenge a Mexican manufacturing process that depletes the ozone layer by failing to spray in sealed chambers (like those required in the U.S.). That's why a flood of Southern California furniture plants moved to Mexico when air quality regulations were tightened in Los Angeles and why they expect to continue manufacturing in Mexico after the treaty is in place.

Subscribe to The American Prospect

In a remarkable leap of logic, NAFTA opponents habitually warn of NAFTA's dangers by pointing to environmental devastation and job loss that has already taken place. Since intentions to negotiate NAFTA were announced, hardly a month has passed without highly publicized pilgrimages of trade unionists, environmentalists, and Democratic politicians to the Texas or California borders. There they highlight sewage and industrial effluent in the Rio Grande, children working in factories that make clothes or electronics for export, or Mexican workers' families storing drinking water in toxic waste barrels.

The pilgrims' conclusion is invariably: reject a free trade agreement with Mexico! It is surely a case of rushing to shut the barn door after the horses have fled. Whether NAFTA is approved or rejected, the movement of industrial employment from the U.S. to Mexico will continue. The trade-weighted average American tariff on imports from Mexico is already too low (4 percent) to impede this relocation.

Lurking behind the misleading debates about NAFTA is a much broader issue. Will we pursue an approach to global commerce and Third World development that defends wages in the industrialized countries while insisting that increased productivity in the Third World translate into increased living standards so that consumption can grow along with production? Or will we permit liberalization, privatization, and deregulated competition to slash wages and hence living standards in both regions? The former path requires debt relief, expanded development lending, protection for "infant industries," increased labor and environmental regulation, domestic content rules, and above all increased purchasing power in the Third World. The latter, laissez-faire path, relies primarily on private capital and the natural tendency of transnational corporations to seek the lowest possible labor costs. NAFTA needs to be understood as a symptom of this broader problem, not as the problem itself.


NAFTA opponents concede that job dislocation will continue with or without a free trade pact. They claim, however, that particular provisions of the agreement will accelerate job loss in vulnerable sectors--apparel, for example. The U.S. and Mexico have an exhaustive agreement under the Multi-Fibre Arrangement (MFA), with numerical quotas established for every conceivable fabric and garment. The U.S. apparel industry has lost 300,000 jobs since 1978, due to labor-saving technology and Third World imports. Apparel imports from Mexico's maquilas have shot up from $360 million in 1986 to $844 million in 1991. If, critics charge, NAFTA supersedes the MFA and quotas are lifted, a greater flood of imports will destroy many thousand more American jobs. In a 1987 Labor Department survey, 32 percent of apparel workers who lost their jobs in plant closings or layoffs in the prior four years had still not found new jobs. Surely, NAFTA opponents claim, apparel is a case where the barn door can be slammed before it's too late.

Not so. MFA quotas for Mexican-assembled apparel are already meaningless. As the U.S. has warmed to Mexico's economic reforms, quotas for Mexican apparel have risen dramatically. While the MFA sets a 6 percent annual growth norm for developing country exports, in 1988 the U.S. granted Mexico a 42 percent increase in apparel quota. In 1990, many garment types were granted additional 25 percent growth while other categorical limits were abolished. A subsequent pact negotiated in 1991 contained more big jumps. Our MFA agreements with Mexico also increase quotas whenever Mexico approaches its ceiling--making the entire quota system almost meaningless. In 1992, the Congressional Office of Technology Assessment surveyed apparel exporters in Mexico and reported that "with one exception, every apparel maker interviewed claimed that they could get all of the quota they needed."

Yet even so, Mexico cannot count on apparel exports as a secure source of foreign exchange. Mexico's free trade strategy can work only if Mexico has privileged access to U.S. markets. But as "free trade" expands, Mexico's competition with other nations for U.S. investment will become cutthroat. Fighting with other nations for shares of a finite U.S. market, Mexico could lose.

An indiscriminate worldwide free trade crusade along with a practice of using American jobs as foreign policy pawns have led to quota concessions for other nations, undermining advantages Mexico expects from favored status. President Bush thanked Turkey (the world's ninth biggest apparel exporter) for its help in the Persian Gulf by doubling Turkey's U.S. quota in men's trousers to 4 million pairs a year. To encourage the breakup of communism in Eastern Europe, President Bush increased apparel quotas for Poland, Hungary, and Czechoslovakia. We supported the Aquino government in the Philippines by granting liberal apparel quotas, and that nation's $500 million apparel export industry is now equal in size to Mexico's.

When the Reagan administration worried about leftist politics in Jamaica, the U.S. promulgated a Caribbean Basin Initiative that virtually eliminated quotas for Caribbean apparel. From 1986 to 1991, while American apparel imports from Mexico grew by 221 percent, imports from other Caribbean countries grew by 245 percent and are now four times the level of apparel imports from Mexico. Guatemala's export sector, for example, has 70,000 apparel workers, up from 2,000 only eight years ago, and is now one-and-a-half times the size of Mexico's maquila zone. In 1986, Guatemala, like Mexico, began an economic stabilization program designed to attract exporters. Like Mexico, Guatemala used a combination of currency devaluation and wage restraint to reduce labor costs in export industries. As a result, Guatemala's apparel wages are now 20 cents an hour, about one-fourth of Mexico's maquila apparel wage. Van Heusen is the largest U.S. employer in Guatemala, employing 1,000 workers and exporting 20,000 shirts a month. Guatemala's apparel exports to the U.S. rose from $22 million to $350 million from 1986 to 1991.

China is already the world's largest textile exporter, with capacity to overwhelm every other nation's garment industry. While human rights issues have lately tempered quota increases, China's apparel exports to the U.S. grew in the mid-1980s by 19 percent annually. With or without NAFTA, Mexico's garment plants will increasingly compete with Chinese exporters having lower wages and superior productivity. In the context of worldwide trade liberalization, it is difficult to argue that NAFTA's lifting Mexican MFA quotas will be a major cause of continued devastation in America's apparel industry.


NAFTA really can't make trade any freer than it already is. NAFTA opponents suggest, however, that the treaty will stimulate greater U.S. job loss by making investment in Mexico more secure. Yet while the agreement contains additional security for investment (guaranteeing, for example, the right to repatriate profits) along with dispute resolution procedures that give U.S. capitalists the property rights of their dreams (without the tort system they abhor), it is not credible to argue that these new legal protections will stimulate vast new investment in Mexico. American firms have shown no sign of withholding investment in Mexico out of fear that property rights are not secure.

U.S. companies considering foreign investment are generally more wary of exchange-rate risk than of political risk. But NAFTA establishes no currency union or even coordinated monetary policy between the U.S. and Mexico, so the treaty itself will not protect the dollar value of peso investments. The peso has fallen from 25 to the dollar in 1981 to 3,100 to the dollar today, and more devaluations are probable. Nonetheless, U.S. direct investment in Mexico continues to grow at the rate of $2.5 billion a year. This flood of cross-border investment continues because the primary purpose of manufacturing in Mexico is export back to the United States. Devaluation may affect the paper value of Mexican assets on American firms' books, but it also cheapens the cost of labor and other Mexican inputs, enhancing export profits. U.S. investors' lack of concern about future peso devaluation and their failure to press for monetary coordination in NAFTA is perhaps the best evidence that, contrary to public propaganda, these firms have little interest in selling to the Mexican market. They don't care about the possible low value of sales in pesos. Only sales in dollars matter, and a cheap peso helps that.

In truth, while Mexico is unlikely under any circumstances to get the $15 billion annually in foreign capital that rapid growth requires, no treaty is needed to cement Mexican hospitality to U.S. investment. Without NAFTA, Mexico has privatized banks, state enterprises, and agricultural land. It now permits 100 percent foreign ownership in even basic industries like cement, glass, iron, steel, and cellulose. It has eliminated most domestic content rules and abolished subsidies to once-protected industries. Indeed, the Mexican government, in its desperate search for foreign capital, reinterprets restrictions too politically charged for NAFTA to repeal. For example, NAFTA fails to remove Mexico's constitutional prohibitions on foreign ownership of oil reserves and on "risk contracts" (where royalty payments are contingent on successful exploration). Yet the Mexican government now permits risk contracts in all but name, labeling them "performance contracts" where payments (not called royalties) to foreign firms reflect the rate at which oil is pumped.

In short, if Mexico wants (or needs) to open its economy to foreign investment, it does not require U.S. treaty consent to do so. And if those who are less hospitable to U.S. influence someday come to power in Mexico, NAFTA can be repudiated with little difficulty. The U.S. army, the International Monetary Fund (IMF), the World Bank, and international capital markets offer much greater disincentives to radical policy change than does a paper treaty.

If NAFTA's opponents exaggerate its costs, NAFTA proponents also engage in flights of fancy in touting the treaty's benefits. They may acknowledge that relocation of American manufacturing to Mexico will continue. But they also allege that the job losses will be offset by gains in U.S. export industries that will now find new market opportunities with Mexico's "80 million consumers."

Leading the pack of economists arguing a case for NAFTA have been MIT's Rudiger Dornbusch and the University of Texas's Sidney Weintraub. Each argues that as Mexican industrialization proceeds, not only will U.S. consumers benefit from the lower cost specialization ("comparative advantage") of trade between the two nations, but competition in Mexico for scarce skilled and semi-skilled manufacturing labor will increase demand for labor, thus raising wages. Economies of scale (from serving U.S. markets) and greater capital investment in technology will lead to increased Mexican productivity and also higher wages. Then Mexican purchasing power will grow, along with Mexicans' appetite for U.S. imports. As Dornbusch testified before a congressional committee, "Mexican wages will rise, the internal market will expand, and a good part of the extra demand will spill over to the U.S. in orders for our export industries."

It is true that U.S. firms are selling more toothpaste, diapers, cameras, clothing, and other consumer products in Mexico since Mexican import restrictions were dropped and tariffs lowered. Yet expectations that such growth can continue are unrealistic. The Mexican middle class is tiny; relatively few Mexicans can shop for any but the most basic U.S. consumer products. Unless incomes of Mexican workers increase dramatically, the Mexican consumer market will provide little long-term benefit to most U.S. exporters.

Mexican incomes are unlikely to grow substantially. Mexico has so many unemployed and underemployed workers that labor markets will not tighten to cause a general increase in wage levels. Mexico needs a 6.5 percent annual growth rate just to provide employment for one million new work force entrants each year, making no dent in the present labor surplus. Yet even optimistic observers hope for less growth than that. The Brady Plan to "solve" Mexico's debt crisis assumed future Mexican growth of 4 percent, assuring that the labor surplus would grow. (In 1991, Mexican GDP grew by 3.6 percent, and 1992 estimated growth was 2.5 percent.)

With greater economic integration, the surplus labor pool could increase rather than shrink. As part of its economic liberalization, Mexico has deregulated agriculture and permitted sale of agricultural lands, withdrawn subsidies from subsistence farmers, and opened Mexico's food and feed markets to greater competition. Many Mexican peasants will be unable to compete with the United States's highly mechanized grain exports once Mexican agricultural protection is fully dismantled. Rural workers will flood industrial labor markets and depress wages, offsetting any tendencies of labor markets to tighten from increased investment.

It is also unlikely that heightened Mexican industrial productivity will lead to higher wages. Mexican automobile engine plants, for example, operate at 80 percent of U.S. productivity rates, yet wages in these plants are only 6 percent of U.S. wage rates. As Walter Russell Mead has documented, wages and productivity diverge in export plants throughout the Third World. In Asia, as nations like Bangladesh and Thailand have hosted increased transnational investment, wages declined while productivity increased. In South Korea, it took nationwide riots in 1987, toppling the Park-Chun dictatorship, before wages of highly productive Korean workers began to inch up. In Mexico, real manufacturing wages fell by 24 percent while industrial productivity increased by 28 percent from 1980 to 1989.

In the United States in the 1980s, real wages also fell as industrial productivity climbed, while in Japan and most European industrialized nations, productivity and wages rose together. In the real world, the relationship between productivity and wages is far looser than in textbooks; wages are affected more by a government's fiscal, monetary, and labor market policies (Mexico's "solidarity pacts," for example).


Experience plainly suggests that while Mexican productivity may grow, wages and purchasing power may not rise, and U.S. exporters may not find the consumer markets they've been promised in Mexico. Nonetheless, a widely cited claim for U.S. export growth from NAFTA came in 1992 from Gary Hufbauer and Jeffrey Schott of the free-trade-oriented Institute for International Economics, who predicted that the treaty will create precisely 130,000 U.S. jobs.

Their claim, subsequently raised to 175,000, was picked up by economists (and, of course, the Bush administration) noted for an utter inability to make accurate predictions about other economic phenomena. The very forecasters who can't predict future growth, unemployment, interest or inflation rates claim to know precisely how many jobs the integration of Mexico and the United States will create.

Free trade promotion has been supported by computer modeling, an academic fad in which complex equations describing observed economic behaviors are fed into computers and "run" with alternate policies. A variety of models floated around Washington as NAFTA was being negotiated. The Hufbauer-Schott version made several dubious assumptions--for example, that dollars invested in Mexico would not otherwise have stayed in the U.S. but would have gone to Asia or the Caribbean. Thus Hufbauer and Schott calculated the new jobs in Mexico that would result from additional U.S. investment there, but declined to calculate American job losses from a corresponding investment decline here. Their assumption may have been valid in some cases (certainly Mexico's openness has attracted some investment from Asia), but when they assumed it would be true in all cases, they guaranteed that their computer analysis would predict exactly the free trade benefits they hoped for. Garbage in, garbage out.

In 1987, the U.S. and Canada negotiated a free trade treaty. Similar computer modeling helped win Canadian support. For example, two Ontario professors, David Cox and Richard Harris, reasoned that Canadian companies would be more efficient with unimpeded access to the larger U.S. market. By adding equations for these "economies of scale," the professors computed that free trade would make Canada's GDP grow by exactly 8.74 percent.

Other computers generated different estimates. A Canadian government commission reported that benefits to Canada's economy would range from 3 percent to 8 percent, depending on the computer model used. Then-U.S. trade negotiator Clayton Yeutter announced that Canadian duty-free exports to the U.S. would grow to $19 billion annually, while U.S. exports to Canada would only be $13.5 billion.

The result was somewhat different. In the first two-and-a-half years of Canada-U.S. free trade, Canada's economy actually shrank by 0.5 percent. Canada lost 90,000 jobs and unemployment jumped from 7.8 percent to 10.5 percent. Cox and Harris's computer predicted employment gains of 62 percent in Canada's transportation equipment (auto) industry during this period; instead, employment declined by 15 percent. Their prediction of 262 percent growth in Canada's clothing industry became, instead, a 33 percent job loss.

These forecasting failures, along with economists' inability to predict other trends, don't faze the modelers who now promote benefits of free trade with Mexico. They have a ready explanation: Canada's job losses result not from free trade but from the higher value of the Canadian dollar and the Canadian recession. Had it not been for these, they aver, free trade would have produced the predicted growth.

They may be right--or wrong. No economy behaves like these models. Currency swings, business cycles, and changes in fiscal or labor policies will overwhelm the isolated effects of a single policy like a trade pact with Mexico.

As imports from developing nations expanded in the 1980s, American manufacturing jobs were lost and real incomes declined. A combination of policies contributed--not only low-wage imports but failure to invest in education, public works, and worker training; a preference for financial speculation over manufacturing; hostility to unions; budget deficits requiring high interest rates and dollar values; and World Bank and IMF policies to depress Third World purchasing power so that these nations' bank debts could more easily be repaid.

Mexican wages have fallen in the last decade, despite employment and productivity growth, because Mexico's need for export earnings to service a $100 billion foreign debt required ever lower wages to attract investors to export plants. Mexican policy could not allow wages to increase, for rising incomes would supply domestic purchasing power to compete with exporters for factory production. If, on the other hand, Mexican incomes could have grown, Mexicans would not only have purchased more U.S. goods but more products from their own plants, leaving less available for export to the U.S. and thus less job destruction here. All told, the U.S. suffered a decline of 1.1 million jobs as a result of lost export sales from developing-nation debt crises of the 1980s.

With or without NAFTA, manufacturing jobs will continue to move south, while Mexican incomes will remain too low to permit many purchases of American consumer goods. We now have a temporary trade surplus with Mexico because of the pent-up demand of Mexico's tiny middle class, because the peso's devaluation has not proceeded rapidly enough to erase an artificially cheap dollar value (a situation that cannot continue for long), and because we export so much industrial machinery to Mexico. But the peso is not strong enough to support a permanent trade deficit, and once our industrial machinery is in place in Mexico, it will likely send a permanent stream of consumer products north, creating a long-term trade deficit for the United States. Nor is there any assurance that as Mexico industrializes it will keep purchasing capital and intermediate goods disproportionately from the United States. The lesson of East Asian development is that nations that establish strong export assembly sectors can use their assembly plants as customer bases for new capital and intermediate goods sectors.

Repudiating NAFTA would simply permit job losses to continue, along with declines in living standards in both countries. So would simple ratification of the treaty. The best opportunity to reverse the trend is to use NAFTA as a lever to negotiate other agreements to reform the Mexican economy so its wages can rise. Mexican incomes are presently too low, not only because debt service obligations compete with consumption but because Mexico's income distribution is so grossly inequitable.


Conventional wisdom has it that Mexico, like much of Latin America, adopted its current economic strategy--openness to foreign investment, wage repression, and export orientation--after its previous experiment with import substitution industrialization (ISI) failed. In response to this failure, Mexico is said to have determined to copy the path to success paved by the Asian "tigers"--Korea, Taiwan, Singapore, and Hong Kong--which experienced rapid growth by using export earnings for purchase of capital goods, debt service, and then for consumer imports to support a rising standard of living.

This conventional wisdom is wrong on two counts. First, Mexico's import substitution strategy did not fail; in many respects it was highly successful. And second, the development strategy followed by Mexico since 1982 bears not the slightest resemblance to that of the Asian tigers.

Free trade proponent Rudiger Dornbusch acknowledges that "in the two decades from the mid-1950s to the mid-1970s, Mexico was a model of financial stability and economic development." Real annual growth in minimum wages was 5.5 percent, and per capita income grew at an annual average rate of 3.6 percent. Indeed, even in the years leading up to the crisis, 1973-1981, when Mexico's borrowing was excessive, the country's per capita income grew at an average annual rate of 2.6 percent, despite rapid population growth.

Much of Mexico's current success as an American low-wage manufacturing center is rooted in now-abandoned import substitution policies. A 1962 policy, for example, prohibited sale in Mexico of autos that did not have Mexican-made engines and transmissions. In 1969, the policy was modified to permit Mexican firms to purchase foreign inputs, provided each firm maintained a positive trade balance--earning the foreign exchange with its own exports.

In August 1981, a year before the beginning of the end of ISI strategies, the Mexican government adopted a similar development plan for the computer industry. Investors in the new computer industry were to receive preferential interest rates and pricing of energy inputs, along with restrictions on competing imports. In return, firms wanting to enter the computer industry would be required to negotiate domestic content agreements with the government. This development strategy has, of course, now been abandoned in keeping with Mexican liberalization, but Mexico is now positioned to send low-wage auto and computer exports to the United States.

Contrary to conventional wisdom, the import substitution strategies Mexico followed prior to liberalization were rather like those of "successful" Asian tigers. South Korea, for example, eschewed market policies that American opinion now assumes Korea pioneered as a path to prosperity for Mexico.

For example, Mexico recently privatized state-owned banks. Yet when Korea's industrial push began in 1961, dictator Park Chung Hee nationalized banks. As Alice Amsden has shown, with control of credit, the government decided which industries to promote, which firms could enter a market, which products they could make, and in what quantity. Government-owned banks gave free credit to protected industries. Such subsidies override market "discipline" and are now anathema to officials in Washington and Mexico City.

Korea's industrialization limited "entrepreneurial risk" that President Salinas now cites as his economy's salvation. Imports competing with Korean goods were banned or subject to exorbitant tariffs. Korea required firms to meet export targets while potential domestic competitors were denied operating licenses.

Today Mexican planners nervously wait to see if free market reforms induce wealthy Mexicans to bring their riches home. If there is any economic explanation for Mexico's pursuit of NAFTA, it is the hope of giving Mexican capitalists confidence that economic ties to the U.S. are irrevocable and that therefore Mexico is a good place for Mexicans to invest. Despite 10 years of economic "reform," more than $50 billion is still held abroad by Mexicans seeking safer returns in places like Tokyo or New York. With some Western interest rates (in Germany, for example) remaining at 10 percent or higher, it is difficult to imagine why Mexican speculative capital would return home if free market reforms were the sole inducement. Korea, on the other hand, has had a different approach to "flight capital": unauthorized transfer of Korean wealth abroad was punishable by death.

A central tenet of Mexico's free market strategy is movement toward an exchange rate that reflects the peso's true purchasing power. Korea, in contrast, manipulated its currency in violation of free market norms, combining currency manipulation with capital controls, credit allocation, domestic content requirements, and selective tariffs--all to promote industrialization and export-led growth.

In 1968, Korea had a per capita income of $180, compared to Mexico's $580. By 1990, Mexico's per capita income was $2,490, while Korea's was $5,400. With market reforms, Mexico's economy is growing at a 2 to 4 percent annual rate. But Korea grew at a 9 percent rate in the 1970s; its new five-year plan is based on 7 percent real growth for 1992 to 1996.


Why, then, did Mexico abandon its import substitution development model, which, though imperfect, shepherded the nation through three decades of rapid growth? The answer, of course, is that the nation's top economic strategists since 1982 have not been in Mexico City but in Washington. Mexico's strategy has been dictated by the U.S. Treasury, the IMF, and the World Bank, which have imposed on Mexico a version of trickle-down, supply-side Reaganomics more extreme than that implemented or even contemplated at home.

Consider: In our recent presidential campaign, a common observation was that no politician could expect election if, even in the face of unsustainable public debt, he or she prescribed the kind of "pain" that Paul Tsongas, Warren Rudman, or Ross Perot demanded--50 cents a gallon tax increases and reduction of social security and Medicare benefits. Yet what was demanded of Mexico? Real minimum wages in urban employment were cut by 47 percent from 1980 to 1989, and average manufacturing wages were slashed by 24 percent. Subsidies to food producers were cut; consumer price subsidies for tortillas, beans, cooking oil, bread, and eggs were reduced in the lowest income urban areas while elsewhere these subsidies were entirely eliminated. Where government-run businesses (parastatals) could not be sold to private investors, the government closed them with minimal adjustment assistance for employees. Government's withdrawal of services was so rapid that a fiscal deficit of 9.4 percent of GNP in 1982 became a surplus of 3.3 percent a year later.

By 1984, subsidies had declined by 43 percent from a year earlier. By 1985, the cost of a basic food basket had risen to 50 percent of the minimum wage, up from 30 percent in 1982. The proportion of households below the "poverty line" went from 37 percent in 1981 to 49 percent in 1989. The goal, however, was achieved: Lowered wages helped Mexico's manufactured exports jump from $1.8 billion in 1982 to $3.3 billion in 1983. By 1989, manufactured exports were worth $11.6 billion.

These policies could not have been implemented by a government that is "democratic" in any meaningful sense of the word. Notwithstanding a ritual of Mexican "elections," it is only because Mexico is an unabashed autocracy that its leaders could impose such pain without provoking widespread rebellion.

The lever Washington used to demand these sacrifices was debt. Mexico was forced to abandon its preferred import substitution strategy because its debt became unmanageable in 1982. There were three reasons for Mexico's virtual default, two of which were beyond its control.

One was the Iran-Iraq war, in which both of those nations (encouraged by the United States) attempted to finance their efforts by flooding world markets with oil, pumped at rates in excess of OPEC quotas. The result was plummeting world oil prices, going from $37 a barrel in 1981 to $8 by 1986. Mexico, not unreasonably at the time, anticipated that prices would stay high, enabling the nation to service its foreign debts as well as maintain an increasing level of domestic investment. With the collapse of world oil prices, Mexico had little alternative but to threaten its international creditors with default. A second factor was the ill-conceived monetary policy adopted by the Federal Reserve's Paul Volcker in 1979. Volcker's strategy was to break American domestic inflation with exorbitant interest rates. Many Mexican loans were pegged to prevailing rates.

The third cause of Mexico's economic crisis was that some of its foreign loans had been inefficiently invested. This factor is heavily emphasized in conventional accounts of the "failure" of import substitution investment, which assert that foreign loans were siphoned off by corrupt officials for either personal gain or private use. These accounts, however, are radically overstated. Foreign capital may not always have been wisely invested--for example, Mexico has not had to add new refinery capacity since 1982, suggesting that there was overinvestment in the petroleum industry before the crisis. But corruption was an insignificant factor. In the absence of exogenous shocks (plummeting oil prices and skyrocketing interest rates), even a nation with Mexico's corruption could have continued to service its debt and grow.

In 1982, Mexican finance minister Jesus Silva Herzog came to Washington threatening default and begging for relief. Instead, he was told to slash living standards and reduce wages--so that taxes could be used for bank payments, not tortilla subsidies, and so that Mexico could attract privately financed export industries to earn dollars for debt service. Mexico then signed an IMF "stabilization" plan, in which banks lent Mexico funds equal to about half the interest due back to the banks in the next two years, in return for Mexican commitments to reorient the nation to free market policies. Still unable to service its debt (though interest payments to foreign creditors rose from 66 percent of Mexico's budget in 1982 to 79 percent by 1988), Mexico sought further relief. In 1985, Treasury Secretary James Baker tried to persuade banks to lend Mexico (and other Third World debtors) more funds for debt repayment. Banks refused; the Baker plan flopped. In 1989, Baker's successor, Nicholas Brady, asked banks to soften repayment terms (for example, by lowering interest rates or forgiving 35 percent of the debt's value), but the Brady plan also won only slight relief. Meanwhile, the IMF and World Bank (both heavily influenced by the U.S.) continued to negotiate agreements with Mexico providing bridge loans and other temporary aid to tide the country over, on condition that Mexico continue to follow free market policies of depressed wages, reduced consumer subsidies, fewer social services, less government participation in the economy, slashed tariffs (Mexico's trade-weighted average tariff on U.S. imports is now only 11 percent) and elimination of other "non-tariff barriers" to foreign penetration of the Mexican economy. In 1986, Mexico joined the GATT, formalizing its renunciation of import substitution development.

Yet despite these "reforms," Mexico's foreign debt (now approximately $120 billion) is higher than it was ten years ago when the crisis began. As a share (in constant dollars) of GDP, the debt is only slightly lower than it was then. The Brady Plan, promoted as the solution to debt crisis, requires $58 billion of Mexican interest payments (and another $18 billion of amortization) during the plan's first five years--1989 to 1994. Despite historically low interest rates, Mexico devoted $9.5 billion in 1990 (28 percent of its export earnings) to foreign interest payments, with another $3 billion going for retirement of principal.

It was after the failure of the Herzog, Baker, and Brady negotiations that President Salinas announced a willingness to sign a "free trade" agreement to symbolize determination to attract U.S. investors to Mexico's low-wage work force. As we have seen, however, whether NAFTA is ratified or rejected will have little impact on the flow of investment from the United States to Mexico. This flow is expanding but will never be sufficient to correct Mexico's capital shortage. Because the structure of trade and investment depresses living standards of working people in both nations, it is very much in the U.S. interest to redefine our international economic policies as well as negotiate anew with Mexico. Whether these negotiations are characterized as separate from NAFTA, parallel to NAFTA, or a renegotiation of NAFTA itself is immaterial and can be left to the convenience of presidential public relations experts. Tactically, the best approach may be Bill Clinton's announcement that he would hold up NAFTA's enabling legislation until new agreements are negotiated. But the essential goal must be a joint development project that can succeed in raising, not depressing, Mexican incomes. Rising Mexican incomes are essential to creating a domestic market for Mexican industrialization, developing reciprocal markets for U.S. exports, and relieving downward pressure on American wages that stems from low-wage Mexican competition.


A 2,000-mile common border does not permit us indifference to Mexican development. Undocumented immigration, common pollution, and the danger of political instability so close to Texas, New Mexico, Arizona, and California require us to manage the relationship, not ignore it. Management has to focus on increasing Mexican growth and incomes. NAFTA won't do it. What will?

In thinking about new negotiations with Mexico, we should keep in mind that there is little meaning to distinctions we make between domestic and international economic policies. As the failure of computer modeling illustrates, it is nearly impossible to distinguish effects of mutually reinforcing policies. For example, we need not worry about industrial relocation to Mexico if our domestic policies aim for low unemployment, rising wages, adequate adjustment and relocation assistance, high investment in leading-edge industries, and regional development planning. Conversely, it helps little to have programs that encourage investments in, say, technology centers in Midwestern cities if the flight of manufacturing jobs to Mexico leaves in its wake unemployment in the rural southeast or in immigrant communities in New York, Miami, and Los Angeles.

Of course we need expanded job training and adjustment assistance for American workers displaced by relocation of manufacturing to Mexico. Yet while this must be part of a new North American strategy, more is needed, especially a re-evaluation of American economic strategy towards developing nations in general and Mexico in particular.

Debt relief and reform of IMF and World Bank development policies must be cornerstones of a new economic relationship. The present downward spiral in Mexico began with the debt crisis; it is foolish to try to end that spiral by addressing every problem except the one that sparked Mexico's decline. If Mexico could spend an additional $10 billion annually on American manufactures instead of debt service, American jobs would be created while Mexico gained both industrial machinery and consumer products--a "win-win" transaction. Every billion dollars earned by Mexico's exports, if then spent on purchases of American consumer and capital goods, would create about 30,000 U.S. jobs. If Mexico's earnings are transferred instead to international bankers, many of those jobs are lost.

There's no principle at risk in forgiving Mexican debt obligations. The Bush administration repaid Egypt for its support of Desert Storm by organizing Western European creditors to forgive $20 billion of Egyptian debt. To encourage Eastern European reformers, we arranged to wipe out $17 billion of Poland's foreign debt. At various times in the last two years, we have pressed allies to let Russia suspend payment on its $65 billion foreign debt -- hoping it could use the cash instead to buy Kansas wheat. The U.S. blows hot and cold on the issue of Russian debt relief, but the issue is never whether debt relief is an improper reward for Soviet profligacy; rather, the issue is solely whether policymakers believe controls are in place to assure that relief will be properly expended (on import of American products and ideology).

Our domestic bankruptcy law permits troubled firms to use revenue for operating expenses rather than onerous debt repayment. Firms whose problems are far less serious than Mexico's use these procedures. Mexico slashed wage rates in half to maintain regular interest payments in the 1980s, but no U.S. bankruptcy judge would require so drastic a wage cut before permitting a firm to ignore creditors. Even complete debt forgiveness would be relatively inexpensive, costing the U.S. much less than the over $100 billion that Mexico now owes. U.S. banks now sell Mexican debt for about 45 cents on the dollar -- guaranteed "Brady bonds" sell for 60 cents. Even if Western nations compensated banks fully (at market rates) for Mexican debt forgiveness, it would cost a total of only about $40 billion (comparable to Western expenses for Polish and Egyptian concessions).

While budget constraints limit our willingness to purchase Mexican debt, the burden could be offset by requiring Mexico (dollar for dollar) to buy American machinery and intermediate goods with funds otherwise designated for debt service. This requirement has historically been attached to foreign aid funds, beginning with the Marshall Plan for Western Europe; it is today precluded not by logic but by ideology. Such a proviso could provide stimulus to underutilized American capacity, resulting in increased income tax revenues and reduced transfer payments (welfare and unemployment), offsetting, to some degree, the budgetary cost.

With the election of Bill Clinton as president, John Major's Great Britain becomes the only remaining industrialized nation that pretends to subscribe to the fanciful economic nostrums we blithely impose on the developing world. And it will be recalled that in September 1992, when Britain was faced with the orthodoxy of repressing its economy to service its debts, it instead dropped out of the European exchange-rate mechanism, a luxury not permitted Mexico (whose debts are denominated in dollars).

Redefinition of IMF and World Bank policies are required so that these institutions, in negotiations with Mexico and other developing nations, cease demanding adoption of laissez-faire policies more extreme than any industrialized nation (including the U.S.) would dare contemplate. Once ideologically motivated officials came to dominate the IMF and World Bank during the Reagan-Bush years, these international institutions have demanded that developing nations adhere to an orthodoxy much more extreme than the institutions' official policies.

The IMF and World Bank need a new development model and practice, adapted from the successful experiences of nations like Korea, Japan, Germany, and the United States, not from textbook laissez-faire theories. Whether in negotiations with the United States or its surrogates, the IMF and World Bank, Mexico should be encouraged to return to some policies that worked in its import substitution period, while avoiding the period's excesses. The nation should be permitted and encouraged to specialize not just in those industries that pay wages too low to remain in the United States.

With an industrial strategy that complements our own, some Mexican industries could be protected or even subsidized. The Korean model is appropriate to avoid inefficiencies that resulted from excessive protectionism of the ISI model; a few designated industries could be expected to earn the protection of their domestic markets by becoming sufficiently competitive to compete internationally.

A continental development strategy should include development assistance from the United States to Mexico. While this aid will be limited by American budgetary pressures, a commitment to common North American development requires a reorientation of American foreign development assistance priorities.

Nowhere in the world has free trade been attempted between economies at such disparate levels of development. When West Germany incorporated East Germany in an economic union two years ago, the German government made direct transfers of wealth to each East German citizen, along with an artificially low exchange rate for East German marks--to avoid uncontrolled immigration of workers from East to West and uncontrolled migration of production from West to East.

The European Community, when incorporating Spain, Portugal, and Greece, recognized that if economic integration were not to pull its prosperous members down to the level of its poorer members, development aid had to be granted from EC coffers to less developed regions. The European Community treaty commits its member nations to "reducing disparities between the various regions and backwardness of the least favored regions." The European Investment Bank and Regional Development Fund support infrastructure projects in the poorer member nations, building roads, upgrading communications, underwriting industrial development. Consequently, wages in the poorer nations have climbed, and wage-driven relocation of manufacturing from northern to southern European nations has been minimal.

In contrast, total official development assistance to Mexico in 1990, from all industrial nations and international agencies, totaled only $140 million, less than one-tenth of one percent of Mexico's GNP--or an average of $1.60 per Mexican citizen.

Mexico reasonably claims that its abysmal environmental record results mainly from lack of resources. In 1980 and 1981, Mexico used 80 percent of its government budget for programs other than interest payments. From 1983 to 1988, the average was 50 percent. Mexico cannot be expected to increase its environmental enforcement budget and infrastructure investments at the same time that the U.S. Treasury, the IMF, and the World Bank are enforcing debt repayment schemes that require Mexico to deregulate its economy, reduce its public sector, and devote whatever resources are available to debt service.


On the other hand, the burden does not belong entirely with the United States and its development agencies. Mexico should be expected to do more on its own behalf, and international assistance cannot be the only source of funds for development.

Tax Reform. Domestic investment funds can be raised in Mexico itself by tax reform--by an increase in taxation, by improving the efficiency of tax collection, and by correcting the excessive regressivity of the Mexican tax system.

Of the nations that the World Bank classifies as "upper middle income," Mexico's tax collections are among the lowest. It collects 14 percent of its GNP in taxes, compared, for example, with Portugal, which collects 35 percent; Venezuela, 17 percent; Uruguay, 27 percent. Mexico has no capital gains tax (except on real estate), not even on speculative profits made on the Mexico City stock exchange in the last three years. Less than 20 percent of Mexico's businesses are registered with the tax collection authorities. Prior to the privatizations of the last decade, most government revenue came not from taxes but from the earnings of parastatal enterprises. Now this source of revenue has disappeared as even profitable state-owned businesses like banks, airlines, and utilities have been sold. Pemex, the petroleum monopoly that remains state-owned, contributed fully one-third of the government's revenue in 1991, yet free-market theorists recommend that Pemex be privatized as well.

Mexican political economist Jorge Castaneda lists six elements of tax reform needed if welfare and public investment goals are to be established: corporate taxes must be increased; taxes on wealth levied; capital gains taxes on financial markets established; collection improved; income taxes enforced on independent professionals and the upper-middle class; and taxation of assets held abroad.

The latter, which would go far toward bringing flight capital home, requires negotiation of effective tax treaties with the United States and other finance-center nations like Britain, Japan, and Germany. Such cooperation is essential for continental development and costs the industrialized nations very little. While Mexican flight capital is of enormous consequence to Mexico, Western nations could easily forego the additional capital they obtain as tax havens for wealthy Latin Americans.

Health and Education Reinvestment. To become a profitable market for U.S. exports, as well as a successful exporter of higher value-added goods, Mexico must be permitted and encouraged by U.S. and international development agencies to reinvest in the education and health of its youth. Mexico's productivity is doomed to lag, offsetting much of the value of increased foreign investment, because the structural adjustment policies the U.S. imposed on the nation required a reduction in spending for education, health, and nutrition. Education spending declined by a third, from 3.8 percent of GDP in 1982 to 2.8 percent in 1983, and it has remained at this lower level. Health spending went from 3.7 to 3.0 percent in the same year, and has also stayed down since. Mexican children who were preschoolers when the economic crisis began in 1982 will be entering the work force within the next five years. They are less nourished and less educated than the previous generation of Mexican workers and so will be less productive. (They will also, incidentally, perform more poorly as immigrants in American schools.)

Minimum Wage Increase. American and international agency negotiators should include in their continental development strategy a gradual increase in the Mexican minimum wage, currently about $.60 an hour. Remember again that this minimum is half its 1982 level and was reduced by Mexican policy at the behest of international creditors, in a bold transfer of wealth from American exporters (and their work forces) to creditor nation bankers.

But policies to make Mexican exports cheap in America also make American exports too expensive in Mexico. American exports to Mexico declined dramatically during the "structural adjustment" of the 1980s. Mexico's annual imports of goods and services dropped from $21 billion to $8 billion from 1982 to 1983. These failed policies to cheapen Mexican export wages should be reversed. The ultimate goal should be a common U.S.-Mexican minimum wage, at the U.S. level. There are several benchmarks by which intermediate goals could be established. A comparison of the wage share of GDP in the U.S. (55 percent) and in Mexico (15 percent) suggests that Mexican wages could triple. If the standard is workers' earnings as a percentage of manufacturing value-added (20 percent in Mexico, 35 percent in the U.S.), an immediate wage increase of 75 percent would be in order. Labor productivity in export industries (roughly equal to U.S. rates) suggests that an even greater increase could be appropriate.

Regardless of what standard is adopted, the rate of increase in the Mexican wage should be rapid enough to provide Mexican workers with significantly increased purchasing power, appropriate to their productivity levels, yet slow enough to continue to entice new investment. Immediate increases could be considerable. Asian nations attract U.S. runaway shops with wage rates much higher than the Mexican standard.

There is precedent for careful incorporation of low-wage areas into the American market. In Puerto Rico, for example, industry wage boards established minimum wages for specific industries. Nonetheless, most minima were reasonably close to the mainland standard. In 1960, the mainland minimum wage was $1 an hour; of 98 different industry minima on the island, 73 were $.70 or higher and only 6 were less than $.50. Gradually, the differential was reduced, until in 1981 parity with the mainland was reached at the $3.35 level. The differential was maintained so that labor-intensive firms would have incentives to invest in Puerto Rico. But the differential was small enough that it did not create a hemorrhage of U.S. firms relocating solely to take advantage of low wages. The strategy worked. From 1960 to 1989, Puerto Rico's real per capita personal income has grown steadily, at an average annual rate of 3.7 percent. Since island industries have been required to pay the full mainland minimum wage in 1981, per capita growth has averaged 3.3 percent. During this same period of Mexican crisis and liberalization, Mexican per capita income growth has declined by an average rate of over one percent a year.

Environmental Standards and Privileged Market Access. A continental development strategy should also include harmonized environmental standards and enforcement procedures. While Mexico's environmental regulations are already close to those of the United States, enforcement has often been nonexistent. The Bush administration, in answering environmental critics, insisted that NAFTA would itself lead to a clean environment by generating wealth that could be dedicated to a national clean-up. But Bush also promised to develop a bi-national environmental plan parallel to but not part of NAFTA. It was an empty promise. The bi-national plan committed U.S. contributions of only $380 million and Mexican contributions of $340 million over the next three years. Responsible estimates of the costs of cleaning up pollution already created by reckless relocation of U.S. manufacturing along the Mexican border range from $18 billion to $50 billion.

The obvious flaw in an approach that aims at upward harmonization of Mexican and U.S. labor and environmental standards is that Mexico competes with other Asian and Latin American developing economies for U.S. investment, based on offers of the lowest labor and regulatory costs. If a Mexican development strategy expects wages and regulatory costs to increase, a likely result would be the flight of international capital to other Third World nations that are not under similar requirements to raise labor and regulatory costs.

The Bush administration's goal was to negotiate worldwide "free trade" with lower wage nations. After NAFTA was ratified, free trade agreements with Chile and Eastern Europe would follow. All the while, the administration pursued an "Enterprise for the Americas" with Latin America that, building on NAFTA and the Caribbean Basin Initiative, offered unrestricted access to U.S. markets in return for Latin nations' agreements to privatize economies, reduce the size of government, and adopt low-wage export platform strategies.

As we have seen in the apparel industry, however, if all developing nations are given privileged access to the U.S. market, none can have it. Clearly, the U.S. cannot be the market of (first and) last resort for every developing economy. A continental development strategy with Mexico can work only if the goal of worldwide free trade is abandoned, and Mexico, perhaps along with the Caribbean Basin nations, is given privileged access to the U.S. market. There are, as suggested earlier, good security reasons for singling out Mexico and the Caribbean in this fashion. Not only need we be more concerned with political instability here than we are with instability elsewhere, but there is ultimately no way of slowing the rate of undocumented migration from Mexico and the Caribbean save with a meaningful development strategy for the sending areas. This was also, of course, one of the rationales for NAFTA, but free trade, with its prospects for furthering the depression of Mexico's urban living standards and disrupting traditional peasant economies in rural areas, will be more likely to stimulate immigration from Mexico than to staunch it.

Neither wage nor environmental harmonization can work unless the North American market is protected from undermining by other nations. Raising Mexico's minimum wage to U.S. levels will do little good if transnational investors can then relocate Mexican plants to the Philippines or Sri Lanka. To prevent such runaways, we need new tariff walls to protect the developing prosperity of North America.

Our free trade area should accept only those third country imports manufactured with wage and environmental standards comparable to those required here. Practically speaking, the protection of harmonization could be accomplished by levy of a "social tariff," taxing third nation imports an amount equal to the difference between the wage paid and that which would be required if an acceptable minimum was enforced, calculated on a similar basis to the new minimum in Mexico. A similar tariff could be levied to represent the costs of environmental responsibility.

Critics may reject U.S. requirements for increased investment in health, education, labor, wage, and environmental standards as an affront to Mexico's sovereignty. But Mexico's present policies of disinvestment in these areas are all consequences of earlier demands by the U.S. and its development agencies. Objection to new policies out of "respect for neutrality" in Mexican domestic policy is a smokescreen. The present NAFTA draft contains numerous Mexican accommodations to demands for changes in domestic policy--Mexico's legal (commercial dispute resolution) system, its intellectual property laws, and its investment rules. Reforms of Mexico's domestic fiscal, welfare, and labor market policies are now more sorely needed.


If the United States negotiates anew with Mexico, we could well succeed in raising Mexico's living standards and ability to purchase U.S. exports. President Salinas, we can assume, is not opposed to higher wages in principle; incomes have been slashed because of bankers' demands, not because a low-wage development strategy was freely chosen. If a new American administration can abolish supply-side economics at home, it could well find a friend in Mexico's leader, now freed from external demands that reactionary remedies be imposed as the policy of his nation as well.

While NAFTA opponents miss the mark with exaggerated predictions that the treaty will accelerate U.S. investment in Mexico (which already proceeds at breakneck speed), they correctly assert that by codifying laissez faire in a treaty, NAFTA would make i

You may also like