NAFTA at Seven

Download PDF

Para leer este informe en español, clic aquí .

Jump to a specific report:

The impact of NAFTA on wages and incomes in Mexico

by Carlos Salas, La Red de Investigadores y Sindicalistas Para Estudios Laborales (RISEL)

Mexico is much changed in the seven years since NAFTA was implemented in 1994. Although Mexico now has a large trade surplus with the U.S., Mexico has also developed a large and growing overall trade deficit with the rest of the world. In fact, Mexico’s net imports from the rest of the world now substantially exceed its net exports to the United States. Official unemployment levels in Mexico are lower now than before NAFTA, but this decline in the official rate simply reflects the absence of unemployment insurance in Mexico. In fact, underemployment and work in low-pay, low-productivity jobs (e.g., unpaid work in family enterprises) actually has grown rapidly since the early 1990s. Furthermore, the normal process of rural-to-urban migration that is typical of developing economies has reversed since the adoption of NAFTA. The rural share of the population increased slightly between 1991 and 1997, as living and working conditions in the cities deteriorated.

Between 1991 and 1998, the share of workers in salaried1 jobs with benefits fell sharply in Mexico. The compensation of the remaining self-employed workers, who include unpaid family workers as well as small business owners, was well above those of the salaried sector in 1991. By 1998, the incomes of salaried workers had fallen 25%, while those of the self-employed had declined 40%. At that point, the average income of the self-employed was substantially lower than that of the salaried labor force. This reflects the growth of low-income employment such as street vending and unpaid family work (for example, in shops and restaurants). After seven years, NAFTA has not delivered the promised benefits to workers in Mexico, and few if any of the agreement’s stated goals has been attained.

Running hard but falling behind
Despite a quick recovery from the 1995 peso crisis and a peak 7% gross domestic product (GDP) growth rate in 2000 (Figure 2-A), NAFTA still has failed to help most workers in Mexico.

Although foreign direct investment (FDI) in Mexico has continued to grow, total investment actually declined between 1994 and 1999 (Table 2-1). The only types of investment that have grown since 1994 are the maquiladora industries, reinvested profits, and stock market investments. Speculative flows of financial capital into stock market investments, in particular, increased, but overall investment in Mexico fell between 1994 and 1999. These inflows help explain the rapid-and perhaps unsustainable-growth in prices on the Mexican stock market in the late 1990s.

Manufacturing exports, as officially reported, have improved rapidly since NAFTA took effect. From 1995 to 1999, these exports grew at an annual rate of 16%, due almost exclusively to “value added” exports in Maquiladora production.2 The total value of these exports increased 19.7% annually, as the average value added of products exported from Mexico decreased (relative to their overall value). However, maquiladora exports contain a substantial share of imported components from the U.S. and other countries, reducing the net benefits of these exports to the Mexican economy and its development. Thus, the export growth and the foreign trade performance of the Mexican economy look better on paper than in reality. But even these benefits disappear when total imports are considered. Total manufacturing imports from the U.S. and the rest of the world grew 18.5% per year between 1995 and 1999, a fact that explains Mexico’s rapidly growing overall foreign trade deficit in this period. In the long run, this process of economic growth with expanding foreign trade deficits could lead to another major currency crisis similar to the one that occurred in 1994 (Blecker 1996).

How strong was employment growth between 1995 and 1999?
Total employment in Mexico grew from 33.9 to 39.1 million jobs over the 1995-99 period (3.7% annually), according to officially reported data (INEGI 1995 and 1999). But these data must be used with some caution, because the sample used for the National Employment Survey changed in 1998. Comparing the 1998 and 1999 data provides a more realistic rate of employment growth. Total employment reported for 1998 was 38.6 million jobs, resulting in an actual rate of growth in 1999 of only 1.2%.

Total employment in Mexico must grow 2.5% in order to fulfill the yearly demand for 1.2 million new jobs (CONAPO 2000). Since GDP grew 3.7% in 1999, these data suggest that GDP should grow at about 7% annually to achieve a sustained 2.5% growth rate in employment and avoid rising unemployment. Yet Mexico has achieved a 7% rate of growth in only one year (2000) in the past decade.

Agricultural employment trends
Agricultural activities are still the most important single source of employment in Mexico. In 1999, agricultural employment (8.2 million workers) accounted for 21% of the total labor force. For the past 10 years, agricultural employment has hovered around eight million. This stability suggests that NAFTA did not lead to a major surge in migration trends from the countryside to the cities. Over the long term, steady growth in corn imports has helped stimulate a general migration process. This doesn’t mean that most campesinos-traditional corn growers-will decide to remain in rural areas in the future. A major increase in rural-to-urban migration process could start sometime in the next decade if corn prices keep falling and no other sources of income generation are provided to campesinos.

Interstate migration patterns, however, remained unchanged in this period, which reinforces the idea that most corn growers still are cultivating their land plots (Nadal 2000). What is more remarkable is that there was a slight increase in the share of the population living in rural areas between 1991 and 1997.

Migration is another major alternative for Mexican workers who cannot find good jobs. Northbound international migration has increased all through the 1990s, and the number of permanent migrants, in particular, has been on the rise (Tuirán 2000). The geographical origin of these migrants is very diverse, as many of the new migrants come from regions with no previous history of migration flows to the United States. At the same time, more migrants are coming from urban areas and are better educated, which provides a
stark contrast with the traditional image of rural, illiterate migrants. This shift in migration patterns is another significant indicator of a decline in the supply of good jobs in Mexico, even for well-educated workers.

Nonagricultural employment trends
Despite the increase in migration to the north, it appears that the rapid growth in Mexico’s potential labor supply has been matched by a seemingly impressive rate of growth in nonagricultural occupations. On average, the number of employed has increased by slightly less than 1.3 million per year. The unemployment rate has, therefore, not shown any upward trend and has remained at a low level, with only short-term fluctuations as economic activity has varied. Unemployment in urban areas remained at very low levels of 2-3% between 1987 and 1999. The only major exception was in 1995, corresponding with the peso crisis, when overall unemployment surged above 6% and reached almost 14% for teenagers. Overall, however, unemployment rates have been low by international standards, rarely exceeding 8% even for young people.

It would be misleading, however, to conclude from such steadily low unemployment measures that Mexico has avoided the difficulties that most market economies face of providing enough jobs. There are, in fact, clear explanations as to why the official unemployment measures are so low. Mexico’s labor force statistics count someone as employed if that person has worked at least one hour in the week before the survey takes place, following ILO standards (Hussmans et al. 1990). Under this definition, a person is counted as employed regardless of whether the person only works half time for no pay in a family business or works full time in a modern manufacturing plant. But Mexico’s low rate of open unemployment is not a statistical distortion-it primarily reflects the workings of a different labor market structure.3 Given that a large proportion of the population has no capacity for saving, and that there is no unemployment insurance, open unemployment in Mexico is, to paraphrase Gunnar Myrdal, a luxury few can afford.

Not surprisingly, unemployment rates are clearly higher for the most educated, who have higher incomes and greater savings capacity. But for those at the bottom of the wage scale, being “employed” does not guarantee an adequate standard of living, especially given the broad definition of what constitutes employment in the Mexican labor market.. Deteriorating labor market conditions in Mexico have thus resulted in a decline in the quality of jobs rather than increases in unemployment rates, as might be the case in other economies with effective social safety nets.

The inability of the Mexican economy to create good quality jobs reflects two primary trends: a virtual halt in the process of urbanization, and the large and growing share of workers holding low-productivity, low-paying jobs in urban areas. While the economy was reducing the relative number of workers occupied in agricultural activities between 1970 and 1990, the past decade witnessed a reversal of this trend. Modernization of the economy, crudely defined as a declining share of rural and agricultural activities in the economy, was stagnant during most of these years. In spite of deficiencies in sampling and comparability of national employment surveys, the available data clearly show that, in the 1990s, the share of the labor force in less-urbanized areas and the share engaged in agricultural activities have both remained roughly constant at around 50% and 20-25%, respectively (INEGI-ENE 1991 and 1997).4

The deteriorating labor market conditions in the most important cities are reflected by an increase in the proportion of workers who are either self-employed or work in businesses with less than five employees. These low productivity jobs usually offer low pay. The share of the self-employed in total employment between 1987 and 1999 is shown in Table 2-2. The most important trend in urban employment in Mexico is the growth in service sector employment, as is happening in most economies. Rapid employment growth (and production) in trade and service industries poses two problems for the Mexican economy. Unlike service sector jobs in developed economies, Mexico’s non-industrial activities do not include a strong and dynamic sector of high value-added services. Even in the case of the growing employment in financial service activities-a process clearly associated with privatization and new investments-a large part of this expansion can be attributed to continued protection and the absence of regulation (but not to the spread of highly competitive, world-class services). Thus, wages and productivity in these industries are low, by world standards.

Mexico’s service sector growth is characterized by extreme heterogeneity, running the gamut from single-person activities such as street vending to stock market brokering using the latest technologies and facilities. Furthermore, unlike the newly industrialized countries of Asia, Mexico’s adoption of an economic strategy that relies on sustained growth in manufacturing exports-facilitated by its close geographic proximity to the U.S.-has not increased the share of manufacturing employment in the economy.

As a result of these trends, the structure of the urban labor landscape has changed in important ways in the 1990s. The most important shift is the diminishing share of regular salaried occupations in total employment. Between 1991 and 1998, the share of salaried employees in total employment decreased by 13 percentage points, from 73.9% to 61.2%. The resulting void was filled by either informal employment activities or simple unemployment. The share of self-employed workers increased by 50%, and the share of workers having unpaid positions as their first occupation doubled (as shown in Table 2-2).

Older salaried workers apparently switched to self-employed occupations, while younger workers were even less fortunate, moving into unpaid positions or becoming unemployed in this period. The share of workers aged 12 to 14 that had unpaid positions jumped from 40% to 60% between 1991 and 1998. The reduction in salaried occupations has cut across most industries. However, there are significant differences between those industries. A high proportion of nonsalaried jobs in the labor market indicates a backward production structure. For example, retail trade, food, transportation, and accommodations have among the largest shares of self-employed and unpaid workers. The high rate of nonsalaried jobs in these industries reflects the large presence of small firms and relative simplicity of the tasks performed by the workers in those jobs. Comparing 1991 and 1998, the loss of salaried occupations was almost completely offset by the growth in self-employed and unpaid workers.

Traditional manufacturing activities show the sharpest relative reductions in the shares of salaried workers, with the modern manufacturing, construction, trade, and communications industries being the next largest losers of salaried jobs. These changes are partially explained by the effects of the 1995 crisis upon traditional types of production in manufacturing and other industries, but they also reflect long-term segmentation trends in labor markets.

The growing share of self-employed workers means that people moved to deteriorating labor occupations. Wages decreased by 27% between 1991 and 1998, while overall hourly income from labor decreased 40%. Thus, labor income for the self-employed was cut in half in this period (Table 2-3).

Average self-employment incomes fell from 17% above salaried worker incomes in 1991 to 19% below in 1998. In real terms, the relative well-being of the self-employed did not decrease as much as suggested by income comparisons,
but this is far from reassuring. Reductions in real wages do not entirely explain the deterioration of labor conditions. During the same period, the share of salaried workers receiving fringe benefits also fell systematically, as shown in Table 2-4.

The maquiladora sector’s employment performance contrasts significantly with that of Mexico’s other large manufacturing plants. The maquiladora sector began as a program for in-bond processing plants, primarily making goods for re-export in Mexico’s northern border cities. These plants employed an industrially inexperienced labor force to perform simple assembly tasks in traditional manufacturing. Maquiladoras have evolved over time, but they have remained largely isolated from the rest of the Mexican economy. Maquiladora employment grew rapidly, from 60,000 workers in 1975 to 420,000 in 1990. The pace of job creation slowed somewhat in the early 1990s, but it accelerated again after the 1994-95 peso devaluation. In 2000, maquiladora industries employed 1.3 million workers, concentrated mostly in electrical and electronic products, auto parts, and apparel and textiles. Employment in those activities accounts for more than 80% of total manufacturing employment in the maquiladora plants (Table 2-5).

Maquiladoras have helped offset weak job creation in other domestic manufacturing industries,5 accounting for about 13% of total manufacturing employment in 1995 and almost 16% in 1999. Maquiladora plants contributed 35% of all new manufacturing employment between 1995 and 1999. Most of the remaining jobs created during this period were in small non-maquiladora plants (Alarcón and Zepeda 1997, 1998).

The 1995 recession’s impact on maquiladora plants was relatively mild, which is not surprising given their nearly complete specialization in export production.6 Maquiladora job growth accelerated between 1995 and 1997, adding 150,000 positions each year during this three-year period. This sum far exceeds the 60,000 jobs added each year between 1987 and 1989. Employment in maquiladora apparel production rose rapidly from 1995 to 1997, a fact closely linked to the relaxation of the Multifibre Agreement quotas after the implementation of NAFTA (O’Day 1997). Maquiladora jobs in electronics and auto part exports expanded as well, in keeping with those industries’ global strategies (Carrillo and Gonzalez 1999).

There were also important regional changes as maquiladora plants were established in cities far from the Mexico-U.S. border. Between 1994 and 1999, the proportion of maquiladora workers in non-border locations increased from 16% to 22% as maquiladora production began shifting southward to sites such as Jalisco, the State of Mexico, Mexico City, Puebla, and Yucatan. Apparel-producing maquiladora plants, in contrast, moved to areas where compliance with labor laws is low, such as the states of Puebla and Morelos.

Declining wages
Most directly employed workers have seen a steady erosion of their wages in the 1990s. In the last decade, the minimum wage in Mexico lost almost 50% of its purchasing power. The minimum wage is set each year through a process that includes consultations between official unions, employers, and the federal government. Currently the minimum wage is just a reference point for the wage bargaining process of wage and salary workers, and wages are usually set above this level in negotiated contracts.

Labor income in industries whose wage bargaining processes are under federal supervision (the so-called salarios contractuales or contractual wages) lost almost more than 21% of their purchasing power between 1993 (the year before NAFTA took effect) and 1999 (Table 2-6). Manufacturing wages also declined by almost 21% in this period, and the purchasing power of the minimum wage fell 17.9% through 1999. The decline in real wages since NAFTA took effect helps explain the decline in labor incomes (see Table 2-3).

The decline in real wages and the lack of access to stable, well-paid jobs are critical problems confronting Mexico’s workforce. While NAFTA has benefited a few sectors of the economy, mostly maquiladora industries and the very wealthy, it has also increased inequality and reduced incomes and job quality for the vast majority of workers in Mexico. In many ways (such as the stagnation of the manufacturing share of employment), the entire process of development has been halted, and in some cases it even may have been reversed. NAFTA has created some of the most important challenges for Mexico’s development in the 21st century. The question that remains is whether Mexico can, under NAFTA, restart its stalled development and find a way to redistribute the benefits of the resulting growth.

1. Most workers in Mexico are paid a daily wage, as opposed to the hourly wage paid in the U.S. These workers are referred to in Mexico’s statistics as “salaried,” or, more literally, “waged” employees. These terms refer to several different methods of payment (both daily and piece-work, for example). Thus, a salaried job in Mexico can be very different from one in the U.S.

2. Under U.S. tariff code provisions (HTS 9802), U.S. firms are allowed to send U.S-made inputs abroad for assembly and then return those semi-finished or finished products to the U.S., paying a tariff only on value added abroad.

3. The condition of open unemployment includes “frictional” unemployment, that is, people who know for sure or firmly believe they will be hired in the near future (Rendon y Salas 1993). For further discussion of measures of Mexico’s unemployment see, for example, Fleck and Sorrentino (1994).

4. The share of less-urbanized areas was 52.6% in 1991 and 53.6% in 1997. The share in agriculture was 23.6% in 1991 and 24.1% in 1997 (derived from INEGI- ENE 1991 and 1997).

5. Prior to the 1994-95 economic crisis, domestic-oriented and export-oriented manufacturing plants were approximately even in terms of employment creation. However, the 1994-95 devaluation of the peso gave exporters a boost, and maquiladora employment rose faster than in domestic-oriented producers.

6. In fact, short-term economic or political events appear to have little effect on maquiladora activities.

Alarcón, Diana, and Eduardo Zepeda. 1998. “Employment trends in the Mexican manufacturing sector.” North American Journal of Economics and Finance. Vol. 9, pp. 125-45.

Blecker, Robert. 1996. “NAFTA, the peso crisis, and the contradictions of the Mexican economic growth strategy.” Center for Economic Policy Analysis, Working Paper No 3. New York, N.Y.: New School for Social Research.

Carrillo, Jorge, and Sergio González. 1999. “Empresas automotores alemanas en México. Relaciones cliente-proveedor.” Cuadernos del Trabajo. No. 17. Mexico: Secretaría del Trabajo y Previsión Social.

CONAPO (Consejo Nacional de Población). 2000. La Situación Demográfica de México 2000. México.

Fleck, Susan, and Constance Sorrentino. 1994. “Employment and unemployment in Mexico’s labor force.” Monthly Labor Review. November, pp. 1-31.

Hussmanns, Ralf, Farhad Mehran, and Vijay Verma. 1
990. Surveys of Economically Active Population, Employment, Unemployment, and Underemployment. Geneva: International Labor Office.

INEGI (Instituto Nacional de Estadística, Geografía e Informática). 1991, 1995, 1998, 1999. Encuesta Nacional de Empleo. Aguascalientes, México.

INEGI. 1994, 1996. Encuesta Nacional de Ingresos y Gastos de los Hogares. Aguascalientes, México.

Nadal, Alejandro. 2000. The Environmental & Social Impacts of Economic Liberalization on Corn Production in Mexico. Gland, Switzerland and Oxford, U.K.: World Wide Fund for Nature and Oxfam Great Britain.

O’Day, Paul. 1997. “ATC phase out-A few big winners, long list of losers.” International Fiber Journal. Vol. 12, February.

Rendón, Teresa, and Carlos Salas. 1993. “El empleo en México en los ochenta: tendencias y cambios.” Comercio Exterior. Vol. 43, No.8, pp. 717-30.

Tuirán, Rodolfo, ed. 2000. La Migración México-Estados. Presente y Futuro. Consejo Nacional de Población, México.

[back to top]