Migration, Free Trade and Regional Integration in North America (OECD Proceedings)

Citation preview

OECD

OECD Proceedings

MIGRATION, FREE TRADE AND REGIONAL INTEGRATION IN NORTH AMERICA

OECD PROCEEDINGS

Migration, Free Trade and Regional Integration in North America

PUBLISHER’S NOTE The following texts are published in their original form to permit faster distribution.

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

Pursuant to Article 1 of the Convention signed in Paris on 14th December 1960, and which came into force on 30th September 1961, the Organisation for Economic Co-operation and Development (OECD) shall promote policies designed: – to achieve the highest sustainable economic growth and employment and a rising standard of living in Member countries, while maintaining financial stability, and thus to contribute to the development of the world economy; – to contribute to sound economic expansion in Member as well as non-member countries in the process of economic development; and – to contribute to the expansion of world trade on a multilateral, non-discriminatory basis in accordance with international obligations. The original Member countries of the OECD are Austria, Belgium, Canada, Denmark, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The following countries became Members subsequently through accession at the dates indicated hereafter: Japan (28th April 1964), Finland (28th January 1969), Australia (7th June 1971), New Zealand (29th May 1973), Mexico (18th May 1994), the Czech Republic (21st December 1995), Hungary (7th May 1996), Poland (22nd November 1996) and Korea (12th December 1996). The Commission of the European Communities takes part in the work of the OECD (Article 13 of the OECD Convention).

 OECD 1998 Permission to reproduce a portion of this work for non-commercial purposes or classroom use should be obtained through the Centre fran¸cais d’exploitation du droit de copie (CFC), 20, rue des Grands-Augustins, 75006 Paris, France, Tel. (33-1) 44 07 47 70, Fax (33-1) 46 34 67 19, for every country except the United States. In the United States permission should be obtained through the Copyright Clearance Center, Customer Service, (508)750-8400, 222 Rosewood Drive, Danvers, MA 01923 USA, or CCC Online: http://www.copyright.com/. All other applications for permission to reproduce or translate all or part of this book should be made to OECD Publications, 2, rue Andr´e-Pascal, 75775 Paris Cedex 16, France.

FOREWORD

The links between migration, free trade and regional integration form the general theme of three OECD regional seminars following the proposals debated at the Conference on Migration and International Co-operation held in Madrid in March 1993 at the initiative of the Canadian and Spanish governments. The first seminar, organised with the Austrian authorities, focused on Central and Eastern Europe (Vienna, February 1996), and the second, organised with the Greek authorities, examined the Mediterranean Basin (Athens, November 1996). This publication contains a selection of the proceedings of the third seminar, which took place in Mexico City in January 1998 with the Mexican authorities and the support of Canada and the United States. The Mexico Seminar provided an overview of economic developments and migration patterns in North America in the light of the recent implementation of the North American Free Trade Agreement (NAFTA). It also offered the opportunity to draw parallels with and make distinctions between North American regional integration and the regionalisation process currently unfolding in Europe. The impact of NAFTA in the future, either directly through labour mobility provisions or indirectly through impacts on individual incentive to emigrate was one of the main issues discussed. The role migration plays on economic development, and in particular its effect on the labour market, was a recurring theme and therefore required attention quite separately from the effects of trade agreements. However, the key question is whether liberalisation of migration should proceed concurrently with liberalisation of trade and economic convergence, or whether it should only come into play after the effects of regional integration have resulted in closer economic convergence. This publication was prepared by the OECD Secretariat and is published on the responsibility of the Secretary-General of the OECD.

3

TABLE OF CONTENTS

Introduction .........................................................................................................................................7 PART I.

REGIONAL INTEGRATION IN NORTH AMERICA

Demographic situation, employment and economic performance in North America Cécile Thoreau and Tania Paracini...............................................................................................17 Globalisation, economic growth and the labour market in North America Norma Samaniego .........................................................................................................................49 PART II.

MIGRATION IN POLICY IN NORTH AMERICA: POLICIES AND SECTORAL ECONOMIC IMPACTS

A. Perspectives on migration policy Migration trends and policies in the United States B. Lindsay Lowell ..........................................................................................................................63 Migration trends and policies in Canada Roderic Beaujot .............................................................................................................................81 The impact of migration on economic development in Mexico Gustavo Verduzco and Kurt Unger .............................................................................................103 B. Sectoral economic impacts Migration and the labour market: sectoral and regional effects in the United States Francisco L. Rivera-Batiz ...........................................................................................................121 Migration and the labour market: sectoral and regional effects in Canada Don J. De Voretz and Samuel A. Laryea .....................................................................................135 Migration and the labour market: sectoral and regional effects in Mexico Rodolfo Cruz Piñeiro ..................................................................................................................155

5

PART III. LINKS BETWEEN FREE TRADE, INVESTMENT AND MIGRATION A. Perspectives on the North American Free Trade Agreement NAFTA, foreign direct investment and economic integration: a United States approach Sidney Weintraub ........................................................................................................................169 NAFTA, foreign direct investment and economic integration: a Canadian approach John Kirton ..................................................................................................................................181 NAFTA, foreign direct investment and economic integration: a Mexican approach Fernando de Mateo Venturini .....................................................................................................195 B. Modelling results: NAFTA and North American migration Applied general equilibrium models: the Mexican experience of NAFTA Horacio Sobarzo Fimbres ...........................................................................................................221 Economic effects of NAFTA: employment and migration modelling results Raúl Hinojosa-Ojeda, Robert McCleery and Fernando De Paolis ...........................................235 PART IV. BROADER PERSPECTIVES ON MIGRATION AND REGIONAL INTEGRATION Migration policies in a free trade area: the issue of convergence with the economic integration process Francisco Alba, Jean-Pierre Garson and El Mouhoub Mouhoud ..............................................261 New directions for managing US-Mexican migration Demetrios Papademetriou ..........................................................................................................279 Annex: Selected Comments ............................................................................................................291

6

INTRODUCTION*

The issue of economic development and migration in North America is set against a background of some similarity but also stark contrast with the other world-regions examined in this seminar series. As in relations between countries examined in both the Vienna and Athens seminars, long land borders (or relatively short sea-crossings) separate countries, which in many cases have very different levels of income and living standards. These conditions provide both the basic motivation for migration and present inherent difficulties in providing full control of migration by admission policies, border control and internal checks. In this regard, the same issue emerges: given the difficulties of immigration control what are the prospects of reducing the basic incentives for immigration through free trade and regional economic integration? Although this is a common theme in the seminar series, other issues emerged which are particular to the case of North America. These relate, in part, to the fundamentally different economic, societal and geographic structure of the region. North America’s chief distinction is that the economy of one country dominates the region, an economy that is furthermore the largest in the world. Canada has living standards on a par with those of the United States but has not much more than 10 per cent of its population, and Mexico has the equivalent of over one third of the US population but significantly lower per capita GDP. These imbalances within the region undoubtedly affect the relative benefits and the politics of regional integration. Unlike other world-regions, North America, but especially the United States and Canada, have been almost wholly populated by the process of immigration within relatively recent times. This provides an open attitude towards migration issues and has generated institutional frameworks and policies that deal in a coherent and comprehensive way with the rewards and challenges presented by migration. It also makes attitudes and issues relating to ethnicity, race and nationality somewhat different. The recognition of jus solis as a basis for granting nationality could be seen as symbolic of such differences. The economic integration of North America has been underway for a considerable amount of time, yet the ratification of NAFTA marks a significant landmark in recent history. Therefore, and not surprisingly, the attention of much of the seminar was focused on the debates surrounding this agreement. Although the agreement was implemented in 1994, such issues are not entirely academic, as judgement as to the success of NAFTA will undoubtedly be a factor in determining the course of future regional integration, whether it leads to deepening NAFTA (to include provisions further * Since it was not possible to publish the papers and discussions of the meeting in their entirety, the Secretariat has made a selection of the reports prepared by the international experts and policy makers invited. However, in order to give an account of the different themes touched upon by the experts in their contributions and in the discussions, the publication includes an introduction drafted by the OECD Secretariat. Reports not included in the publication are identified with an asterisk in the general introduction.

7

facilitating movement, etc.), to new regional agreements of a different nature, or alternately to slowing down the integration process. North America has devoted extensive literature to the effects of regional integration and to debates about the impacts of immigration. In this regard, the seminar, of which selected papers appear in this volume, offered rich evidence on which to draw conclusions about the effects of immigration and its links with regional economic integration. Part I of these proceedings presents an overview of the Canadian, Mexican and US economies and an introduction to some of the main themes covered in the seminar. Part II reviews US and Canadian immigration policy and looks at the impacts of immigration on these countries as well as the impact of emigration on the Mexican economy. Part III examines NAFTA and the current thinking about its effects, both present and future. Part IV presents papers which offer broader perspectives, making comparisons and contrasts with European integration, and placing migration policy within a wider context of regional integration in North America. Part I.

Regional Integration in North America

By way of background, the first paper in this section, by Cécile Thoreau and Tania Paracini, provides an overview of the Canadian, US and Mexican economies. They underscore the significant volume of economic links within the region in the form of trade and foreign direct investment. These illustrate the dependence of both Canada and Mexico on the United States in their external economic relations. In addition, the paper presents an overview of international migration within the region, pointing to the potential gains that agreements such as NAFTA could have in reducing some of the underlying incentives to migrate through reductions in economic disparities within the region. Norma Samaniego provides a broader introduction to the issues surrounding regional economic integration and migration in North America. Samaniego stresses the wide disparities between the North American economies, which make regionalisation in this part of the world distinct from other regional experiences, like that of the European Union. She also takes the view that the expected effects of NAFTA were somewhat exaggerated in the light of current evidence. The prospect of economic integration acting as an informal substitute for migration policy (or controls) is seen as unlikely for the immediate future and, it is argued, may never be achieved unless the increased benefits of integration provide for convergence of economic outcomes. These issues are tackled by a number of other papers and address themes that emerged regularly throughout the seminar. Carlos Marichal* introduced another perspective, arguing that a purely economic account of international migration, based on disparities in living standards and incomes, falls short of accurately representing the complexity of the forces that drive real-world migration patterns. This is especially true of migration between Mexico and the United States, where political boundaries post-date traditions of migratory movement and where mature chains of migration, based on habit and social considerations as well as economic motivation, have developed. In this regard, Marichal emphasises the need for public policy in Mexico to evaluate strategies that might be effective in modifying the local conditions that induce strong migration flows, both to other regions within Mexico and to the United States. The complexity of migration flows between the United States and Mexico was also stressed by Jorge Santibañez Romellón,* who examined data on monthly flows across the border around the time of the implementation of Operation Gatekeeper, of California’s approval of Proposition 187, and of the Peso devaluation in late 1994.

8

Part II. Migration in Policy in North America: Policies and Sectoral Economic Impacts Of critical importance to understanding links between regional integration and migration in North America are the migration policies adopted by the United States and Canada. Lindsay Lowell provides an overview of current US policy and describes the characteristics of current immigrant inflows. Special attention is given to the Immigration Act of 1990 which adopted a series of measures, the net result of which was an increase in the minimum admissions targets of about 40 per cent. However, Lowell also indicates that the 1996 Welfare Reform Act and other legislation seeking to reduce the benefits and government services provided to immigrants may dampen future admissions by reducing the incentives to immigrate. Lowell’s paper also points out the large number of unauthorised entries, citing an estimate of 200 000 to 300 000 per year. Such a figure suggests that there are limits to the degree of control that can be provided by admission policies, border controls and internal checks. Migration policy and the characteristics of immigrants to Canada are described in the paper by Roderic Beaujot. Canadian immigration policy, as compared to that of other countries, tends to be more explicit about its broader objectives. Beaujot discusses the visible shift in stance emerging from the 1994 “Strategy for immigration and citizenship”, which suggested that greater emphasis should be placed on admitting economically independent immigrants as a result of concerns about costs and other economic issues. In examining the papers by Lowell and Beaujot it is evident that there are a number of areas in which both Canada and the United States share common ground with regard to policy and the characteristics of their immigrant intakes. Although both countries have established admissions systems which include entry on the basis of skills, education, and employability, the majority of entries in both countries arrive as dependants or as humanitarian cases. This reflects the fact that economic concerns are only one of the factors governing immigration policy, with the implication that even if economic assessment points to a shift in intakes, the policy response may well be diluted once other commitments and concerns are taken into account. In addition, both countries share concerns about immigrant welfare and the potential discrimination faced by some immigrant groups, which are indicated by their low labour force participation, high unemployment, and concentration in low-wage job sectors. Other concerns, largely shared by the native population, address the impact of immigrants on the labour market and the economy as a whole. In this regard, both the United States and Canada appear to be increasing the responsibilities assumed by the sponsors of migrants. The United States has further initiated (often controversial) measures to reduce access to social welfare benefits for some groups of immigrants. And Canada has altered the mix of its immigrant intake to include a greater proportion of economic migrants. That being said, there are no signs that either country is likely to impose significant reductions in total intakes in the future, or for that matter that intakes are likely to be massively increased. Another group of papers examine in greater depth the impact of immigration on the Canadian and US economies. A paper by George Vernez* looks at the diverse and often conflicting evidence about the economic impact of migration on the US economy. Despite this, he suggests that a consensus of opinion is forming on a number of points. First, immigration generally increases aggregate economic output but there are re-distribution effects from labour to capital that are many times those of the overall net effect. In addition, the short-term opportunity cost to natives in the form of higher taxes to cover the costs of public services can be as large, if not larger than the gain in per capita GDP. Finally, he argues that there is also a consensus that the economic effects are highly sensitive to the

9

composition of immigrants, with their level of education being particularly important in determining their contribution to per capita GDP. Francisco L. Rivera-Batiz examines the impact of immigration on the US labour market. He stresses the regional concentration of immigrants in the United States and the concentration of migrants in lower-skilled occupations, implying uneven impacts on the US labour market, especially in the shortrun. In economy-wide terms, he suggests that there has been little impact on wages, not only because the effects on labour supply are judged to be small but also because immigration itself increases the demand for labour. However, he does suggest that the short-run employment opportunities of natives may be affected due to institutional rigidities and employers favouring low-wage immigrant labour, especially where the migrant influx is substantial. Finally, he rejects the argument that inflows of immigrants in the 1980s displaced high-school dropouts from their local labour markets and provides evidence that, if anything, areas of high immigration were associated with the out-migration of the highly-skilled rather than the low-skilled. The paper by Don DeVoretz and Samuel Laryea discusses the labour market impacts of immigration in Canada. They address issues similar to those examined by Rivera-Batiz: concerns about the displacement of workers and the possibility of negative impacts on wages. They present econometric evidence based on Canadian data that reflect the broad conclusions drawn by Rivera-Batiz for the United States, i.e. at the economy-wide level there is little support for displacement effects within Canada and any effect on wages is judged to be minimal. The final issue tackled in the paper is concern about “brain drain” from Canada to the United States. While there is data to support this view, the authors also point out that, in net terms, Canada probably continues to be a recipient of human capital. The potential for brain drain is one issue that Canada may share more with Mexico than with the United States. As Mexico is essentially a country of emigration, the issues surrounding migration are quite different from those of recipient countries in North America. Rodolfo Cruz Piñeiro examines the structural changes that have been taking place in Mexico. He stresses the changing patterns of internal migration and employment structure and their implications for migration to the United States. He also describes the shifts in the structure of employment of Mexicans in the United States. The paper by Gustavo Verduzco and Kurt Unger offers further perspectives on the economic impacts of emigration. They stress the fact that low-skilled agricultural labour from certain regions of Mexico remains the backbone of emigration to the United States, while also noting that there is substantial emigration from urban centres that has not been adequately documented or interpreted. In terms of the economic effects of emigration, Verduzco and Unger focus on the impact of remittances, which are estimated at about US$ 3.7 billion in 1995, a relatively small percentage of total GDP but extremely important for certain regions and localities. In this regard, the paper cites evidence from studies of local communities that provide insights as to variations in the importance of remittances and also the variety of ways in which they are used. The authors suggest that the degree to which remittances are channelled into investment, rather than current consumption, is of crucial importance in determining whether this financial link with the United States represents a form of dependence or a source of independent development and growth. Part III. Links Between Free Trade, Investment and Migration NAFTA is the most recent landmark in the regional integration of North America, consisting of provisions for greater freedom of trade in goods and services and greater mobility of capital. The

10

textbook assessment of such agreements is that they yield aggregate economic benefits, through greater exploitation of comparative advantage and more efficient allocation of investment. However, in the shorter-run, these benefits are sometimes difficult to perceive, not least in those sectors of the economy for which such agreements represent increased competition and perhaps ultimately the need for dramatic structural change. Such fears are highlighted in the papers by Sidney Weintraub and John Kirton who examine NAFTA from US and Canadian perspectives. Weintraub stresses the intense controversy surrounding the approval of NAFTA in the United States. He examines a number of anti-NAFTA claims, including links between NAFTA and the Mexico’s “peso crisis” in 1994 and also fears about job loss in the United States through intensified competition with Mexico. Although the paper argues that such fears are largely unfounded, Weintraub concludes that they continue to dominate US thinking about freetrade and cites the continued difficulties encountered by the Clinton Administration to secure fasttrack authority for future trade negotiations. While the paper by Kirton addresses similar issues, he argues that if fears not unlike those expressed in the United States were initially aroused about NAFTA’s effects on investment and jobs in Canada, controversy over the Agreement has declined over time as little evidence of negative impacts has emerged. The final section of the paper suggests that currently there is a very high level of acceptance of the Agreement amongst the general public, which implies that, if anything, Canada is in a position to consider deepening or enlarging regional agreements about free-trade and capital mobility. While the obstacles to NAFTA in the United States and Canada appear to have often been of a political nature, those in Mexico were arguably more institutional. The paper by Fernando de Mateo discusses the increasing restrictions imposed on trade and FDI in Mexico over much of the twentieth century. Within that context, he traces the unilateral measures that Mexico has taken in recent years towards liberalisation, up to and including NAFTA. He concludes that these represent a significant reversal of policy, from import-substitution towards a more open market. Much of the evidence about of the effects of NAFTA is based on a partial assessment of the economic implications of free-trade and greater mobility of capital, such as evidence of increased trade and investment flows. These inevitably provide a limited perspective, as they do not offer any assessment of how these more immediate effects will reverberate throughout the economy. One means of assessing the broader implications of the NAFTA is through computable general equilibrium (CGE) models, which describe the economy as a set of interacting markets. Using quantitative estimates of the parameters and initial economic conditions (hence “computable”), the full effects of policy initiatives such as free-trade agreements can be simulated by comparing the economic outcomes of scenarios which either relax or enforce different initial conditions. The paper by Horacio Sobarzo Fimbres provides an assessment of general equilibrium models used to evaluate NAFTA. This is followed by the paper of Raúl Hinojosa-Ojeda which reports the results of a CGE model focusing, not only on general economic effects, but also on the effects of migration and migration policy. The results of Hinojosa’s simulations suggest a number of interesting conclusions. As other research has found, the effects of free-trade alone appear to have small impacts on economic outcomes and migration flows, in comparison with a combination of free-trade and increased foreign direct investment. In addition, Hinojosa finds that tightening restrictive policies towards undocumented migrants, such as were implied by Proposition 187 in California, could lead to increased rather than decreased overall migration. One reason for this is that Proposition 187 may

11

force those undocumented workers remaining in California to accept lower wages due to their more precarious position and may thus result in an increased demand for their services. Whilst serving to illustrate the full range of consequences across markets of alternative scenarios of free-trade, investment, and migration, CGE models of the type modelled by Hinojosa fail to capture some important dynamic trends. For example, a paper was presented by Rodolfo Tuirán* which includes estimates of a statistical model establishing the links between the migration rate (net migration per head of the Mexican population) and various economic variables. Applying this model to future economic scenarios, such as high-performance in Mexico and low-performance in the United States (and vice versa), he establishes a likely range of the migration rate for the future. However, when applied to population projections it is found that even under the most optimistic scenario with regard to convergence between the US and Mexican economies, net migration is projected to increase due to growth in the Mexican population. His results also suggest that variation in net migration is unlikely to make a large difference in the size of the Mexican-born population in the United States, due to the significant part played by natural increase in the growth of this population. Part IV. Broader Perspectives on Migration and Regional Integration The paper by Francisco Alba, Jean-Pierre Garson and El-Mouhoub Mouhoud compares the process of regionalisation and migration in the United States with that in Europe. The development of regional integration and migration policy within Europe presents a number of scenarios against which the North American experience can be judged. Some areas of Europe are in a similar position to that of North America, with a high degree of trade but disparities in income and living standards that create pressures making it difficult to control flows of international migrants. Elsewhere in Europe, there are examples of countries entering the European Union that are making the transition towards liberalisation of trade, capital movement and labour movement within the space of a few years. The paper stresses that the short-term impacts of free-trade agreements can lead to increased international migration unless they are accompanied by measures that stimulate investment in the less-developed countries of the region. The authors also suggest that even in the longer-term, the formation of regional blocks should not be viewed as an unqualified success from the point of view of solving large and unwanted imbalances in international migration. They conclude that a liberalisation of migration policy is highly unlikely in the near future, partly because economic integration can largely take place without the free-movement of persons and because the down-side risks of liberalised movement probably serve to keep such a move off the political agenda. Demetrios Papademetriou addresses the political realities of immigration policy given that the prospects of free-movement in North America are limited for at least the immediate future. Within this context, Papademetriou emphasises the need to find areas of active co-operation in migration policy, motivated by the prospect of constructive mutual gain. He also points out the wider benefits of co-operative immigration policy, inasmuch as it promotes a sense of cohesion and dialogue within the region and therefore contributes to an overall sense of political and economic unity, which is necessary for further regional co-operation and development.

12

Conclusion As is evident from this brief summary of the papers presented at the seminar, complex and diverse issues emerge when the subject of regional economic integration is combined with the controversies and debates surrounding many aspects of international migration. In an attempt to summarise the main findings of the seminar, the following conclusions can be drawn. The prospect of a complete removal of the incentives to migrate is extremely slim in the mediumterm. Only in the long-run (and conditional on a sustained period of economic convergence) will the incomes of prospective migrants in Mexico perhaps begin to dissuade large numbers of them from spending at least part of their economically active lives elsewhere. However, this view must be qualified by the unique features of Mexican emigration which, some argue, is not wholly driven by straightforward economic incentive. On this view, migration may continue despite convergence of economic conditions as a result of its unusual degree of maturity and endogeneity. Regional economic integration via trade liberalisation would appear to be somewhat impotent as a catalyst for economic growth and convergence, unless accompanied by technological catch-up, physical infrastructure development and investment in human capital. For the moment, the noticeable effects of NAFTA on the North American economy have been rather less than those predicted by the Agreement’s proponents in the lead-up to the Agreement. It must, however, be borne in mind that given the relatively short time-span in which the Agreement has been in place, such assessments are based on modest amounts of evidence and are subject to a broad range of speculation. What is probably of equal importance is the overall perception of NAFTA, and in this respect, the subsequent role it will play as a reference for discussions about wider or deeper free-trade agreements in the region. Given that over the medium-term economic convergence is unlikely to develop to an extent sufficient to reduce emigration to levels acceptable to the other partners in the Agreement, the prospect of totally free circulation in the region is necessarily remote (this leaving aside the host of other concerns that the anticipation of free circulation is likely to provoke). Therefore, the issue of immigration from Mexico will remain part of the broader policy concerns about immigration, especially for the United States. In this regard, it should be noted that concerns expressed about the impact of immigrants on the economy, the government budget and labour markets appear, at the national level, to be largely unfounded. However, the effects are unevenly distributed -- geographically, sectorally and socially -- which in part explains the intense controversy and the divergent stances taken by different sectors of the community on migration issues. It appears that some of the concerns are being addressed; both Canada and the United States have made changes towards altering the mix of immigrant intakes in an effort to reduce the fiscal burden of immigrants and to increase their overall contribution to these countries’ economies. To summarise, trade liberalisation does not automatically, or immediately, provide the conditions under which migration flows are significantly reduced. Any marked decrease would appear to be dependent on a number of additional elements which together might eventually result in the reduction in the incentive to emigrate. Only then might policy-makers entertain the possibility of freecirculation within the region. As a result, restrictions on movement of some sort are likely to continue into the future. Within this context, the need for balanced migration policies which take account of concerns about the economic consequences and social welfare of both sending and receiving countries will remain on the agenda for the foreseeable future. As economic integration within North America continues, migration policy will have to adjust, both to meet the new challenges and to reap the potential rewards offered by the interaction and exchange of individuals between countries. 13

PART I

REGIONAL INTEGRATION IN NORTH AMERICA

15

DEMOGRAPHIC SITUATION, EMPLOYMENT AND ECONOMIC PERFORMANCE IN NORTH AMERICA

by 1 Cécile Thoreau and Tania Paracini International Migration Unit Directorate for Education, Employment, Labour and Social Affairs, OECD With a market containing 387 million inhabitants and boasting a GDP of US$9 330 billion in 1996, the North American Free Trade zone presents a profile comparable to that of the European Union. However, the three member countries do not share the same socio-economic characteristics: for example, per capita GDP in Canada and the United States is respectively seven and eight times that of Mexico. Moreover, Mexico’s GDP is only 4.3 per cent that of the United States and about half that of Canada (Table 1). Nor are the economic situation and performance of the three countries the same. The first part of this report will present a number of macro-economic (e.g. growth and foreign trade) and demographic differences, along with an analysis of the labour market and wages. The second part will look at regional integration between the three countries: interdependence built up by trade, foreign direct investment, and migration. These factors will be examined together with the impact of NAFTA on employment and migration. 1.

North America: a contrasting economic landscape

An overview of the economic situation The GDP by value and per capita of the United States (in 1996, US$6 146 billion and US$23 377 respectively) places it among the most developed economies. Since 1991, real GDP has steadily increased; growth in 1997 figured at 3.8 per cent. The United States, well endowed with industrial raw materials and energy resources, is also a major producer and exporter of agricultural products, particularly cereals. However, the US economy is focused mainly on services; this sector accounts for almost 60 per cent of GDP and employs 70 per cent of the labour force. In 1991, Canada’s per capita GDP was over US$20 000, placing it in second place, just after the United States, among the countries of the G7. Since then, the gap between the two countries has grown due to the weak growth of Canada’s per capita GDP. Although annual growth in GDP was close to 4 per cent in 1994, it managed only 1.5 per cent in 1996, a year marked by a steep decline in Canadian automobile exports to the United States. Canada is one of the world’s leading mining economies. The agricultural sector is also relatively dynamic, although it employs no more than 3 per cent of the workforce. Certain branches of the 17

manufacturing sector are currently experiencing difficulties: these include textiles, clothing and footwear (in decline since the late 1980s), and the automobile industry which has been suffering from the effects of Japanese competition. The peso crisis of December 1994 interrupted the economic lift-off which had been taking place in Mexico since the end of the 1980s: the consumer price index accelerated sharply in 1995 (34 per cent as against 7.5 per cent in the previous year). As a consequence of the drop off of investment and private consumption, overall domestic demand dropped by almost 14 per cent during the year. A rigorous stabilisation plan was implemented in 1995. The economy has since witnessed a decline in inflation during the second half of 1996 and a revival of domestic demand, which grew by 6.5 per cent. This recovery has been relatively rapid; it took almost eight years to emerge from the 1982 crisis. Undoubtedly, the existence of NAFTA and the relatively healthy budgetary situation since the beginning of the 1990s have encouraged a more rapid return of foreign capital to Mexico after the 1995 crisis. In Mexico, the agricultural sector accounts for only 5 per cent of GDP but employs 20 per cent of the labour force; it is still beset by low levels of productivity. The industrial landscape is very uneven with modern, competitive, export-oriented sectors such as the automobile industry existing alongside less dynamic sectors that produce mostly non-durable consumer goods and services, particularly in the informal sector. Infrastructure, telecommunications, and the transport sector are suffering from a lag in development that the current process of deregulation and an increasing involvement of the private sector are attempting to remedy. Moreover, low tax receipts due to the size of the informal sector, are being 2 further limited by the progressive suppression of customs duties. The tax base, calculated as a percentage of GDP, is one of the lowest of the OECD Member countries. The insufficiency of tax receipts restricts the government’s ability to respond to the essential needs of the population in terms of education, health care, and nutrition for the poorest inhabitants. Features of foreign trade Having been structurally in deficit, the US current account recovered somewhat towards the end of the 1980s. Since 1991, it has undergone a further decline linked to a trade deficit of US$191 billion in 1996 (Chart 1). This deficit, reflecting an insufficient level of savings due to the heavy public sector deficit, can be explained by several factors whose relative importance varies according to the period under consideration. These factors include, inter alia, the declining competitiveness of exports caused by an over-valued dollar between 1980 and 1985. Furthermore, the heightened participation of Asian countries in international trade reduced US market share both domestically and internationally in sectors such as non-durable consumer goods marked by low levels of capital intensity, textiles, and electrical and electronic equipment. This competition prompted the United States to relocate, or even cease, this type of production whilst at the same time re-deploying its industrial production towards goods with high value-added. The Canadian current account registered a US$3 billion surplus in 1996, after having posted deficits since 1988 (Chart 1). Unlike the United States, this recent recovery is attributable to a positive trade balance (US$30 billion in 1996); the improvement in Canada’s current account has also held steady throughout the last ten years compared with the size and variation of that of the United States.

18

Following the process of freeing up trade restrictions which began in 1987, Mexico’s current account registered a deficit which steadily increased until 1994. In 1996, the current account deficit stood at only US$1.9 billion, or 0.6 per cent of GDP (Chart 1). Since 1995, the trade balance has returned to a surplus (almost US$7 billion). This improvement is attributable mainly to the recovery in exports that occurred after depreciation of the peso and to weak domestic demand. Looking at the ratio of foreign trade (imports plus exports of goods and services) to real GDP (Chart 2), the United States is clearly less dependent on foreign trade than either Mexico or Canada. In Canada, this degree of openness has been growing since 1991; in the case of Mexico, this has only been the case since 1994. Turning now to the composition of the trade in goods in North America (Table 2), it is clearly dominated by transport equipment; for all three countries, it represents almost 45 per cent of all trade; manufactured products account for an additional quarter. Concerning the latter, imports take a larger share than exports in the United States and Mexico; whereas in Canada the situation is reversed. The United States exports relatively more food (including drinks and tobacco) and chemical products than it imports. In Canada, non-food raw materials account for over 22 per cent of exports, compared with 7 per cent of imports, reflecting its rich endowment of fuel and mineral resources. In Mexico, food products, and particularly mineral fuels and lubricants, account for a larger share of exports than imports. Finally, following manufactured products (including transport machinery and equipment), chemical products are the principal goods imported by Mexico and Canada; in the United States, this position is occupied by fuels and lubricants. The demographic situation and shifts in employment and income Overall population growth has fallen substantially in Canada and the United States over the last four decades; it is now about 1 per cent, compared with rates in the early 1950s of 2.6 per cent in Canada and 1.7 per cent in the United States. The rate of natural increase has also fallen considerably during the same period, and the proportion of people aged over 65 has grown, particularly in Canada (Table 3). In both countries, especially in Canada, migration is an important contributor to the overall increase in the population. Demographic transition came to Mexico late: during the first half of the 1970s the fertility rate was still above six children. This figure has since fallen sharply (Table 4). People of working age now account for a growing percentage of the total population. Moreover, over the same period, life expectancy has risen by over 10 years. Notwithstanding, the rate of natural increase is still more than 20 per 1 000. The labour force is likely to continue rising over the medium term, while the proportion of people aged 15-24 will probably not stabilise until around the year 2000. Each of the three countries has experienced very contrasting changes in terms of employment, although generally speaking the North American labour market is marked by characteristics that often place it in opposition with the European market. For example, employment is increasing at a higher rate in North America, and less time is spent in a given job on average than in Europe. In fact, the pace of employment growth (1.8 per cent for the entire North American region, compared with 0.1 per cent for the European Union) is one of the highest in the OECD area (Table 5). Employment in Canada and the United States has been rising faster than the labour force for 15 years (this is not true of the European Union as a whole). Unemployment and the frequency of long-term 19

unemployment is generally lower than in European Member countries of the OECD, and the ratio of people unemployed for over 12 months to the overall unemployment figure has even begun to fall in recent years. Since the early 1980s, supply and demand adjustment in the North American labour market has usually been made through wage flexibility. Workers have not been identically affected. The resultant widening of income distribution is still visible in the United States and Mexico, but it appears to have been on the wane in Canada since the early 1990s. A clear split exists in the employment situation in Canada and the United States on the one hand, and Mexico on the other. Developments in employment and wages in Canada and the United States The Canadian and US labour markets are notable for relatively high participation rates (Chart 3). During the last two decades, their labour forces have grown more rapidly than the working-age population, with the increase in female participation rates more than compensating for the fall in male employment. In 1996, the participation rate for both men and women was close to 80 per cent in the United States and 76 per cent in Canada (compared with under 67 per cent in the European Union). In addition to the increase in participation rates, greater net migration rates have also contributed to increasing the labour force. In Canada, the upturn in employment has been slower than in previous cycles, and it has been accompanied by an expansion in “involuntarily” part-time employment rather than in full-time jobs. In the United States, by contrast, job creation has mainly focused on skilled employment. Demand for skilled workers has outstripped supply. As a consequence, the relative real wage levels of this category of worker have witnessed significant gains. In the manufacturing sector, productivity gains have been greatest in high technology industries; this has been particularly apparent in the United States (Table 6). In Canada and the United States, the service sector has created relatively more employment than the manufacturing sector (Table 7). Among providers of business services, temporary employment agencies have undergone sustained expansion, climbing from 400 000 employees at the end of 1982 to 2.1 million in 1995. Unemployment in the United States hovers around 5.5 per cent, a figure below its natural rate. Shifts in the rate of unemployment in Canada follow movements in the United States, where a higher rate (9.6 per cent in 1996) is much closer to levels observed in European Member countries of the OECD. From the early 1990s onwards, increases in the duration of unemployment can explain in part differences between the Canadian and US unemployment rates. Since that time, real wages in the private sector in the United States and Canada have risen slowly, and at a more gentle pace than in Europe. In the United States, real wages earned by full-time employees fell by over 3 per cent between 1985 and 1995, and rose only slightly in Canada (Table 8). That said, these figures mask the fact that wage changes diverge according to worker category. For example, women’s wages have risen while real wages received by men, particularly in the United States, have fallen. Moreover, the level of real wages earned by young people (aged 20-24) has fallen in comparison with those in the 25-54 years cohort. Low wages continue to rise more slowly than high wages. Indeed, the former have dropped by over 7 per cent in ten years in the United States, while the latter grew by 3 per cent; the average wage increase exceeds, then, that of the median wage. While recession in the early 1990s in the United States was accompanied by stabilisation in the wage distribution, the subsequent economic upturn has led to a further widening of the wage gap. However, the most recent data for the United States point both to an acceleration in the growth of real

20

wages for the first time in many decades and to a rise in low wages -- due in a likelihood to the recent fall in the unemployment rate. Developments in employment and wages in Mexico The labour market situation and the slow pace of economic development in Mexico pose a number of problems for employment and income. For about 15 years, the extent and rapidity of both the deregulation and trade liberalisation processes have had a major impact on the labour market. They have not, however, led to an increase in registered unemployment; which only just exceeded 6 per cent during the recession years. Yet, this low unemployment rate should not be taken as an indicator of a well functioning labour market. The absence of an unemployment benefit system, together with the very low income levels of a broad section of the population obliges a large number of people to accept any kind of work simply to survive. In Mexico, the role played by the unemployment rate in labour market adjustment is slight; other mechanisms such as the relative flexibility of labour costs, rural and informal sectors, and international emigration take this shock-absorbing role. A notable feature of the Mexican labour market is sustained and substantial growth in the labour force: over 4 per cent a year during the 1970s and 1980s, and 3 per cent in the early 1990s. This trend results from the marked growth of the working aged population and the remarkable rise in the participation rate. The female participation rate, however, is still one of the lowest in the OECD area (Chart 3). According to official data employment grew in the same proportion as the labour force until the early 1990s. However, due to the stagnation in wage earning employment during the years following the 1982 crisis, a significant proportion of the labour force was absorbed by the informal sector. This phenomenon is also associated with the acceleration in the growth of the service sector during the early 1980s. Employment in agriculture, which still accounts for almost half of all employment, has contracted at a reduced pace since 1980 as a result of the sector’s role in absorbing surplus labour. Meanwhile, a marked contrast has emerged between the most dynamic sectors (particularly in maquiladora industries) and the rest of the economy. This contrast became prominent following the liberalisation of trade in the 1980s, a period during which Mexico increasingly looked abroad because of weak domestic demand. Following the 1994/95 crisis, growth in manufacturing employment has occurred almost exclusively in export-oriented firms (maquiladoras) (Table 5). These firms have adapted to demand by orienting their production from textiles to the manufacture of fabricated metal products and by hiring more skilled employees. It has been reported that 2 400 such firms employ some 600 000 workers, nearly all of whom are Mexican, as required by the legislation. However, these enterprises have made little impact on the Mexican economy since intermediary goods are imported and production is mostly for export. Wage changes have broadly followed fluctuations in the economy’s overall performance: real wages fell by 40 per cent between 1981 and 1988. In 1996, real wages in the manufacturing sector fell by over 10 per cent for the second consecutive year; the fall should be lower over the first half of 1997. The gap between wages in the maquiladoras and the rest of the manufacturing sector reflects the nature of work in the former; it is still often unskilled and largely performed by women. In recent years, however, wage reductions in the maquiladoras have been smaller than elsewhere in the economy as a result of relatively higher gains in productivity. The largest reductions occurred in the construction sector (almost 32 per cent during the period 1995-96). Productivity in the manufacturing industry has risen only slightly since 1980. The sector was profoundly affected by the 1982 crisis and has since undergone major structural change. Later in the 1980s, investment again began to grow. This development was accompanied by higher employment. After 1992, the opening of the economy 21

led to pressure to modernise and rationalise the manufacturing sector. As a consequence, employment has continually declined, particularly in branches like the basic metal industries (Chart 4) where potential productivity gains are the most significant. The Mexican labour market has demonstrated a remarkable capacity for adaptation. However, most adjustments which have taken place have been to the benefit of the informal sector: both the opening of the economy and the rise in the real exchange rate up to 1994 have had a negative effect on employment creation in the formal sector. Job creation has largely centred on low-productivity activities in the small business sector. The problem of low incomes, then, is further compounded by the low quality of new employment creation. Access to employment is also highly unequal. The proportion of low-productivity and inadequately paid jobs continues to grow. Moreover, the nature of work fails both to make optimum use of human capital and to provide social benefit entitlement. 2.

North America: interdependence and regional integration

As NAFTA has only recently come into force its impacts are difficult to identify. The peso crisis, moreover, further complicates evaluation. That said, NAFTA’s long term impacts on Mexico are likely to be more significant than those on the United States and Canada, due to the developmental gap. NAFTA has led to the inclusion of Mexico in the previously existing free trade area covering the United States and Canada. This agreement is neither a customs union nor a fortiori a common market. Regional integration between the three countries is expected to take place via trade and capital flows. There is no provision in NAFTA for the free movement of people. There nevertheless exists important links between the labour markets of all three countries. Indeed, the volume of earlier migration flows fostered regional integration long before the free trade agreement was signed. Asymmetrical trade flows Though characterised by marked asymmetry, trade was already well established between the three countries before the implementation of NAFTA. Indeed, trade in goods conducted by Canada and Mexico with the United States accounts in both countries for over two thirds of the total imports and exports. By contrast, US trade with Canada and Mexico combined represents less than one third of its total trade in goods (Table 9). Developments in imports and exports do not always move in the same direction: recent trends in trade flows point to an increasing level of trade from Mexico to Canada whilst that towards the United States is likely to remain steady. NAFTA will facilitate, and therefore likely increase, access of US products to the Mexican market. Inter alia, intermediate chemical products, high quality metal products, capital goods and electronic goods could well penetrate the Mexican market more deeply as the country develops and modernises. In Mexico, NAFTA is perceived as a further stage of the policy of economic liberalisation and increased openness. Mexico’s membership in GATT (WTO) and its involvement in a number of trade agreements and treaties with countries in Latin America, Asia, and more recently with the European Union have also enabled it to extend its trading activities. NAFTA will benefit the export branch of Mexico’s agriculture sector and sectors such as automobiles and labour-intensive consumer goods, including textiles, leather products, furniture, and electronic goods. Vulnerable sectors include certain agricultural products (food producing, etc.), capital goods, high quality consumer goods, financial services, and transport. 22

NAFTA is not expected to greatly destabilise Canada’s position in the US market. There is the possibility of losing market share in automobiles, energy, and textiles since these are sectors in which Mexico specialises. Canada enjoys comparative advantages different from those of the United States; NAFTA should allow Canada to exploit these advantages further and so consolidate its position. For instance, Canadian finance houses are less stretched than their US counterparts and should therefore be able to profit more effectively from liberalisation of the Mexican banking sector. Sectors in which Canadian companies currently outperform US companies on the export market include telecommunications, transport, and high-tech electronics. Moreover, Canada could well gain market share from Mexican firms in petrochemicals, non-ferrous metal products, and transport equipment. The United States and Canada also consider NAFTA as a means of breaking into Latin American markets via new agreements that bring the two hemispheres together. In this regard, expanding trade with Mexico can be seen as a first step. Direct investment flows Since the early 1990s, US outflows of foreign direct investment (FDI) considerably exceeded inflows (Table 10). The 1995 outflow was three times greater than the yearly average over the period 1988-92, and was again substantially greater than the inflow. Canadian FDI flows are substantially less than those of the United States (Table 10). Since the late 1980s, outflows and inflows have generally balanced out, except in 1985 when a sharp rise in inflow and a substantial fall in outflow occurred simultaneously. The liberalisation of Mexico’s FDI policy accompanied trade liberalisation. FDI inflows are of vital importance to Mexico: the country is reliant mainly on FDI to sustain economic growth, preserve the stability of the macro-economic environment, and encourage the pursuit of financial and structural reforms. In contrast to volatile portfolio investment flows, FDI flows assist the consolidation of economic development and the maintenance of advantageous and profitable conditions for national and foreign firms operating in Mexico. The combined effects of NAFTA and 1993 legislation concerning foreign investment gave FDI a remarkable boost (Table 10), while the peso crisis at the end of 1994 caused a substantially decline in 1995 inflow. This decline is unlikely to have persisted, indeed, an examination of the capital account reveals a significant improvement since the second half of 1996. The relative importance of the origin and destination countries of FDI flows varies markedly according to the country under consideration and over time. For a long time, the principal beneficiary region of US FDI has been the European Union (Table 11a): in 1995, it received half of US FDI (12 per cent went to the United Kingdom). That same year, Canada and Mexico together received only 11 per cent. As for the flow of FDI into the United States (Table 12a), the European Union is the major player (72 per cent, half of which came from the United Kingdom) while Canada and Mexico’s contribution is slight at 7 per cent. Figures relating to Canada’s FDI inflows and outflows illustrate the degree of its economic integration with the United States (FDI flows between Canada and Mexico are residual): the United States receives over a half of Canada’s FDI, and it provides two thirds of Canada’s FDI inflows (Tables 11b and 12b).

23

Irrespective of the year considered, two thirds of Mexico’s FDI inflows come from its partners in NAFTA, and mainly from the United States; in 1995, the United States accounted for over 60 per cent of this inflow (Table 12c). This predominance is unsurprising given the size of the US economy and the increasing integration of the two economies. Canada’s share is, however, gradually increasing from its very low base. The leading position of Mexico’s NAFTA partners is also due to the maintenance of provisions discriminating against investors outside of the NAFTA regime. Analysis of the sector distribution of FDI in North America reveals that the service and the secondary sectors received broadly equal shares (Table 10); this is true of both inflows and outflows, and applies to all countries. Such a statement needs to be qualified in the case of Canada, because the breakdown of data is insufficiently detailed. The importance of the Mexican manufacturing sector can be explained by the fact that it was the first to be touched by the liberalisation of FDI. Moreover, this highly dynamic sector, largely orientated towards the export market offers significant prospects for growth: in 1995, it accounted for 46 per cent of FDI, coming a close second to the service sector. The proportion of FDI directed towards the service sector is set to increase: measures designed to broaden opportunities for acquiring stakes in Mexican banks and other financial institutions under state control have recently been implemented. Mexican emigration as a factor in regional integration The United States remains the principal focus for migration flows in North America. Powerful and traditional links exist between the United States and Mexico while flows between Canada and Mexico are slight. Immigrants to Canada are principally of European and Asian origin (Table 13). Broadly speaking, legal migration inflows to the United States from Canada and Mexico are different both in volume and nature: migration between Canada and the United States and vice versa is mainly temporary and largely concerns highly-skilled workers, representing for each country a major source of that labour category; migration between the United States and Mexico is much more asymmetrical and includes a larger proportion of permanent migration from Mexico (Table 14). During the 1970s, almost 65 000 Mexicans were admitted annually into the United States as permanent residents (Table 15). This figure more than doubled during the 1980s, and increased further in the early 1990s as a result of the regularisation procedure introduced under the Immigration Reform and Control Act (IRCA) which came into effect in 1988. Between 1991 and 1996, the United States legally accepted over 1.5 million Mexicans, of whom almost two thirds were regularised under the provisions of the IRCA procedure (Table 15). Legal Mexican immigration reached its peak in 1991 when almost 950 000 permits were granted. After falling back, it grew again in the wake of the 1994 crisis. The volume of migration flows and the fact that large numbers of Mexican migrants are already well settled in the United States can only have strengthened regional integration in North America and the interdependence between its labour markets. Migration and labour market adjustment The presence of illegal immigrants is a sensitive issue both politically and socially. A series of measures have been implemented by the United States in an attempt to regulate these flows. The IRCA, enacted in 1986, was supposed to stem unauthorised immigration. It seems, a decade later, that the scheme has not worked. The approach used at the time sought to regularise illegal immigrants in certain circumstances and to increase the legal penalties facing employers who hire 24

illegal workers. The ineffectiveness of measures aimed at employers, and the difficulties experienced in monitoring them, caused the programme’s utility to be called into question. Another component of the policy entailed strengthening control measures at the border to combat irregular immigration, yet to date results appear to be unsatisfactory. Indeed, both migrants and traffickers show great skill in adapting to extensions of border control measures. In October 1996, the population residing illegally in the United States was an estimated 4.6-5.4 million, that is to say 19-22 per cent of the total immigrant population. More than half of all illegal immigrants are likely to have come from Mexico. Since 1988, approximately 150 000 Mexicans per year are thought to have entered the country without immigration visas. Moreover, the proportion of those who enter legally and then overstay is likely to have increased. The Mexican component of the population residing illegally in the United States, however is among least likely to have entered the country with a temporary visa; this may be because the authorities are more suspicious that they may intend to prolong their stay. The volume of the migration flow between Mexico and the United States can be explained by factors affecting the labour market situation in both countries. Since the 1970s, emigration from Mexico (along with internal migration) has helped to limit the amount of excess labour supply resulting from the high growth rate of the working-age population. The fall in real income brought about by the 1982 crisis sharpened wage differentials between the two countries and increased the incentive to emigrate. Moreover, demand for unskilled labour in the United States also encouraged the entry of a large number of Mexicans, particularly as networks of compatriots had developed in the United States, thereby facilitating the job search of new entrants. The development of the maquiladoras in the northern region of the country offers a similar spring board. Although NAFTA facilitates the movement of specific categories of worker (e.g. business people, investors and intra-company transferees) within North America, the United States accepts almost exclusively Canadian workers (34 500 in the 1996 fiscal year, compared to only 250 Mexicans). With regard to professionals, the United States maintains a quota for Mexico (5 500 workers); although only a few hundred have entered under this category. Hence, despite still being faced with a restrictive migration policy, the legal inflow of highly qualified Mexicans to the United States has continued to grow in recent years. By contrast, only 4 200 specialists from Canada were recruited in 1996 even though no limit is placed on their numbers (Table 16). Employment, migration and NAFTA: the challenges The US government published a report in July 1997 stating that the Agreement had had a small but positive effect on employment and incomes. This positive judgement is rather controversial: the argument has been put forward that the Agreement has failed in that the overall trade deficit with Canada and Mexico has slipped further since it came into force, and production capacity has been transferred to Mexico where labour costs are six times lower, indeed, 11 times lower in the maquiladoras. Mexico’s labour cost advantage has been further reinforced by the 1995 depreciation of the peso. The gap between Mexican labour costs in the manufacturing sector and those of its two North American partners are likely to widen as a function of differences in productivity levels (Chart 5). Given that production relocation would have taken place in any case as a natural consequence of this labour cost gap, care must be taken to avoid overstating NAFTA’s impacts on job cuts in the United States. The elimination of positions in certain sectors have doubtless been compensated by employment creation elsewhere, though this is less easily observable since it is more dispersed. In

25

fact, relocation may well be beneficial to the United States in the long-term by virtue of the proximity of the two markets and secondary economic effects. For some, NAFTA is also seen as a means to reduce the volume of illegal labour flows. Although an actual reduction in flows has yet to be attained, it is felt that promoting stable economic growth will ultimately have this effect; the development of high-performing economic zones in Mexico should, over the long term, have the result of reducing the incentive to emigrate. Internal migration in Mexico, which used to take the form of a rural exodus, has steadily changed in nature. It now tends to be directed towards industrial centres along the northern border where the maquiladoras are located. Growth in employment in these areas is much higher than in the rest of the manufacturing sector. Against a backdrop of prolonged economic growth in the United States, migration flows from Mexico persist. However, although total unemployment is low, the employment situation is not uniformly favourable: as in Canada, unskilled workers still face difficulty finding positions. Even though a significant proportion of Mexican labour is likely to be competing with native workers, it would be incorrect to overstate the role that Mexican emigration plays in exacerbating tensions in the US labour market. In any event, the United States will sooner or later have to face the problem of retraining less-skilled workers.

NOTES

1.

Cécile Thoreau is a Statistical Assistant in the International Migration Unit of the OECD. Tania Paracini was a consultant to the OECD, in the same unit, when this paper was written. The authors would like to thank Mrs. Bénédicte Larre, Administrator in the Department of Economic Studies and Mr. JeanPierre Garson, Principal Administrator in the Directorate for Education, Employment, Labour and Social Affairs for the useful ideas and comments which they generously contributed.

2.

The informal sector is characterised by modes of production that are low in capital intensity and are very flexible. It employs an unskilled workforce whose remuneration is much lower than that earned elsewhere in the economy. The firms are typically family owned, small in size, located in urban areas, and largely engaged in the production of services and non-durable consumer goods.

26

BIBLIOGRAPHY

COMMISSION FOR LABOR COOPERATION (NAALC) (1997), North American Labor Markets, a Comparative Profile. HINOJOSA OJEDA, R. (1994), “The North American Free Trade Agreement and migration”, in Migration and Development, New Partnerships for Co-operation, OECD, Paris. OECD (1996a), Trade Liberalisation Policies in Mexico, Paris. OECD (1996b), OECD Economic Surveys, United States, Paris. OECD (1996c), OECD Economic Surveys, Canada, Paris. OECD (1996d), International Direct Investment Statistics Yearbook 1996, Paris. OECD (1997a), OECD Economic Surveys, Mexico, Paris. OECD (1997b), OECD Economic Surveys, United States, Paris. OECD (1997c), Economic Outlook, Paris. OECD (1997d), Employment Outlook, Paris. OECD (1997e), Trends in International Migration, Paris. PHAM, K.S. and VEGANZONES, M-A. (1993), “La zone de libre-échange nord-américaine: trois stratégies pour un accord”, in Économie et statistique, No. 264, 1993-94. STATISTICS CANADA (1993), “Mexico’s Demographic Challenges (An overview)”, Report on the Demographic Situation in Canada 1993, Current Demographic Analysis, Ottawa. STATISTICS CANADA (1996), Report on the Demographic Situation in Canada 1996, Current Demographic Analysis, Ottawa.

27

Chart 1. Trade balance and current balance

Trade balance US

Current balance US

Trade balance Canada

Current balance Canada

Trade balance Mexico

Current balance Mexico

40 20 0 -20 -40 -60 -80 -100 Thousands US $ -120 -140 -160 -180 -200 1987

1988

1989

1990

1991

1992

Source : OECD, Main Economic Indicators.

28

1993

1994

1995

1996

Chart 2. Ratio of foreign trade1 in goods to GDP United States Mexico

Canada

75% 70% 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 1987

1988

1989

1990

1991

1992

1993

1994

1995

1. Exports plus imports. Sources: OECD Foreign Trade database and National Acounts, OECD, 1997.

Chart 3. Participation rate and unemployment rate by sex in North America and in the European Union, 1996 Percentages Participation rate Unemployment rate Canada

United States

Mexico

EU

Canada

United States

Mexico

EU

14

100

12

80

10 60

8

40

6 4

20

2

0

0 Men

Men

Women

Source: Labour Force Statistics, OECD, 1997.

29

Women

Chart 4. Relative manufacturing productivity levels in North America, 1990 et 1995 Value added per employed person, United States = 100 Canada 1990

Mexico

1995

1990

120

120

100

100

80

80

60

60

40

40

20

20

0

1995

0 Total

1

2

3

4

5

Manufacturing industries: 1. Food, beverages, tobacco 2. Textiles, footwear, leather 3. Wood and wood products 4. Paper, printing

6

7

8

Total

5. 6. 7. 8.

1

2

3

4

Chemicals Non met. mineral products Basic metals Fabr. metals and machinery

Source : OECD STAN database.

30

5

6

7

8

Chart 5. Annual labour cost per employee in some manufacturing industries Thousands US$ United States

Mexico

Canada

Total manufacturing

Textile, apparel and leather

50

50

40

40

30

30

20

20

10

10

0

0 1980 81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

1980 81

82

83

84

Chemical products

85

86

87

88

89

90

91

92

93

94

95

91

92

93

94

95

Basic metal industries

50

50

40

40

30

30

20

20

10

10

0

0 1980 81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

1980 81

Source: OECD STAN database.

31

82

83

84

85

86

87

88

89

90

Table 1. Overview on the North American demographic and economic situations, 1996

Population (thousands) 2

Density of population (pop./km ) Annual increase rate of the population (per 1 000) 1 Net migration (% of total increase rate)

1

GDP (Price levels and exchange rates of 1990) (millions US$) GDP per head (Price levels and exchange rates of 1990) (US$) GDP per head (Current PPP and current prices) (US$)

Canada

Mexico

United States

29 680.

92 718

269 444

3.

47

28

11.8 58.0

19.7 -44.0

8.7 35.3

612 200 20 679 21 031

283 500 2 990 7 383

6 146 700 23 377 26 438

Consumer prices (Change between 1996 and 1995) (%) Consumer prices index 1990=100

1.5

34.4

2.9

113.5

301.7

120.0

Exchange rate (US Cents/National currency)

73.01

12.74

-

211 067 224 031 435 098 73.1

77 007 93 362 170 368 58.8

966 677 789 773 1 756 450 24.1

21.4

14.6

17.6

13 676 75.9 46.1 9.7

34 108 64.5 36.8 3.8

126 707 79.3 47.0 5.4

54.9 4.65 38.5

31.1 0.38 27.2

60.2 5.15 39.3

Foreign trade (1995) Imports (Price levels and exchange rates of 1990) Exports (Price levels and exchange rates of 1990) Total (Price levels and exchange rates of 1990) Foreign trade (% of GDP) Gross fixed capital formation (% of GDP) Employment (thousands) Participation rate (%) Employment/total population (%) Unemployment rate (%) Unit labour cost (Compensation cost as a % of GDP) (1995) Minimum wages per hour (US$ 1997) % of non-supervisory workers (manufacturing sector)

1. 1990 for Mexico. Sources: National Accounts, OECD, 1997; Employment Outlook, OECD, 1997; Labour Force Statistics, OECD, 1997.

32

Table 2. Foreign trade by commodities, 1993 and 1995 (Percentages) United States Imports NAFTA 1993 1995

Mexico Exports

Total

NAFTA

Canada

Imports Total

NAFTA

Exports Total

NAFTA

Imports Total

NAFTA

Exports Total

NAFTA

Total

1995 1993 1995 1995 1993 1995 1995 1993 1995 1995 1993 1995 1995 1993 1995 1995

33

Food and live animals

5.1

4.7

3.8

5.5

4.7

7.5

6.4

4.7

4.2

6.7

6.8

6.9

5.2

4.6

5.0

4.6

4.1

6.3

Beverages and tobacco

1.0

0.5

0.7

0.2

0.2

1.5

0.3

0.1

0.2

0.7

0.6

0.7

0.2

0.2

0.5

1.0

0.5

0.5

Inedible crude materials, except fuel Mineral fuels, lubricants and related materials Animal and vegetable oils, fats and waxes Chemicals and related products Manufactured goods

6.3

6.0

2.9

3.7

4.0

6.3

4.6

4.5

4.2

2.0

1.9

2.5

3.2

3.6

3.6

8.1

8.0

13.1

11.2

9.6

8.2

1.7

1.7

1.9

2.7

2.3

2.1

12.9

9.5

10.3

1.5

1.3

3.6

12.2

10.5

9.3

0.2

0.2

0.2

0.2

0.3

0.5

0.5

0.7

0.8

0.1

0.1

0.1

0.1

0.1

0.1

0.2

0.2

0.3

4.3

4.7

5.5

8.8

9.2

11.1

8.6

9.1

9.9

3.7

2.4

4.9

8.7

8.9

8.1

5.2

5.9

6.1

13.8

14.8

12.2

13.2

13.9

9.1

16.9

17.2

16.4

10.2

9.1

11.2

12.7

13.3

13.0

16.5

18.0

17.5

Machinery and transport equipment Miscellaneous manufactured goods Commodities and transactions not elsewhere classified

45.7

46.4

46.4

51.5

52.1

46.9

37.4

43.3

44.6

52.2

57.7

52.4

54.8

55.6

51.5

44.7

45.0

39.4

6.9

7.7

16.8

11.3

10.8

11.3

12.4

12.7

12.1

11.3

11.7

10.7

10.6

10.2

11.7

4.8

5.5

5.2

5.7

5.4

3.3

3.9

3.1

3.9

10.1

5.3

5.5

0.2

0.3

0.3

3.1

2.4

2.9

2.7

2.2

2.5

Total

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

100

Note: Percentages are calculated by dividing exports (resp. imports) of the corresponding commodity by total exports (resp. imports) of the geographic area (Total or NAFTA). Source: OECD Foreign Trade database.

Table 3. Structure and components of the change in total population in North America, 1970-1996

Canada Average population (thousands) Children aged 0-15 (% of total population) Persons aged 65 and over (% of total population) Annual increase rate of the population (per 1 000) Annual natural increase rate (per 1 000) Annual net migration rate (per 1 000) Fertility rate Mexico

1

1970-79

1980-89

1990-95

1996

22 988 26.5 8.5

25 882 21.5 10.2

28 712 20.5 10.0

29 964 20.0 12.2

12.1

12.1

13.1

11.8

8.6 3.5

8.3 3.8

6.7 6.4

.. ..

2.0

1.7

1.7

1.6

66 957 43.0 3.9

81 250 38.3 4.8

87 855 36.7 4.7

.. .. ..

..

19.7

20.4

..

.. ..

28.3 -8.6

.. ..

6.5

4.2

3.1

2.8

215 026 25.4 10.5

237 482 21.8 11.9

256 655 21.9 12.6

265 557 21.7 12.8

10.5

9.9

10.1

8.7

6.5 1.9

7.1 2.8

6.9 3.2

5.7 3.1

2.0

1.8

2.1

2.0

2

Average population (thousands) Children aged 0-15 (% of total population) Persons aged 65 and over (% of total population) Annual increase rate of the population (per 1 000) Annual natural increase rate (per 1 000) Annual net migration rate (per 1 000) Fertility rate

1

United States Average population (thousands) Children aged 0-15 (% of total population) Persons aged 65 and over (% of total population) Annual increase rate of the population (per 1 000) Annual natural increase rate (per 1 000) Annual net migration rate (per 1 000) Fertility rate

1

1. Average over the last 5 years of the corresponding period. 1990-1994 instead of 1990-1995; projections for the period 1995-99 instead of 1996. 2. Data respectively refer to 1980, 1990 and 1992-1995. Sources: Labour Force Statistics , OECD, 1997; United Nations World Population Prospects.

34

Table 4. Demographic indicators, actual and projected changes, 1950-2009

Canada

Mexico

35 U.S.

Crude death rate (per 1 000)1 8.7

Life expectancy at birth1 69.1

Fertility rate1

7.7

Crude birth rate (per 1 000)1 27.8

7.6

24.6

7.7

71.4

3.6

65.0

8.5

16.0

7.4

73.1

2.0

68.3

10.2

14.2

7.0

75.9

1.7

14.0

67.8

11.5

14.2

7.1

78.5

1.7

29 931

13.2

67.7

12.3

11.9

7.4

78.9

1.6

2000-04

31 153

13.1

68.4

12.8

11.5

7.8

79.4

1.7

2005-09

32 317

13.1

69.0

13.4

11.5

8.1

79.9

1.8

1950-59

31 487

18.0

52.0

4.5

45.3

17.0

50.7

6.9

1960-69

42 681

18.2

49.6

4.5

44.6

12.2

58.5

6.8

1970-79

58 238

19.1

49.7

4.0

43.2

9.5

62.6

6.5

1980-89

74 663

20.7

53.9

3.8

31.9

6.4

67.7

4.2

1990-94

86 391

22.1

58.7

4.1

27.0

5.2

71.5

3.1

1995-99

94 268

21.2

61.0

4.5

24.6

5.1

72.5

2.8

2000-04

101 830

19.6

62.9

4.9

22.2

5.1

73.4

2.5

2005-09

108 884

18.6

64.7

5.5

20.1

5.2

74.3

2.3

1950-59

169 933

13.5

62.1

8.7

24.3

9.5

69.0

3.5

1960-69

197 967

15.4

60.3

9.5

21.9

9.4

70.0

3.3

1970-79

219 160

18.4

64.1

10.4

15.7

9.2

71.3

2.0

1980-89

240 785

16.7

66.3

11.8

15.7

8.6

74.5

1.8

1990-94

259 383

13.9

65.5

12.5

15.5

8.8

75.9

2.1

1995-99

271 553

13.3

65.6

12.5

13.8

8.7

76.7

2.0

2000-04

281 815

13.7

66.7

12.4

13.2

8.7

77.4

2.0

2005-09

292 173

14.1

67.5

12.6

13.5

8.7

78.0

2.0

aged 15-24 (%)

aged 15-64 (%)

aged 65 and over(%)

1950-59

Average total population (Thousands) 15 558

14.7

60.5

1960-69

19 484

16.1

59.4

1970-79

22 941

19.4

1980-89

25 886

17.6

1990-94

28 465

1995-99

1. Average over the last 5 years of the corresponding period. Source: United Nations, World Population Prospects 1950-2050 (The 1996 revision).

3.7

Table 5. Employment and labour force growth in some OECD countries, 1984-1998 (Annual percentage change) Employment Level

Average

1995 (000s)

1984-1994

1995

1996

Labour force Projections 1997

1998

Level

Average

1995 (000s)

1984-1994

1995

1996

Projections 1997

1998

North America1

153 159

1.6

1.5

1.8

2.3

1.3

162 982

1.4

1.3

1.5

1.9

1.3

Canada

13 508

1.5

1.6

1.3

1.8

2.0

14 929

1.4

0.7

1.5

1.5

1.7

Mexico2

14 752

..

1.9

5.0

3.0

2.4

15 749

..

4.7

4.1

1.9

2.1

Except maquiladoras

..

..

-9.1

2.2

5.1

..

..

..

..

..

..

..

Maquiladoras3

..

..

11.2

16.4

19.9

..

..

..

..

..

..

..

1.2

1.9

1.1

Of which: In manufacturing sector

36

United States

124 899

1.6

1.5

1.4

2.3

1.0

132 304

1.4

1.0

France

146 306 22 444

0.3 0.2

0.5 0.9

0.1 -0.2

0.4 0.2

0.8 1.0

164 681 25 374

0.5 0.5

0.1

0.3 0.8

0.3 0.5

0.4 0.5

34 868

0.6

-0.3

-1.2

-0.9

0.4

38 480

0.7

-0.5

-0.1

-0.1

0.2

EU Germany United Kingdom Total OECD

0.1

25 820

0.5

0.8

0.5

1.3

0.7

28 111

0.3

-0.3

-0.3

-0.1

0.2

443 973

1.1

1.1

1.0

1.3

1.1

480 333

1.1

0.7

1.0

1.1

0.9

1. 1984-1994 averages exclude Mexico. 2. Data based on the National Survey of Urban Employment. 3. Data based on Estadistica de la Industria Maquiladora de Exportacion. Employment Outlook, OECD, 1997; Instituto Nacional de Estadística, Geografía e Informática Sources: Estadistica de la Industria Maquiladora de Exportacion.

(INEGI), Encuesta Industrial Mensual,

Table 6. Production and employment changes by sector in North America, 1980 and 1995 United States Structure of value added (% of total value added) Manufacturing Services 1 Manufacturing industries Manufacturing value added (%) High-technology Medium-high-technology Medium-low-technology Low-technology Total manufacturing sector Employment in manufacturing industries (%) High-technology Medium-high-technology Medium-low-technology Low-technology Total manufacturing sector Unit labour costs (Total manufacturing=100 for each country and year) High-technology Medium-high-technology Medium-low-technology Low-technology Labour productivity (Total manufacturing=100 for each country and year) High-technology Medium-high-technology Medium-low-technology Low-technology Services Services value added (%) 1 Electricity, gas and water Construction Wholesale and retail trade, hotels and restaurants Transports, storage and communications Finance, insurance, real estate and business services Community, social and personal services Total services Employment in services industries (%) 2 Wholesale and retail trade, hotels and restaurants Transports, storage and communications Finance, insurance, real estate and business services Community, social and personal services Total services

Canada

Mexico

1980

1995

1980

1995

1980

1995

21.6 59.6

18.5 68.1

17.9 48.2

17.9 51.4

18.0 55.9

19.1 61.5

10.5 32.6 26.4 30.5 100.0

15.8 32.6 21.4 30.1 100.0

6.5 24.7 25.3 43.5 100.0

9.2 31.3 18.7 40.8 100.0

5.2 24.9 25.1 44.8 100.0

7.8 28.0 23.0 41.2 100.0

11.4 28.2 25.0 35.4 100.0

11.1 27.4 24.2 37.4 100.0

6.2 23.1 25.2 45.5 100.0

8.0 26.7 22.6 42.7 100.0

.. .. 20.8 48.2 100.0

.. .. 19.8 46.8 100.0

101 126 107 76

85 105 107 99

84 111 126 86

90 91 113 103

.. .. .. ..

.. .. .. ..

111 91 100 104

154 118 90 77

125 94 86 107

128 117 89 90

.. .. .. ..

.. .. .. ..

4.4 8.1 28.1

4.3 5.6 24.6

5.5 13.7 26.0

6.2 10.6 23.3

1.7 10.8 49.6

2.4 7.2 32.6

10.7 34.7

8.9 39.9

13.2 32.4

11.4 36.7

10.7 14.9

13.9 26.7

14.1 100.0

16.8 100.0

9.2 100.0

11.7 100.0

15.1 100.0

17.3 100.0

42.4

37.8

48.3

45.7

..

..

8.9 21.8

7.2 25.6

15.7 20.4

12.1 23.4

.. ..

.. ..

26.9 100.0

29.5 100.0

15.7 100.0

18.7 100.0

..

..

1. United States and Mexico: 1993 instead of 1995; Canada: 1992 instead of 1995. 2. United States: 1993 instead of 1995; Canada: 199 4 instead of 1995. Source: Science, Technology and Industry, Scoreboard of indicators, OECD, 1997.

37

Table 7. Employment by sector of activity, 1980 and 1996 (Percentages) Canada Agriculture Mining and quarrying Manufacturing Electricity, gas and water Construction Wholesale and retail trade Transports Financing and business services Community, social and personal services All activities All activities (Thousands)

Mexico

United States

1980

1996

1990

1995

1980

1995

5.4

4.1

22.6

23.5

3.6

2.8

1.8 19.7 1.2 5.8 22.6 7.3 9.6 26.6

1.2 15.2 1.1 5.3 23.8 6.4 12.6 30.3

0.4 19.2 1.4 6.8 16.2 4.5 1.5 17.4

0.7 15.2 0.2 5.4 23.2 4.5 3.4 23.7

1.0 22.1 1.2 6.3 21.5 5.7 8.4 30.4

0.4 16.2 0.9 6.3 22.1 5.4 11.2 34.6

100.0 11 082

100.0 13 676

100.0 23 403

100.0 32 385

100.0 99 303

100.0 126 708

Source: Labour Force Statistics, OECD, 1997.

38

1

Table 8. Real earnings growth for different groups of workers over the past five and ten years (Percentage changes)

Earnings of full-time workers

39

Australia 1995 Austria 1995 Belgium 1994 Canada 1995 Denmark 1993 Finland 1995 France 1994 6 Germany 1994 Italy 1993 Japan 1995 Korea 1995 Netherlands 1994 New Zealand 1994 Sweden 1994 Switzerland 1996 United Kingdom 1996 United States 1995

Compensation per employee (national accounts) Past Past ten five years years 4.4 -1.9 5.5 17.9 14.5 23.5 0.1 3.0 5.3 9.6 4.9 22.7 5.8 10.2 4.1 14.1 10.3 20.1 2.6 13.4 27.9 91.8 3.9 7.3 -3.4 1.5 1.5 15.1 3.3 15.1 5.1 15.7 0.9

2.2

Total

Men

3

Women

Youth

2

4

Prime-age

Low-paid

5

High-paid

Past five years 5.5 8.0 9.9 0.7 0.1 4.6 2.6 9.9 0.8 4.5 43.5 3.3 -0.6 -2.3 3.0 8.5

Past ten years 1.8 .. 16.9 3.8 5.3 21.5 7.2 21.0 10.4 17.5 116.3 9.3 -2.8 9.3 .. 23.2

Past five years 5.8 7.0 8.0 -1.4 0.0 4.8 2.1 7.6 3.1 3.3 38.5 2.7 -1.3 -2.0 3.9 7.8

Past ten years 2.7 .. 15.3 1.5 .. 21.9 6.7 19.7 12.4 15.8 100.2 8.4 -4.0 10.8 .. 21.9

Past five years 6.6 8.5 14.1 6.5 2.7 5.4 4.4 15.7 2.5 9.9 50.7 7.7 5.8 -0.2 6.2 11.7

Past ten years 3.9 .. 25.8 14.1 .. 22.1 10.0 26.1 12.6 24.7 149.1 17.1 6.0 10.0 .. 33.4

Past five years 2.3 .. 6.9 .. .. 3.8 1.1 9.6 .. 6.2 41.0 .. .. -9.6 -3.8 1.6

Past ten years -4.8 .. 17.9 .. .. 23.1 1.1 19.5 .. 17.0 132.8 .. .. 4.2 .. 13.4

Past five years 7.9 .. 8.6 .. .. 2.9 1.1 3.0 .. 1.4 41.0 .. .. -3.3 1.8 6.0

Past ten years 1.6 .. 16.3 .. .. 19.1 1.7 10.9 .. 11.8 91.2 .. .. 6.5 .. 18.9

Past five years 8.4 3.6 8.1 .. .. 8.8 3.1 30.8 -11.1 11.4 .. 3.5 0.3 -5.1 3.9 4.9

Past ten years 0.8 .. 15.7 .. .. 26.9 4.0 59.6 7.4 24.3 .. 8.3 -4.4 3.4 .. 13.8

Past five years 12.6 10.1 13.3 .. .. 2.0 3.4 11.7 0.5 5.9 .. 2.7 3.4 -1.8 5.2 9.1

Past ten years 7.7 .. 20.3 .. .. 18.5 10.2 21.5 20.0 19.9 .. 9.9 0.3 11.8 .. 24.9

-0.9

-3.1

-4.8

-6.3

0.2

3.7

-8.2

-11.0

-2.8

-4.8

-7.4

-7.2

-2.1

3.1

.. Data not available. 1. All nominal wage series have been deflated by each country’s consumer price index. The latest year to which the data refer is shown in parentheses. For the following countries, the data for earnings growth refer to a different period than indicated but have been expressed in terms of a standard five-yearly or ten-yearly rate of change: for Italy and New Zealand, the column "past 5 years" refers to the past six years; for Belgium and Finland, the column "past 10 years" refers to the past nine years; and for the Netherlands, it refers to the past eight years. 2. The data for Austria also include part-time workers. 3. Youth refer to 21-25 year olds for France. 4. Prime-age workers refer to workers aged 31-40 for France, 35-39 for Korea, and 35-44 for the Netherlands and Sweden. 5. For Austria, high-paid earnings correspond to 8th decile earnings. 6. All data refer to western Germany only. Source: Employment Outlook, OECD, 1997.

Table 9. North American partners’ share of foreign trade in goods, 1993-1995 (Percentages) Imports

United States

40

Canada

Mexico

Exports

Trade partners

1993

1994

1995

1993

1994

1995

Canada Mexico

18.8

19.1

19.2

20.9

21.5

20.7

6.8

7.3

8.1

9.2

10.2

8.2

NAFTA

25.6

26.5

27.4

30.1

31.7

28.9

United States

67.0

67.7

66.8

80.2

81.1

79.1

Mexico

2.2

2.2

2.4

0.4

0.5

0.4

NAFTA

69.2

70.0

69.2

80.7

81.6

79.6

United States

69.9

69.0

74.2

91.1

84.2

82.1

Canada

1.8

2.0

1.9

1.7

2.6

2.6

NAFTA

71.7

71.0

76.1

92.7

86.8

84.7

Source: OECD Foreign Trade database.

Table 10. Foreign direct investment flows in North America by sector 1988-1995 (Thousands of US$) a. Sectoral distribution of North American direct investments inflows FDI recipient countries United States 1988-92 Primary

444

1993

1994

911

Canada %

1995 2

1988-92

5

10

609

1993

1994

620

Mexico %

1988-92

-391

-4

135

1995

-120

1993

1994

48

%

1995

192

139

2

Secondary

20

13

22

26

43

608

1

3

2

25

2

4

7

3

46

Tertiary

22

29

24

28

47

1

1

1

2

21

966

2

4

4

52

-

-

-

-

-

2

1

2

6

57

-

-

-

-

-

43

43

49

60

100

5

4

7

10

100

3

7

12

8

100

Unallocated

*

Total

41

b. Sectoral distribution of North American direct investments outflows FDI investor countries United States

Primary Secondary Tertiary Unallocated Total

*

Canada % 1988-92

1993

1994

1995

%

502

1 786

2 761

2 077

36

47

505

619

1 496

455

8

49

1 888

2 070

1 198

702

12

-

-

1 612

1 182

1 787

2 534

44

77 247 53 078 93 406

100

4 507

5 657

7 241

5 768

100

1988-92

1993

1994

1995

- 540

5 149

2 840

3 729

4

14 319

18 522 25 533 43 520

18 580

53 576 24 705 46 157

32 359

-

-

- Means nil or negligible. * The sectoral analysis of Canada’s FDI flows is restrained due to the volume of unallocated flows. Note: Data are based on national sources and therefore are not strictly comparable from one country to another. For Mexico, data are from the Directorate General of Foreign Investment (SECOFI) and from the Directorate General of Economic Research (Bank of Mexico); for Canada, data are from Statistics Canada, Canada’s International Investment Position; for the United States, from United States Department of Commerce, Bureau of Economic Analysis: Survey of Current Business. Source: International Direct Investment Statistics Yearbook , OECD, 1996.

Table 11. United States and Canada: Foreign direct investment outflows by recipient country and region, 1988-1995 (Millions US $) a. United States: Flows of direct investment abroad Receiving regions and countries

1988-92

1993

1994

1995

19 591

42 812

22 852

42 878

(% of 1995 total) 46

560 1 938

1 625 4 566

2 522 968

1 583 4 222

2 5

13 860

38 167

17 823

46 335

50

5 335 518 1 772 2 222

25 355 1398 4 263 -495

1 920 3 212 1 846 2 770

11 624 7 134 2 481 5 954

12 8 3 6

3 823

6 100

9 614

9 880

11

2 246 1 578

3 584 2 516

6 287 3 327

7 767 2 113

8 2

32 359

77 247

53 078

93 406

100

1988-92

1993

1994

1995

OECD

3 620

3 692

4 938

3 580

(% of 1995 total) 62

European Union (*)

1 144

2 364

1 233

943

16

799

868

142

-86

-1

2 126

731

2 735

2 947

51

2 126 -

731 -

2 735 -

2 947 -

51 0

4 507

5 657

7 241

5 768

100

OECD Of which:

Japan Switzerland European Union (*) Of which:

United Kingdom Netherlands Germany France NAFTA Of which:

Canada Mexico Total

b. Canada: Flows of direct investment abroad Receiving regions and countries

Of which:

United Kingdom NAFTA Of which:

United States Mexico Total

(*) 15 Members of the EU. Note Data are based on national sources and therefore are not strictly comparable from one country to another. For Canada, data are from Statistics Canada, Canada’s International Investment Position and for the United States, from United States Department of Commerce, Bureau of Economic Analysis: Survey of Current Business. Source: International Direct Investment Statistics Yearbook , OECD, 1996.

42

Table 12. Foreign direct investment inflows in North America by investor country and region, 1988-1995 (Millions US$) a. Foreign direct investment to the United States Investor countries and regions OECD Of which: Japan Switzerland European Union (*) Of which: United Kingdom Netherlands Germany France NAFTA Of which: Canada Mexico Total

1988-92 40 442

1993 40 090

1994 44 695

1995 62 243

% 102

14 653 1 453 21 723

1 058 2 875 32 964

7 654 4 680 24 814

5 252 7 961 43 576

9 13 72

9 090 5 393 2 000 3 231 1 634

13 232 2 967 7 698 6 778 3 687

11 123 -2 272 6 551 3 987 5 307

22 081 -184 8 117 3 719 4 036

36 0 13 6 7

1 381 253 43 537

3 799 -112 43 534

4 031 1 276 49 903

4 489 -453 60 848

7 -1 100

1988-92

1993

1994

1995

%

OECD

4 896

4 433

6 573

10 262

95

European Union (*) Of which: United Kingdom

1 699

404

-859

2 458

23

676

278

-1 544

-10

0

NAFTA Of which: United States Mexico

2 032

4 009

6 783

7 219

67

2 032 -

4 009 -

6 783 -

7 219 -

67 0

Total

5 332

4 853

7 090

10 777

100

b. Foreign direct investment to Canada Investor countries and regions

c. Foreign direct investment to Mexico Investor countries and regions

1988-92

1993

1994

1995

%

OECD Of which: Switzerland European Union (*) Of which: Germany NAFTA Of which: United States Canada

4 555

6 689

10 824

5 321

63

144 921

112 1 099

123 2 724

407 1 897

5 23

300 2 076

93 5 318

267 6 853

877 5 621

10 67

2 002 74

5 142 176

6 361 492

5 228 393

62 5

Total

3 724

7 007

12 382

8 428

100

(*) 15 Members of the EU. Note Data are based on national sources and therefore are not strictly comparable from one country to another. For Mexico, data are from the Directorate General of Foreign Investment (SECOFI) and from the Directorate General of Economic Research (Bank of Mexico); for Canada, data are from Statistics Canada, Canada’s International Investment Position and for the United States, from United States Department of Commerce, Bureau of Economic Analysis: Survey of Current Business. Source: International Direct Investment Statistics Yearbook , OECD, 1996.

43

Table 13. Inflows of immigrants to Canada and the United States by country and region of origin, 1980-1996 Average annual flows (Thousands)

1980-1989

1990-1995

1996

Average annual flows divided by the stock of residents of the same origin at the beginning of the period (per 1 000) 1980-1989

1990-1995

1996

1

2

CANADA Asia and Pacific

Of which: Hong Kong India China Europe Of which: Bosnia Herzegovina Africa and Middle East Americas Of which United States Total

50.0

113.3

124.1

93.2

106.4

116.5

10.6 7.7 3.3 35.1

33.8 15.0 11.3 45.2

29.9 21.2 17.5 39.7

181.1 70.5 63.5 13.7

221.9 86.4 71.5 19.1

195.9 121.9 111.0 16.8

.. 20.3 17.4

4.7 36.7 36.4

5.1 36.1 24.2

.. 199.1 29.9

.. 221.0 51.9

.. 217.2 34.5

7.9 126.0

6.6 231.6

5.8 224.1

26.3 32.8

26.5 53.3

23.2 51.6

UNITED STATES Asia Of which Philippines Vietnam China India North and Central America Of which Mexico Canada Europe Of which Former USSR South America Of which Colombia Total

3

271.5

328.8

307.8

106.9

66.0

..

47.4 39.6 26.6 25.4 233.2

59.4 54.1 43.1 37.0 559.6

55.9 42.1 41.7 44.9 340.5

94.5 171.4 93.1 123.1 54.2

65.1 99.6 81.4 82.3 80.2

48.0 56.8 52.1 59.3 ..

103.1 11.6 66.5

361.2 15.3 140.1

163.6 15.8 147.6

46.9 13.8 12.9

84.0 20.5 32.2

24.5 .. ..

6.9 41.0

50.4 61.3

62.8 61.8

17.0 73.1

126.4 59.1

.. ..

11.2 633.2

15.3 1 127.8

14.3 915.9

77.7 45.0

53.4 57.1

.. 37.3

1. For Canada : divided by the stock at the 1991 Census; for the United States, divided by the estimates of the stock in March 1996 (CPS). 2. The figures include both principal applicants and their accompanying dependants, if any. Figures include backlog clearance. 3. Data refer to fiscal years (October to September of the given year). From 1989 to 1996 data include respectively 478 814, 880 372, 1 123 162, 163 342, 24 278, 6 022, 4 267 and 4 635 immigrants who obtained a permanent residence permit following legalization under the 1986 Immigration Reform and Control Act. Source: Trends in International Migration , OECD, 1997.

44

Table 14. Inflows of temporary workers to Canada and the United States by country and region of origin, 1981-1996 Average annual flows (Thousands) 1981-90

1

1991

1992

1993

1994

1995

1996

2

CANADA United States Philippines China Japan United Kingdom Sri Lanka Jamaica Mexico France Somalia Other countries Total

48.2

47.3

46.8

46.4

45.8

..

..

21.5 14.2 7.4 10.8 3.1 8.5 6.3 6.3 2.1 131.1 259.4

26.4 11.7 8.8 10.0 4.3 7.5 6.6 6.3 .. 104.8 233.8

24.4 9.3 9.8 6.7 8.5 6.0 5.7 5.5 .. 107.7 230.4

17.5 11.2 9.6 8.0 7.1 6.6 5.8 5.4 3.0 65.1 185.6

10.8 9.7 9.0 7.2 6.8 6.0 5.5 5.4 2.8 63.8 172.9

.. .. .. .. .. .. .. .. .. .. ..

.. .. .. .. .. .. .. .. .. .. ..

UNITED STATES

3

Europe

28.8

52.9

53.5

52.7

60.7

63.4

76.0

Of which: United Kingdom North America Canada Mexico Caribbean Asia Of which: India Philippines South America Oceania Africa Central America Other and not defined Total

11.4

19.0

19.5

18.5

20.6

21.8

25.2

34.7 12.4 8.8 13.5 16.6

51.9 19.4 14.0 18.5 46.6

53.3 23.4 14.0 15.9 49.9

59.0 27.8 16.8 14.4 48.8

67.5 36.0 17.8 13.7 56.6

65.3 34.1 19.2 12.0 64.6

73.2 37.2 26.7 9.3 70.2

2.5 4.6 4.2 1.8 1.3 0.9 1.2 89.6

5.4 9.9 9.4 3.8 2.5 2.0 0.5 169.6

8.7 11.0 9.7 4.3 2.7 1.7 0.6 175.8

12.7 10.0 11.0 4.8 3.5 1.8 0.7 182.3

18.5 11.6 13.6 5.2 4.5 1.9 0.9 210.8

24.1 12.6 14.6 5.1 4.9 2.1 0.5 220.7

31.4 7.0 18.5 7.0 5.0 2.2 2.4 254.4

1. 1989-1990 for Canada. 2. Work permits delivered. 3. Including trainees, excluding intra-company transferees and treaty traders/investors. Data refer to fiscal years (October to September of the given year). Figures may be overestimated because of multiple entries by the same person. Source: Trends in International Migration , OECD, 1997.

45

Table 15. Inflows of Canadian-born and Mexican-born immigrants to the United States, 1960-1996 Canadian-born immigrants Average annual flows (in thousands) 1961-70 1971-80 1981-90 1991-96 1989 1990 1991 1992 1993 1994 1995 1996

41.3 17.0 15.7 15.1 12.1 16.8 13.5 15.2 17.2 16.1 12.9 15.8

Mexican-born immigrants

Per 1 000 Average annual Canadian-born flows 1 residents (in thousands) .. 20.9 18.6 20.3 .. .. .. .. .. 23.7 19.1 24.0

45.4 64.0 165.6 275.2 405.2 679.1 946.2 213.8 126.6 111.4 89.9 163.6

Per 1 000 Mexican1 born residents

.. 84.3 75.3 64.0 .. .. .. .. .. 17.8 13.4 24.5

Of which: Persons who benefited from the IRCA program (% of total) 58.1 63.3 83.6 91.7 94.4 57.2 13.8 3.9 3.3 2.2

1. Average annual inflows per 1 000 residents born in Canada/Mexico and already present in the U.S. at the beginning of the period. Source: Trends in International Migration , OECD, 1997.

46

Table 16. Inflows of North American temporary workers 1 to the United States by admission category (1992 and 1996 fiscal years) Canadian-born 1992

Foreign government officials, spouses and children Students Vocational students Spouses and children of students International representatives, spouses and children Temporary workers

Mexican-born 1996

1992

1996

Numbers

%

Numbers

%

Numbers

%

Numbers

%

750

1.0

445

0.5

2 959

5.4

2 533

2.8

12 182

15.8

9 341

10.3

8 347

15.2

10 887

12.1

10 178

13.2

10 386

11.5

14 006

25.4

26 536

29.5

734

1.0

371

0.4

839

1.5

1 027

1.1

1 523

2.0

1 820

2.0

1 180

2.1

1 314

1.5

51 780

67.1

68 323

75.3

27 730

50.4

47 521

52.9

30

-

23

-

123

0.2

73

0.1

4 633

6.0

4 192

4.6

5 077

9.2

5 273

5.9

644

0.8

127

0.1

5 210

9.5

8 833

9.8

4 014

5.2

1 738

1.9

2 906

5.3

5 539

6.2

89

0.1

74

0.1

58

0.1

141

0.2

13 958

18.1

34 438

38.0

-

-

243

0.3

44

0.1

481

0.5

24

-

171

0.2

44

0.1

207

0.2

48

0.1

65

0.1

Of which: Registered nurses Professionals Temporary agricultural workers Temporary non-agricultural workers Industrial trainees Professional workers: North American Free Trade Agreement Workers with extraordinary ability Workers accompanying the workers with extraordinary ability Athletes and entertainers International cultural exchange Workers in non profit religious organisations Intra-company transferees Treaty traders and investors and dependants Spouses and children of temporary workers Spouses and children of intracompany transferees Exchange visitors

786

1.0

1 522

1.7

369

0.7

5 463

6.1

5 467

7.1

3 698

4.1

3 962

7.2

4 461

5.0

398

0.5

595

0.7

145

0.3

512

0.6

5 664

7.3

7 037

7.8

1 845

3.4

4 759

5.3

2 924

3.8

2 620

2.9

184

0.3

980

1.1

1 825

2.4

2 231

2.5

1 050

1.9

1 965

2.2

3 830

5.0

4 015

4.4

1 272

2.3

2 787

3.1

5 467

7.1

3 698

4.1

3 962

7.2

4 461

5.0

1 027

1.3

602

0.7

966

1.8

1 142

1.3

519

0.7

478

0.5

447

0.8

395

0.4

Spouses and children of exchange visitors Fiancé(es) and children of fiancé(es) of US citizens NATO officials, their spouses and children Total of the above categories

417

0.5

547

0.6

82

0.1

258

0.3

77 147

100.

90 686

100.

55 061

100.

89 818

100.

Temporary visitors for business

15 322

11 471

285 238

309 141

Temporary visitors for pleasure

13 946

7 370

1 192 415

915 918

716

646

17 114

13 858

Transit aliens

1. Non-immigrants are visitors, transit aliens and persons who hold a temporary residence permit. Figures may be overestimated because of multiple entries by the same person. Sources: United States Department of Justice, Statistical Yearbooks of the Immigration and Naturalization Service.

47

GLOBALISATION, ECONOMIC GROWTH AND THE LABOUR MARKET IN NORTH AMERICA

by Norma Samaniego General Director for Santa Fe Consultants, Mexico City

1.

Introduction

Wherever there is sharp economic disparity between two adjacent countries there is potential for economically induced migration. This potential can be expanded through labour force pressure and lack of employment opportunities in the lower-income country, combined with demand for migrant labour in the higher-income country. These circumstances have been long present in North America, particularly along the border between Mexico and the United States, dividing a highly developed economy and a less developed one. Migration is presently one of the most delicate issues of the bilateral relation between Mexico and the United States. It has been a recurrent focus of restrictive policy measures, but its root causes, namely, economic development differentials and labour market structure and behaviour, have been largely neglected by actions aimed at reducing flows. Migration in North America needs to be examined from a new perspective: as occurring within a region of the world that is rapidly moving towards a global economy. What have been the effects of globalisation in North America on economic growth and labour markets in the region? What are the expected effects of regional integration on economic convergence and migration potential? 2.

Globalisation in North America

The global economy is shaping a new world context. Factors of production, such as natural resources, capital, technology, and sometimes labour, as well as goods and services, are moving across national and regional boundaries, creating increased interdependence, new economic and political balances, and deep domestic structural change. The three North American countries are a part of this new reality. Transition has not been smooth. The challenge of globalisation for countries that have previously pursued policies aimed at protecting domestic industry from foreign competition or creating an industrial infrastructure has been huge. Deep reductions in barriers to imports and capital flows, along with far reaching domestic reforms have had to be implemented. In addition, globalisation has substantially reduced the room for manoeuvre of government policies and magnified the effects of financial imbalances. 49

Due to the different roles the three North American economies play in the world economy as a result of their broad economic differences and trade policies, the road to globalisation taken by each has been unique in terms of national strategies and structural adjustment needs. At the same time, they have coincided with the need for a regional trade area. For the United States, globalisation meant assuming a different and more competitive role in global markets. Since the end of the cold war, and the emergence of new players in the economic scene in Asia, different economic balances have emerged.

− Fifty years ago, every country had to trade with the United States, where almost two thirds of the world’s industrial production capacity was concentrated. As the world recovered from World War II, the US share of the world’s GDP declined. By the end of the 1980s, the United States was no longer the world’s largest market, a position henceforth occupied by the European Community, while Japan and a group of newly industrialised countries were busy creating new trade associations. − The United States was already relatively open in terms of trade barriers. It needed to raise its productivity, form productive associations, enhance its international competitiveness, and open up markets for its goods, services, and capital. Canada also understood its need to find a new role to play in the world economic scene. It rapidly moved towards a more open economy, increasing its export share of GDP from an average 15.5 per cent in the 1960s to 27 per cent in the early 1990s (Wolff, 1997). Meanwhile, it signed two strategic trade agreements with its North American neighbours. The first one was a bilateral agreement with the United States, which aimed at establishing a more predictable trade relationship with its primary trading partner. The second one, NAFTA, continued the evolution towards stronger trade links within the North American geographical economic region by its inclusion of Mexico. In Mexico, the globalisation process has entailed the transition from an import substitution development strategy, which had been in place since the 1950s, to an export oriented one. This implied rapid trade liberalisation and profound economic reform. 3.

The regional integration approach

The transition from national economies to a one-world economy is not easy to achieve. Regional trading blocs have emerged as a first step to a truly global economy, or as a response to increased constraints resulting from globalisation. But regional integration experiences differ substantially from one another, as do the problems encountered along the way and the chances for success. The approach to regional integration in North America differs sharply from that in Europe. The European Community was shaped as a common market. It envisages free flow of labour, a monetary union, provides for social policy co-ordination, and is guided by a far reaching political vision. It was promoted by three countries (France, Germany and Italy) that were at similar stages of economic development. This approach provided a background for balance and consensus, but also posed delicate foreign and domestic policy issues. Free trade agreements in North America followed a totally different pattern. Both the free trade agreement between the United States and Canada, and the North American Free Trade Agreement, establish formal trade relationships between the world’s largest economy and much smaller ones. 50

They do not aim at a political union or even at a common market, but at a rather different outcome: a large free trade area for goods, services, and investment. One could hardly think of an economic agreement between countries with such extreme diversity as those represented by NAFTA. The sharp contrasts between NAFTA partners make this agreement radically different from the integration approach followed by countries within the EEC and closer to those existing between the Community and developing nations in the Mediterranean or the central European economies in transition. 4.

Economic and labour market differentials within the region

Canada, the United States and Mexico not only differ sharply in their gross domestic products, per capita GDP, and relative export and import shares, but also in the size of their populations and in their role in the world economy. Contrasts can also be found in their legal and political institutions, in labour market structure and trends, and in their historical attitudes toward free trade. In terms of 1996 GDP, the size of the US economy is 12.7 times that of Canada and 22.5 times that of 1 the Mexican economy. Although per capita GDP differentials are lower between Canada and the United States (US$18 915 and US$26 438), they are very high when considering Mexico’s GDP per 2 capita (US$2 946). The relative shares of international trade show a different degree of exposure and dependence on the world market. Canada’s export share of GDP averaged 27 per cent in the early 1990s, Mexico’s 24.2 per cent, and that of the United States 7.9 per cent over the same period. All three countries raised their import shares over the past 25 years, although there was a sharp decline in the mid-1980s in Mexico following the 1982 crisis, and a strong opening up by 1987. Regional trade among the three economies had a clearly different impact on each of the trade partners. Bilateral trade with the United States was highly important for Mexico and Canada, in terms of its relative share of their GDPs, whereas US trade with Mexico and Canada, though two of its three major trading partners, represented only a small portion of US GDP in 1993. This continues to be the case. Labour force growth Labour markets, one of the most sensitive issues connected with free trade, also reflect the highly visible differences between Mexico, Canada, and the United States. On the supply side, Mexico has experienced significantly greater population growth among the working aged, at a rate of 2.5 per cent over the last 20 years, compared to 1.5 per cent for Canada, and 1.1 per cent for the United States (OECD, 1995). Participation rates in the labour force have increased in all three economies over the last three decades as a result of a significant growth of female participation rates. Nonetheless, women’s participation rates in Mexico still lag behind those of Canada and the United States. This implies that in the near future further increases in labour force participation rates may be expected in Mexico. Combined with the highly dynamic natural growth rate of the working age population, this growing participation in the labour force in Mexico will result in persistent pressures on the labour market in the medium term. 51

A higher proportion of total growth in the labour force in Canada and the United States can be attributed to immigration. In 1986 the share of foreign born in the labour force in Canada was 21.9 per cent. In the United States, the figure was 9.3 per cent in 1990 (OECD, 1994). Labour demand and unemployment On the demand side, and in sharp contrast with the EEC economies, employment has been growing at a relatively fast rate during the past three decades in all three nations. A large share of existing unemployment is only of a short duration and there are frequent inflows and outflows in the unemployed population. In both the United States and Canada, employment shares in the services sectors have been increasing, while there has been a persistent decline in the shares of agriculture and of the manufacturing industries in total employment. In Mexico, the persistent decline in agricultural employment has been accompanied by rising shares of employment in the industrial and services sectors, with these latter showing a faster growth rate over the last two decades. Unemployment rates have been higher in Canada, following a trend more similar to that of European countries over the last two decades than to the United States or Mexico (OECD, 1996). Yet, lower unemployment indicators in a context that does not provide for unemployment insurance does not provide a complete measure of labour market imbalances. A new category for precarious forms of employment needs to be taken into consideration. Precarious employment, comprising the informally self-employed and employment in family small production units has, since the 1980s, been growing in Mexico faster than salaried jobs in the formal sector of the economy. These sharp differentials in development, economy size and labour behaviour have been historically associated with strong migration flows within the North American region, particularly along the USMexican border. 5.

Globalisation, economic growth and the labour market

Globalisation has taken place in very different contexts, requiring specific strategies and raising different concerns in Mexico, Canada, and the United States. The transition from an import substitution policy to an exports oriented one in Mexico Mexico adopted a growth model based upon import substitution, which helped the economy to maintain a high and uninterrupted rate of growth with a notable degree of price stability for over twenty years (Samaniego, 1997). Between 1950 and 1970, the average growth rate of GDP was 6.6 per cent, with an average annual inflation rate of 4.9 per cent. The occupational structure by sectors also underwent changes in those years, consistent with the pattern of other developing countries. The import substitution model began to show signs of exhaustion in the early 1970s, when the phase of rapid import substitution of consumer goods, based on simple technologies and aimed at the domestic market, had run its course. The oil boom of the late 1980s gave a short boost to the

52

domestic economy and employment, but came to an abrupt end in 1982 when the fall in international oil prices, followed by difficulties meeting debt payments, gave way to a severe economic crisis. The economic crisis of the 1980s prompted a decade of dramatic structural change, accompanied by a unilateral process of trade liberalisation. The long duration of the effects of the crisis, along with the structural adjustment process, brought about a sharp decline in the average rate of growth of GDP during the 1980s (1.8 per cent as compared of 6.6 per cent for the two previous decades). The 1995 crisis During the first half of the 1990s, the economy showed a modest growth rate after a decade of economic stagnation. Trade liberalisation continued and in 1994 Mexico entered the North American Free Trade Agreement (NAFTA) with Canada and the United States. Nevertheless, a semi-fixed exchange rate, combined with an explosive availability of foreign capital and political shocks, resulted in a new devaluation of the peso and a more profound economic crisis ensued in 1995 (Gil Díaz and Carstens, 1996). The effects of the crisis were particularly pronounced in the labour market. Indicators of formal employment for salaried workers in the private sector reveal a 6.4 per cent decrease of total labour force in the formal sector over ten months. Job loss was generalised across the economy, unlike what happened during the 1982 crisis, when employment decline only reached the manufacturing sector. This time though, economic recovery came quicker than with the 1980s’ crisis. By October 1996 the peak level of formal employment had been surpassed and the economy was growing at the highest rate in ten years. This, of course, does not take into account labour force growth over the period of the crisis, which resulted in a significant expansion of open unemployment rates, a considerable increase in precarious forms of self-employment and mounting labour market pressures. 6.

Recent labour market trends and concerns in the United States and Canada

Over the past two decades some of the salient labour market trends in the United States have been: an increasing income disparity between more and less skilled workers, with decreasing real earnings of the less skilled, and a decline of the relative share of manufacturing jobs in the economy. In Canada these trends have been accompanied by a comparatively high rate of unemployment. Trade with the developing countries has been frequently blamed for these trends. The surging exports from Asian developing economies and the debate over NAFTA attracted a great deal of attention to these issues. Frustration with the rising earnings gap in the United States and high unemployment levels in Canada often led to the conclusion that globalisation and imports from lowwage countries are negatively affecting earnings or demand for unskilled workers in the importing countries. These concerns have motivated a proliferation of empirical studies on this issue. Most of the recent studies shows that trade with less developed countries can only explain a small part of the decline in wages of unskilled workers. Both Canada and the United States continue to buy most of their imports from other developed countries with similar wages. Imports, other than oil, from less developed countries accounted for 2.8 per cent of GDP in the United States in 1990 (Krueger, 1997).

53

GLOBALISATION and increased trade with emerging economies can only explain a small proportion of the decline in the relative share of manufacturing employment. Most studies have found that the proportion of income spent on them in developed countries has declined. Others claim that employment problems in developed economies can be attributed to the mobility of capital to less developed economies. However, during the 1980s there was virtually no North-South net investment. Interest payments and debt repayments were consistently higher than new investment (Krueger, 1997). A review of the empirical literature suggests that trade accounts for only a small proportion of the observed decline in wages for unskilled workers. Most studies conclude that skill-biased technology is the main reason, although the effects of trade and technology might be inter-related. In many countries, the nature of labour market institutions plays an important role in determining whether the adjustment process will affect employment instead of wages. 7.

NAFTA: Employment and economic growth

The short run impact of NAFTA has been difficult to isolate since accurate measurement of its effects would require disentangling other events not directly related to the agreement, such as the severe recession in Mexico, the depreciation of the Mexican peso, and the opening up of the three economies under the Uruguay Round. As a number of studies anticipated, due to the relative sizes of the three economies and to their regional trade shares, the estimated impact of NAFTA on the US and Canadian economies in the first three years has been minimal in the aggregate, while there has been a larger impact on the Mexican economy. A recent evaluation (President of the United States, 1997) found that during the first three years of its implementation, NAFTA had a “modest positive effect on US net exports, income, investment and jobs supported by exports”. Moreover, at current levels of near full employment in the US and a rapid rate of job creation, small changes in trade have not appreciably affected total employment, but only shifted jobs between sectors and caused reductions of earnings in some industries. The impact of NAFTA on the wages of US manual workers in some labour-intensive sectors has been minimal since existing barriers to trade were already small. NAFTA was identified as affecting, whether positively or negatively, industries that employ a total of 3.6 per cent of the non-farm labour force (Office of the United States Trade Representative, 1996). Industries with increases in hours worked outnumbered those with decreases, and industries with earnings reductions outnumbered industries with earnings increases. Another study (Hinojosa et. al., 1996) found that “the overall negative or positive employment impacts of US-Mexico trade have not been significantly affected by the liberalisation of tariffs due to NAFTA. The most important negative effect on employment has been the decline of US exports due to the peso crisis, not the liberalisation of trade due to NAFTA”. Although trade with Canada and Mexico represents 30 per cent of US trade flows (President of the United States, 1997), two-way trade with both countries since NAFTA was signed has grown more rapidly than U.S. trade with the rest of the world. Moving beyond its quantitative effects, other researchers propose that NAFTA should be evaluated in terms of the qualitative change it has brought 54

in the way Mexican and American manufacturing sectors relate to each other, based on growing complementarity. In the case of Mexico, the effects of NAFTA and the prior opening up of the economy cannot be isolated from the adverse effects of the peso crisis of 1995. Mexican GDP decreased by 6.9 per cent, wages plummeted by more than 20 per cent, open unemployment surged from 3.7 per cent to 6.3 per cent, and more than 1 million jobs were lost in 1995. The manufacturing sector, which had grown by a modest 3.6 per cent between the first quarters of 1987 and 1990, decreased by 23.8 per cent between that period and the end of 1995. In sharp contrast, the maquiladora -- sub-contracting industry -- sector showed an increase of 187.3 per cent over the whole period (Alberro, 1997). Formal employment also shows sharp differentials. After a rapid recovery, the total number of jobs in the private formal sector increased by 8.4 per cent between December 1994 and June 1997. Labour intensive export-oriented industries such as apparel and electronic products, which have been booming for years, show employment increases of 55.6 per cent 3 and 39.1 per cent over the same period. Exports have played an important role in Mexico’s economic recovery. The sectors showing the highest rates of growth after 1995 were those where regional integration and opening up were comparatively advanced. Mexican industry is increasingly regrouping itself into two segments: the export-oriented sector and the one that produces for the domestic market. Unfortunately, recovery has not yet reached the average Mexican, whose real wages are still under 1994 levels, and many industries oriented towards domestic demand are still facing an adverse situation. During the last decade, income distribution worsened and precarious employment grew faster than jobs in the formal sector. 8.

Conclusions

The 1980s were marked by the long term effects of the financial crisis of 1982 in Mexico, together with the radical adjustment process towards a different development strategy, resulting in near null economic growth over the decade, very low net formal employment growth, and a persistent decrease in average real wages. A rapidly increasing labour force supply over the same period translated into a significant growth of precarious forms of employment, mainly self-employment, small family productive units in the non formal sector, and increasing potential for emigration. The 1990s opened new prospects for growth, but only after the economy had undergone costly adjustments towards a new export-oriented strategy, and a regional free trade agreement was in place. However, the effects of the 1995 financial crisis cancelled out much of the advance made in employment growth during the first half of the decade. This time though, recovery was more rapid. By October 1996, the peak level of formal employment had been surpassed and the economy was growing at the highest rate in ten years. At the same time, a ten-month period of no net employment creation, coupled with a growing labour force, has considerably increased labour market imbalances, which translates into precarious employment and migration pressures. The fastest growing sectors of the economy over the last three years have been those related to exports, particularly in sub-contracting industries, which create employment for the low skilled labour force, some of whom have been displaced from rural areas. More recently, other manufacturing industries, such as automobile and auto parts are also showing significant growth. These activities 55

generate more employment in input-producing related industries, but domestic related industries, trade, and services still lag behind in employment creation. In the short term, globalisation strategies that require deep structural adjustment processes often imply costly economic growth and employment sacrifice. Regional economic integration can increase trade, accelerate capital inflow and technology transfer, and increase trade in lower income countries, thus mitigating the cost of the transition. However, if no provisions for a more even distribution of those effects are in place, expanding trade on the basis of different factor endowments can lead to a deepening of domestic income disparities, constituting a trap for countries whose specialisations remain traditional and rely heavily on unskilled labour. Trade is not an automatic substitute for factor mobility. If it is not accompanied by measures to promote technological catching up, physical infrastructure, human capital investment, and a more even development, it can lead to a lack of convergence in the medium term. In this context, the only alternative to reducing the incentive to emigrate in the long-term is to promote sustainable and more even development with employment creation.

NOTES

1.

On an exchange rate US current dollar basis (US International Trade Commission, 1997).

2.

At 1995 current dollars (OECD, 1997).

3.

An indicator of formal employment in the private sector is the total number of salaried workers covered by the IMSS (Mexican Social Security Institute).

56

BIBLIOGRAPHY

ALBERRO, J.L. (1997), “Growing Inequality and Productive Restructuring: Productivity and Income Performance of the Mexican Economy 1970-1996,” North American Seminar on Incomes and Productivity, North American Commission on Labor Cooperation, Dallas, Texas. GIL DÍAZ, F. and CARSTENS, A. (1996), “Some hypothesis related to the Mexican 1994-1995 crisis”, Serie Documentos de Investigación, (Research Papers Series), No. 9601, Banco de México, Mexico. HINOJOSA OJEDA R., DOWDS, C., McCLEERY, R., ROBINSON, S., RUNSTEN, D. and WOLFF, C. (1996), North American Integration Three Years after NAFTA: A Framework for Tracking, Modeling and Internet Accessing the National and Regional Labor Market Impacts, UCLA, NAID Center. KRUEGER, A. (1997), “Labor Market Developments, Trade and Trade Agreements”, North American Seminar on Incomes and Productivity, Dallas, Texas. OECD (1994), The OECD Jobs Study. Evidence and Explanations. Part I: Labor Market Trends and Underlying Forces of Change, Paris. OECD (1995), Economic Outlook, Paris. OECD (1997), “OECD in Figures”, Statistics in OECD Member Countries, Paris. OFFICE OF THE UNITED STATES TRADE REPRESENTATIVE (1996), North American Free Trade Agreement, Information Package, Washington, D.C. PRESIDENT OF THE UNITED STATES (1997), Study on the Operation and Effects of the North American Free Trade Agreement, Washington , D.C. SAMANIEGO, N. (1997), “El Mercado de Trabajo en México”, El Economista Mexicano, (“The Labor Market in Mexico”, The Mexican Economist), Vol. I, No. 2, Mexico. THUROW, L. (1996), The Future of Capitalism, William Morrow and Company Inc., New York. US INTERNATIONAL TRADE COMMISSION (1997), The Impact of the North American Free Trade on the US, Economy and Industries: a Three Year Review, Washington, D.C. WOLFF, E. (1997), “Per capita income and relative productivity performance in Canada, Mexico and the US”, North American Seminar on Incomes and Productivity, North American Commission on Labor Cooperation, Dallas, Texas.

57

PART II

MIGRATION IN POLICY IN NORTH AMERICA: POLICIES AND SECTORAL ECONOMIC IMPACTS

59

A. PERSPECTIVES ON MIGRATION POLICY

Migration trends and policies in the United States B. Lindsay Lowell Migration trends and policies in Canada Roderic Beaujot The impact of migration on economic development in Mexico Gustavo Verduzco and Kurt Unger

61

MIGRATION TRENDS AND POLICIES IN THE UNITED STATES

by B. Lindsay Lowell United States Department of Labor

1.

Introduction

Immigration to the United States divides into historical phases with distinct characteristics. Initially, the flow of immigrants to the United States was largely unregulated. The first restrictions barring “illiterates” and “public charges” were established under the 1882 Chinese Exclusion Act. The Immigration Act of 1924 banned most “Orientals” and enacted limited national origin quotas against those not from north-western Europe. With only minor adjustments, notably the Immigration Act of 1952, nationally selective exclusion remained in force for four decades. The Immigration Act of 1965 marked a significant change in that policy. The old system was overhauled, yielding a complex system that permitted equal numbers of visa applicants from all sending countries under numerically limited “preference categories” based either on certain skills (i.e. occupations) or on immediate family ties (i.e. spouse, children, and siblings). Otherwise, entry was allowed for unlimited numbers of immigrants based upon immediate family relationships to US citizens (i.e. spouse, children, and parents). Thus, the underlying conventions of the 1965 Act were equity for applicants from all sending countries and family reunification. The unexpected result of the new system was a multiplier effect in which new immigrants sponsored applicants from previously under represented developing countries. Mexicans and other Latin Americans, who have a long history of migration to the United States, make up approximately 40 per cent of today’s immigrants. Asians, who were either banned or restricted prior to the 1965 Act, now make up just under 40 per cent. At the same time, a large number of immigrants enter through unofficial channels. The Immigration Reform and Control Act of 1986 (IRCA) imposed legal work authorisation requirements for new US employees. In reality, there has been little effective reduction of unauthorised migration and, ironically, IRCA’s legalisation program solidifies family and community networks that may increase unauthorised flows. Unfortunately, the failure to develop effective strategies to control unlawful immigration has blurred the public perception of the distinction between legal and illegal immigrants. Nevertheless, IRCA’s passage cleared the way for reform of the immigration system and anticipated the substantial increases in immigration levels that later occurred with passage of the Immigration Act of 1990 (IMMACT). Overall, the levels of immigration to the United States have been increasing steadily since the 1965 Act. Consider that the number of entries grew 42 per cent between 1970 and 1980 (from 373 000 to 531 000), a more modest 24 per cent through 1990 (656 000), but 23 per cent in only 63

2 years by 1992 (810 000 -- the first year IMMACT was in effect). These figures do not include about 2.5 million IRCA legalisations, nor do they indicate the demand for future immigration that these newly legal sponsors will generate. 2.

The Immigration Act of 1990 and admissions today

The Immigration Act of 1990 (IMMACT) attempted to balance a number of competing interests. First, its admission policies created limits on legal immigration by establishing annual numerical targets for total immigration. Second, it created a guaranteed minimum number of visas for close family members to accommodate increases in the number of immediate relatives of US citizens seeking entry. Third, it addresses present and future labour market needs by increasing the number of persons admitted for employment reasons, with higher priority being given to professionals and highly-skilled persons. Fourth, it created a “diversity” category of admissions for persons from nations that have not recently sent many immigrants to the United States. As intended, post-IMMACT admissions are significantly greater than admission numbers in the 1980s. IMMACT authorised a increase of about 40 per cent in the number of admissions compared to previously established levels. Since IMMACT went into effect in fiscal year (FY) 1992, the number of admissions in subsequent years has followed an up and down course. Admissions first increased in FY 1993, only to drop back down to FY 1992 levels by FY 1994; and they dropped yet again in FY 1995. However, much of this latter drop is explained, not by effective declines in the number of admittable persons, but because of administrative delays in processing applications. By FY 1996, the number of new immigrants admitted hit an all time post-World War II high, partly due to a catch-up in processing applicants from the previous year and partly due to an historical upward trend (Table 1). Future admission levels under IMMACT may continue to grow for a number of reasons. The backlogs of those waiting for legal entry in the family-based classes of admission remain very large, and are growing in some cases. It is unlikely that applications for family preferences will be undersubscribed in the foreseeable future. Continuing backlogs ensure that available family quotas, as well as any unused employment numbers, will be filled. A pierceable cap for the immediate relatives of citizens permits growth in this category. Given the pace with which immigrants are naturalising, growth in this category can also be expected as newly naturalised citizens petition for the admission of their immediate relatives. The number of admissions available for employment-based immigrants has been under subscribed, but demand in some categories appears to be increasing. Employment-based admissions remain substantially below IMMACT’s numerical limit of 140 000.1 Among skilled classes of admission, only India is oversubscribed with nearly a two-year wait for admission for professionals with advanced degrees and skilled workers. There is a backlog in the category of unskilled workers, which remains heavily oversubscribed as of FY 1997 with just less than a seven-year wait for admission for all nations. Continued demand, along with changes implemented earlier in the regulations governing temporary admissions, may lead to an increase in employment-based admissions. The total backlog of family applicants, which figured at 3.5 million at the start of FY 1997, was essentially unchanged from the previous year. Of this total, 42 per cent are adult brothers and sisters of US citizens. Their backlog numbers continue to increase, and there is now an average wait of over six years before they are admissible (Bureau of Consular Affairs, 1997). The spouses and children of legal permanent residents (LPRs) constitute 29 per cent of the family-based backlog. The Department

64

of State concludes that “the long waiting list assures continued over subscription and a considerable delay between the filing of a petition and the applicant’s turn for a visa issuance. ”2 In principle, the recent surge in the naturalisation of potential sponsors could reduce the backlog of spouses and children of LPRs. Sponsors who have been naturalised can bring their spouses and children in under the unlimited citizen reunification category. However, this process will not decrease the backlog in an expeditious fashion. Surprisingly, early indications are that the large volume of naturalisations since 1995 have not resulted in increased numbers of relatives of US citizens in the family preference total (Bureau of Consular Affairs, 1997). At the same time, new provisions adopted in the 1996 Welfare Reform Act and the Illegal Immigration Reform and Immigration Reform Act may dampen future admissions. This latter act, in particular, requires all family members to be sponsored by a US petitioner whose income meets high standards, i.e. 125 per cent of the poverty level. Sponsors must sign legally binding affidavits by which they pledge to provide any financial support needed by the new immigrants. In addition, the welfare reform legislation bars non-citizens from most social safety-net programs. Some US family members may be unwilling or unable to take on these new financial responsibilities. 3.

Origins abroad and US destinations

Asia and North America remain the sending regions with the largest share of immigrants. Mexico remains the largest single sending country, and its share of immigrants increased from an average of 12 per cent in the 1980s to 18 per cent in FY 1996. The effects of IRCA, which legalised about two million formerly unauthorised Mexican residents, certainly explain this trend. Even though the special admission category for the spouses of legalised immigrants’ dependants has been discontinued, Mexico benefits from the IMMACT’s removal of per country limits on the limited spouse and children class of admission (Table 2). In 1995, IMMACT’s permanent “diversity” class of admission program replaced a transitional class of admission for adversely affected countries.3 Under today’s diversity program, countries are not eligible for the 55 000 available diversity visas if they have sent 50 000 (or more) numerically-limited immigrants over the preceding five years. The share of immigrants from non-traditional sending countries has increased under the permanent diversity program. Of those admitted under the program, 42 per cent came from Europe in FY 1995 and FY 1996, whereas Europeans made up only 15 per cent of immigrants in all other classes of admission. At the same time, Africans made up 35 per cent of diversity immigrants, but only 0.7 per cent of immigrants in all other classes of admission. Immigrants in FY 1996 remain destined for just a few states: about two-thirds intend to reside in California, New York, Texas, Florida, and New Jersey. One-quarter of admissions intend to reside in California alone, and one-seventh in New York. New York City retains its place as the pre-eminent immigrant city with 14 per cent of immigrants intending to go there. Among other top destination cities, about 10 per cent of immigrants choose Los Angeles, and San Francisco comes in a close third with about 9 per cent of the total. There has been very little change in these leading destinations since IMMACT. The increased numbers and greater diversity of admissions have not markedly changed the choice of destinations.

65

4.

Demographic characteristics of new immigrants

Over the past three decades, the age of immigrants at the time of admission has risen, but the pace of this trend has not changed recently. Overall, immigrants are younger than the US population. More than half of family-based immigrants are under 30 years of age, reflecting the predominance of spouses and children among them. They tend to be either younger or older than employment-based immigrants. While employment-based immigrants have just as many minors (aged 15 and under) as other groups, more than 40 per cent of them are 30-44 years old, which are highly productive working years. In contrast, humanitarian admissions tend to be quite a bit older than other immigrants, due in large part to Soviet refugees in FY 1996 (Table 3). Females represented 54 per cent of admissions in FY 1996. Interestingly, there had been an essentially even balance of men and women throughout the 1980s. In 1991, just prior to IMMACT, women accounted for only 34 per cent of admissions, but that reflected the predominance of men among those who had gained admission through IRCA’s legalisation provisions. The increased share of females in the 1990s parallels the historical tendency toward increased female immigration throughout much of the post-World War II period. In FY 1996, family admissions were predominantly female (57 per cent) and employment admissions were evenly balanced by gender. Diversity (45 per cent) and humanitarian (48 per cent) admissions, in contrast, had more male immigrants. 5.

Skill composition of new immigrants

The Immigration and Naturalization Service (INS) admissions data includes only the most basic information on immigrants and tells us nothing about such vital skills as English ability and education. The New Immigrant Survey (NIS) of FY 1996 is the first detailed sample of “green card” admissions ever undertaken (Jasso et. al.,1997). It is a first attempt to appraise the value of such special purpose surveys, especially the feasibility of a longitudinal immigrant survey (Figure 1). Our knowledge of English ability in the US Census and most standard surveys is based on a selfreported five-level ranking from speaks “very well” to speaks “not well at all.” The NIS finds that employment-based immigrants report the greatest English ability, with 70 per cent speaking at least fairly well and less than 10 per cent speaking very little or no English (the remainder report an “average” speaking ability). About 37 per cent of family-based admissions report speaking English at least fairly well and an almost equal proportion report speaking little or no English. The humanitarian admissions trail the furthest behind in the English language ability: only 16 per cent speak English at least fairly well, while more than 50 per cent speak little or no English. According to the 1990 Census, 47 per cent of all foreign-born residents over the age of five reported that they did not speak English at the scale’s highest level of ability, i.e. “very well”. The years of schooling completed by immigrants is, perhaps, one of the most critical measures of skill level. The NIS provides information for the first time about the education levels of immigrants upon legal admission (Figure 2). Immigrants differ significantly in their educational attainment by class of admission. Fully 46 per cent of employment-based admissions have completed four years of university-level education or a graduate degree. This figure includes beneficiaries (spouses and children), demonstrating that well-educated employment-based immigrants tend to have welleducated spouses. In contrast, just 17 per cent of family-based immigrants have completed university-level education, while 42 per cent have less than a secondary education. The humanitarian

66

classes of admission are less educated than US natives or employment-based admissions, but are better educated than family admissions. Again, the large number of Soviet refugees entering in FY 1996 may partly explain this advantage over the family immigrants. About 21 per cent have less than a secondary education, while about 19 per cent have university and higher degrees. Ultimately, the English and educational skills that immigrants have are reflected in the occupations in which they work. The INS admissions data, which we use here, has only crude occupational classifications. It imperfectly captures the difference between immigrants who adjust into legal permanent residence status after working in a US job for several years, and those who report an occupation upon admission, which tells us more about what the immigrant did in the country of origin than what he or she will do in the United States. Note also that 65 per cent of all immigrants in FY 1996 reported no occupation or being a “home maker,” reflecting the fact that children, parents, and spouses make up a large share of all admissions and most do not work at the time of entry (Figure 3). Nevertheless, occupational status faithfully reflects the legal requirements of the admission class: the proportion of immigrants not reporting an occupation is greater among family and humanitarian admissions, about 70 per cent of total immigrants in each category. By way of comparison, only about half of all employment and diversity admissions have no reported occupation.4 The skills which immigrants bring to the United States is reflected in the type of occupations they fill. Family and humanitarian immigrants are primarily blue-collar workers. In contrast, employment and diversity immigrants are predominantly white-collar workers. These differences between the major classes of admission have changed only slightly over the course of the past three decades. 6.

The impact of the 1990 Immigration Act

Thus, we can see that there are strong differences between the classes of admission in terms of their skills and occupation. Nevertheless, these figures tell us little about how policy, and the Immigration Act of 1990 in particular, has affected the number and skill distribution of today’s immigrants. In order to gauge this, simple linear projections for FY 1996 were made for all the regular admission categories using FYs 1972-1991 data (Greenwood and Ziel, 1997). This gives us a picture of what immigration today might have looked like if past trends had continued to be unaffected by IMMACT (Table 4). The total number of regular admissions in FY 1996 was 720 314, which, when compared to the projected figure of 426 929, represents an increase of 69 per cent over projected levels. PostIMMACT admissions were greater than the projected figure because of the legislated increases for non-exempt family, employment, and diversity admissions. Numerically exempt admissions for citizen reunification also continued to grow between 1992 and 1996. Of immigrants who reported an occupation, the actual admissions in FY 1996 were 221 731, which represents an increase of 34 per cent over the projected of 165 234. Clearly, the greatest growth in post-IMMACT immigration occurred among immigrants who did not work at the time of admission. The 34 per cent increase in working immigrants was still less than the increase among non-working immigrants, who experienced a 91 per cent increase of actual over projected. Perhaps, this has to do with the relatively greater growth of family-based immigrants -- who tend not to report an occupation upon admission -- since IMMACT was implemented.

67

As might be anticipated, IMMACT’s new requirements led the way toward growth in skilled occupations. For example, the number of health professionals was projected to be 10 244. The actual number was 85 per cent greater at 18 985. Interestingly, projections not shown here show that the greatest growth in professional occupations occurs among the family members of employment-based admissions. This demonstrates that when principals with more skills are admitted, they bring with them spouses who are likewise more skilled than in the past. At the same time, there is a greater than average increase in the actual number of operators, fabricators, and labourers. There were 53 936 admissions in these occupations compared to the projected 37 702. This particular occupational shift may be due to the overall increase in the family-based admissions, who tend to hold lesser skilled occupations. The NIS survey permits a final look at the implications of admission class for immigration and its role in the US economy. The median earnings of all immigrants admitted in 1996 was $15 600 for men and $11 960 for women. These amounts are lower than median US earnings, but represent a substantial gain for most immigrants. Compared to earnings in their last country of residence, men experienced a 59 per cent increase and women a 45 increase in earnings upon admission to the United States. As might be expected, differences in earnings among admission classes are substantial. Employment-based admissions earn a median income of $36 400 on the date of their admission to legal permanent residence status, while the siblings or spouses of legal permanent residents earn $11 750, and the spouses of citizens earn $18 200. 7.

Immigrants’ effects on the economy

An independent evaluation of immigration by a panel of well-known social scientists at the National Research Council (NRC, 1997) found that immigration has a largely positive economic impact on the national level. This national-level assessment substantiates the belief that immigrants contribute to the US economy. Migration has played a particularly apparent role in mitigating inner city collapse, one of the greatest problems faced by the United States. Cities seldom shrink to prosperity: as population declines, so does quality of life. New immigrants have moved in to take the place of middle-class natives leaving inner cities. They have renewed countless neighbourhoods from New York City to Los Angeles. It is estimated, for example, that one half of Southern California’s current small business boom is fuelled by immigrant communities. The NRC panel estimates that immigrants may directly add as much as $10 billion to the national economy each year. Business owners and investors appear to benefit most from the immigrant labour force, since recent newcomers are often more willing to work for lower wages than other US workers. At the same time, most immigrant groups tend to earn as much as natives after a decade. Immigrants are also willing to take jobs that many natives will no longer take at any wage. Research suggests that “capital” returns are higher in cities with concentrations of immigrants (Greenwood and Hunt, 1995). However, the NRC panel’s findings also confirm the assertion that immigration is associated with tangible costs to certain sectors of the labour market and certain communities. Workers with less than 12 years of education are the most adversely affected by low-skilled immigrant workers. Immigrants may have reduced the wages of secondary school dropouts (who represent about one tenth of the workforce) by 5 per cent nation-wide over the past 15 years (Borjas et.al., 1992). Those workers most at risk in our restructuring economy are those who directly compete with today’s lowskilled immigrant, i.e. low-skilled workers in production and service jobs. It is most often Latin American and Asian immigrants working in those labour sectors who experience the greatest

68

competition.5 In the area of lesser skilled work, however, the NRC finds evidence of wage competition as a result of labour demand. Research also indicates that at the national level the foreign-born pay more in taxes than they receive in various sources of federal services. At the local level, however, immigrants tend to be a net fiscal cost because the taxes that go to the federal government often do not make their way back to the locality. Research on unauthorised immigrant populations residing in urban and rural settings finds the clearest examples of state and local fiscal costs (Taylor et al., 1997). The imbalance is greatest, once again, in communities with large foreign-born populations, especially where there is a large number of school age children, which is the single greatest cost associated with foreign-born households. Whereas research on the costs of legal immigrants finds either no or little cost, with the exception of states with exceptionally high immigrant concentrations, practically all of the available research concludes that “illegal aliens generate more in public costs than they contribute in revenues to government” (GAO, 1995; Clark, R, et. al, 1994). 8.

The foreign-born and labour market skills

Education remains the greatest engine of personal advancement, productivity, and social integration in the United States. Over time, the nation as a whole benefits from investment in the education of immigrants and their children in English and in skills for a modern economy. As immigrants gain in experience, skill, and education, they generate more revenue over their working life. The NRC study projected the future balance of fiscal costs for today’s immigrants and their children through the middle of the next century, an exercise requiring several assumptions about the future, but one that amply demonstrates that immigrants with less than a secondary education represent a likely long-term fiscal cost to the nation. Immigrants who complete secondary education and beyond represent a likely and sizeable fiscal contribution. Immigrants today tend to be more educated than their predecessors from the same country three decades ago. At the same time, natives in the United States complete more years of education than they did in the past, thus widening the gap between immigrant and native education levels (Borjas, 1995). Low educational attainment in the recent immigrant pool “is directly attributable to illegal immigrants and refugees, not to legal immigrants” (Fix, M. and Passel, J.S, 1994). About 40 per cent of all foreign-born adults counted in the 1990 Census had less than a secondary education, almost twice the proportion for US natives. The secondary school drop out rate was disproportionately due to refugee and unauthorised populations. At the other end of the spectrum, one quarter of all foreignborn adults had completed a university education or more, while just a little more than the one-fifth of native adults had. Limiting the inquiry to legally-admitted immigrants, it is estimated that fully one third had completed a university education or more. Once in school in the United States, immigrant children tend to do very well. For immigrant students who have enrolled in secondary schools by the tenth grade, completion rates are greater than for natives. Likewise, these immigrants are more likely to prepare for and to continuously attend and complete a university education. Immigrant children whose parents are both foreign-born tend to have lower drop out rates and higher scholastic achievement than natives or children with only one foreign-born parent. Discouragingly, time in the United States tends to reduce the educational success of immigrant children, suggesting that assimilating to “native norms” may, in some contexts, be detrimental. Research finds that immigrant and second-generation children often experience such

69

“segmented assimilation,” i.e. those who live in certain inner city or less advantaged contexts tend to be most likely to adopt poor outlooks on education (Portes, 1995). Ultimately, much of the economic success and civic participation of immigrants over their lifetime has to do with the level of education they achieve (Shoeni et al., 1996). In general, Asian immigrant students tend to do best in secondary and university education, while Hispanics do the poorest, with whites and blacks falling in between. The 1990 Census shows, for example, that 87 per cent of immigrant children attended secondary school as compared to 93 per cent of natives. This difference is due to Hispanic, and primarily Mexican, immigrant youth. More than one quarter of Mexican immigrant youth ages 15-17 were not in school in 1990, which may be attributed to their simply not ever attending school in the first place (Vernez and Abrahamse, 1996). By age 15, the average youth in Mexico has not been in school for two years. Differences in culture, the education of parents, and the need to work to support their families explain much of this pattern. 9.

Population growth

Each year there are about 800 000 legal admissions and an additional estimated 200 000 to 300 000 unauthorised entries, but the net annual increase of the foreign-born population is only about 700 000 each year due to return migration and mortality (Edmonston and Passel, 1994). In 1996, the foreignborn population was 24.6 million, representing 9.3 per cent of the total US population. Recent arrivals make up a large share of the resident immigrant population: about 28 per cent of the 1996 foreign-born population arrived after 1990, and an additional 35 per cent during the 1980s. It is estimated that international migration makes up somewhere between one quarter and one third of annual net US population increase (Edmonston and Passel, 1994). The NRC report found that without immigration since 1950, the US population would have been 14 per cent smaller than its 1995 size and considerably older on average. A projection into the middle of the next century indicates that there will be an increase from the present of 124 million persons, with immigrants being responsible for about two-thirds (82 million) of that increase, yielding a total population of 387 million. Much of the increase occurs through large immigrant families; the NRC projection of 82 million includes some 45 million new immigrants. In fact, if the number of new immigrants admitted in future was half that of today’s levels, it would reduce the projected population by only about 10 per cent. 10.

Conclusions

The United States remains one of the world’s premiere immigration nations and will likely remain the largest in terms of sheer numbers. Academic researchers have tended to focus less on the overall numbers involved and more on skill composition. The Immigration Act of 1990 has increased both the number of new immigrants and has improved the skill level of the flow somewhat. Nevertheless, the foreign-born stock in the US possesses fewer skills than the average US resident, largely as a result of unauthorised and humanitarian immigrants. Newly admitted legal immigrants, by contrast, tend to be better educated than US residents primarily due to employment-based admissions and to Asian and European admissions. The primary economic benefit to the United States has come in the form of providing willing workers at both the lower and upper tiers of the labour market. At the same time, the system has placed increasingly visible strains on the public sector (especially in education) and on sectors of the labour market where large numbers of low-skilled immigrants are found.

70

There are a number of developments that warrant observation. Immigrants are sure to be affected by 1996 legislative changes in the provisions of welfare benefits that, at this point, bar primarily poor children and families that had previously been able to qualify for government aid in food purchases -- humanitarian admissions are exempted from these exclusions. New “deeming” provisions place financial responsibility for new immigrants on their sponsors. There are several research projects under way to assess whether or not legal immigrants will be adversely affected by these restrictions. One avenue available for bypassing these restrictions is naturalisation.6 For various reasons, from the changes in welfare eligibility just discussed to fears about their status in a seemingly more antiimmigrant environment, the number of naturalisations has been steadily increasing. There were over one million naturalisations in FY 1996, or a 135 per cent increase over 1995. The numbers are up again in 1997 and look as though they will remain high in the coming years (despite an inefficient administrative process that takes months and sometimes years to complete).7 These new citizens now have access to welfare benefits and eligibility to sponsor immediate relatives to immigrate; however, the social and political implications of increased numbers of naturalisations are unknown. On other fronts, there continue to be major initiatives on border and interior enforcement. There were, for example, 1.6 million apprehensions of illegally entering persons in FY 1996, though preliminary 1997 figures appear lower. Increased personnel, border fences, and more sophisticated technologies are being employed along the border with Mexico from San Diego to El Paso. These new efforts have led some illegal migrants seeking entry to United States to attempt passage at less well patrolled areas that are often much more difficult and more dangerous. It remains to be seen whether or not substantially lower illegal entries result from these efforts in the longer term. For the moment, indirect estimates of the unauthorised population as of 1996 indicate that about 275 000 net new illegal entrants take up residence in the United States each year, a number roughly equivalent to the growth of this population over the past decade. 8 The INS removed about 69 000 criminal and non-criminal aliens from within the United States in 1996 and projects an additional 93 000 removals for FY 1997. These increased removals reflect improved detection strategies, particularly with regard to criminal aliens who are discovered by the INS while serving prison sentences. Of course, even these stepped-up removals reflect only a small portion of the backlog of those who have orders to be removed, as well as being only a fraction of the five million unauthorised residents of the United States. But the new level of removals seem appropriate in the case of criminals, and may serve to change the perception that once in the United States the chances of removal are slight. Elsewhere, new “expedited removal” regulations are in place at the ports of entry for those seeking asylum or using fraudulent documents. These new expedited removal procedures bear close observation to ensure that legitimate claims are not being inappropriately rejected.

71

NOTES

1.

Subtracting out the one-time admissions under the Chinese Student Protection Act and the unskilled preference category, the admissions of needed skilled-workers has gone from 107 266 in FY 1992 to 73 239 in FY 1995 and back up to 105 111 in FY 1996. The increase of admissions in FY 1996, once again, partly reflects a catch-up from administrative processing delays in the previous year.

2.

The waiting time until admission for the spouses and children of legal permanent residents in the backlog has continued to grow since IMMACT to almost four and one half years. The priority date for admission advances little each month, meaning longer and longer waits for new applicants.

3.

Ireland (35 percent) and Poland (38 per cent) benefited most from the transitional diversity program during FY 1992-1996, although the share of Irish immigrants never reached the mandated 40 per cent.

4.

The NIS survey finds that about 40 per cent of non-exempt family immigrants are not employed. On the other hand, more than 95 per cent of employment-based immigrants are employed. The INS admission figures for “no occupation” include children and persons who are unemployed, retired, or for whom no information is given.

5.

Immigration also has measurable adverse effects on the working conditions of unskilled AfricanAmerican workers, although these impacts at the national level are small (Bean et al., 1997). Minority effects appear to be found only when low-skilled minority workers compete with low-skilled immigrants in particular industries under conditions of economic stress. Impacts on minorities tend to be found in cities like Los Angeles and New York, and in certain industrial “niches” where immigrants -- many of whom are unauthorised -- dominate the pool of workers (Waldinger 1996; Waldinger and Bozorgmehr 1997). In other states with concentrations of immigrants like New Jersey, there appears to be little impact, which is perhaps due to the fact that the immigrant population has relatively high levels of education and diverse industrial employment (Espenshade, 1997).

6.

The welfare legislation of 1996 restricts the number of years that citizens can qualify for benefits and requires states to move welfare recipients into the workforce.

7.

The recent slow-down in the processing of naturalisation applications has to do primarily with stricter compliance with administrative protocols. Some of this is due to the INS’s greater scrutiny of applications for fraud and ineligibility (due to criminal activity, etc.).

8.

The INS is undertaking pilot programs in places of employment that make it easier for employers to access work authorisation information about new hires. The result of these efforts may lead to more effective enforcement against illegal employment nation-wide, but once again the outcome remains to be seen. For 1996, the INS estimates that there were 5 million unauthorised residents in the United States (60 per cent of whom entered illegally and 40 per cent of whom overstayed their visas).

72

BIBLIOGRAPHY

BEAN, F.D. (1997) (ed.), “Immigration and its Impact on US Minorities”, Forthcoming proceedings of a conference, The Urban Institute, Washington D.C., May. BORJAS, G. (1995), “Assimilation and Changes in Cohort Quality Revisited: What Happened to Immigrant Earnings in the 1980s”, Journal of Labour Economics, No. 13(2), pp. 201-245. BORJAS, G.J., FREEMAN, R.B. and KATZ, L.F. (1992), “On the Labor Market Effects of Immigration and Trade”, in Immigration and the Work Force, Borjas, G.J. and Freeman, R.B. (eds.), University of Chicago Press, Chicago, pp. 213-244. BUREAU OF CONSULAR AFFAIRS (1997), Visa Bulletin, Department of State, No. 73(7). CLARK, R., PASSEL, J.S., ZIMMERMANN, W.N. and FIX, M.E. (1994), Fiscal Impacts of Undocumented Aliens: Selected Estimates for Seven States, The Urban Institute, Washington, D.C. EDMONSTON, B. and PASSEL, J.S. (1994), Immigration and Ethnicity: America’s Newest Arrivals, The Urban Institute Press, Washington, D.C. ESPENSHADE, T. (1997), Keys to Successful Immigration: Experience, The Urban Institute Press, Washington, D.C.

The Integration of

Implications of the New Jersey

FIX, M. and PASSEL, J.S. (1994), Immigration and Immigrants: Setting the Record Straight, The Urban Institute Press, Washington, D.C. GAO (1995), Illegal Aliens: National Cost Estimates Vary Widely, United States General Accounting Office, Washington D.C. GREENWOOD, J. and HUNT, G.L. (1995), “Economic Effects of Immigrants on Native and Foreign-Born Workers: Complementarity, Substitutability and Other Channels of Influence”, Southern Economic Journal, No. 61(4), pp. 1976-1997. GREENWOOD, M. and ZIEL, F.A. (1997), “The Impact of the Immigration Act of 1990 on US Immigration”, Research Paper, US Commission on Immigration Reform. JASSO, G., MASSEY, D.S., ROSENZWEIG, M.R. and SMITH, J.P. (1997), “The New Immigrant Survey [NIS] Pilot Study: Preliminary Results”, Paper presented at the Joint Meeting of the Public Health Conference on Records and Statistics and the Data Users Conference, Washington, D.C., July.

73

NATIONAL RESEARCH COUNCIL (1997), The New Americans: Economic, Demographic and Fiscal Effects of Immigration, National Academy Press, Washington, D.C. PORTES, A. (1995) (ed.), The Economic Sociology of Immigration: Essays on Networks, Ethnicity and Entrepreneurship, Russell Sage Foundation, New York. SCHOENI R., McCARTHY, K.F. and VERNEZ, G. (1996), The Mixed Economic Progress of Immigrants, RAND, Santa Monica. TAYLOR, E., MARTIN, P. and FIX, M. (1997), Poverty Amidst Prosperity: Immigration and the Changing Face of Rural California, The Urban Institute Press, Washington, D.C. VERNEZ, G. and ABRAHAMSE, A. (1996), How Immigrants Fare in US Education, RAND, Santa Monica. WALDINGER, R. (1996), Still the Promised City? African Americans and New Immigrants in Post-industrial New York, Harvard University Press, Cambridge. WALDINGER, R. and BOZORGMEHR, M. (1996), Ethnic Los Angeles, Russell Sage Foundation, New York.

74

Table 1. Immigrant admissions by major category (Fiscal years 1992-96) TOTAL • SUBJECT TO THE NUMERICAL CAP Family-based immigrants Immediate relatives of US citizens Children born abroad to alien residents Family-sponsored immigrants Legalization dependents Employment-based preference Total employment, 1st preference Aliens with extraordinary ability Outstanding professors or researchers Multinational executives or managers Spouses and children of 1st preference Total employment, 2nd preference Members of the professions holding advanced degrees or persons of exceptional ability Spouses and children of 2nd preference Total employment, 3rd preference Skilled workers Baccalaureate holders Spouses and children of the above Chinese Student Protection Act Principals Spouses and children Other workers (unskilled workers) Spouses and children of unskilled workers Total employment, 4th preference Special immigrants Spouses and children of 4th preference Total employment, 5th preference Employment creation not targeted area Spouses and children Employment creation targeted area Spouses and children Total Pre-1992 3rd and 6th preference Principals Spouses and children Diversity programs • NOT SUBJECT TO THE NUMERICAL CAP Amerasians Cuban/Haitian entrants Parolees Soviets and Indochinese Refugees and asylees Refugee adjustments Asylee adjustments Registered nurses and their families Registry entered prior to 1/1/72 Other

1992 810 635 655 541 502 995 235 484 2 116 213 123 52 272 116 198 5 456 261 319 1 446 3 430 58 401 27 503

1993 880 014 719 701 539 209 255 059 2 030 226 776 55 344 147 012 21 114 1 259 1 676 5 088 13 091 29 468 13 801

1994 798 394 662 029 497 682 249 764 1 883 211 961 34 074 123 291 21 053 1 313 1 809 4 975 12 956 14 432 6 807

1995 716 194 593 234 460 653 220 360 1 894 238 122 277 85 336 17 339 1 194 1 617 3 922 10 606 10 475 4 952

1996 909 959 771 604 595 540 350 192 1 658 293 751 184 117 346 27 501 2 060 2 633 6 354 16 454 18 462 8 870

30 898 47 568 12 257 4 192 22 187 4 017 4 915 4 063 1 768 2 295 59 21 32 3 3 651 227 424 36 348 155 094

15 667 87 689 12 813 9 560 28 434 26 915 26 852 63 4 405 5 562 8 158 3 576 4 582 583 159 311 37 76 33 480 160 313

7 625 76 956 10 139 7 732 28 398 21 297 21 008 289 4 136 5 254 10 406 4 647 5 759 444 106 190 51 97 41 056 136 365

5 523 50 245 9 094 5 792 23 262 4 213 4 134 79 3 636 4 248 6 737 2 929 3 808 540 95 190 79 176 47 245 122 960

9 592 62 756 16 001 5 507 28 998 401 373 28 6 010 5 839 7 844 3 494 4 350 936 143 301 152 340 58 718 138 323

17 253 99 13 661 117 037 106 379 10 658 3 572 1 293 2 179

11 116 62 15 772 127 343 115 539 11 804 2 178 938 2 904

2 822 47 8 253 121 434 115 451 5 983 304 667 2 838

939 42 3 120 114 632 106 795 7 837 69 466 3 692

954 29 2 283 128 367 118 345 10 022 16 356 6 318

Source: Immigration and Naturalization Service, Statistics Division.

75

Table 2. Leading country of birth and State of destination: Immigrant admissions (Fiscal year 1996) Admissions

Percent of total

Cumulative percent

159 731 55 778 44 781 42 006 41 662 39 516 26 415 21 051 19 646 19 029

17.6 6.1 4.9 4.6 4.6 4.3 2.9 2.3 2.2 2.1

17.6 23.7 28.6 33.2 37.8 42.1 45.0 47.4 49.5 51.6

199 221 153 731 82 229 79 067 63 162 42 154 23 017 21 329 20 683 18 718

22.0 16.9 9.0 8.7 6.9 4.6 2.5 2.3 2.3 2.1

21.9 38.8 47.8 56.5 63.5 68.1 70.6 73.0 75.2 77.3

Country of birth Mexico Philippines India Vietnam Mainland China Dominican Republic Cuba Ukraine Russia Jamaica State of destination California New York Texas Florida New Jersey Illinois Massachusetts Virginia Maryland Washington

Source: Immigration and Naturalization Service, Public Use Admissions Data.

Table 3. Major age groups by class of admission: Immigrant admissions - fiscal year 1996 (Percentages) Age group 15 years or less 16 - 29 years 30 - 44 years 45 - 64 years 65 years or more Total

Total

Family

Employment

Diversity

Humanitarian

22 31 27 15 5 100

23 34 23 14 5 100

20 23 44 12 100

22 33 34 10 1 100

20 27 24 21 8 100

Source: Immigration and Naturalization Service, Public Use Admissions Data.

76

Table 4. Immigrant admissions in Fiscal year 1996 by occupation: predicted and actual Occupation

Predicted

Actual

Relative growth (percent)

WHITE-COLLAR Professional, technical, and kindred health professionals Other professionals Technical and specialty Executives Sales Administrative support

10 244 9 231 22 115 20 283 12 943 19 437

18 986 19 477 33 117 30 702 13 002 19 807

85.3 111.0 49.7 51.4 0.5 1.9

BLUE-COLLAR Precision production Operators, fabricators and laborers OTHER

21 028 37 702 12 251

20 116 53 936 12 588

-4.3 43.1 2.8

WITH OCCUPATION (Total)

165 234

221 731

34.2

NO OCCUPATION (Total)

261 694

498 583

90.5

GRAND TOTAL

426 928

720 314

68.7

Note: Changes due to the Immigration Act of 1990. Predicted numbers in Fiscal year 1996 are based on linear projections (from the years 1972 to 1991), and are kept within numerical limits on nonexempt categories. Humanitarian admissions are not included.

Source: Greenwood and Ziel (1997).

77

Figure 1.

Figure 2.

78

Figure 3.

79

MIGRATION TRENDS AND POLICIES IN CANADA

by Roderic Beaujot University of Western Ontario, Canada The importance of immigration for Canada is best symbolised by the policy requirement for a yearly “Statement to Parliament” in the form of an Annual Immigration Plan. This plan must include an estimate of the total number of immigrants for the next year. While the Immigration Act covers the key issues involved in the administration of immigration, it also requires an annual review of the situation, and it leaves the crucial question of the number of immigrants along with their composition by class (i.e. independent, family, or refugee) to this yearly statement. This makes these annual statements to Parliament a useful source of information on the evolving trends in immigration policy. After reviewing policy issues, focusing on classes of immigrants, this paper will summarise migration trends in terms of overall numbers, and in terms of composition according to demographic, economic, and cultural characteristics. 1.

Planned and actual immigration levels

In Canada, the number of admissions is not subject to limits or quotas. The number accepted simply reflects those who meet the relevant requirements for entry. As a result, the Annual Immigration Plan merely indicates what is expected to happen. It is important for the Ministry to appear to be in control of immigration, and consequently the plan needs to anticipate as best as possible what will in effect happen over the coming year. Bearing these qualifications in mind, it is interesting to compare the anticipated levels and actual arrivals during the period that the current Immigration Act has been in effect, i.e. since 1978 (Table 1). The announced levels, which have ranged from 85 000 to 250 000, have closely matched 1 actual arrivals. The greatest consistency occurred during the 1990s which may have been a function of 1993 changes in the Immigration Act that sought to achieve better management of immigration, partly by establishing priorities within the various classes. Trends in the planned immigration levels can also be interpreted in the context of the political parties that have formed the government. The Liberal party was in power until 1984 and it sought to bring immigration down during the recession of the early 1980s, especially as high levels of unemployment persisted. The lowest immigration level was reached in 1985 when the recession was over and the Conservative party had formed the government. The Conservative party clearly indicated that it felt the Liberals had made a mistake in reducing immigration to such low levels and it soon anticipated levels of 250 000 which were maintained throughout the recession of the early 1990s. While there has been less change with the return of the Liberal party since 1993, both the announced and actual arrivals have been slightly lower than they had been during the Conservative regime. 81

Immigration has not been a major issue in elections, but it has been discussed. The parties typically 2 indicate their orientation regarding levels — which do not differ significantly. In fact, political discussions of immigration are sensitive because arguments for lower levels can be seen as racist opposition to an immigration inflow that is largely non-European. Consequently, the policy statements tend to focus only on the positive side of immigration, and discussions soon turn to means of public education that would enable the population to appreciate these positive aspects. This relative political consensus, however, somewhat masks the ambivalence of public opinion. In a series of eight polls taken in 1995 to 1997, there are as many who think that there are “too many” immigrants in Canada as those who say that the level is “about right”, and only about 6 to 9 per cent who think there are “too few” (Palmer, 1997). The commissioned analysis concludes that support for current levels policy is weak Negative opinions concerning immigration focused largely on the economic situation, especially high unemployment, while respondents are more likely to express an appreciation for the cultural contribution of immigrants. It was also noted that attitudes are more favourable among persons who are more educated, younger, employed, and living in larger urban centres. Given the sensitivity surrounding racial or ethnic questions, it is difficult to properly gauge the cultural determinants affecting views towards immigration. While many may be concerned about the change that immigration is bringing to the country, the politically correct response is to focus on the richness of ethnic diversity. It may be for these reasons that the concerns that are raised largely focus on economic questions, and unemployment in particular. Actually, there has been a strong correlation between unemployment and the immigration level of the following year during the period 1946-89. In a time series analysis, the change in unemployment explains 40 per cent of the variation in immigration arrivals (Veugelers and Klassen, 1994). The association is even stronger over the period 1978-89, when annual levels were specifically set. However, these authors observe a de-linking of immigration levels from unemployment in the subsequent period. In the 1990s, immigration did not decline with the recession of the early part of the decade nor during the subsequent period of high unemployment. Veugelers and Klassen (1994) speculate that demographic considerations may be playing a larger role than economic ones. Given the context of low population growth, however, Foot (1994) and Stafford (1993) propose that political considerations have played an important role, with parties sensitive to the influence of groups representing interests in the immigration area. Alternatively, Simmons (1994) argues that globalisation has changed the dynamics of the situation, especially in terms of the reduced power of labour interests. 2.

Objectives underlying immigration

It is difficult to disentangle policy trends, partly because a number of issues and concerns are involved in immigration. Several of these are specifically stated in the 1978 Immigration Act. The objectives stated there have not been touched in subsequent revisions of the Act. They identify the primary objective of addressing Canada’s demographic and economic needs while supporting family reunification, respecting standards for non-discrimination, and promoting the integration of new arrivals through the co-operative efforts of government and non-government agencies. In addition, the Act recognises Canada’s obligation to accommodate refugees and other humanitarian immigrants 3 while selecting out persons likely to engage in criminal activities. The 1994 “Strategy for immigration and citizenship” is the product of an extensive set of consultations carried out in that year. According to the summaries of these consultations, the majority of experts agreed that immigration should continue to play a role in shaping Canada’s future, 82

but some concerns were raised regarding costs and economic questions: integration is necessary but also expensive, the independent class should be a larger proportion of the total, there should be a reduction in the category of family class sponsorships, humanitarian programs should be better managed to reduce costs, and unemployment levels should be more explicitly incorporated into planning (Citizenship and Immigration, 1994c). The language of the annual statements to parliament over the last several years also makes a strong case for immigration, associating it with “nation 4 building” and a “stronger Canada”. Nevertheless, in his review of “Canadian immigration policy since Confederation,” Whitaker (1991) is able to say that the successive governments have never been able to clearly define the need for immigrants. Immigration has been linked to evolving short-term conditions, yet immigrants are largely permanent additions to the population. Moreover, the statements to parliament have not acknowledged some of the significant concerns raised by the Royal Commission on the Economic Union and Development Prospects for Canada. These important changes in Canada’s racial and ethnic composition will continue to transform our economic and political life in the coming decades. They are also likely to generate a certain amount of social conflict, and future generations of Canadians will need to invent new policies and techniques for coping with the stresses of a vibrant and dynamic multi-cultural society. While Canada gains by following an occupationally selective system of immigration, this course runs the risk of making us more dependent on immigration for certain types of skilled workers. Canadians may find that particular occupations become difficult to enter, for training and apprenticeship programs may become scarce if employers can rely on pretrained immigrants. The creation of a harmonious multi-cultural and multi-racial society will require a high degree of tolerance and civility from all Canadians, reinforced by policies aimed at preventing foreseeable conflicts. The political viability of a less restrictive immigration policy will depend on our capacity to deal with the domestic challenges which are likely to flow from it. (Royal Commission on the Economic Union and Development Prospects for Canada, 1985) Besides the effect that immigration may have on training programs within the country, as suggested above, some macro-level studies conclude that immigration has a rather small effect on the economy. In particular, the major study by the Economic Council of Canada (1991) found that the efficiency gains achieved by the introduction of 100 000 more immigrants only increased real incomes by some one per cent, while the tax and dependency benefits lowered costs by 0.3 per cent of per capita income. It was concluded that the case for immigration should not be made in economic terms, but on humanitarian and social grounds, in terms of the diversity that helps make Canada a more interesting and exciting society. In this vein, the edited collection by DeVoretz (1995), entitled Diminishing Returns: The Economics of Canada’s Recent Immigration Policy, suggests that while the economic returns from immigration are still positive, they are diminishing and differ significantly by region. In terms of demographics, as will be further elaborated below, immigration clearly adds to population growth, but it does not significantly mitigate population ageing, and it accentuates inequalities in regional distribution.

83

3.

Immigration by main categories

While the 1978 Immigration Act solidified the separate treatment of independent, family, and refugee classes of immigrants, these actually have a longer tradition. In the immediate post-war period, Prime Minister MacKenzie King wanted to strengthen immigration without offending Quebec which had been frustrated with the fact that immigration had not served the interests of French Canada. At the height of the conscription crisis in 1944, the Quebec Legislative Assembly had indicated its opposition to mass immigration. However, the French could hardly object to family reunification, and consequently it was a significant way to circumvent this opposition. The family relationships that could be used were defined broadly, to the point that it was widely suggested in the 1950s that for every Italian who came another 49 gained the possibility of admission. Over the years, the family sponsorship categories have become more limited. As of 1993, the eligible categories are as follows (Citizenship and Immigration, 1997d): − spouse or fiancé(e); − dependent children including children under 19 who are to be adopted; − parents and grandparents; − brothers or sisters, nephews or nieces and grandchildren who are orphans, unmarried and under 19; − any other relative providing that the sponsor has none of the above and has no family in Canada. Other relatives who are applying as independent applicants benefit from 5 points (out of the 70 needed to qualify) for having relatives in Canada. Another change introduced in 1997 makes the financial responsibilities for family sponsorship more stringent. A study of family sponsorships in Toronto had found that 14 per cent were on social assistance (Citizenship and Immigration, 1997a). There are also minimum financial requirements for prospective sponsors. However, these are quite low, amounting to a combined $40 000 income for a couple with two children in a large city who are sponsoring two grandparents. In addition, the financial criteria do not apply to people sponsoring spouses or children under 18. Studies indicate that sponsorship agreements are more likely to breakdown when the sponsored relative(s) live separately from the sponsors (Thomas, 1996). This may occur through subsequent separation or divorce. Other cases involve single parents who bring their young dependent children to Canada and who may consequently become less able to maintain full-time employment. The category of “independent” immigrant was initially defined in 1967 when a points system determined whether people had the necessary qualifications. In the immediate post-war period, there was a preference for immigrants from Britain, France, and the United States. In 1962, it was decided that place of origin should not be the basis of selection and that other qualifications should be used. These were first established in 1967 based on age, education, language, occupation, skills, and adaptability. Various changes have been made over the years, including the use of the points system to restrict the arrival of immigrants in the early 1980s by requiring that applicants have a job opening in Canada for which there was no eligible Canadian. In other words, the points system can be used to disqualify applicants by requiring that they have at least one point in a given area. As of 1994, the points system requires that applicants obtain at least 70 points in areas ranging from education and experience to personal suitability (Table 2).

84

The occupational criteria has been broadened over time. It was found that using specific occupations led to frustrations as people were not able to find employment in the occupation for which they had been selected. It was argued that selecting well-qualified and well-educated applicants, rather than using specific occupational qualifications, would provide greater flexibility (Royal Commission, 1985). The occupational demand list now contains some 180 broad occupational groups, covering about 2 400 eligible occupations. In addition, persons who are self-employed can obtain a bonus of up to 30 points. Other recent changes have placed more emphasis on language, and the maximum points are now obtained for age category 25-35 which is younger than in previous selections. Since 1984, there has been special treatment for business immigrants either as entrepreneurs or investors. While other applicants need to reach 70 points, business immigrants need only obtain 25 points. Entrepreneurs need to be starting or managing a business that will employ at least one person besides themselves and their dependants, while investors need to have a net worth of at least $500 000 of which they will invest about $350 000 in Canada (Citizenship and Immigration, 1997d). The Canadian experience with refugees also has a long history. Displaced persons from Europe were accepted in the post-war period and subsequently persons from the Hungarian revolution, and Czechs, Slovaks, US draft dodgers and deserters, Tibetans, Ugandan expelees of Asian ancestry, Chileans, and Indochinese (Ziegler, 1988). These refugee arrivals involved special arrangements in response to specific circumstances, but since the 1978 Immigration Act, refugee admissions are expected on an annual basis. Initially, refugees were selected abroad, but since the early 1980s many have sought refugee status on arrival in Canada. While Canada is geographically removed from refugee producing areas, it is often viewed as an ideal location because it not only grants asylum but also admits refugees as immigrants. A 1985 ruling determining that refugee claimants were subject to the legal protection of the Charger of Rights and Freedoms has required various legal arrangements in order to determine the legitimacy of claims for refugee status. Over the period 1978-96, refugees constituted about 15 per cent of immigrants. The arrival of refugees in the mid-1980s significantly outpaced the ability to handle claims. In 1989, an Immigration Refugee Board was established as a quasi-judicial administrative tribunal. A “backlog clearance” that started in 1989 managed to complete the 95 000 cases of the 1986-88 period by June 1993 (Ruddick, 1996). In 1995, there were a total of 25 900 claims, with another 12 800 in the first half of 1996. The acceptance rate is considerably higher than that of any other country, about 45 to 50 per cent (Shenstone, 1997). In addition, the removal of persons who were not accepted has been very slow, to the point where this matter was underscored in the 1997 report of Canada’s Auditor General. Shenstone (1997) notes that this situation may turn public opinion against refugees and even immigrants in general. He also suggests that there is “some hypocrisy, perhaps unavoidable, in our policy of vaunting our welcome to genuine refugees who manage to reach Canada while doing our level best through control measures overseas to prevent irregulars (who necessarily include most refugee claimants) from getting here”. This prevention includes imposing visa requirements on countries from which refugee claimants are arriving, as occurred most recently for the Romany (Gypsies) from the Czech Republic. Large numbers of admissions through claims made in Canada also undermines the potential to accept more persons who are in refugee camps abroad, for whom the refugee status is easier to determine. With regards to admissions by class, there was a general orientation through the 1980s that sought to increase the relative size of the independent class of immigrants who are specifically selected for their skills and other qualifications (Table 3). However, the reductions in immigration in the mid-1980s necessarily affected the skilled worker categories, since they were required to already have a job offer in Canada. In 1994, the Government specifically undertook to raise the independent class to 52 per 85

cent of the total for 1997-99 (Citizenship and Immigration, 1994a). This target was already reached in 1996, partly as a function of the stronger management that was introduced in 1993. If the breakdown by class for the year 1996 is considered in greater detail, one notes that 52 per cent of the family class involves spouses and fiancés, with another 8 per cent involving children, and 36 per cent involving parents or grandparents (Table 4). Clearly, the family class is largely limited to very close relatives. To the refugee class must be added the “deferred removal” category since these persons were admitted after their removal had been delayed for a period of at least three years. It should be noted that 66 per cent of the refugee class were already on Canadian soil. The independent class is broken down into skilled workers, business class and the special category of “live-in caregiver” (which is not subject to the points system). The business class accounts for 18 per cent of the total number of independent immigrants. However, it should be remembered that only the principal applicant is subject to the points system. With each business immigrant there are 2.6 dependants, and with each skilled worker there are 1.3 dependants. While the independent class (including business, skilled workers and live-in caregivers) amounted to 56 per cent of 1996 immigrants, only 21.4 per cent of total immigrants had been selected through the points system. The proportion of immigrants selected through the points system has increased slightly to 18.8 per cent in 1995, an initial drop from 19.0 per cent in 1988 to 16.1 in 1994 (Beaujot, 1991; Citizenship and Immigration, 1997c). In short, families of immigrants do not arrive exclusively through the family class of admissions, but also as the dependants of principal applicants. 4.

Immigration and socio-demographic trends

The trends in immigration show considerable annual variations (Figure 1 and Table 5). Taking the period since 1850, four phases can be distinguished. The period 1850-96 during which world trade suffered a long period of depression can be described as one of low immigration, in fact of net outward migration, as there was more movement toward the industrialising states of New England (Beaujot, 1991). The first wave of post-Confederation immigration rose from a low point of 17 000 in 1896 to reach 400 000 in 1913. In fact, the numbers for the years 1910-13 have never been surpassed. A phase of relatively low immigration followed with two world wars and the depression period of the 1930s, although immigration was higher in the 1920s than in the remaining part of this period. The fourth phase brought a second wave of post-Confederation immigration after 1945. While there are fluctuations during the post-war period, it appears that the 1990s have brought higher levels of migration and that immigration has made a larger contribution to population growth. In the period 1951-91, net migration accounted for about a quarter of population growth, but this reached 51 per cent in the period 1991-96. While in some regards the period before the first world war remains unique, it is interesting to observe that only the six consecutive years 1909-14 had levels above 150 000 while there have now been eleven years with such levels (1987-96). Using the symbolic figure of 200 000, there were four consecutive years in the earlier period (1910-13) but eight with these levels in more recent years (1990-97). This means that the 1996 census shows a large number of recent immigrants. In fact, the number who arrived in the period 1991-96 (1 166 thousand) was almost equal to that of the ten-year period 1981-90 (1 092 thousand) (Table 5). Despite fluctuations in immigration, the total number of immigrants as a percentage of the population has been much more stable. After remaining at 22 per cent in the censuses from 1911 to 1931, the percentage of foreign born has increased slowly from 14.7 in 1951 to 17.4 per cent in 1991 (Figure 2). There is, however, much more variation in the geographic distribution of immigrants over provinces

86

(Figure 3) and census metropolitan areas (Figure 4). About a quarter of the populations of Ontario and British Columbia are comprised of immigrants, compared to under 5 per cent for the Atlantic provinces. The immigration of the post-war period has been largely a metropolitan phenomenon, but it has mostly affected the metropolitan areas west of Quebec, plus Montreal. In this regard, two metropolitan areas are unique: Toronto has 41.9 per cent foreign born and Vancouver has 34.9 per cent. In fact, a fifth of the 1996 populations of Toronto and Vancouver consist of immigrants who arrived after 1981 (Statistics Canada, 1997b). The three largest metropolitan areas, Toronto, Montreal, and Vancouver; have the largest numbers of immigrants, but the proportion of foreign born in Montreal is very close to the national average. Immigrants to the Province of Quebec are very concentrated in its only metropolitan area, Montreal. In most ways, the differences represented by immigrant populations tend to lessen over time (Beaujot, 1998). For instance, their fertility and mortality rates come to resemble those of the Canadian born, as do their economic characteristics. Even the visibility of minorities lessens over time as styles of dress and speech become more similar over the course of residence in the receiving country; this is especially true of the second generation. Age and sex characteristics of immigrants also reveal important information. Earlier periods involved a higher proportion of male immigrants, but the masculinity ratio of persons arriving over the period 1981 to 1990 was 0.97. This is very similar to the ratio for the entire Canadian population, which was also 0.97 in 1991 (Beaujot, 1998). Of greater significance, is the fact that women are more likely to arrive as dependants than as principal applicants. This means that they are less likely to have been selected through the points system. In terms of age, it is sometimes suggested that immigration might help overcome population ageing. The impact of immigration on the age structure can best be appreciated by comparing the median age of immigrants at arrival to that of the Canadian population. The median age of immigrants has been relatively stable, averaging 25 years for each year between 1956 and 1976, then increasing to 27 years in 1981-86, 28 years in 1986-90 and 30 in 1994 (Beaujot, 1997; Citizenship and Immigration, 1997b). The median age of the entire population rose from 26.3 in 1961 to 35.3 in 1996. The age of both immigrant arrivals and the receiving population has risen, but arrivals remain younger on average. However, the overall impact is rather small. For instance, in projections based on the 1991 census, an immigration level of 330 000 per year produces a median age of 40.7 in 2026, compared to a median of 43.7 with an immigration level of 150 000 (Statistics Canada, 1994). As of the 1991 census, non-permanent residents have been included as a separate population count. This category is largely comprised of persons on temporary work permits, foreign students, and persons awaiting the determination of their refugee status. Estimates dating back to 1977 indicate that the numbers have increased from 97 000 in 1977 to a high of 448 000 in 1990, but have since declined to 259 000 in 1996 (Statistics Canada, 1997a; see also Dumas and Bélanger, 1997). The current levels therefore represent about one year’s immigrant arrivals. The censuses obtained lower figures, with 223 400 in 1991 and 166 715 in 1996. It is likely that some registered themselves as immigrants while others were missed by the census. The specific category of temporary entry under FTA and NAFTA has involved small numbers, increasing from 6 223 persons in 1992 to 9 985 in 1995. Some 80 per cent of these are professionals, with most of the remainder being intra-company transfers (Ruddick, 1996).

87

Another way to appreciate the demographic impact of immigration is through population projections. There have now been five generations of projections by Statistics Canada following the censuses. The national level assumptions for fertility and immigration are shown in Table 6. As can be seen, except for the projection following the 1991 census, the immigration assumptions have not involved a very wide spread. The projections from the 1986 census indicated that with a fertility constant of 1.7 births per woman, the natural increase would become negative around 2020, and that population would start to decline after 2026 with immigration levels of 140 000. The decline would be delayed until 2035 with immigration levels of 200 000. These results and subsequent analyses on questions like labour force growth may have been part of the motives behind raising immigration levels in the late 1980s. Indeed, the projections based on the 1991 census indicate that a total fertility rate of 1.5 and immigration of 150 000 would show population decline only after 2033. The combination of fertility at 1.7 and immigration at 250 000 shows continued growth to the end of the projection period, that is 2041. 5.

Immigration and socio-cultural composition

For the first half of the century, immigration to Canada was largely of European origin, with British immigrants comprising more than half of all arrivals (Beaujot and McQuillan, 1982). When Prime Minister Mackenzie King set the course of post-war immigration policy, he proposed that immigration should not change the “character of the population.” That policy largely remained in effect until discrimination by place of origin was removed in 1962. However, the subsequent change was slow, and it was not until 1975 that half of all immigrants were coming from Asia, Latin America, and Africa. The 1996 census indicates that a new threshold has been reached, with more than half of the immigrant stock being of non-European origin (Statistics Canada, 1997b). Compared to their presence in the Canadian born population, English and French ethnic origins are now much less represented among immigrants (Badets and Chui, 1994). This is especially true of French ethnicity: among persons who declared a single ethnic origin, only 1.7 per cent of foreign born indicated French, compared to 27.1 per cent of Canadian born. More important for the cultural make-up of the country, immigration has largely contributed to the relative strength of the English language. While a variety of factors contribute to the changing language composition, including differential natural increase and language transfer, immigration is the main component introducing a shift toward English at the national level (Lachapelle, 1988). Immigrants to Quebec continue to associate with the English language to a larger extent than the Canadian born population of the region, although since the mid-1970s, immigrants who arrived and stayed in Quebec are more likely to associate with the French than the English language (Beaujot, 1998). At the time of arrival, significant numbers of immigrants know neither of the official languages. This was the case of 30 per cent of arrivals in the 1969-77 period and of more than 45 per cent of those in the 1978-90 period (Thomas, 1992). Among arrivals for 1996, 51.5 per cent knew English, 4.3 per cent French, 2.9 per cent both, and 41.1 per cent neither language (Citizenship and Immigration, 1997c). Over time, most immigrants become adept in at least one of the official languages. Yet, the 1991 census showed that 5.8 per cent of male and 10.7 per cent of female immigrants who had arrived in the period 1981-85 did not know either language well enough to carry on a conversation (Beaujot, 1998).

88

Canadian statistics also follow numbers of visible minorities, defined as non-white, non-Caucasian, and non-aboriginal. This component amounted to 4.7 per cent in the 1981 census and 11.2 in 1996. The differential in relative growth is basically a function of immigration, with projections indicating around 20 per cent visible minorities by 2016, assuming current trends (Dai and George, 1996). However, these projections assume the retention of visible minority status, both over generations and after inter-marriage. Specifically considering the population born in the United States and Mexico, there is slight decline in the former and an increase in the latter (Table 7). While the population exchange with the United States has been significant, Samuel et. al. (1995) conclude that there has not been a significant level of population exchange with Mexico. Mexico does not figure among the 26 countries from which Canada received at least 2 000 immigrants in both 1994 and 1995 (Dumas and Bélanger, 1997). Nonetheless, the period 1980-92 showed considerable growth in non-permanent persons from Mexico, while the United States showed a slight downward trend over the period 1981-95 (Dumas and Bélanger, 1997). 6.

Immigration and socio-economic status

The extensive studies made on the relative socio-economic status of the immigrant population point to clear income disadvantages at the time of arrival. The key question thus becomes the extent to which these are overcome. In terms of education, immigrants are more concentrated at the two extremes of the distribution. However, the greater concentration at higher levels of education means that on average the foreign born are more educated, and immigrants of the 1981-91 period have higher average education than the Canadian born (Badets and Chui, 1994). Counting both principal applicants and dependants, 67.4 per cent of 1996 immigrants intended to work at the time of arrival (Citizenship and Immigration, 1997c). Compared to the Canadian born, both men and women immigrants have lower labour force participation at ages 25-44 but higher participation at ages 45-64. In each age cohort, immigrants who had arrived in the ten years preceding the 1991 census had higher levels of unemployment than non-immigrants, but unemployment rates were typically lower for cohorts with longer periods of residence (Citizenship and Immigration, 1994c). The occupations of immigrants are not particularly concentrated in any one area. For instance, in 1991, 41 per cent of immigrant arrivals intended to join the top ten occupations held by 57 per cent of the total labour. Both in terms of total income and average income, adjusted for age, the overall averages for men and women between the Canadian born and foreign born populations in 1991 are very close (Beaujot, 1998). However, this result hides considerable differences by arrival cohort and place of origin. Groups that arrived before 1971 for men and before 1976 for women have average incomes higher than that of the Canadian born population, while more recent immigrant cohorts have a disadvantage. Another interesting change has taken place over these same cohorts: the average income of immigrants from Europe and the United States moved from being less than to being more than that of immigrants from other continents. There is a strong disadvantage for post-1976 cohorts who are not from Europe and the United States, that is, who are largely visible minorities. For persons admitted in the 1980s, the income disadvantage was more than 20 per cent at the time of the 1991 census.

89

A comparison of the 1961 and 1971 censuses showed a very encouraging outcome for post-war immigrants (the 1946-60 cohort). In the majority of age-sex groups, the average income in 1961 was lower than that of the Canadian born, but by 1971 these groups had largely exceeded the averages of their Canadian-born counterparts (Richmond and Kalbach, 1980). Few similar transitions occurred over the 1971, 1981, and 1986 censuses (Beaujot and Rappak, 1990). It would appear that the immigrants of cohorts since 1976, despite their educational advantage, may not reach the average income of persons born in Canada. The situation is more positive for immigrants from Europe and the United States, who exceed the Canadian born average except in the 1986-89 cohort for men and the 1976-89 cohorts for women. For the majority of recent immigrants, that is those from Asia, Latin America, and Africa, who are largely visible minorities, the 1976-89 cohorts show serious disadvantages at the time of the 1991 census. On the other hand, in the cohorts that preceded 1970, the immigrants from these continents have an average income that is typically superior to the average for immigrants from Europe and the United States. It is worth noting that the region of residence in Canada, along with other socio-cultural characteristics, has important implications for these income differences. The disadvantage of recent cohorts, especially those of non-European origin, is greater among residents of metropolitan areas (Beaujot, 1997). Except for the most recent arrivals, immigrants living in non-metropolitan areas had average incomes that compared favourably to the Canadian-born in the same areas. Other results show that the advantages of immigrants of European origin apply to the British but not to the Italians. At the same time, the disadvantage of non-European immigrants does not apply to Chinese arrivals of the 1961-80 period. Also, immigrants who speak one of the national languages at home have average levels of income that compare favourably to that of the Canadian born. Overall, immigration enhances Canadian averages as far as socio-economic status is concerned, but differences are small. More recent cohorts are more educated, but there is also a higher proportion of immigrants with lower levels of education. The cohort differences by labour force participation and average income clearly show that the strong negative differential at the time of arrival is reduced over time. However, the income disadvantages of the more recent cohorts are not being reduced as quickly, especially for those who are not from Europe or the United States. It is not even clear that these cohorts, in spite of their educational advantages, will ever arrive at the average income levels of the Canadian born. 7.

Conclusion

International migration plays an important role in the evolution of societies (Piché, 1997). In a globalising world, migration is both an outcome of development and a possible support for further development. Consequently, it is important for receiving countries to benefit from the diverse resources that immigrants represent, while maintaining social cohesion by successfully integrating them (Citizenship and Immigration, 1996). In the Canadian case, immigration has even played an important demographic role, constituting more than a fifth of the nation’s population growth over this century, and half of the growth of the 1991-96 period. It has also accentuated the concentration of foreign born in the largest metropolitan areas. Immigration clearly brings cultural diversity, although this diversity decreases over time as integration occurs. In this regard, it is possible to say that immigration has worked better for English Canada than for Quebec. The foreign born have been less likely to integrate with the French speaking majority of Quebec. Persons of third languages retain their languages longer in Quebec, partly as a function of the larger potential for ethnic separateness provided by French-English tensions in the province. Compared to the immigrants in English 90

Canada, those in Quebec maintain a greater linguistic distinctiveness from the French majority of the province. From an economic perspective, it can be argued that the immigration of the post-war period, 1945-65, helped to overcome some of the significant skill shortages that resulted from the relatively underdeveloped education and apprenticeship system. However, this was less the case in the period 1965-85 when the education levels of the Canadian population increased and there was significant labour force growth associated with the entry of baby boomers and women. The 1990s have seen slower labour force growth, which can in part be attributed to immigration. In spite of predictions to the contrary, there have been difficulties absorbing the smaller “baby bust” cohorts into the labour force, and unemployment has persisted at high levels. It is possibly for these reasons that studies on the economic impact of immigration have come to show neutral or less substantial consequences. For instance, the net transfers to the public treasury are found to be higher for 1966-70 immigrants than for those of 1981-90 (Akbari, 1995). In a time series analysis, Marr and Siklos (1995) find a small but statistically significant relationship between increased immigration and higher unemployment. DeVoretz (1995) further proposes that expanding immigration in the early 1990s benefited British Columbia, but that it exacerbated the unemployment situation in central Canada. As the Economic Council of Canada (1991) concluded, it is especially in socio-cultural terms that a case for immigration should be made. As long as integration can ensure equal opportunity, immigration brings diversity, richness, and contact with a broader world. However, it is also important to maintain a vigilant watch for possible socio-cultural disintegration, and to be ready to ensure that immigration levels do not move beyond an optimal level.

91

NOTES

1.

Taking the mid-point when a range was used, there were twelve years when the actual arrivals were higher than anticipated (1979, 1980, 1987-93, and 1995-97) and seven years when the they were lower than anticipated (1981-86 and 1994). In the early 1980s the actual arrivals were even lower than the declining anticipated levels, while in the late 1980s the arrivals increased more rapidly than anticipated levels. Partly through the use of a range, and with the exception of 1994, the announced and actual levels have not differed extensively in the 1990s.

2.

For instance, in the 1993 election, all but the Reform party indicated that levels should be at 250 000 or more. The Reform party proposed the figure of 150 000, which is close to the long-term post-war average. In its policy book for the 1996 election, the governing Liberal party used the figure of 1 per cent of the receiving population, which would mean 300 000. However, in the context of the evolving economic situation, it has reduced the numbers from the higher levels of the early 1990s. The Reform party is the Official Opposition since 1996, but it has not extensively publicised its policy orientation towards lower immigration levels. For instance, when the 1998 levels were announced, the newspapers did not carry the Reform party alternative.

3.

“It is hereby declared that Canadian immigration policy ... shall be designed and administered in such a manner as to promote the domestic and international interests of Canada recognising the need: a) to support the attainment of such demographic goals as may be established by the Government of Canada from time to time in respect of the size, rate of growth, structure and geographic distribution of the Canadian population; b) to enrich and strengthen the cultural and social fabric of Canada, taking into account the federal and bilingual character of Canada; c) to facilitate the reunion in Canada of Canadian citizens and permanent residents with their close relatives from abroad; d) to encourage and facilitate the adaptation of persons who have been granted admission as permanent residents to Canadian society by promoting co-operation between the Government of Canada and other levels of government and non-government agencies in Canada with respect thereto; e) to facilitate the entry of visitors into Canada for the purpose of fostering trade and commerce, tourism, cultural and scientific activities and international understanding; f) to ensure that any person who seeks admission to Canada on either a permanent or temporary basis is subject to standards of admission that do not discriminate on grounds of race, national or ethnic origin, colour, religion or sex; g) to fulfil Canada’s international legal obligations with respect to refugees and to uphold its humanitarian tradition with respect to the displaced and the persecuted; h) to foster the development of a strong and viable economy and the prosperity of all regions in Canada;

92

i) to maintain and protect the health, safety and good order of Canadian society; and to promote international order and justice by denying the use of Canadian territory to persons who are likely to engage in criminal activity.” 4.

“It is, perhaps more than any other area of public policy, closely tied to our history, to the development of our values, our cultural diversity, the regional and economic dynamics of our nation. It will, in the long run, help to determine the character of our country in the years ahead” (Citizenship and Immigration, 1994b). “Immigration is among the most compelling issues with which we as a country are faced. It speaks to who we are. It reflects our history as a nation and it heralds our future as a society. Immigration policy is much more than an annual ritual of announcing new levels for the coming twelve months. It is the vehicle that will carry Canada into the 21st century and beyond. Our challenge is to ensure that the path we choose today leads to a sustained and improved prosperity for all Canadians” (Citizenship and Immigration, 1994c: iv). “Immigration is necessary for Canada’s growth. At the same time, we must be sensitive to the need to balance our demographic and economic needs with our capacity to settle and absorb immigrants. I think that the data and studies speak for themselves. Immigrant investors injected nearly $606 million in investment capital into our country last year. Current studies show that immigrants account for about half of Canada’s labour force growth. This contribution is expected to increase as we move into the 21st century. I know we’ve all heard people say that immigrants take jobs away from Canadians. That is one myth that must be laid to rest. Evidence suggests the contrary. Immigrants create at least as many jobs as they take. Immigrants and newcomers also enhance Canada’s ability to expand into global markets and to build trade links throughout the world” (Statement made by Lucienne Robillard accompanying the 1997 plan). “Our 1998 Immigration Plan is based on the humanitarian values of welcoming and caring that characterise us. It also reflects our desire to make the most of the undeniable economic and cultural benefits that newcomers bring to this country. ... [Immigration will] help achieve our objectives of economic prosperity and social development. [The numbers represent those] that Canada can support ... and also reflect the attraction that our thriving economy has for immigrants. ... The economic, social and cultural contribution of immigration continues to be vital for the future of our country and will be a great advantage in the years to come, not least of all because of the ageing populations we are seeing in all G-7 countries” (Citizenship and Immigration, 1997a).

93

BIBLIOGRAPHY

AKBARI, A.H. (1995), “The impact of immigrants on Canada’s Treasury, circa 1990”, in D. DeVoretz (ed.), Diminishing Returns: The Economics of Canada’s Recent Immigration Policy. Toronto, C.D. Howe, pp. 113-127. BADETS, J. and CHUI, W.L. (1994), Canada’s Changing Immigrant Population, Ottawa, Statistics Canada Cat. No. 96-311. BEAUJOT, R. (1991), Population Change in Canada: Toronto, Oxford.

The Challenges of Policy Adaptation,

BEAUJOT, R. (1997), “Comportements démographiques et statut socio-économique des immigrants canadiens”, in J.L. Rallu, Y. Courbage and V. Piché (eds.), Old and New minorities, Monrouge, Editions John Libbey Eurotext, pp. 147-164. BEAUJOT, R. (forthcoming) “Immigration and demographic structures”, in S. Halli and L. Driedger (eds.), The Immigration Challenge 2000. BEAUJOT, R. and McQUILLAN, K. (1982), Growth and Dualism: The Demographic Development of Canadian Society, Toronto, Gage. BEAUJOT, R. and RAPPAK, J.P. (1988), “The role of immigration in changing socio-demographic structures”, Report for Review of Demography, Health and Welfare, Ottawa. BEAUJOT, R. and RAPPAK, J.P. (1990), “The evolution of immigrant cohorts”, in Shiva S. Halli et al. (eds.), Ethnic Demography, Carleton University Press, Ottawa, pp. 111-139. st CITIZENSHIP AND IMMIGRATION (1994a), Into the 21 Century: A Strategy for Immigration and Citizenship, Ottawa.

CITIZENSHIP AND IMMIGRATION (1994b), Annual Report to Parliament: Immigration Plan 1994, Ottawa. CITIZENSHIP AND IMMIGRATION (1994c), A Broader Vision: Immigration and Citizenship Plan 1995-2000, Ottawa. CITIZENSHIP AND IMMIGRATION (1994d), “Report to SOPEMI on immigration to Canada”, Ottawa. CITIZENSHIP AND IMMIGRATION (1996), "International migration, security and economic issues: A Canadian perspective", Paper presented at NATO Seminar on Economic Aspects of the Impact of Migrations and Refugees on State Security, Warsaw, September. 94

CITIZENSHIP AND IMMIGRATION (1997a), A Stronger Canada: 1998 Annual Immigration Plan, Ottawa. CITIZENSHIP AND IMMIGRATION (1997b), Citizenship and Immigration Statistics 1994, Ottawa. CITIZENSHIP AND IMMIGRATION (1997c), Facts and Figures 1996, Ottawa. CITIZENSHIP AND IMMIGRATION (1997d), You Asked About ... Immigration and Citizenship, Ottawa. DAI, S.Y. and GEORGE, M.V. (1996), Projections of Visible Minority Population Groups, Canada, Provinces and Regions, 1991-2016, Ottawa, Statistics Canada Cat. No. 91-541. DEVORETZ, D.J. (1995), “New issues, new evidence, and new immigration policies for the TwentyFirst Century”, in D.J. DeVoretz (ed.), Diminishing Returns: The Economics of Canada’s Recent Immigration Policy, Toronto, C.D. Howe, pp. 1-30. DUMAS, J. and BÉLANGER, A. (1996), Report on the Demographic Situation in Canada 1995, Ottawa, Statistics Canada Cat. No. 91-209. DUMAS, J. and BÉLANGER, A. (1997), Report on the Demographic Situation in Canada 1996, Ottawa, Statistics Canada Cat. No. 91-209. ECONOMIC COUNCIL OF CANADA (1991), New Faces in the Crowd: Economic and Social Impacts of Immigration, Study No. 22-171, Ottawa. FOOT, D.K. (1994), “Canada’s unemployment-immigration linkage: demographic, economic and political influences”, Canadian Journal of Sociology, No. 19, Vol. 4, pp. 513-524. LACHAPELLE, R. (1988), “L’immigration et le caractère ethnolinguistique du Canada et du Québec”, Statistique Canada, Direction des études analytiques, Document de recherche No. 15. MARR, W.L. and SIKLOS, P.L. (1995), “Immigration and unemployment: a Canadian macroeconomic perspective”, in D. DeVoretz (ed.), Diminishing Returns: The Economics of Canada’s Recent Immigration Policy, Toronto, C.D. Howe, pp. 293-330. PALMER, D. L. (1997), “Canadians’ attitudes toward immigration: November and December 1996, and February 1997 surveys”, Report prepared for Strategic Policy, Planning and Research Branch, Citizenship and Immigration Canada, June. PICHÉ, V. (1997), “Les grandes migrations internationales”, Relations, April. RICHMOND, A. and KALBACH, W. (1980), Factors in the Adjustment of Immigrants and their Descendants, Ottawa, Statistics Canada. ROYAL COMMISSION ON THE ECONOMIC UNION AND DEVELOPMENT PROSPECTS FOR CANADA (1985), Report, Ottawa, Supply and Services Canada. RUDDICK, E. (1996), “SOPEMI report for Canada”, OECD, Paris.

95

SAMUEL, T.J., GUTIERREZ, R. and VAZQUEZ, G. (1995), “International migration between Canada and Mexico”, Canadian Studies in Population, No. 22, Vol. 1, pp. 49-66. st SHENSTONE, M. (1997), "World population growth and movement: towards the 21 century", Report prepared for Department of Foreign Affairs and International Trade, June.

SIMMONS, A.B. (1994), “Canadian immigration policy in the early 1990s”, Canadian Journal of Sociology, No. 19, Vol. 4, pp. 525-534. STAFFORD, J. (1993), “Welcome but why: Recent changes in Canadian immigration policy”, Annual Review of Canadian Studies, No. 22, Vol. 2, pp. 235-258. STATISTICS CANADA (1992), Immigration and Citizenship, Ottawa, Statistics Canada Cat. No. 93-316. STATISTICS CANADA (1994), Population Projections for Canada and the Provinces and Territories, 1993-2016, Ottawa, Statistics Canada Cat. No. 91-520. STATISTICS CANADA (1997a), Annual Demographic Statistics, 1996, Ottawa, Statistics Canada Cat. No. 91-213. STATISTICS CANADA (1997b), The Daily, November 4, Ottawa. TEPPER, E. (1987), “Demographic change and pluralism”, Canadian Studies in Population, No. 14, Vol. 2, pp. 223-235. THOMAS, D. (1992), “The social integration of immigrants in Canada”, in S. Globerman (ed.), The Immigration Dilemma, Vancouver, The Fraser Institute, pp. 211-260. THOMAS, D. (1996), "The social welfare implications of immigrant family sponsorship default", Paper presented at Symposium on Immigration and Integration, Winnipeg, October. VEUGELERS, J. and KLASSEN, T.R. (1994), “Continuity and change in Canada’s unemploymentimmigration linkage (1946-1993)”, Canadian Journal of Sociology, No. 19, Vol. 3, pp. 351-370. WHITAKER, R. (1991), “La politique canadienne d’immigration depuis la Confédération”, Société historique du Canada, Ottawa. ZIEGLER, E. (1988), "Refugee movements and policy in Canada", Report for Review of Demography, Health and Welfare, Ottawa.

96

Figure 1. Immigration - Historical perspectives (1860 - 1996) 450 000 400 000 350 000 300 000 250 000 200 000 150 000 100 000 50 000 0 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990

Source: Citizenship and Immigration, 1997 c, p. 2.

Figure 2. Immigrants as a percentage of Canada’s population 1901-1996 25.0

20.0 1994 - 17.4 15.0

%

10.0

5.0

0.0 1901

1911

1921

1931

1941

1951

1961

1971

1981

Source: Statistics Canada, 1997 b, The Daily, 4 November 1997, pp. 2, 4, 8.

97

1991

Figure 3. Percentage of immigrants in provinces and territories, 1996 Ontario British Columbia Alberta Manitoba Yukon Territory Quebec

Canada 17.4%

Saskatchewan Northwest Territories Nova Scotia Prince Edward Island New Brunswick Newfoundland 0

5

10

15

20

25

30

% Source: Statistics Canada, 1997 b, The Daily, 4 November 1997, pp. 2, 4, 8. Figure 4.

Figure 4. Percentage of immigrants in census metropolitan areas, 1996 Toronto Vancouver Hamilton Kitchener Calgary Windsor London Victoria Edmondo

St. Catharines-Niagara Montreal Winnipeg Oshawa Ottawa-Hull Thunder Bay

Canada 17.4

Regina Saskatoon Sudbury Halifax Sherbrooke Saint John St. John’s Quebec Trois-Rivières Chicoutimi-Jonquière 0

5

10

15

20

%

25

30

Source: Statistics Canada, 1997 b, The Daily, 4 November 1997, pp. 2, 4, 8.

98

35

40

45

Table 1. Planned and actual immigration levels, 1979-1998, Canada (Thousands) Year

Annual plan

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Actual Level

100 120 130-140 130-135 105-110 90-95 85-90 105-115 115-125 125-135 150-160 165-175 220 250 250 250 190-215 195-220 195-220 200-225

112.1 143.1 128.6 121.1 89.2 88.2 84.3 99.2 152.1 161.9 192.0 214.2 230.8 252.8 255.8 223.9 212.5 224.0 1 215.0

1. Estimates.

Sources: Citizenship and Immigration, 1997b; Annual immigration plans.

Citizenship and Immigration, 1997c;

Table 2. Immigration point system, Canada Factor

Maximum units

Education Specific vocational preparation Experience Occupational demand Arranged employment or designated occupation Demographic factor Age Knowledge of official language(s) Personal suitability

Source: Citizenship and Immigration, 1994 b.

99

16 18 8 10 10 10 10 15 10

Beaujot, 1991;

Table 3. Immigration by Class, 1966, 1970-1996, Canada (Percentages and numbers) 1

Year

Family Class (%)

1966 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

34.2 21.8 27.4 27.1 22.6 24.8 34.1 40.7 44.7 52.8 41.7 35.7 39.7 41.3 54.6 49.7 45.7 42.5 35.2 31.7 31.7 34.3 37.4 39.5 44.1 45.3 36.5 30.2

1. 2. 3.

Independent Class (%) 65.8 78.2 72.6 68.7 76.1 74.4 62.9 51.4 48.9 42.3 33.4 36.1 48.7 44.8 29.7 33.0 34.5 38.1 50.6 51.7 49.1 47.2 39.4 40.0 43.9 45.6 50.2 55.5

2

Refugee Class (%) 4.2 1.3 0.8 3.0 7.9 6.4 4.9 24.9 28.2 11.6 14.0 15.7 17.4 19.9 19.3 14.2 16.6 19.3 18.5 23.2 20.6 11.9 9.1 13.4 14.2

3

Total (Numbers) 194 742 147 713 121 900 122 006 184 200 218 465 187 881 149 429 114 914 86 313 112 096 143 117 128 618 121 147 89 157 88 239 84 302 99 219 152 098 161 929 192 001 214 230 230 781 252 842 252 137 223 875 212 491 224 050

Family class includes the retired for 1994-96. Independent class includes assisted relatives who receive extra points by having relatives in Canada, as well as business immigrants and live-in caregivers. Also includes retirees for 1981-93. Refugee class includes those designated as equivalent to refugees.

Sources: Royal Commission on the Economic Union and Development Prospects for Canada, 1985; Dumas and Bélanger, 1996; Citizenship and Immigration, 1997 b.

100

Table 4. Immigration by class and characteristics, 1996, Canada Principal Applicants Family Class Spouse Fiancés Son or daughter Parent or grandparent Other Retired Refugees Government Sponsored Privately Sponsored Landed in Canada Dependants Abroad Deferred Removal Independent Entrepreneur Self-employed Investor Skilled Worker Live-in Caregiver Total

49 178 27 276 3 602 4 276 11 841 2 139 44 18 089 3 397 1 388 10 702 238 2 364 51 629 3 174 1 380 1 597 41 898 3 580 118 896

Source: Citizenship and Immigration, 1997 c.

101

Dependants

Total

18 546 4 352 287 1 108 12 576 122 101 13 835 4 564 1 653 6 514 155 949 72 773 8 673 2 975 4 564 55 644 917 105 154

67 724 31 628 3 889 5 384 24 417 2 261 145 31 924 7 961 3 041 17 216 393 3 313 124 402 11 847 4 355 6 161 97 542 4 497 224 050

Table 5. Immigration, emigration and contribution to population growth, Canada, 1851-1996 Immigration (Thousands) 1851-61 1861-71 1871-81 1881-91 1891-1901 1901-11 1911-21 1921-31 1931-41 1941-51 1951-61 1961-71 1971-81 1981-91 1991-96

Emigration (Thousands)

352 260 350 680 250 1 550 1 400 1 200 149 548 1 543 1 429 1 429 1 381 1 166

Contribution to population growth (Percentage)

23.0 -32.6 -8.5 -28.7 -24.2 44.1 19.7 14.5 -8.1 7.9 25.5 21.7 28.6 27.7 50.9

170 410 404 826 380 740 1 089 970 241 379 463 707 636 490 228

Sources: Beaujot and Rappak, 1988; Statistics Canada, 1997 a.

Table 6. Fertility and immigration assumptions, Canada Assumptions following:

Total fertility rate

Immigration (Thousands)

1971 census 1976 census 1981 census 1986 census 1991 census

1.8, 2.2, 2.6 1.7, 2.1 1.4, 1.7, 2.2 1.2, 1.7, 2.1 1.5, 1.7, 1.9

120 and 160 125, 150 and 175 100 and 150 140 and 200 150, 250 and 330

Source: Statistics Canada, 1994.

Table 7. Non-natives of North American origin Persons born in the United States Immigrants 1986 1991 1996

282 025 249 075 244 695

Non-permanent .. 18 155 16 375

Source: 1986-96 censuses; Ruddick, 1996; Samuel et al., 1995.

102

Persons born in Mexico Immigrants 13 845 19 400 27 480

Non-permanent .. 2 635 2 605

THE IMPACT OF MIGRATION ON ECONOMIC DEVELOPMENT 1 IN MEXICO

by Gustavo Verduzco and Kurt Unger El Colegio de México

1.

Introduction

Migration impacts can be defined as the effects produced by the movement of people between two places. Given the nature of Mexican migration throughout the entire twentieth century, which has involved movement in response to labour demand rather than for religious or political motives, migration will be approached in this study as an essentially economic phenomenon. These population movements are usually seen as a positive adjustment by which labour supports and complements a better use and distribution of capital in the receiving region. Contemporary labour migration invariably occurs towards regions experiencing successful economic performance. The sending region, however, usually also benefits from remittances sent back by migrants as well as from the experience they gain working in a more developed economy. Emigration may also alleviate unemployment at the point of origin. Migration between countries is usually more complicated than movements between regions within a country due to the existence of migration policies which create additional barriers to movement and the existence of a greater range of cultural and economic differences between origin and destination, creating a different set of migration issues compared to internal movement. A number of factors have had a broad influence on US-Mexican migration. First, as part of its response to economic pressures stemming from globalisation, the United States has implemented a variety of sectoral and regional strategies in recent decades. One of these strategies has involved the hiring of low skilled labour at very low salaries in order to remain competitive in certain agricultural activities (this is particularly evident in California) and in some services (Escobar, Bean and Weintraub, 1996). Second, in Mexico trade liberalisation, structural adjustment, and privatisation during the last decade have resulted in a huge increase in levels of unemployment and have triggered the search for forms of complementary income to finance basic family survival (Cortés and Rubalcava, 1995). Structural reforms have resulted in the successful recovery of certain agricultural activities, in particular those for export, and have also led to the channelling of substantial new investment into subcontracting (maquiladoras) and a few other industrial and service activities. These trends will shortly manifest themselves in the form of higher local demand for labour in Mexico, even if in some cases (as in the first direct effect of maquiladoras) this comes at the expense of employment in the United States.

103

Finally, the North American region has taken several policy initiatives which deepen economic links, most notably the US-Canadian free trade agreement and its subsequent extension to include Mexico through NAFTA. This joint initiative can be seen as reflecting a common strategy to improve global competitiveness. It is clear that the economic ties between the three NAFTA countries have been substantially and rapidly strengthened as a result. This is most visible in trade and investment flows. This paper focuses on structural changes in migration patterns that have been taking place in Mexico and the issue of remittances. Section 2 examines the history and character of migration flows to the United States: Section 3 examines the regional and local sources of Mexican migration and Section 4 focuses on what is often considered the most important consequence of Mexican emigration, the impact of remittances on Mexican communities. 2.

History and character of migration flows

Migration flows have varied in size, intensity, and characteristics over the course of the century. While these changes are specific to the period and context in which they occur, they also reflect a general dynamic of labour interaction that has become embedded in the economic structure of the two countries. Early in this century, a labour shortage in the United States resulting from the First World War led to a sizeable demand for Mexican workers, especially in agriculture and railroad construction (Hall and Coerver, 1990). The Second World War once again created a demand for low-cost Mexican labour (Fisher, 1953), inciting the US government to issue 4.6 million legal work contracts to Mexicans in the United States -- the so called Programa Bracero. In spite of the relatively high number of contracts (equivalent to approximately 209 000 a year), those opting to reside permanently in the United States were in the minority, representing only 12 per cent of the total number of contracts granted. This is possibly explained by the fact that during this period, Mexico was enjoying very high rates of growth -- the period of the “Mexican miracle” -- which translated into better living standards for many in Mexico as a result of improvements in education, health, and other social benefits, in spite of high rates of population growth. Mexican labour in the United States continues to be concentrated in agriculture, in the production and packaging of fruits and vegetables, in gardening, and in a few other areas, including the clothing industry, construction, and certain services for buildings. While evidence indicates that recent Mexican migrants are predominantly in unskilled and low paying jobs, there are groups of Mexican migrants who engaged in more highly-skilled and better paid employment in the United States. Temporary versus permanent migration Patterns of migration are extremely diverse, however a broad classification into three categories provides a useful framework. First there are migrants who spend no more than one or two spells of employment in the United States in their working lives, and may therefore be called temporary migrants. When these spells of employment are more frequent, often conducted on a regular basis then the migration is of a cyclical nature. Finally, in some regions the mass exodus of families through permanent migration has drastically altered the character of the locality. These different types of migration -- temporary, cyclical, or permanent -- have different impacts on both the sending and receiving countries.

104

Though there are some preliminary attempts to distinguish the impact of cyclical and permanent migrants, in practice it is extremely difficult to separately estimate the effects of each group. According to the EMIF (Migration Survey to the Northern Border) survey, the annual flow of labour migrants travelling to the United States in 1993 and 1994 was around 800 000 individuals (Corona and Tuirán, 1996). The vast majority of these individuals are cyclical migrants, whose impacts are different in many respects from those residing permanently in the United States. The legal status of migrants also has important repercussions. While legal status is itself important in terms of volume of illegal flows and policy responses to them, it has less quantifiable effects on, for example, the worker’s own perception of his labour competitiveness. In this regard the Mexican debt crisis of 1982 coupled with IRCA in 1986 resulted in 2.24 million Mexican immigrants residing in the United States acquiring legal status and quite probably into permanent settlement during the period 1986-92. IRCA’s sudden transformation of a significant portion of the undocumented working population into legal immigrants may have a number of effects. Unfortunately, it is difficult to distinguish the impacts of one group from those of the other -- legal versus undocumented, cyclical versus permanent -- though these differences may be considerable. 3.

Regional and local sources of Mexican migration

While national level impacts may be small in many respects, migration plays an extremely important role in a number of regional and local economies. This is as valid for the sending country as for the recipient states and counties in which flows concentrate. The duration of the migration pattern also influences its’ impact on both sides of the border. On the Mexican side, intense migration is heavily concentrated in a few localities (municipios). These are mostly located in five states: Jalisco, Michoacán, Zacatecas, Durango, and Oaxaca. Furthermore, more than half of these municipios (57 per cent) are rural localities with less than 20 000 inhabitants, and very few of them (only 1.7 per cent) have more than 50 000 inhabitants. The predominantly rural origin of Mexican migrants explains their employment in the least skilled jobs in the United States. Their situation is very different from that of Mexicans who migrate within Mexico, these latter tending to be more urban in origin. Municipalities of origin Of the 2 428 municipalities in the country, 62 per cent show some degree of “migratory activity” to the United States, although 18 per cent have extremely low levels (Table 1). Looking at migration flows by region, we find that in the south-eastern states (Chiapas, Tabasco, Campeche, Yucatán, and Quintana Roo), 66.5 per cent of municipalities do not participate in the flow of migrant workers to the United States, despite the fact that these regions include some very poor communities which might be expected to have much to gain from migration to the United States. In the central states (Oaxaca, Veracruz, Guerrero, Puebla, Tlaxcala, Hidalgo, Querétaro, State of Mexico and Morelos), 33.8 per cent of the municipalities do not participate, while in the border states only 10.2 per cent do not participate. In the western and northern states, practically all the municipalities have some participation in migration.

105

Table 1. Migration intensity by region Percentage of municipalities showing activity within various bands of intensity 1 (Percentages and total numbers) Migration activity level % of EAP 0 0.1-1.0 1.1-6.9 7.0-24.9 25.0 + Total number

Region (Percentage of municipalities within region)

National average

Border

North

West

Center

Southeast

10.2 11.3 49.4 26.7 2.2

5.5 8.2 39.0 28.5 18.7

2.5 1.9 19.0 59.4 17.0

39.7 33.8 18.0 7.2 1.1

66.5 13.8 14.2 4.5 0.8

38.0 18.3 22.0 16.9 4.5

273

182

313

815

817

2 404

1. Migratory activity: percentage of migrants out of the economically active population (EAP) of the municipality. Source: National Survey of Demographic Dynamics (ENADID 1987-1992).

Looking next at the intensity of migration activity in Mexico’s municipalities, we discover that only 4.5 per cent have high or intense activity. These are the municipalities that most likely contribute to cyclical migration. Among the municipalities with high migratory activity, 80 per cent are municipalities with a population of less than 20 000 inhabitants and only one of them (Jerez, Zacatecas) has more than 50 000 inhabitants, which means they are very rural municipalities. By contrast, we find that 43 per cent of the country’s municipalities have low migratory activity, and 17 per cent medium. Among these groups, 56 per cent are municipalities with less than 20 000 inhabitants and only 17 per cent have more than 50 000 inhabitants, which confirms the rural 2 origins of the most intense emigration to the United States. Of the municipalities with high or intense migratory activity (109 municipalities), the great majority (88 per cent) are situated in nine states which also experience the most migration: Jalisco, Michoacán, Guanajuato, Zacatecas, Durango, Chihuahua, San Luis Potosí, Guerrero, and Oaxaca. Among these municipalities, 48 per cent are found in only three states: Jalisco, Michoacán, and Zacatecas. In summary, we find that the intensity of migration is far from uniform throughout the country, and that strong migration is limited to relatively few of the country’s municipalities. We therefore need to be very careful when generalising interpretations of the phenomenon of migration to Mexico as a whole. While migration is dispersed, touching most regions of the country, intense activity is geographically limited. Rural versus urban origins Although rural municipalities predominate over urban ones as the source of the most intense migration, there is migration, albeit less intensive, from practically every municipality that has urban populations. The Survey on Migration to the Northern Border (EMIF) reported that 56 per cent of 106

total flows came from urban localities of more than 15 000 inhabitants (Corona, 1996). In fact, urban migrants form a substantial component of total flows to the United States. Furthermore, the impact of these more highly skilled workers is likely to be different from that of their unskilled rural 3 counterparts. However, to-date most studies of migration to the United States have focused almost entirely on rural populations. The lack of substantial research in the area of urban migration limits our ability to understand the logic of the labour movements of an urban work force drawn to seeking work abroad, and even further limits our ability to assess the specific impacts of these workers in other than the most general of ways. Assessments of the impact of migration A number of points should be borne in mind in examining studies of the impact of emigration. First, the study of this issue has not received a very high level of attention in Mexico, much less than immigration issues have received in the United States. Second, as mentioned above, the majority of studies deal with migrants and migrations from the rural sector, which, although consistent with the greater presence of the phenomenon in those places, bypasses the question of impacts on the diverse urban localities where in fact migration also exists. Third, most of the studies, particularly the better known ones, consider only the two states with the most migration in the country (Michoacán and Jalisco): several of them focus only on the communities within these states’ municipalities where there is the greatest intensity of migration in the entire country. While this allows a ‘case-study’ approach, it nevertheless entails certain limitations, especially in the light of evidence that emigration is becoming more widely spread throughout the country. Among the studies carried out, the community studies of anthropologists or of an anthropological nature are the most common. These studies do not usually have the specific intention of studying migration, but rather the general community life of campesino villages. They have, however, been forced to consider migration to the United States because of its magnitude and the importance it plays in the lives of these communities. In addition, there are studies of a socio-demographic nature whose information comes mainly from surveys applied to the population, sometimes of a single community or city, or to areas or groups of communities or cities. In some cases these studies have used samplesurvey data and in this respect have provided quantitative insights into the process of migration. Studies specifically addressing the impacts of migration tend to focus on the issue of remittances, i.e. money sent back or brought by migrants themselves to their communities. Some researchers, have also focused on the topic of the non economic effects of migration, although few have delved deeply into this matter. The authors frequently attempt to interpret the effects of both monetary remittances and work experience in the overall process of change in the communities studied. Here too, however, we should be cautious, since the authors do not always give an account of socioeconomic structure of the community, therefore making their results difficult to put into context. Despite the limitations and inadequacies described above, research carried out in Mexico by both Mexican and US researchers does contain some useful information and insight on the impact of emigration. In the remainder of this paper we focus on what is generally regarded as the most important aspect of this issue, monetary remittances and their effect on economic development.

107

4.

Monetary remittances and economic development

The topic of remittances has perhaps received the most attention among the economic effects of migration. This is partly because monetary remittances can be quantitatively measured and play a fundamental role in the dynamics of migration. Wages in the United States typically compare favourably with those in Mexico, allowing savings of relatively high value in peso-terms to be sent home or brought back by immigrant workers. Although the importance of monetary remittances has been known since the 1920s (Taylor, 1931), it was only a few years ago that researchers began to seriously work on the topic and to try to record the amounts in some way. The estimates of remittances reported in the literature generally derive from two sources: government estimates and estimates based on specific community studies. Drawing on data from community research projects, researchers have frequently attempted to make more global calculations of the remittances sent by all Mexican migrants (Díez-Canedo, 1984; García y Griego and de los Ríos, 1985; Massey and Parrado, 1994). However, such estimates are not generally regarded as being very reliable and such calculations have been useful only as preliminary approximations where more accurate estimates were not available. Recently, there have been more complete estimates based on information from surveys with greater coverage, such as the survey carried out by CENIET, and the Survey on Migration to Mexico’s Northern Border (Corona, 1996). The Banco de México (the Central Bank) has designed a complex methodology together with a strategy of very broad coverage in order to be able to measure the dollar inflow to Mexico of remittances sent by Mexican migrants working in the United States. For 1995, the Banco de México estimates total remittances at the equivalent of US$3.7 billion. These calculations have been replicated by other researches and similar estimates found. Viewed from a national standpoint, the importance of remittances may seem only moderate in comparison with earnings from exports of goods, since the remittances sent by migrants represented barely 5 per cent of the total value of exports in 1995. They were nonetheless equivalent to 57 per cent of foreign investment for that year. It should be recalled that remittances over the previous decade had a much greater relative importance. During that period, remittances were in fourth place in foreign-currency inflows from abroad. Some estimates also show that a large proportion of total remittances (64.7 per cent) are sent by permanent migrants. The decline in the relative importance of remittances accords with the notion that permanent migrants reduce their remittances over time, suggesting that the importance of remittances may decline further in the future. At present, however, remittances continue to play a significant role in the Mexican economy, and given the geographic concentration of the phenomenon of migration, it is necessary to evaluate the impact of remittance inflows from a regional point of view. From such a perspective, the relative importance of remittances is obviously greater, particularly if we bear in mind that the majority of sending municipalities are rural. Since migration is mainly concentrated in the states of the west and to a lesser extent in those of the north, the economic benefits of migration are the most apparent in those areas. In these areas it is not uncommon for remittances to form a large proportion of family income. Although much of the income from remittances is spent on consumption goods it has also been used along with other forms of assistance from migrants to provide support for agricultural activities, as well as for public works

108

such as parks, schools, churches, roads, street paving, the introduction of drainage, etc., in a great many communities in these regions. There are also indirect effects associated with remittances. These can be calculated, although in a partial and limited manner, on the basis of economic simulation exercises. J.E. Taylor (1996) calculates that the greatest multiplying effect of remittances takes place in rural households whose consumption and spending patterns favour goods produced domestically with labour-intensive technologies: by contrast, when remittances reach urban households, the multiplying effects for Mexico are less because consumption patterns in those localities include a larger number of imported goods. Taylor calculates that as a result of multiplying effects, each dollar sent to Mexico by migrants could translate into an additional US$0.30 to US$0.40 in the income of farmers with small landholdings. The most direct impact of remittances in dollars obviously occurs in the households where they are received and depends on the amounts sent and the circumstances of the receiving household. The community as a whole may also benefit, depending on the number of households with migrants and the numbers of their members working abroad. There will also be greater benefit to the community if the initial expenditure and subsequent expenditure through multiplier effects is on local goods and services or investment projects. Similarly, the impact will depend on whether the migration pattern in the community is occasional or cyclical, that is, whether households have members migrating with a regular annual or biannual frequency, or whether migration takes place only occasionally, once or twice throughout people’s lifetimes. As mentioned previously, evidence suggests that the remittances sent by those settled permanently overseas tends to decline over time. The following presents a summary of community-level studies and serves to illustrate the variation in the importance that remittances play at the local level as well as the diversity of uses to which they are put by communities. Paul Taylor (1931) reported that during the twenties there was a mass migration from Arandas, Jalisco, which included young men of all social classes except landholders and wealthy merchants, though not excluding their sons. In 1923, the income from the 200 “northerners” from the town was 75 000 pesos, or the equivalent of US$375 per person (Martínez, 1985), an amount that was not negligible in those days. Robert V. Kemper (1995) and Castile (1974) affirm for the Michoacán towns of Tzintzuntzan and Cherán, respectively, that almost all the young men in the two communities enrolled at some point in the Bracero Program, although many later changed their destination from the United States to Mexico City. Currently, Kemper states, 200 families from Tzintzuntzan are living in Mexico City and around 150 are living in the United States. In addition, 362 persons are living temporarily outside the community in Mexico City, Morelia, and the United States. Of all the migrants from four communities in Michoacán and Jalisco studied by Massey et al. (1987), between 15 and 25 per cent had left to work in the United States at least four times, i.e. were “cyclical” migrants. In contrast, Jorge Durand (1994) finds that in the 11 communities in his study (in Michoacán, Jalisco and Guanajuato), on average 54 per cent of the migrants had only made one journey in their entire life. Something similar is observed by Gustavo López (1986) for the town of Tangancícuaro in Michoacán, where he notes that approximately 60 per cent of the adult population have experience of migration.

109

In Los Altos (the highlands) of Jalisco, Juan Luis Orozco (1992) identifies three different migration profiles in his communities: in “Corralillos,” 36 per cent of the sons had gone to work in the United States, while in the village of “Los Dolores” 21.4 per cent had left and in “El Refugio” only 4.2 per cent. In the case of Zamora, in Michoacán, it is reported that 14 per cent of the city’s working population had migration experience, but only 19 per cent of these were “cyclical” migrants who had migrated at least four times (Verduzco, 1992). These data give us an idea of the ways in which migration profiles vary across communities. One of the reasons that research results on remittances are often at odds with each other is due to their lack of attention to geographic and temporal specificity. Although we can talk of overall quantities and average monthly or annual amounts, such figures do not reflect the very individualised ways in which remittances actually reach the communities of origin. For example, Orozco (op. cit.) made a calculation based on data from his three communities and found that the income from remittances reached US$86 464, representing 13.9 per cent of the total income of the communities, or the equivalent of US$371 a year per family. But he finds that actual amounts differ a great deal depending on the community. Most of the inflow is received by Corralillos, where the annual income is US$548 per family, equivalent to 28.4 per cent of the income. Meanwhile in El Refugio it is only US$60 per family, equivalent to only 4.3 per cent of the income. Obviously, the differences between these quantities have much to do with each town’s particular migration patterns. Massey and Parrado (1994) found similar behaviour in their study of 22 communities in the states of Michoacán, Jalisco, Guanajuato, and Nayarit, since the remittances calculated as a percentage of aggregate income varied considerably. Similarly, for Gómez Farías (Michoacán) López (op. cit.) calculated that remittances ranged from between US$80 to US$300 a month per family. In the village of Tepusco (municipality of Villa Hidalgo, Jalisco), almost 100 per cent of external family income comes from remittances, and surpasses the amount of public spending in the municipality (Arroyo et. al., 1991). Moreover, in Colotlán, Jalisco, the same authors point out that remittances are the economic base of the area, once again amounting to much more than municipal spending. In contrast, in the area of Ocotlán, also in Jalisco, which is an industrialised region, there is less migration and those who do migrate to the United States have among the lowest levels of education. Consequently, the amount received there in the form of remittances is relatively insignificant. Pointing to this high degree of variation, the authors contend that interpretations made from averaged estimated remittances must be limited (1991). The majority of families in their study received only small quantities, more than a third did not receive anything at all, and only a few received larger amounts. Their study has the additional value of focusing on the “area of rejection” of 4 Jalisco and of including 69 communities. Furthermore, Papail and Arroyo (1996) mention that between 27 per cent and 32 per cent of migrants from the four medium-sized cities of Jalisco were 5 unable to send money during their last migration to the United States. The possibilities for saving and sending remittances is determined, in part, by the migrant’s means of arrival in the United States. Those families who do not already have savings available to finance the transit of their migrating members must borrow from either Mexican or US sources, and usually at high interest. Fonseca and Moreno (1984) illustrate this point particularly well, citing numerous individual accounts. Similarly, it has been shown that once in the United States, workers may incur high living-costs, particularly those who work isolated in the fields or who are undocumented and have to depend on intermediaries to obtain food, clothing and other types of goods in order not to place themselves at risk through public exposure (Fonseca and Moreno, op. cit.). Moreover, the transit costs for undocumented migrants often include fees charged by “coyotes” to cross the border, 110

and bribes paid to Mexican officials upon their return to the country (Cornelius, 1978; López, 1986; Massey et al., 1987). Finally, there is considerable range in the ability of migrants to find work in the United States. On the one hand there are workers who, because of past migration experiences, contacts in networks, or the type of work in which they engage, manage to quickly find a job and begin saving money to return to their region of origin. On the other hand, a substantial share of migrants do not manage to send or bring back any savings to their families. Furthermore, there are some activities, especially those related to agriculture, that do not make it possible to have work continuously throughout the year. In fact, this is a point on which further research should be carried out in order to grasp the complexities of the economics lying behind this type of labour experience. The relation between clandestine forms of arrival and labour integration may also reveal specific patterns. Dependence or development? Although there is some literature describing the “counterproductive” effects of remittances (conspicuous consumption, inflation, etc.) it is generally believed that migrants’ contributions help to improve living conditions in their regions of origin. Communities with more migration are often visibly different to those with low levels of migration, as Rionda (1986) has observed for Copándaro in Michoacán, and Gustavo López (1986) for Gómez Farías. The former commonly have better housing, as well as good urban services, or at least relatively better services than in neighbouring communities. Although researchers have shown that remittances are spent mainly on consumption goods, they are also used to enlarge and embellish houses as well as to improve them with electricity and water and drainage services. In addition, the “northern money” often goes towards supporting the community’s urban infrastructure works and may be used to pay community debts. The degree to which remittances are used for investment purposes varies from one community to another. In some, there seem to be no productive changes, or else changes that do occur are perceived as being of minor local or regional importance. Here, the migration process appears to have established a dependence on remittances. Such is the case of places like Huecorio (Dinerman, 1982), Jaripo (Fonseca and Moreno, 1984), Gómez Farías (López, 1986), Chamitlán (Massey et al., 1987) Las Animas (Goldring, 1990), Corralillos (Orozco, 1992), Tepusco and Colotlán (Arroyo et al., 1991). In other communities, remittances appear to have positively supported processes of change. Remittances can serve as a viable source of investment capital, as when braceros’ remittances were used to start furniture workshops in Cuanajo, Michoacán. In Ihuatzio, also in Michoacán, Dinerman (op. cit.) describes how remittances supported the introduction of handicrafts, an activity that did not exist a few years earlier, providing a new economic alternative for some of the inhabitants. In Chilchota, Michoacán, something similar occurred with the different crafts that emerged there (Ramírez, 1986). Rionda (1986) recounts how remittances supported agricultural mechanisation in Copándaro, Michoacán, as well as the development of lentil cultivation. Michoacán is now the country’s main producer of lentils. Other activities, including animal husbandry (cattle and hogs) have also been introduced into the local economy. A number of studies suggest that the productive investment may have reduced migration in the Jalisco region as a whole. Orozco (op. cit.) reports that in the town of Los Dolores, productive changes and the improvement of living standards have been very noticeable as compared with the performance of

111

the other two communities he studied. Where in the past there was a pattern of intense migration in Los Dolores, there is now very little. Arroyo, De León and Valenzuela (op. cit.) point out in their study of Jalisco’s rural area of “rejection” that the most productive investment from remittances occurs in Tepatitlán, where it is directed to agriculture and other types of business. In another study, but this time of the urban area of Jalisco, Papail and Arroyo (op. cit.) report that in Tepatitlán, remittances used to support poultry farming have helped establish the city and its surroundings as a significant producer of eggs and poultry, supplying around 30 per cent of the country’s total production. In the city of Zamora, Michoacán, which is a subregional city in the north-west of the state, Verduzco (1992) studied the changes in the region’s rural and urban sectors, as well as the characteristics of its diverse migration flows, both internal and international. The study suggests that owing to intense agricultural development and the promotion of urban activities, migration to the United States has probably diminished over time. Verduzco also suggests that the current migrant pattern is different from that of other periods. Papail and Arroyo (1996) suggest something similar for the medium-sized cities of Jalisco which they studied. In addition to these instances of economic development in communities and sometimes in regions, 6 there are other accomplishments that, while perhaps more limited, are no less important. In the studies of Arandas (Taylor, 1931), Jaripo, Gómez Farías, Chavinda (Alarcón, 1995), Huecorio (Belshaw, 1967), Alvaro Obregón (Trigueros, 1994), Cherán (Beals, 1946) among others, it appears very clear that at various times, but especially in the case of ejido lands during the 1940s, remittances served as a useful source of income and wealth in the development of agriculture, especially when other sources of capital, such as bank credits, were limited. From another perspective, recent works by Taylor (1996) and Yúnez (1997), based on the social accountability matrices of various communities, offer preliminary but convincing conclusions that emphasise the need to distinguish the effects of migration according to the differences between localities. For example, Yúnez’s micro-multisectoral study of rural localities with different characteristics establishes the differentiated nature of the impacts of foreign currency inflows, 7 whether from remittances or from exports. The extreme cases in his study illustrate the point: Napizaro, Michoacán is the locality that is most dependent, in relative terms, on remittances and therefore the one that maximises the multiplying effects of a change in the inflow of foreign exchange 8 (whether because of devaluation or increased hiring), while Concordia, Coahuila has a more balanced economy that depends to a lesser extent on remittances and that gathers more limited multiplying effects from a change in foreign currency (Yúnez, 1997). The implementation of this methodology presupposes a good deal of technical sophistication together with high costs associated with gathering the information from each community, but it is undoubtedly a promising course of research. These diverse effects and changes reported at the community level in Mexico correspond to those reported in other parts of the world with similar migration experiences. Remittances appear to be a vehicle for furthering the development of sending regions, provided the channelling of remittances is accompanied by other conditions favourable to economic growth. Such conditions include concentrating other private and public resources, notably for infrastructure, in those same areas. Otherwise, remittances are more likely to be used for basic consumption, resulting in low multiplier effects (Martin, 1990).

112

This relation between remittances and development has specific policy implications for both the United States and Mexico. While the concentration of remittances in specific regions ensures the multiplying effects that reduce dependence on remittances in the future, and consequently reduce migration pressures, it further makes sense for Mexico to concentrate investment in infrastructure development in those same regions. The downside of such a policy is that it focuses mainly on regions with the greatest development potential, neglecting other regions that may be in equal or 9 greater need of state support. 5.

Conclusion

In addition to providing quantitative data on remittances, researchers unanimously report that migrant workers use part of their savings from working in the United States to improve the quality of life in their regions of origin. They improve their homes, whether by decorating them, building additions to them, or equipping them with basic utility services like electricity, water, and drainage. It is also common to find that localities with migrants benefit from their contributions by channelling monetary support to improve urban infrastructure and provide diverse services, including health, religious, educational, and entertainment. In towns with migrant populations, it is frequent to find acknowledgements from residents to their fellow townspeople for their contributions to the building of a school, a church, or a health clinic, or for having helped introduce drinking water, etc. Money from remittances is also often productively invested in activities like agriculture, mainly in the acquisition of machinery or implements, or through other types of improvements or small technical innovations. As is evident from this review of evidence on the economic and social impacts of emigration, the majority of studies examine selected communities. These provide valuable insights into individual experiences and emphasis the heterogeneity across communities in patterns of migration and remittances. However it is clear that research is still needed to provide more comprehensive accounts of migration patterns and the inflows of remittances and their effects on the economy at regional and national level.

113

NOTES

1.

Limitations of space dictate that this paper be presented here in its shorter form.

2.

Note that in different information sources, localities of more than 15 000 inhabitants are commonly situated in the “urban” category. To avoid confusion as to what is considered urban, it would be preferable to talk of the number of inhabitants of those places, since ultimately what we want is to have an approximate indicator of levels of socio-economic differentiation in those places.

3.

In fact, this is an approximation, since although there is migration from those places, we do not know how many migrants come from the urban localities that may exist in the municipality.

4.

This was a very extensive study which was representative of a “rural area of high rejection” in Jalisco. It comprised 26 municipalities, 69 localities, 576 families, and 3 739 persons.

5.

This study was based on statistically representative samples for the four medium-sized cities of Jalisco.

6.

The perceptions of different authors tend to be contrasting. For some, the benefits stemming from the establishment of new small businesses are considerable. See, for example, the summary review of studies that record the capitalisation of remittances into own businesses, in Taylor (1996). Others do not consider them significant. As obvious as it may seem, opinions will depend largely on the moment or phase in the development of the community in which the observation is made.

7.

These differences coincide with what has been reported by other authors on similar topics. Durand and Massey, for example, find major differences between communities in terms of their percentages of productive expenditure (reported in Taylor et al., 1996).

8.

Two-thirds of remittances have come from Mexicans resident in the United States, which leads one to suppose that given observed trends these will gradually diminish in the future. This drop at the community level is also reported by Zendejas (1995) for the town of Erícuaro, Michoacán.

9.

The dilemma is well expressed by Martin (1990): “The neglect strategy makes emigration areas even more dependent on an external labour market (Mexico's central highlands); but the subsidise-theremittances strategy increases regional inequalities within emigration nations”. Certain isolated evidence shows the Federal Government's indifference towards states with migration. For Zacatecas, for example, it is estimated that federal outlays are equivalent to less than one-third of the inflows from remittances.

114

BIBLIOGRAPHY

ALARCÓN, R. (1995), Immigrants or transnational workers? The settlement process among Mexicans in rural California, Report for the California Institute for Rural Studies, University of California, Berkeley, CA. ARROYO ALEJANDRE, J., DE LEÓN ARIAS, A. and BASILIA VALENZUELA VARELA, M. (1991), Migración rural hacia Estados Unidos. Un estudio regional en Jalisco, Consejo Nacional para la Cultura y las Artes, México. BEALS, R. (1946), Cherán: A Sierra Tarascan Village, Smithsonian Institute of Social Anthropology, Washington, D.C. BELSHAW, M. (1967), A Village Economy: Land and People of Huecorio, Columbia University Press, New York. CASTILE, G.P. (1974), Cherán: La adaptación de una comunidad tradicional de Michoacán, Instituto Nacional Indigenista, México. CORNELIUS, W.A. (1978), Mexican Migration to the United States: Causes, Consequences and US Responses, Center for International Studies, Massachussetts Institute of Technology, Cambridge, Massachussetts. CORONA, R. (1996), La migración indocumentada hacia Estados Unidos, Consejo Nacional de Población, México. CORONA, R. and TUIRAN, R. (1996), “Estimacion del Saldo Neto de la Migracion Internacional de Residentes Mexicanos en el Quinquenio 1990-95”, Background paper prepared for México/United States Binational Study on Migration. CORTES, F. and RUBALCAVA, R.M. (1995), El ingreso de los hogares, INEGI-COLMEXIISUNAM, México. DÍEZ-CANEDO, J. (1984), La migración indocumentada de México a los Estados Unidos -- un nuevo enfoque, Fondo de Cultura Económica, México. DINERMAN, I.R. (1982), Migrants and stay-at-homes: A comparative study of rural migration from Michoacan, México, Monograph Series 5, Center for United States-Mexican Studies-University of California, San Diego, CA. DURAND, J. (1994), Más allá de la línea. Patrones migratorios entre México y Estados Unidos, Consejo Nacional para la Cultura y las Artes, México.

115

ESCOBAR, A., BEAN, F. and WEINTRAUB, S.(1996), The dynamics of Mexican Emigration, paper presented at the Policy Workshop on Emigration Dynamics in México, Central America and the Caribbean, San José, Costa Rica. FISHER, LLOYD H.(1953), The Harvest Labor Market, Harvard University Press, Cambridge. FONSECA, O. and MORENO, L.(1984), Jaripo pueblo de migrantes, Centro de Estudios de la Revolución “Lázaro Cárdenas”, A.C., Jiquilpan, Michoacán, México. GARCÍA Y GRIEGO, M. and GÍNER DE LOS RÍOS, F. (1985), “¿Es vulnerable la economía mexicana a la aplicación de políticas migratorias estadounidenses?”, in México-Estados Unidos 1984, El colegio de México, México. GOLDRING, L. (1990), “Development and Migration: A Comparative Analysis of Two Mexican Migrant Circuits”, in Commission for the Study of International Migration and Cooperative Economic Development, Working paper No. 37, Washington, D.C. HALL, L. and COERVER, D. (1990), Revolution on The Border: The United States and México, 1910-20, University of New México Press, Albuquerque. KEMPER, R.V. (1995), “Comunidad y migración: el caso del pueblo de Tzintzuntzan, Michoacán, 1988-94”, in Relaciones, No. 61-62, El Colegio de Michoacán. LÓPEZ, G. (1986), “Iangancícuaro: Población y migración”, in Estudios Michoacanos I, Zamora, El Colegio de Michoacán y el Gobierno del Estado de Michoacán, Michoacán. MARTIN, Philip L. (1990),”Harvest of Confusion: Immigration Reform and California Agriculture”, International Migration Review, No. 24, Spring 1990, pp. 69-95. MARTÍNEZ S.T. (1985), “Los impactos políticos y económicos de los emigrados en Jalisco: El caso de Arandas, Jalisco”, in Desarrollo rural en Jalisco: Contradicciones y perspectivas, El Colegio de Jalisco y CONACYT, Guadalajara, Jalisco. MASSEY, D., RAFAEL ALARCON, S., DURAND, J., and GONZALEZ, H. (1987), Return to Atzlan: The Social Process of International Migration from Western México, University of California Press, Berkeley CA. MASSEY, D. and E. PARRADO (1994),”Migradollars: The Remittances and Savings of Mexican Migrants to the USA”, Population Research and Policy Review,Vol. 13, No. 1. OROZCO, J.L. (1992), El negocio de los ilegales -- ¿ganancias para quién?, Editorial AgataInstituto de Filosofía, ITESO, Guadalajara, Jalisco, México. PAPAIL, J. and ARROYO ALEJANDRE, J. (1996), Migración mexicana a Estados Unidos y desarrollo regional en Jalisco, Universidad de Guadalajara, Guadalajara, Jalisco, México. RAMÍREZ, L.A. (1986), “Agricultura campesina y migración: el impacto de un cultivo comercial en un pueblo de migrantes”, in Relaciones -- Estudios de historia y sociedad, No. 26, Primavera 1986, Zamora, Michoacán, México.

116

RIONDA, L. (1986), “Zacapu: continuidad y escisión social en Copándaro”, in Estudios Michoacanos II, El Colegio de Michoacán -- Gobierno del Estado de Michoacán, Zamora, Michoacán, México. TAYLOR, J.E. (ed.) (1996), “International migration and economic development: A micro economywide analysis”, in OECD, Development Strategy, Employment and Migration: Insights from Models, Paris. TAYLOR, P.S. (1931), Mexican Labour in the United States, University of California Press, Vol. VII, No. 1, Bethelhem, Pennsylvania, Berkeley. TRIGUEROS, P. (1994), “Sorgo, campesinado y migrantes. El papel de la migración internacional en la reproducción de una comunidad campesina que adoptó la modernización de la agricultura”, Tésis para optar al grado de Doctor, El Colegio de México, México. VERDUZCO, G. (1992), Una ciudad agrícola: Zamora -- Del porfiriato a la agricultura de exportación, El Colegio de México/El Colegio de Michoacán, México. VERDUZCO, G. (1997a), “Los trabajadores Mexicanos in la Agricultura de California: Evolucion de Cultivos y Demanda Laboral”, Background paper prepared for México/United States Binational Study on Migration. VERDUZCO, G. (1997b), Los Impactos de la Migracion en México, Team Report to México/United States Binational Study on Migration. YÚNEZ, A. (1997), Impactos de los cambios económicos en el agro mexicano y en la migración: un análisis micro-multisectorial, Trabajo preparado para la Comisión Binacional para el estudio de la Migración de México a los Estados Unidos. ZENDEJAS, R.S. (1995), Migración de mexicanos a los Estados Unidos y su impacto político en los poblados rurales de origen, ponencia presentada en el XX Congreso de la Asoc. Latinoam de Sociología, México, 2-6 October.

117

B. SECTORAL ECONOMIC IMPACTS

Migration and the labour market: sectoral and regional effects in the United States Francisco L Rivera-Batiz Migration and the labour market: sectoral and regional effects in Canada Don J. DeVoretz and Samuel A. Laryea Migration and the labour market: sectoral and regional effects in Mexico Rodolfo Cruz Piñeiro

119

MIGRATION AND THE LABOUR MARKET: SECTORAL AND REGIONAL EFFECTS IN THE UNITED STATES

by Francisco L. Rivera-Batiz Columbia University

1.

Introduction

The issue of immigration stands out as one of the most debated ones in the US policy arena this century. Despite the passing of major immigration reform legislation every few years and the emergence of many additional proposals from legislators and academics every year, the issue of the economic consequences of immigration continues to exert considerable impact almost daily in the country’s public life. In recent years, a growing number of policy makers and academics have claimed that immigration has strong negative effects on the US labour market. These views have affected the general legislative tone in the country, leading to a number of policy measures which have affected immigrants negatively. New federal legislation, for example, has stripped legal immigrants of a variety of government entitlements that continue to be available to citizens. And illegal immigrants have been growingly subject to the wrath of federal, state and local policy makers who often blame them for all kinds of ills, from crime to budget deficits. Despite the increased visibility of this negative perspective on immigrants, a large portion of the US population, and sometimes the majority, is in disagreement with it. The fact is that hostility towards immigrants varies substantially, depending on the overall state of the economy, the region being considered, and the socio-economic status of the person. While during periods of recession and greater unemployment public opinion turns against immigrants, prosperity and greater economic growth substantially diminish anti-immigrant sentiments. There was, for instance, an upsurge of antiimmigrant sentiment in the early 1980s, when the US economy was in deep recession. Indeed, a 1982 national poll found 66 per cent of the population preferring that the number of legal immigrants in the country be decreased, with only 23 per cent stating that the existing number of immigrants was acceptable (Moore, 1986). Similarly, in 1990, during a deep recession, dozens of anti-immigrant incidents surfaced in the country, from fishermen in Texas harassing Vietnamese competitors, to the Florida supermarket operator who suspended some cashiers because they were speaking Spanish at work. Still, after the economy recovered, open public incidents against immigrants subsided. In 1994, a Gallup Poll Organization surveyed New Jersey residents and asked them a set of questions about immigration. When asked whether they preferred that the number of immigrants in New Jersey change or not, 51.4 per cent said immigration should be kept at the present level, compared to 39.6 per cent who felt that it should be reduced (Espenshade, 1997).

121

The ambivalence towards immigrants is also reflected in academia. Despite the vocal clamours of a few social scientists who have raised serious concerns about the effects of immigration, there is widespread disagreement on the issue. In a 1985 survey, 82 per cent of all social scientists interviewed responded that immigration this century has been either favourable or slightly favourable towards US economic growth, Furthermore, between 60 and 70 per cent believed that recent immigrants have about the same impact as past immigrants. Close to 40 per cent declared that the number of legal immigrants should be increased, compared to 20 per cent who felt the number should be decreased and 43 per cent who felt there should be no change (Moore, 1986). These figures suggest that, at the present time, the US public as well as the academic community do not uniformly perceive current immigration as a negative force, as some policy makers have argued. Rather, people are divided on the issue of whether immigration benefits or hurts them. It is the purpose of this paper to present the available evidence on the labour market impact of immigrants in the United States, especially as it relates to the regional and sectoral effects of immigration. The paper presents the key issues of controversy in the field, with each section dedicated to a specific area of debate. Section 2 presents the regional distribution of immigrants, showing that immigration is predominantly a regional phenomenon that affects to a significant extent only a small portion of the US population. Section 3 then examines the occupational and sectoral distribution of immigrants in the United States. Section 4 proceeds to discuss the issue of whether immigrants cause a reduction in the wages and employment opportunities of natives; and Section 5 examines whether immigration displaces native-born workers, inducing them to out-migrate from regions and sectors of high immigration. The last section provides a summary of the paper’s conclusions. 2.

The regional distribution of immigrants

Despite the great concerns expressed by policy makers and some academics about the economic impact of immigration on the US labour market, the fact is that immigration directly affects only a small portion of the US population. The great majority of the population in the country faces comparatively little contact with immigrants in their local labour markets. The 1990 Census of Population provides information on the regional distribution of the foreign born in the United States (Table 1). There were close to 20 million persons born outside the United States residing in the country in 1990. This constitutes about 8 per cent of the 248 million persons counted by the Census. The foreign-born population is not evenly dispersed across the United States. Immigrants tend to crowd into a small number of areas. As a result, close to three-quarters of all the foreign-born in the country reside in only six states: California, New York, Texas, Florida, New Jersey and Illinois. In total, these states are home to a little over 14 million foreign-born persons, which accounts for about 17 per cent of the overall population in the six states. The remaining immigrants are dispersed over the other 44 states of the Union, constituting only about 3 per cent of the population in these states. Since there were 163 million residing in these states, close to 66 per cent of the US population in effect encountered a very small or no presence of foreigners in its local labour market in 1990. The regional concentration of immigrants in the United States continues to the present. The distribution of recent immigrants admitted to the country during the 1996 fiscal year, according to their intended place of residence, is shown in Table 2. The same states that appear in Table 1 also

122

appear in Table 2 as the leading states receiving immigrants. The ten states shown in Table 2 account for close to 80 per cent of all immigrants, leaving the remaining 40 states to receive only a small portion of them. The figures in Table 2 depict the regional distribution of legal immigrants. Undocumented migration, however, is as concentrated as legal immigration. According to the most recent estimates of the illegal immigrant population in the United States by the Immigration and Naturalization Service (INS), there were approximately 5 million undocumented immigrants in the United States in October 1996. Some 4 million, or 80 per cent, reside in just six states: California, New York, Texas, Florida, New Jersey, and Illinois -- the same states where legal immigrants cluster (Table 3). Even within the states where they cluster, immigrants tend to concentrate in a few urban areas. As Table 4 indicates, close to 90 per cent of all immigrants admitted to the state of New York in 1996 intended to locate in or near New York City, even though the city accounts for only slightly over 40 per cent of the state’s population. In Florida, half of all immigrants intended to locate in Miami. And in California, Los Angeles is the intended residence of almost one out of every three immigrants in the state. The flow of immigrants has made the foreign-born a major population component of these urban areas. In 1990, the foreign-born accounted for close to 30 per cent of the population residing in the Los Angeles–Long Beach–Anaheim consolidated metropolitan area, 26 per cent of the Miami–Ft. Lauderdale metropolitan area, and 17 per cent of the New York–Newark–Jersey City consolidated metropolitan area. At the same time, besides these and another eight or nine major metropolitan areas, the remaining part of the nation (which includes most of its population and the vast majority of its territory) is host to a comparatively small number of immigrants, making the foreign-born between 3 and 4 per cent of the population in these areas. The heavy regional and urban concentration of immigrants suggests that the direct impact of immigrants occurs mostly on a limited set of local labour markets. Local impacts, however, can spread over time into those regions that do not receive many immigrants, a topic that will be examined in a later section. 3.

The industrial and occupational distribution of immigrants

The overall distribution of employment by industry is very similar for immigrants and natives, though immigrants tend to be somewhat over represented in the agriculture, manufacturing and trade sectors (Table 5). At the same time, immigrants tend to be under represented in the public sector, in transport and communications, and in the FIRE sectors (Finance, Insurance and Real Estate). Although the overall sectoral distribution presented in Table 5 does not display major differences between natives and immigrants, there are significant differences in the jobs that immigrants and natives take within each sector. Immigrants tend to have significant over representation in blue-collar jobs (Table 6). Among natives, 26.5 per cent of employed workers were in blue-collar jobs in 1990. Among the overall foreign-born population, the proportion of blue-collar workers was 40.2 per cent. And among recent immigrants (foreign-born persons who moved to the United States between 1980 and 1990), the proportion of blue-collar workers was 50.7 per cent in 1990. The greater proportion of blue-collar workers among immigrants is partially associated with the lower educational attainment of immigrants when compared to the native born. In 1990, 40 per cent of both the overall foreign-born population and recent immigrants had not been able to complete a high school education (Table 7). This is quite high compared to the 23 per cent among the native-born. 123

Given the greater proportion of the foreign born without a high school diploma, one of the key issues in the immigration debate is the extent to which immigration may have deteriorated the wage and employment conditions of high school dropouts in the United States. As a result of immigration, the increased supply of workers with less than a high school education is substantial at the low end of the schooling distribution (Table 8). Recent immigrants comprised 15.4 per cent of the total population with less than five years of education. The immigrant workforce, as a fraction of the overall workforce, tends to decline as one increases schooling levels. The combination of a comparatively large inflow of unskilled immigrants with the regional concentration of immigration noted earlier has led some to imply that recent immigration may have caused major negative effects on the labour market conditions facing high school dropouts in the United States. 4.

The impact of immigrants on employment and earnings

One of the best-known labour market phenomena in the United States over the last twenty years has been the rising rate of return to education. This pattern has been documented across sectors and regions in the country. One question that logically emerges is whether immigration has had any part in causing the earnings of unskilled workers to drop relative to those of highly-educated workers. Given the significant portion of the unskilled labour force accounted for by recent immigrants, some economists and many social observers have concluded that the answer to this question is yes. What is the theory and evidence on this issue? The simplest and most popular approach to examining the labour market effects of immigration is based on a simple labour market equilibrium model. In this model, an influx of immigrants increases labour supply in a particular occupation, sector or region, and, given labour demand, it generates a surplus of workers in the labour market that pushes down the wages of native workers. The estimates of the impact of immigrants on wages are usually carried out in two simple stages: first, the author identifies the increased relative supply of labour induced by immigrants in the labour market and, second, he or she calculates how much the increased relative supply affects wages. The latter computation involves the use of a so-called wage elasticity of demand for labour, which simply shows how a proportional increase in labour supply acts to reduce wages in a labour market. Based on this methodology, a number of authors have suggested that unskilled immigrants (specifically high school dropouts) have had a sharply negative effect on the wages of unskilled workers in the sectors, occupations and regions where they locate. For instance, based on a report released in May 1997 by a commission assembled by the National Academy of Sciences, Harvard economists George Borjas and Richard Freeman -- who participated in the commission -- concluded that “44 per cent of the decline in the real wages of high school dropouts from 1980 to 1995 resulted from immigration” (Borjas and Freeman, 1997). This provides a more recent numerical estimate for a similar result established by Borjas, Freeman and Lawrence Katz (also of Harvard) in 1992, indicating that: “immigration (in the 1980s) ... adversely affected the relative earnings of American high-school dropouts”. The key assumption of these studies, though, is that while immigration increases the labour supply of unskilled labour, the labour demand for educated labour stays unchanged. However, labour demand can be expected to increase for various reasons. For one, the influx of immigrants consists of both unskilled and skilled immigrants. Insofar as skilled labour is a complement of unskilled labour, the

124

increased supply of skilled labour will raise the demand for unskilled labour, which tends to push upwards the wage rate of unskilled labour. But even if the accompanying influx of skilled labour to any particular local labour market is small, there are other reasons why labour demand may still rise. In most economies, increased labour supplies of both immigrant and natives are absorbed by the economy without significant reductions in wages. The US economy has absorbed millions and millions of workers during this century, yet earnings and living standards have generally risen. The explanation is that increased labour supplies tend to generate other mechanisms in the economy that increase the demand for labour and, therefore, employment. Any influx that significantly reduces wages in a particular labour market also tends to attract industries into that labour market. Given the cheaper labour, these industries are labour-intensive and raise the demand for native labour in addition to absorbing the immigrants. Over time, then, the downward pressure on wages exerted by immigration is reversed by an increased demand for labour. This mechanism works especially well if the region of immigration has also been the subject of a growing economy. In that case, the institutional processes required to shift capital to a particular location are already in place. On the other hand, in labour markets in decline, where capital has been flowing out and local work has disappeared [in William Julius Wilson’s (1996) words], it is less likely that the increased immigration will be associated with a counteracting employment growth. In the latter case, one may then observe the wages and employment opportunities of unskilled natives being negatively affected by the immigrants, especially if the increased flow of unskilled immigrants is substantial. The empirical work on the labour market impact of immigration following this more complex approach suggests that the impact of immigration during the 1980s on the wages of high school dropouts has been insignificant (see Altonji and Card, 1991; Friedberg and Hunt,1995; Rivera-Batiz and Sechzer, 1991; and Gang and Rivera-Batiz, 1994). As David Card concludes in a thorough analysis of the issue: “increases in the population shares of different skill groups are associated with generally small changes in the relative wage structure” (Card, 1997). At the same time, the recent evidence on the impact of immigration has found significant negative effects of unskilled immigration on the employment opportunities of very unskilled natives, especially in metropolitan areas subject to large influxes of unskilled workers, such as Los Angeles and Miami. As noted earlier, factors such as the condition of the local economy, and institutional forces (such as employer discrimination) which may slow down or restrict the creation of unskilled jobs in areas of high immigration, can result in an increase in unemployment among native high school dropouts (see Card, 1997; and Ong and Valenzuela,1996). 5.

The regional impact of immigration

The last section focused on discussing the effects of immigration on local labour markets. In recent years, however, a number of scholars have suggested that immigration into a particular region may have effects that spill over into other regions. Given that immigration directly affects only a small part of the US labour market, the implication of spillover effects significantly increases the scope of the impact of immigration. The best-known regional effect associated with immigration is that, by inducing a drop in wages and or employment opportunities in a particular region, immigration induces natives to out- migrate to other regions. This phenomenon has been associated particularly with the out-migration of natives 125

with low levels of educational attainment in response to the increased immigration of unskilled labour. For instance, as Randall Fiber notes: “It is clear that there is a strong relation between the arrival of immigrants in a local labour market and the mobility patterns of native workers. The higher the concentration of recent immigrants in an area, the less attractive that area appears to have been for native workers ... the mobility responses of native workers to immigrant arrivals are concentrated among low-skilled and less educated natives” (Filer, 1992). The perception that substantial numbers of unskilled natives have indeed fled high-immigration regions is clear in the following statement: “blue-collar workers, facing competition for fewer and fewer relatively well-paying jobs, have fled cities with the highest levels of immigration” (Beck, 1997). This is the same conclusion arrived-at by William Frey in his analysis of geographical mobility patterns in the 1980s: “In the 1980s, more than two-thirds of minority immigrants were directed to only seven states -- led by California, New York and Texas ... these immigrants ... represent increased competition with native residents for jobs and housing, since most states that received large numbers of immigrants also lost white internal migrants to other parts of the country ” (Frey, 1995). Despite the various assertions about the induced out-migration of unskilled workers in response to the inflow of competing, unskilled immigrants, hardly any evidence has been brought to bear on this issue. Once one looks at the data, it becomes clear that the evidence does not support the view that greater immigration flows have caused unskilled workers to out-migrate from those areas. During the late 1980s, the out-migrants from California, Texas and New York -- the states with the greatest influx of immigrants, especially unskilled immigrants -- were overwhelmingly skilled and with a relatively high level of education (Table 9). These were professional and technical workers and administrators, not operators and labourers. These data thus contradict the statements noted earlier. This is, however, the same result that is obtained by Card (1997), who concludes that: “inflows of new immigrants to individual labour markets over the 1985-90 period did not generate large offsetting mobility flows by natives or earlier immigrants in the same skill group ”. 6.

Conclusions

This paper has examined the regional and sectoral effects of immigration in the United States. The following conclusions stand out: 1. Immigration to the United States is highly concentrated into a few states. The remaining part of the country is largely unaffected by immigration. 2. Immigration is highly concentrated into a few urban areas. The remaining part of the country (which constitutes the great majority of the population in the United States) is not significantly affected by immigration. 3. Although the industrial distribution of immigrants does not differ in major terms from that of the native-born population, there are major differences in the jobs that immigrants and natives take within each sector. 4. Immigrants tend to have a greater proportion of blue-collar workers than natives, and a larger proportion of workers who have not completed high school. Immigrants who 126

arrived in the country during the 1980s account for as much as 15.7 per cent of workers whose highest level of education is elementary school. 5. Although the supply of workers with less than a high school education has been increased by immigration, both theory and empirical evidence suggest that there has been very little, if any, impact of immigration on the wages of high-school dropouts. While it is true that the labour supply of these workers is raised by immigration, labour demand is also increased by immigration, resulting in the absence of major wage effects. 6. Although immigration may not reduce the employment opportunities of natives in the long-run, over the short-run, institutional rigidities and employer discrimination may act to raise the unemployment of those persons who compete with unskilled immigrants, especially if the migrant influx is substantial. 7. There is no evidence that the inflows of immigrants into the United States in the 1980s caused the exodus of native high-school dropouts from those labour markets. In fact, the data shows that the states with the greatest immigrant inflows were also the states from which highly-educated, skilled workers -- not unskilled labour -- out-migrated.

127

BIBLIOGRAPHY

ALTONJI, J. and CARD, D. (1991), “The effects of immigration on the labor market outcomes of less-skilled natives”, in J. Abowd and R. Freeman (eds.), Immigration, Trade and Labor, the University of Chicago Press, Chicago. BARRINGER, F. (1990), “A land of immigrants gets uneasy about immigration”, The New York Times, October 14. BECK, R. (1997), “Unfettered immigration has costs, too”, The New York Times, January 11. BORJAS, G.J. and FREEMAN, R.B. (1997), “Findings we never found”, The New York Times, December 10. BORJAS, G.J., FREEMAN, R.B. and KATZ, L. (1992), “On the labor market effects of immigration and trade”, in G.J. Borjas and R.B. Freeman (eds.), Immigration and the Work Force: Economic Consequences for the United States and Source Areas, The University of Chicago Press, Chicago, pp. 213-244. CARD, D. (1990), “The impact of the Marilee Boatlift on the Miami labor market”, Industrial and Labor relations Review, pp. 245-257, January. CARD, D. (1997), Immigrant Inflows, Native Outflows, and the Local Labor Market Impacts of Higher Immigration, National Bureau of Economic research, Working Paper No. 5927, Cambridge, Massachusetts, February. CHISWICK, B. and SULLIVAN, T. (1995), “The new immigrants”, in R. Farley (ed.), State of the Union: America in the 1990s, Social Trends, Russell Sage Foundation, New York, pp. 210-270. ESPENSHADE, T.J. (1997), “Taking the pulse of public opinion toward immigrants”, in T.J. Espenshade (ed.), Keys to Successful Immigration: Implications of the New Jersey Experience, The Urban Institute Press, Washington, D.C., pp. 89-116. FILER, R.K. (1992), “The effects of immigrant arrivals on migratory patterns of native workers”, in G.J. Borjas and R.B. Freeman (eds.), Immigration and the Work Force: Economic Consequences for the United States and Source Areas, The University of Chicago Press, Chicago, pp. 245-269. FREY, W.H. (1995), “The new geography of population shifts: trends towards Balkanization”, in R. Farley (ed.), State of the Union: America in the 1990s, Social Trends, Russell Sage Foundation, New York, pp. 271-336.

128

FREY, W.H. and TILOVE, J. (1995), “Immigrants in, native whites out”, The New York Times Magazine, August 20, pp. 44-45. FRIEDBERG, R. and HUNT, J. (1995), “The impact of immigrants on host country wages, employment and growth”, Journal of Economic Perspectives, pp. 23-44, Spring. GANG, I.N. and RIVERA-BATIZ, F. (1994), “Labor market effects of immigration in the United States and Europe: substitution vs. complementarity”, Journal of Population Economics, pp. 157-75. HUDDLE, D.L. (1994), “Can we afford so many unskilled immigrants?”, The Wall Street Journal, January 26. IMMIGRATION AND NATURALIZATION SERVICE (1997), Annual Statistical Yearbook, US Government Printing Office. MOORE, S. (1986), “Social scientists’ views on immigrants and US immigration policy: a postscript”, The Annals of the American Academy of Political and Social Science, Vol. 487, September, pp. 213-217. ONG, P. and VALENZUELA, A. (1996), “The labor market: immigrant effects and racial disparities”, in R. Waldinger and M. Bozorgmehr (eds.), Ethnic Los Angeles, Russell Sage Foundation, New York. RIVERA-BATIZ, F. and SECHZER S. (1991), “Substitution and complementarity between immigrant and native labor in the United States”, in F. Rivera-Batiz, S. Sechzer and I. Gang (eds.), US Immigration Policy reform in the 1980s: A Preliminary Assessment, Praeger Publishers, New York, pp. 89-116. US DEPARTMENT OF COMMERCE (1994), 1990 US Census of Population and Housing. WILSON, W.J. (1996), When Work Disappears: The World of the New Urban Poor, Vintage Books, New York.

129

Table 1. Regional distribution of the foreign-born population, 1990, United States Overall state population (millions)

state population as a % of total U.S. population 12.0

Foreign-born population in the state (millions) 6.4

Foreign-born in the state as a % of all foreign-born 32.9

Foreign-born as a % of total state population 21.7

California

29.8

New York

18.0

7.2

2.9

14.6

15.9

Texas

17.0

6.8

1.5

7.8

9.0

Florida

12.9

5.2

1.7

8.5

12.9

New Jersey

7.7

3.1

1.0

4.9

12.5

Illinois

11.4

4.6

1.0

4.8

8.3

Total for six states listed above Total for other states

85.4

34.3

14.4

73.4

16.9

163.3

65.7

5.2

26.6

3.2

All states

248.7

100.0

19.6

100.0

7.9

Source: US Department of Commerce, 1990 US Census of Population and Housing.

Table 2. Regional distribution of recent immigrants, 1996, United States Number of Immigrants

Percentage of Total

California

201 529

22.0

New York

154 095

16.8

Texas

83 385

9.1

Florida

79 461

8.7

New Jersey

63 303

6.9

Illinois

42 517

4.6

Massachusetts

23 085

2.5

Virginia

21 375

2.3

Maryland

20 732

2.3

Washington

18 833

2.1

Top 10 states

708 315

77.3

All states

915 900

100.0

Source: Immigration and Naturalization Service (1997).

130

Table 3. Regional distribution of undocumented population, 1996, United States

California

Number of Immigrants (thousands) 2 000

Percentage of Total

22.0

Texas

700

9.1

New York

540

6.8

Florida

350

8.7

New Jersey

135

6.9

Illinois

290

4.6

Top 6 states

4 015

80.3

All states

5 000

100.0

Source: Immigration and Naturalization Service (1997).

Table 4. Intrastate distribution of recent immigrants, 1996, United States Number of Immigrants California Los Angeles

Percentage of total

201 529 64 285

100.0 31.9

San Diego

18 226

9.0

San Francisco

18 171

9.0

Orange County

17 580

8.7

Oakland

15 759

7.8

San Jose

13 854

6.9

Riverside

10 314

5.1

Total for 7 largest recipient Standard Metropolitan Statistical Area (SMSA)

158 189

78.4

New York state New York City

154 095 133 168

100.0 86.4

10 594

6.9

143 762

93.3

83 385 21 387

100.0 25.6

15 915

19.1

Nassau-Suffolk Total for 2 largest recipient SMSAs Texas Houston Dallas

8 701

10.5

Total for 3 largest recipient SMSAs

El Paso

46 003

55.2

Florida Miami

79 461 41 527

100.0 52.2

10 290

13.0

51 817

65.2

Fort Lauderdale Total for 2 largest recipient SMSAs

Source: Immigration and Naturalization Service (1997).

131

Table 5. The industrial distribution of immigrants and natives, 1990, United States (Percentages) Industrial Sector Agriculture/Mining Construction

Natives 3.3

Foreign-Born 4.3

Recent Immigrants 5.2

6.2

6.4

7.1

17.4

20.8

20.8

7.3

5.3

4.3

Trade

21.1

22.5

25.0

2

7.0

6.1

4.7

32.7

32.2

31.2

5.0

2.4

1.5

100.0

100.0

100.0

Manufacturing Transport and Communications FIRE

Services Public administration Total

1

1. Recent immigrants include foreign-born persons who entered the United States between 1980 and 1990. 2. Finance, Insurance and Real Estate.

Source: US Department of Commerce, 1990 Census of Population and Housing.

Table 6. The occupational distribution of immigrants and natives, 1990 Employed persons 16 years and over (Percentages) Occupation

Natives

White collar

73.5

59.8

49.3

Blue collar

26.5

40.2

50.7

100.0

100.0

100.0

Total

Foreign-Born

Recent Immigrants

Source: US Department of Commerce, 1990 US Census of Population and Housing.

Table 7. The Educational distribution of immigrants and natives, 1990 Persons 25 years and over (Percentages) Level of Schooling

Natives

Foreign-Born

Recent Immigrants

Less than high school

23.0

41.2

40.6

High school diploma

31.0

19.6

17.9

Some college

25.6

18.8

17.8

College or more

20.4

20.4

23.7

100.0

100.0

100.0

Total

Source: US Department of Commerce, 1990 US Census of Population and Housing.

132

Table 8. The impact of recent immigrants on the supply of educational categories, 1990 Persons 25 years and over 1

Educational category

Recent immigrants as a % of all persons in the educational category 15.4

Less than 5 years of schooling 5 to 8 years of schooling

5.7

9 to 12 years of schooling

3.3

12 years of schooling

1.9

Some college

2.3

College or more

3.8

1. Recent immigrants include foreign-born persons who moved to the United States between 1980 and 1990.

Source: US Department of Commerce, 1990 Census of Population and Housing.

Table 9. Comparative educational attainment of out-migrants, 1990 California, Texas and New York Persons 25 years of age or older (Percentages) Level of schooling

Out-migrants, 1985-90

Overall population

California Less than high school

14.6

22.3

High school diploma

26.7

22.2

Some college

34.9

33.8

College or more

23.8

21.7

100.0

100.0

Less than high school

16.0

25.2

High school diploma

27.8

26.7

Some college

30.6

29.7

College or more

25.6

18.4

100.0

100.0

Less than high school

13.2

19.8

High school diploma

22.3

31.1

Some college

30.9

27.0

Total Texas

Total New York

College or more Total

33.6

22.2

100.0

100.0

Source: US Department of Commerce, 1990 US Census of Population and Housing.

133

MIGRATION AND THE LABOUR MARKET: SECTORAL AND REGIONAL EFFECTS IN CANADA

by Don J. De Voretz and Samuel A. Laryea Simon Fraser University, Burnaby

1.

Introduction

The underlying Canadian historical framework that allows us to analyse the labour market impact of Canadian immigration can be attributed to Dales (1964, 1966), who first merged the concepts of North American free trade and immigration. Dales argued that during Canada’s first major growth period, circa 1896-1910, the combined forces of trade, immigration into Canada, and Canadian emigration to the United States led to extensive economic growth -- i.e. greater gross domestic product but lower income per capita.1 At the turn of the century, according to Dales’ historical view, British and European immigration to Canada became increasingly urban and unskilled, causing labour 2 displacement and depressing wages for urban resident Canadians. Given free factor mobility to the United States this resulted in emigration of Canadian residents to the United States and lower wages for the remaining Canadian residents. The Harberger triangle rationalises this view (Figure 1). Given a fixed amount of capital and technology, as we then add immigrants to the Canadian labour force, the pre-immigration wage falls, resulting in a redistribution of income as capitalists gain income. This lower wage sends skilled resident Canadians who earned the higher pre-immigration wage to the United States. Unemployment is not an outcome in the Dales model because in his historical period wages were not sticky downward and free mobility across North America allowed the labour market to clear at any wage. Clearly, Figure 1 is a static case. If complementary capital enters the Canadian economy with the immigrant flow, then Figure 2 results, with greater employment, higher wages, and greater returns to capitalists. In such a case, immigration induces growth in GDP and per capita income or wages. The crucial question is what figure applies to modern day Canada with its recent high levels of immigration. Figure 3 depicts the recent historical fluctuations in immigration levels. It is clear that Canada’s modern immigration flows are subject to yearly fluctuations. The post-war yearly flows have ranged from a high of nearly 300 000 (1958) to a low of 85 000 (1985). These levels partially reflect policy responses to Canada’s changing macro-economic conditions. Econometric analysis has argued that these historical fluctuations were partially a response to changes in the Canadian unemployment rate (Marr and Siklos, 1995) while others argued that political considerations determined these levels 135

3 (Green and Green, 1995). Even subject to modern day policy restrictions, the most recent immigration inflows conform to some of Dales’ historical pre-conditions for wage compression and labour displacement. The 1996 Canadian Census reveals that 85 per cent of all immigrants -- and 93 per cent of those who arrived between 1991-96 (1.03 million) -- lived in a census metropolitan area. In fact, 71 per cent of post 1991 immigrants moved to either Toronto (40 per cent), Vancouver (18 per cent), or Montreal (13 per cent). The result of this rapid urban immigration has left Toronto and Vancouver with 42 per cent and 35 per cent of their respective populations foreign-born in 1996. In contrast, only 17 per cent of Canada’s population was foreign-born at that time. Clearly, Dales’ first condition -- urbanised immigration -- is being met in the modern Canadian immigration experience.

Dales’ second condition, namely free trade with the United States, has once again appeared with the advent of both the FTA and later NAFTA. Under the latter agreement and changes in the 1990 United States Immigration Act, substantial Canadian emigration to the United States has again appeared (DeVoretz and Laryea, 1997). Could it now be possible that Dales’ view of wage compression or labour displacement has again reappeared? Before attempting to answer this question we must note substantial modern day differences. First, unlike Dales’ historical period, some present day Canadian immigrants possess substantial financial and human capital (Coulson and DeVoretz, 1993) which implies that labour markets in some regions or sectors may have gained (Figure 2) while others may have experienced wage compression and labour displacement (Figure 1). Next, current Canadian immigration policy differs markedly from Dales’ historical period under the 1910 Immigration Act, which essentially only excluded unhealthy or criminal elements (Green and Green, 1995). In contrast, Canada’s 1976 Immigration Act and subsequent revisions has led to three separate entry categories: refugee, family class, and economic, with the latter consisting of entrepreneurs, investors, and “points” assessed independent migrants. The criteria for entry under the independent class for the points-assessed components are set out in Table 1. Those who were economically assessed must meet human and/or financial capital and labour market 4 criteria. The central question remains: how frequently was this points-based selection system used throughout our study period? During the 1968-76 period, or prior to the current 1978 Immigration Act, the majority of principal immigrant applicants (73 per cent) were screened via a point system for their economic suitability. In contrast, in the 1980s, the majority (71 per cent) of immigrants were not economically assessed. During the 1980s, immigrants either entered as reunited family members (53 per cent) or refugees (18 per cent). By 1995, the distribution again changed with over 50 per cent 5 of the newer immigrants entering via the economic class. These fluctuating policy conditions, from predominately economic admissions (1967-80) to largely family class entrants (1980s) to once again a majority of economic class entrants (post-1994) must be kept in mind when we later review the econometric literature of the late 1980s and 1990s. The supposition is that Figure 1, with its attendant labour displacement and earnings degradation, would appear whenever the majority of immigrants are not economically assessed and vice-versa. Canada’s labour markets, especially for immigrants, are both widely separated geographically and diverse in terms of ethnicity, source countries of immigrants, and size of the resident foreign-born stock. Again, keeping in mind the Dales paradigm, we would expect less human capital and lower wages for the foreign-born population if any city met the pre-conditions for Figure 1. We now turn to census information to test for these stylised facts across Canada’s major recipient cities.

136

A quick perusal of Tables 2-4 indicates that both differences in labour market performance and attributes arise between the foreign-born and Canadian populations in each city. In Vancouver, the foreign-born head of household was 3 years older, was more likely to be married, but worked 2.2 weeks less per year than his Canadian-born counterpart. The foreign-born head of household in Vancouver earned C$ 5 010 annually less than his Canadian-born cohort, perhaps owing to fewer weeks worked and greater employment in low skilled occupational groups. In Toronto (Table 3) and Montreal (Table 4), the foreign-born populations obtain similar demographic values as in Vancouver, with similar wage gaps of C$ 5 100 and C$ 5 448 arising between the economically active foreignborn and Canadian-born heads of households. Two points are clear: there is little variance in the (uncontrolled) earnings performance of immigrants across Canada’s three major immigrant receiving areas; the representative foreign-born male head of household earned 16-18 per cent less earnings in 1991, regardless of the destination city. In sum, several of Dales’ pre-conditions for employment displacement and earnings degradation appear in Canada’s major urban centres circa 1991. First, the Canadian-born urban population has more human capital with greater (i.e. 16-18 per cent) earnings than the foreign-born. The foreignborn earnings shortfall in turn lowered the average income in these three cities during that time. The major point still remains: do these conditions result in Canadian unemployment and/or earnings decline for the Canadian-born? We now turn to these issues. 2.

Labour displacement

To analyse the displacement (or wage compression) issue, elasticities of substitution between immigrant labour and Canadian-born labour must be computed from a production function which has been specified for all relevant Canadian industries. To clarify the estimating procedure, a theoretical exposition of the displacement phenomenon follows. Consider an immigrant receiving country that produces a single, non-exported output by means of two inputs: capital and homogeneous labour. The left panel of Figure 4 presents a situation in which 6 the world supply of labour is perfectly elastic at wage rate (We). The right panel shows the labour market in the immigrant receiving country, i.e. Canada. If labour were to seek its maximum earnings, with negligible transportation and other costs, and with no institutional impediments, then (od) workers would migrate to Canada. Thus, the Canadian labour supply would consist entirely of the foreign born and its wage rate would fall to the world equilibrium level, or (We). On the other hand, if Canada closed its borders, the wage rate in Canada would be (Wc). In addition, if independent events such as a rise in product price or increased complementary capital appeared in the economy, then these supply shifts induced by immigration could be offset by labour demand shifts. For example, with a partially open immigration policy (i.e. supply shifts to Sc’) the original equilibrium wage (Wc) would rise with a shift in the demand curve equal to (Dc’). This would result in no domestic labour displacement and a rising wage rate. Under a more realistic scenario for Canada, immigrants arrive under a binding quota of (ab=a’b’) workers, thus an increase in the labour supply from (Sc) to (Sc’) would result. In particular, this restraint on immigration appears appropriate for Canada after 1978 (Akbar and DeVoretz 1992). Now, this limited increase in immigration (ab) would have two important labour market consequences. First, the domestic wage rate would fall to (Wc’) and total employment would rise from (oe) to (ob). However, domestic employment declines from (oe) to (oa). Thus, immigrants displace domestic workers by the amount (ae). Second, when the wage rate falls from (Wc) to (Wc’), 137

labour earnings change from (oxze) to (otrb), of which (otsa) accrues to Canadian-born workers and (asrb) to immigrants. The earnings of Canadian-born workers have fallen from (oxze) to (otsa). On the other hand, returns to non-labour inputs or capital have risen from (xyz) to (tyr). Hence, in this intermediate case, Canadian capital owners benefit from immigration while others, i.e. native-born labour, are injured. Finally, the existing wage differential of (Wc’-We), relative to the immigrant’s opportunity cost, is no doubt substantial. Given such a substantial wage differential, a queue would form with immigrant applications exceeding yearly available slots. The actual wage and employment changes resulting from any immigration flow depend upon the domestic elasticities of labour demand and supply, the magnitude of the quota, and other assumptions implicitly embedded in Figure 4. For example, the more inelastic the demand and supply relationships, the greater will be the reduction of domestic wages due to any given amount of immigration. Moreover, the displacement effect will be greater, the more elastic is the labour supply and the less elastic is labour demand. We offer the following production function to estimate the relevant cross elasticities for modern day Canada. Yi = f(Ki, Ei, BIi, Dii)

(1)

where Yi = Value added in industry i. Ki = Capital stock in industry i. Ei = Employed labour in industry i aged 15 years and over who were born in Canada. BIi = Employed labour in industry i aged 15 years and over who were born abroad and migrated to Canada prior to 1981. DIi = Employed labour in industry i aged 15 years and over who were born abroad and migrated to Canada after 1981. The above production function was estimated in translog form for 125 Canadian manufacturing and non-manufacturing industries using 1986 data (DeVoretz, 1989). In terms of employment, these 125 industrial groupings represent approximately 93 per cent of the Canadian labour force in 1985. The estimated elasticities of complementarity are reported in Table 5. For convenience, pre-1981 immigrants are termed earlier immigrants, and those arriving later termed recent immigrants. This distinction between immigrant vintages is an attempt to explicitly recognise the hypothesis that the two immigrant pools are drawn from different populations. It is observed from Table 5 that the elasticities of substitution between Canadian-born workers and earlier immigrants as well as between Canadian-born workers and recent immigrants, are negative. This negative elasticity value implies that immigrant labour substitutes for Canadian-born labour. However, the corresponding t-values indicate that the elasticity coefficients are not statistically significant at the 0.05 level. Hence, the hypothesis that there is no displacement of Canadian-born workers by immigrants can be accepted for both the earlier and recent immigrant flows. Furthermore, it is also important to note that both earlier and recent immigrants affect the employment of the Canadian-born to the same extent (the elasticity values are identical). Hence, the hypotheses that the pre- and post-1981 immigrant pool have differential substitution effects with respect to the Canadian-born labour force must be rejected. Thus, circa 1986, there is no evidence to support Dales’ displacement hypothesis, at least for the entire economy.

138

Foreign-born intensive industries Critics are quick to note that Dales’ hypothesis applies to individual industries which contain a high concentration of immigrants. The most obvious tactic to detect any significant substitution effect is to select those industries with a high concentration of foreign-born under the supposition that more foreign-born in the labour force would cause more substitution. A highly concentrated foreign-born industry is defined as any 3 digit S.I.C. industry group with a greater than 23 per cent foreign-born labour force, since this is one standard deviation greater than the average. Several features of these selected industries are important to note. First, although many of the selected 59 industries are characterised by an unskilled labour force using labour intensive techniques (e.g. bakeries, clothing, and food processing), many groupings are highly skilled (e.g. Universities, machine shops, metal stamping). Thus, no generalisation appears a priori for this sub-set of industries other than the pre-selected degree of concentration of foreign-born workers. Why a heavy concentration of foreign-born workers in this diverse set of industries would cause displacement is open to speculation. One implicit hypothesis is that the job vacancy criteria under the 1978 Immigration Act may have been inappropriately applied (DeVoretz 1995), or that members of the family reunification class entered some or all of these foreign-born labour intensive industries. Table 6 reports the calculated elasticities of complementarity and associated test statistics for the foreign-born intensive industrial groupings. One result is obvious. In these industries both recent and earlier immigrants are significant substitutes for Canadian-born labour and the Dales’ hypothesis holds. However, capital is still not a significant complementary input to either recent or earlier immigrants. Finally, old and more recent immigrants are not substitutes for one another. These findings are in sharp contrast to the Canadian economy-wide results reported in Table 5. Given the above results, the actual degree of displacement between Canadian-born and foreign-born workers can be calculated for this subset of 59 foreign-born intensive industries. For example, in the meat and poultry, clothing, and university industry groups, the absolute marginal displacement of Canadian-born workers for each one per cent rise in their immigrant labour force is respectively 214, 582, 268 Canadian-born. Clearly, the degree of displacement depends directly upon the labour intensive nature of the industry and the absolute number of Canadian-born workers in the industry. In general, across the pre-selected 59 industries a 1 per cent rise in foreign-born labour would have reduced Canadian-born employment in 1980 by 2 543 workers. This section posed one central question: do immigrants displace Canadian-born workers? First, economy-wide there is no modern evidence circa 1986 that the post-war stock of immigrants significantly displaced Canadian-born workers. In addition, this lack of substitution was invariant to date of arrival. Also, economy-wide, immigrants did not require a significant amount of physical capital upon entry, while an expansion of the Canadian-born labour force did. This lack of economywide capital complementarity for immigrants, we believe, was a result of the average high human capital content embedded in Canadian immigrants upon arrival from 1967 to the 1990s (Coulson and DeVoretz, 1993). However, in the foreign labour intensive industries, significant labour substitution occurred between the foreign-born and Canadian-born. A combination of factors, including a greater than average foreign labour content, and a large share of value-added attributed to physical capital, led to labour

139

displacement in this sector. Equally important, under these conditions it is found that in this foreign intensive sub-sector, recent immigrants required a significant increase in physical capital. Moreover, the foreign intensive industrial sub-group does not conform to the stereotypical view of immigrant entry level industries. In fact, these industries include firms which use unskilled labour intensively, and other firms that use highly skilled human capital intensively. These two types of immigrant streams reflect the two broad components of the overall immigrant flows that resulted from the post1967 policy regimes. Immigrant policies after 1967 simultaneously selected immigrants with a greater level of human capital (1967-73), while later policies (post 1978) expanded the family reunification class, thus reducing human capital (Coulson and DeVoretz 1993). This indicates the inherent policy dilemma of attempting to simultaneously achieve humanitarian and economic goals while avoiding labour displacement. 3.

Wage compression

In addition to labour displacement, Figure 2 indicates that immigration can lead to higher or lower wages for the resident labour force. For example, two mutually exclusive outcomes for wage and labour displacement appear in Figure 2. First, immigration under an expanding labour demand curve may lead to greater native-born wages and employment due to the accompanying complementary human capital in the migration flow. In the opposite case, both earnings degradation and labour displacement occur in the native-born labour force if immigrants have little complementary human capital. We now turn to the econometric results which sort through these wage outcomes. Most existing studies on the effects of immigration on wages, for example Lalonde and Topel (1991), Altonji and Card (1991), and Roy (1997), rely on variations in immigration levels across cities or Census Metropolitan Areas (CMAs) to identify the consequent change in relative wages of immigrants and the native-born. Friedberg and Hunt (1995) refer to this approach as cross-section differencing. A problem with cross-section differencing is that, in the presence of free trade within the recipient country and coupled with capital or labour mobility, the result may be factor price equalisation. Thus an uneven distribution of immigrants across the country may not result (in the long run) in cross-section wage differences, since wages may be equalised by flows in goods or factors. Moreover, because immigrants are likely to be the most mobile of workers, as Newbold (1996) has established for Canada, they will probably move to those regions whose demand shocks have led to higher wages. Thus, an endogeneity problem ensues, prompting a naïve econometrician to possibly conclude that greater immigrant densities will lead to higher wages. To overcome these difficulties, Laryea (1997) employs an age-cohort technique developed by Suen to estimate the effects of immigration on wages. This approach estimates a two-stage Constant Elasticity of Substitution (CES) model that aggregates immigrant groups by age cohorts and aggregates cohorts into effective labour, which is used to study the substitution relationships between age cohorts and between immigrant groups. One advantage of this immigrant cohort size approach is that immigrant age cohorts are not mobile at any one point in time. Laryea estimated the following reduced form equations from an aggregated two-staged CES production function. The equation estimated in the first stage is summarised as follows: log wijk = α + Xijk γ + βj + (ρ2 - 1) log Njk

140

(2)

In the second stage of the estimation procedure the following equation was estimated: log wijk = α + Xijk γ + (ρ1 - ρ2) Log Mj + (ρ2 - 1) log Njk

(3)

where wijk = wage rate of individual i in age cohort j and immigrant group k. Xijk = an array of demographic characteristics (e.g. education, experience, marital status etc.). γ, ρ1, ρ2 = vector of parameters to be estimated. Mj = Labour supply from age cohort j. Njk = Number of workers from immigrant group k in age cohort j. The above equations were estimated by Laryea using data from the public use sample tape of individual records from the 1991 Canadian Census. The sample was further classified in eight fiveyear cohorts ranging from the 25-29 year-old age group to the 60-64 year-old age cohort. Individuals were also classified into four groups based on birth status: Canadians, early immigrants (those who immigrated to Canada before 1970), middle immigrants (those who immigrated between 1971 and 1980), and recent immigrants (those who immigrated to Canada after 1981). The raw count of the number of persons in these 8 x 4 subgroups gives the variable Njk used in the wage model. After the initial base runs, Laryea also conducted a simulation exercise involving a 20 per cent increase in the number of recent immigrants to ascertain the impact on wages. The results are summarised in the Table 7. The results show that the wage impacts of a 20 per cent increase in recent immigration levels on the native-born and other immigrant vintages are minimal. The wage decreases associated with this hypothetical inflow are no more than 1 per cent, ranging from -0.037 per cent for the native born and early immigrants in the 60-64 year-old age cohort, to a high of -0.102 per cent for recent immigrants in the 40-44 year-old age cohort. Part of the reason why these wage impacts are so small can be attributed to the relatively small percentages of immigrants making up the total labour force. For example, recent immigrants constitute only 3.8 per cent of the effective labour supply of the 40-44 year-old age cohort, which in turn makes up 18.9 per cent of the effective aggregate labour supply. Thus, economy wide, it appears that the wage impacts of immigration flows are minimal and have no adverse impacts on Canadian labour markets. We conclude again that Dales’ wage compression thesis does not hold in modern day Canada, at least not economy-wide. Wage impacts by industries: a panel analysis The absence of significant wage impacts of immigration flows economy-wide can mask the potential outcomes in the various industries in the economy where Dales’ thesis may hold. Estimating wage impacts across industries is also very important because certain industries serve as entry points for immigrants, and immigrants can potentially suppress wages of native-born workers in those industries. Laryea estimated the following random effects model to examine the impact of foreignborn labour on native-born wages by industry:

141

wit = α + β′ Xit + ειτ + ui

(4)

where wit = hourly wage rate of native-born worker i in year t. Xit = Set of exogenous variables including the proportion of foreign-born workers in various industries between 1988 and 1990. ειτ = Τraditional error term unique to each observation. ui = error term representing the extent to which the intercept of the ith cross-sectional unit differs from the overall intercept.

The above model was estimated using panel data from the 1988-90 Labour Market Activity Survey (LMAS). The model was estimated for the total sample and then by gender to address the outstanding literature, which suggests that females of foreign birth may in particular suffer a double negative effect. Laryea (1997) shows that looking at the total sample, and then the male and female subsamples separately, immigration had a positive impact on the wages of Canadians. The estimated wage elasticities suggest that a 1 per cent increase in the overall share of foreign-born labour results in a 1.1 per cent, 1.3 per cent and 1.4 per cent increase in wages for all Canadians, Canadian males, and Canadian females, respectively. However, when the data was disaggregated by industry, wage suppression was detected in the primary, transportation and storage, and retail and wholesale trade industries. This finding was detected in all three samples. The elasticities ranged from a low of 0.6 per cent in the primary industries for the female sample, to a high of 5.9 per cent in the transportation and storage industries for the male sample. Again, Dales’ hypothesis of low wage compression is found to hold within a limited part of the Canadian economy. 4.

Brain drain

As noted earlier, Dales argued that a third labour market adjustment was open to Canadians who felt the impact of immigration, namely movement to the United States. In fact, Canada has experienced periodic large scale movements, in particular, the highly trained to the United States. If Figure 1 holds for an industry or region in Canada, then increased immigration in the absence of a complementary capital inflow will lead to one of three possible outcomes for the resident labour force under contemporary conditions. Two outcomes have been noted above, namely labour displacement (unemployment) or wage compression. Obviously emigration of displaced or discouraged Canadian resident workers is the third possibility in times of open borders. Two contemporary periods of emigration to the United States have occurred in the post-war period. Parai (1965) documented the pre-1965 movement to the United States. For the 1980s and the 1990s, DeVoretz and Laryea (1997) describe the modern movement of Canadians to the United States. We concentrate on describing the latter flow to highlight the combined effect of trade, immigration, and emigration in a North American regional setting. The trends in the post-1980 Canadian emigration flows to the United States are summarised in Table 8. For the pre-FTA period, or 1982-89, the gross permanent Canadian flows to the United States were 13 940 professionals and 7 883 managerial workers. In addition, in the non-collegegraduate category, 2 951 skilled workers and 8 104 unskilled workers emigrated to the United States between 1982-89.

142

The corresponding immigrant numbers by skilled category for the shorter post-NAFTA 1990-95 period are 15 242 professionals, 10 229 managers, and 2 369 and 13 764 skilled and unskilled migrants, respectively. Comparing 1982-89 to post-FTA 1990-1995, one observes a significant increase in the gross flow of Canadians to the United States in all categories but the skilled occupations. The professional occupations recorded an increase of about 9.3 per cent. The growth rates for emigration flows in the managerial and unskilled occupations grew by 29.8 per cent and 69.8 per cent, respectively, after 1989. On the other hand, the skilled occupations experienced a decrease of almost 19.7 per cent between the periods. Two other perspectives are available to put this outflow in context. First, DeVoretz and Laryea (1997) have estimated the Canadian tax payer subsidy inherent in the post-1982 outflow of Canadians to the United States as C$ 9.1 billion (1993), or the equivalent of one major Canadian university operating solely to provide the United States graduates for this outflow. Secondly, this outflow can be described in terms of the ratio of Canadian graduates to the Canadian emigrant outflow to the United States. In the 1991-93 period, approximately 14 per cent of Canada’s highly trained graduates emigrated to the United States (DeVoretz and Laryea, 1997). For the nursing and managerial occupations in particular, the ratios emigrants to graduates were 19 and 40 per cent, respectively. This trans-border movement is of course part of world-wide flow, with Canada receiving a substantial gross human capital inflow (C$ 42.8 billion 1993) in the 1980s from the rest of the world (Coulson and DeVoretz, 1993). No doubt, for the post-1982 period, Canada is a net recipient of educated immigrants, however, this inflow occurs in the context of a substantial outflow to the United States as predicted by Dales. 5.

Conclusion

Does Dales’ gloomy thesis of wage compression, labour displacement, and emigration hold for modern free trading Canada? For limited sectors and types of human capital, a portion of each aspect of Dales’ thesis holds. In 59 out of 125 industries, labour displacement was significant, while in a smaller set of specific industries, wage compression appeared. Finally, emigration to the United States in some highly trained professions has re-appeared in the midst of large scale immigration inflows.

143

NOTES

1.

Given a neo-classical production function with no accompanying capital or technical change, recent immigration to Canada would lead to unemployment or lower wages or both for the resident stock of labour.

2.

Several Canadian studies in this historical period either depict immigrants as having little permanent impact (Pope, 1968), or as causing a substantial negative effect (Dales) on wages and jobs, or as actually impeding growth (Chambers and Gorden, 1966) by raising wages. No consensus has yet been reached on these conflicting interpretations of the role of early twentieth-century Canadian immigration in the development process. However, a choice of any of these historical views ultimately leaves an intellectual legacy that depicts the economic impact of fin de siècle Canadian immigration as negative or trivial.

3.

Marr and Siklos argue that for the post-war period, immigration levels reacted inversely to prior unemployment levels. They find limited evidence for two way Granger causality and dismiss the possibility of immigration levels affecting Canadian unemployment.

4.

Entrepreneurs and investors did not have to accumulate points but were required to provide employment and/or a minimum amount of capital (C$ 250 000-350 000).

5.

It should be noted that the economic category includes the principal applicants’ immediate family members.

6.

Perfect elasticity implies that within a relevant range, all the labour demanded will be supplied at the prevailing wage We.

144

BIBLIOGRAPHY

AKBAR, S. and DEVORETZ, D.J. (1992), Canada’s Demand for Highly Trained Immigrants: 1976-86, World Development, Vol. 21, No. 1. AKBARI, A. and DEVORETZ, D.J. (1992), “The Substitutability of Foreign-Born Labour in Canadian Production: Circa 1980”, Canadian Journal of Economics, No. 25, pp. 604-614. ALTONJI, J.G. and CARD, D. (1991), “The Effects of Immigration on the Labour Market Outcomes of Less-Skilled Natives”, in Abowd, J. and Freeman, R. (eds.), Immigration, Trade and the Labour Market, University of Chicago Press, Chicago. BICHA, K.D. (1965), “The Plains Farmer and the Prairie Province Frontier, 1897-1914”, Journal of Economic History, Vol. 25, June, pp. 263-267. CHAMBERS, E. and GORDEN, D. (1966), “Primary Products and Economic Growth”, Journal of Political Economy, August, Vol. 75, pp. 876-880. COULSON, R.G. and DEVORETZ, D.J. (1993), “Human Capital Content of Canadian Immigration 1966-1987”, Canadian Public Policy, No. 19, pp. 357-366. DALES, J. (1964), “The Cost of Protectionism with High International Mobility of Factors”, Canadian Journal of Economics and Political Science, No. 30, November, pp. 512-525. DALES, J. (1966), The Protective Tariff in Canada’s Development, University of Toronto Press, Toronto. DEVORETZ, D.J. (1989), Immigration and Employment Effect, Institute for Research on Public Policy, 89B, 3 November, p. 32. DEVORETZ, D.J. (1995), Diminishing Returns: The Economics of Canada’s Immigration Policy, C.D. Howe Institute, Toronto and Laurier Institution, Vancouver. DEVORETZ, D.J. and LARYEA, S. (1997), “Are We Losing It? Canada’s Brain Drain to the United States: New Evidence from the 1990s”, C.D. Howe Institute, Toronto and Laurier Institution, Vancouver. FRIEDBERG, R.M. and HUNT, J. (1995), “The Impact of Immigration on Host Country Wages, Employment and Growth”, Journal of Economic Perspectives, No. 9, pp. 23-44.

145

GREEN, A.G. and GREEN, D.A. (1995), “Canadian Immigration Policy: The Effectiveness of the Point - System and Other Instruments”, Canadian Journal of Economics, Vol. 28, No. 4b, pp. 1006-1041. HIEBERT, D. (1997), “The Colour of Work: Labour Market Segmentation in Montreal, Toronto and Vancouver, 1991”, Working Paper No. 97-02, Vancouver Centre of Excellence: Research on Immigration and Integration in the Metropolis (RIIM), Vancouver. LALONDE, R.J. and TOPEL, R.H. (1991), “Labour Market Adjustment to Increased Immigration,” in Abowd, J. and Freeman, R. (eds.), Immigration, Trade and the Labour Market, University of Chicago Press, Chicago. LARYEA, S.A. (1997), “The Impact of Foreign-Born Labour on Wage Rates in Canada”, Unpublished Ph.d. Dissertation, Department of Economics, Simon Fraser University. MARR, W. and SIKLOS, P (1995), “Immigration and Unemployment: A Canadian Macroeconomic Perspective”, in Devoretz, D. (ed.), Diminishing Returns: The Economics of Canada’s Immigration Policy, C.D. Howe Institute, Toronto and Laurier Institution, Vancouver. NEWBOLD, K.B. (1996), “Internal Migration of the Foreign-Born in Canada”, International Migration Review, No. 30, pp. 728-747. POPE, D. (1968), “Empire Migration to Canada, Australia and New Zealand”, Australia Economic Papers, Vol. VII, December, pp. 167-188. PARAI, L. (1965), Immigration and Emigration of Professional and Skilled Manpower during the Post War Period, Economic Council of Canada, Special Study No. 1, Ottawa Queens Printer, Ottawa. PENDAKUR, K. and PENDAKUR, R. (1996), “The Colour of Money: Earnings Differentials Among Ethnic Groups in Canada”, Working Paper, No. 96-03, Vancouver Centre of Excellence: Research on Immigration and Integration in the Metropolis, Vancouver. ROY, A.S. (1997), “Job Displacement Effects of Canadian Immigrants by Country of Origin and Occupation”, in International Migration Review, No. 31, pp. 150-161.

146

Table 1. Canada’s points system circa 1992 Category

Potential Points

Long-term Education Age Occupational Demand Occupation skill Experience Personal Suitability

12 10 10 15 8 10

Short-term English/French Arranged Employment Levels Control

15 10 10

Total

100

Source: Green and Green, 1995, p. 338.

Table 2. Social and economic attributes of Vancouver’s population by birth status, 1991 Variable Age Education

Canadian-born

1

All Foreign-born

2

Econ. 3 Canadian

Econ. 4 Foreign-born

31.87 (21.18)5

43.26 (19.18)

38.95 (9.84)

42.04 (10.19)

12.79 (2.87)

12.38 (4.12)

13.49 (2.69)

13.29 (3.69)

46.6 26.6 26.7

47.0 23.6 29.4

35.9 31.4 32.6

36.4 28.3 35.4

2.94 (1.44)

2.94 (1.46)

2.57 (1.30)

3.05 (1.38)

Highest level of education Elementary Non-University University Family size Married

35.2

59.5

56.3

73.3

18 365.46 (21 743.41)

14 902.82 (20 018.0)

28 462.45 (22 793.78)

23 452.41 (21 449.38)

Hours worked

23.07 (21.01)

21.15 (21.30)

33.74 (17.77)

33.13 (18.11)

Weeks worked

40.9 (16.14)

39.59 (16.90)

44.79 (13.13)

42.53 (14.92)

78.4 24.2 29.1 46.7

82.0 24.0 28.3 47.7

86.5 29.5 31.5 39.0

86.8 26.5 29.7 43.8

Wages and salaries

Full-time work Prof. occupation Skilled occupation Low-skilled occupation

1. 2. 3. 4. 5.

All persons in all ages born in Canada. All persons in all ages born outside Canada. All persons aged 25-65 born in Canada and in the labour force. All persons aged 25-65 born outside Canada and in the labour force. Figures in parentheses are the standard deviations.

Source: Laryea, 1997, p. 15.

147

Table 3. Social and economic attributes of Toronto’s population by birth status, 1991 Variable

Age

Canadian-born

1

All Foreign-born

2

Economicallyactive 3 Canadian

Economicallyactive 4 Foreign-born

29.30 5 (20.74)

42.11 (18.29)

38.70 (10.44)

41.74 (10.32)

13.05 (3.05)

11.97 (4.31)

13.77 (2.87)

12.85 (3.95)

Elementary Non-University

48.1 22.1

54.0 20.9

36.8 26.5

44.3 24.9

University

29.8

25.1

36.7

30.8

Family size

3.14 (1.42)

3.01 (1.46)

2.68 (1.30)

3.09 (1.38)

33.4

60.5

59.2

73.4

21 171.89 (24 500.17)

18 003.51 (21 686.42)

32 601.61 (25 590.70)

26 646.88 (22 400.23)

Hours worked

24.64 (20.89)

22.90 (21.12)

35.74 (16.51)

34.09 (17.21)

Weeks worked

42.13 (15.68)

42.12 (15.68)

46.62 (11.46)

44.62 (13.58)

33.4 28.7 26.4 44.8

86.9 23.4 26.4 50.2

89.4 35.7 28.2 36.1

91.3 25.6 27.5 46.9

Education Highest level of education

Married Wages and salaries

Full-time work Prof. occupation Skilled occupation Low-skilled occupation 1. 2. 3. 4. 5.

All persons in all ages born in Canada. All persons in all ages born outside Canada. All persons aged 25-65 born in Canada and in the labour force. All persons aged 25-65 born outside Canada and in the labour force. Figures in parentheses are the standard deviations.

Source: Laryea, 1997, p. 16.

148

Table 4. Social and economic attributes of Montreal’s population by birth status, 1991 Variable

Age

1

All Foreign-born

2

Economicallyactive 3 Canadian

Economicallyactive 4 Foreign-born

33.73 5 (20.93)

42.73 (19.08)

39.53 (10.11)

42.10 (10.27)

11.96 (3.75)

11.39 (4.74)

12.98 (3.35)

12.39 (4.47)

56.1 21.7 22.2

55.8 17.0 27.2

46.3 23.8 29.9

46.0 19.1 34.9

2.90 (1.34)

3.02 (1.50)

2.67 (1.23)

3.10 (1.41)

34.4

58.1

52.4

70.5

16 279.00 (19 793.12)

12 808.00 (18 941.54)

26 260.61 (20 639.23)

20 812.17 (21 157.92)

Hours worked

21.31 (20.40)

19.03 (21.08)

32.95 (16.76)

31.53 (18.84)

Weeks worked

41.51 (15.96)

39.65 (16.98)

44.95 (13.15)

41.98 (15.56)

81.5 25.1 28.4 46.5

85.9 26.7 24.9 48.4

88.0 29.4 30.2 40.4

89.8 28.8 25.8 45.4

Education Highest level of education Elementary Non-University University Family size Married Wages and salaries

Full-time work Prof. occupation Skilled occupation Low-skilled occupation 1. 2. 3. 4. 5.

Canadian-born

All persons in all ages born in Canada. All persons in all ages born outside Canada. All persons aged 25-65 born in Canada and in the labour force. All persons aged 25-65 born outside Canada and in the labour force. Figures in parentheses are the standard deviations.

Source: Laryea, 1997, p. 18.

149

Table 5. Elasticities of factor complementarities, 107 industries, 1986 Elasticity between 2

Native-born - earlier immigrants 3 Native-born - recent immigrants Native-born - capital Recent immigrants - capital Earlier - recent immigrants

Value

Std. Error

-0.97 -0.96 0.91 1.03 1.18

0.48 1.05 1.02 0.97 0.39

1

t-value

-0.47 1.01 1.93 ** 1.00 0.47

1. A single asterisk indicates significance at the .05 level of significance. 2. Those who arrived before 1981. 3. Those who arrived after 1981.

Source: DeVoretz, 1989.

Table 6. Elasticities of factor complementary: above average foreign-born intensive industries, 1986 Elasticity between 2

Native-born - earlier immigrants 3 Native-born - recent immigrants Native-born - capital Recent immigrants - capital Earlier - recent immigrants

1

Value

Std. Error

t-value

-0.67 -0.63 0.79 2.2 0.23

9.2 1.5 0.92 0.68 2.0

-6.3* 0.97 0.73 1.51 0.46

1. A single asterisk indicates significance at the .05 level of significance. 2. Those who arrived before 1981. 3. Those who arrived after 1981.

Source: DeVoretz, 1989.

Table 7. Impact on wages of a 20 per cent increase in the number of recent immigrants (Percent of wage increase or decrease) 25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

Native-born

-0.079

-0.047

-0.076

-0.102

-0.085

-0.070

-0.047

-0.037

Early immigrant

-0.079

-0.047

-0.076

-0.102

-0.085

-0.070

-0.047

-0.037

Recent immigrant

-0.086

-0.069

-0.095

-0.120

-0.103

-0.088

-0.066

-0.056

Source: Laryea, 1997.

150

Table 8. Canadian emigration to the United States by occupational group, 1982-1995 Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 3 1982-1989 3 1990-1995 1. 2. 3.

Professionals

Managerial

1 690 1 627 1 628 1 757 1 751 1 848 1 867 1 772 2 493 2 080 2 384 2 916 2 929 2 440 13 940 15 242

831 914 996 928 971 1 122 934 1 187 1 751 1 327 1 853 2 022 1 861 1 415 7 883 10 229

1

Skilled 264 343 368 378 336 383 380 499 752 539 322 318 262 176 2 951 2 369

2

Unskilled 664 900 933 1 097 1 127 1 143 1 111 1 129 3 571 2 709 2 082 2 092 1 798 1 512 8 104 13 764

These include workers in precision production, craft and repair occupations. These include operators, fabricators, labourers, sales, administrative support, farming, forestry, fishing and service occupations. Represents cumulative total flows for the respective years.

Source: US Immigration and Naturalization Service, Demographic Statistics Branch, Statistical Yearbooks, passim 1983-96.

151

Figure 1: Static Harberger Triangle with immigration

Figure 2: Dynamic Harberger Triangle with immigration

152

Figure 3. Canada’s historical immigration flows Number of Immigrants

Figure 4. Labour market displacement model

153

1995

1992

1989

1986

1983

1980

1977

1974

1971

1968

1965

1962

1959

1956

1953

1950

1947

300,000 250,000 200,000 150,000 100,000 50,000 0

MIGRATION AND THE LABOUR MARKET: SECTORAL AND REGIONAL EFFECTS IN MEXICO

by Rodolfo Cruz Piñeiro El Colegio de la Frontera Norte

1.

Introduction

Immigration flows from Mexico to the United States remained relatively homogeneous until the 1980s. Mexicans who crossed the border heading to the United States were typically men who travelled without their families. They came from rural areas of central and western Mexico and worked in temporary agricultural jobs, especially in the Southwest of the United States. Since the work was largely seasonal and temporary, migration tended to be cyclical, with workers crossing the border on a regular basis. However, in recent years the characteristics of this labour migration have gradually changed. First, there has been an important increase in the number of documented and undocumented migrants living in the United States. Second, the socio-demographic and labour characteristics of migrants have changed. Migrants are increasingly to be found working in the urban areas, and are participating in urban economic activities that nonetheless require poor qualifications and skills (services, construction and commerce). These changes are in part a reflection of structural changes within Mexico. The patterns of population in Mexico show a similar tendency towards urban concentration, and the employment structure in Mexico has also gone through important changes in recent years that have directly affected the labour and salary conditions of the Mexican labour force. There is no doubt that these changes have had direct and indirect impact on migration between Mexico and the United States. The main objective of this paper is to analyse the principal employment changes in Mexican labour markets and their connection to Mexican migration to the United States. This issue is considered in three parts: i) characteristics of internal migration flows in Mexico, chiefly from rural regions to urban centres; ii) changes in the employment structure in Mexico from a sectoral perspective; and iii) differences in sector insertion of migrant labourers who work in both Mexico and the United States. In preparing this analysis, we have drawn heavily on the data made available by a number of recent studies taking place along the northern border, notably the Survey of Migration at the Mexican 1 Northern Border (EMIF), which is conducted with the support of the Secretaria del Trabajo y Previsión Social (Ministry of Labour), the Consejo Nacional de Población (CONAPO), and El Colegio de la Frontera Norte (COLEF). In addition, data was taken from the study conducted by the 155

Binational Commission, and from the results of the Encuesta Nacional de Empleo (ENE), carried out by the Secretaria de Trabajo y Previsión Social, the Instituto Nacional de Estadística, Geografía e Informática (INEGI), the Encuesta Nacional de Empleo Urbano (ENEU), and by Mexican Census. 2.

New Mexican internal migration flows

Economic and social changes in Mexico have in great part determined the redistribution of the country’s population, changing the level and nature of internal migration flows. Within Mexico, migration between states has grown enormously: while in 1950 around 3.5 million (13 per cent of the total population) of people lived in a different state from the one in which they were born, by 1970 these number had increased to 7.5 million (14 per cent of the total population), and in 1990 to 15.4 million (18 per cent of the total population) (Corona and Tuirán, 1994). Accelerated urbanisation is perhaps one of the clearest features of internal migration in Mexico. It has been estimated that 35 per cent of total urban growth between 1960 and 1990 was caused by rural migration (Partida Bush, 1994). In recent years there are signs that the pattern of urbanisation is changing. For example, Browning and Corona (1995) document the transition of Mexico City from the greatest receiving city in the country to a sending city. This change is attributed to several factors, such as the increase in the cost of living, the incapacity of the labour force to absorb inflows, as well as a high level of criminality and environmental pollution. Other recent studies show that the northern border cities, the medium cities of the northern centre of Mexico, and Guadalajara and Monterrey are receiving an increasing number of people from Mexico City. There is also evidence that the percentage of international migrants heading to the United States from Mexico City has significantly increased. These new forms of internal migration, therefore, seem to also affect Mexican migration to the United States to some degree. 3.

Changes in employment structure in Mexico

During the import substitution industrialisation era, Mexican cities were able to absorb and offer employment opportunities to increasing numbers of workers. However, throughout the 1980s and 1990s, the Mexican urban system has been less capable of providing job opportunities. This may, in part, be a result of the transition away from the import-substitution model in favour of greater openness to foreign commerce. In addition, austerity measures implemented with the objective of reducing or restructuring the Mexican external debt directly affected the level of employment and wages. A large number of salaries, and of particular note, the minimum wage, were reduced by a half. These processes affected the volume, direction, and characteristics of both internal and international Mexican migration as well as the sectoral composition and distribution of employment. The Mexican population census and other national surveys show that from 1950 to 1995 the percentage of the population occupied in the agricultural sector steadily decreased, dropping from 58 per cent in 1950 to 25 per cent in 1995 (see Table 1). In addition, studies show that from 1970 to 1990 non-wage jobs grew by 63 per cent in this sector (Rendón and Salas, 1995). The National Employment Survey (ENE) also shows that from 1991 to 1995 the number of workers active in the agricultural sector kept decreasing as the number of non-wage workers increased (most of which were non-salaried family workers). Several authors agree that the increasing mechanisation of entrepreneurial agriculture and transformations in farming products have been the principal causes of decreasing employment in the 156

agricultural sector. In recent years, this process seems to be intensifying as a result of the opening-up of commerce and the lack of available credit. The labour force that works in the industrial sector increased from 15 per cent to 21 per cent of the total labour force over the period 1950-79 (see Table 1). However, during the 1980s and 1990s, the proportion of workers in this sector, with respect to the total labour force, significantly decreased. This decrease is due, in part, to technology advances and to the closing of factories caused by the widespread opening-up of markets that began during the 1980s (García, 1996). The creation of jobs in the manufacturing sector has also recently decreased. According to data made available by the Mexican Economic Census, while the number of people occupied in the sectors of commerce, manufacturing, and services increased to 4.25 million, 75 per cent of these new jobs belonged to the tertiary sector, especially in services. Only 25 per cent belonged to activities in the manufacturing sector. A significant proportion (40 per cent) of the wage jobs created by the manufacturing sector were in the Industria Maquiladora de Exportación (export processing industries). During the NAFTA discussions, there was a great deal of debate focusing on the growth of the maquiladora industry, and whether this growth would stop or stimulate the migration of Mexicans to the United States (Kopinak, 1997). The maquiladora industry in Mexico has grown considerably. The increase in the number of plants and personnel occupied within this industry over the period 1990-97 is shown in Table 2. Over this period the number of maquiladora plants grew 59 per cent. The increase was greatest from 1994 to 1997, after the devaluation of the peso (see Table 2). The proportion of maquiladora plants and employment located in the border states is a decreasing proportion of total maquiladora activity. In 1990, the border states accounted for 86 per cent of the total number of plants and 90 per cent of the jobs of this sector, but in 1997 these shares had dropped to 77 per cent and 80 per cent, respectively. However, even though the proportion is decreasing, the maquiladora industry obviously remains highly concentrated in the northern border cities. Although the maquiladora sector has grown enormously, it employs only a small proportion of the labour force. In 1995, the total number of workers in the maquiladora industry represented 12 per cent of the labour force employed in the manufacturing, mining, and energy sectors, and 2 per cent of the total of the total labour force in Mexico. As mentioned earlier, there have been discussions of the possible impact of the maquiladora industry on Mexican migration to the United States. Some authors believe that the rapid expansion of the maquiladora may reduce undocumented migration. Others argue that the maquiladora industry attracts people to the northern border cities, which in turn stimulates international migration as greater numbers locate closer to the border. It can be argued, however, that there possibly is little impact in either direction on migration flows. First, although the maquiladora industry is creating new jobs, it is a sector with a very small employment potential when compared to the great number of Mexicans who are seeking a job. Second, the EMIF, the survey of migration in the Mexican Northern border, reports that of the total migrants who come from the south of Mexico to the northern cities, only around 35 per cent would be interested in crossing to the United States, and only 31 per cent would like to work there. Finally, the EMIF also reports that of the Mexicans returning from the United States to live in Mexico, less than 4 per cent worked in the border city prior to migrating to the United States. Most of these return workers stated that they had worked in the construction and commerce (retail) sectors in Mexico; only a very small percentage had worked in the maquiladora industry. 157

The vast majority of workers who crossed the border (around 90 per cent) stayed less than seven days in border cities prior to crossing. Overall there has been a deterioration in employment possibilities in Mexico. The National Employment Survey reports an increase of 1.6 million of jobs during 1991-93, of which 48 per cent belonged to commerce and 41 per cent to services. A substantial number of these newly created jobs are low-paid and involve poor labour conditions. Thirty-seven percent of those who entered economic activity between 1991 and 1995 did it through the retail commerce sector (García, 1996). From 1950 to 1979 the number of paid workers went from 47 per cent to 63 per cent of the total labour force in Mexico. Since the 1980s these paid workers have lost importance, and in 1995 they represented only 57 per cent of the labour force (García, 1996). Other signs of an increasing proportion of unstable employment conditions are found in the increase in economic participation among older women, often with little education, who are married with children. They often work in low-wage jobs with no social benefits and poor labour conditions (Cruz Piñeiro, 1995). In sum, recent years have seen significant changes in the Mexican economy resulting in shifts in the structure and composition of the labour force: The principal changes observed from the end of the 1980s have been: − the manufacture sector reduced its capacity to create new jobs; − the importance of agriculture as a force to create employment has continued to decrease; − the share of wage-based employment in total employment stopped growing; − the shift towards service-sector employment has intensified; − small-scale activities increased in the cities; − a regional redistribution of economic activity took place; − precarious employment grew in urban areas; − the proportion of women in the labour force increased. 4.

Sectoral absorption of Mexican labour migrants

During the 1980s and 1990s, Mexican emigration rose to unprecedented levels in absolute terms (Bean et al.,1990; Corona, 1993). The Binational Study on Migration points out that the total size of the resident Mexican-born population in the United States in 1996 was within the range of 7 to 7.3 million persons, of which 4.7 to 4.9 million are legal residents and 2.3 to 2.4 million are unauthorised residents (Binational Study on Migration between Mexico and the United States, 1997). According to the Binational Study on Migration and EMIF reports, there has been a decline in the rate of the back-and-forth labour movement. One reason suggested for this is that immigrants are deciding to settle permanently in the United States or, at least, prolonging their stay, possibly as a result of the increasing difficulties associated with crossing the border and/or the higher cost of migrating to the United States. A further cause may be the increasing number of migrants who work in urban jobs, which are typically not seasonal, therefore requiring workers to stay for longer periods of time. These changes may also be linked to the processes of economic integration, productive restructuring, and labour flexibility that characterise the current age of the globalisation of capitalism (Sassen, 1990). 158

Mexican migration to the United States is very diverse. In order to better analyse these flows we distinguished two groups of migrants according to their permanent place of residence. Using primarily EMIF data, we were able to identify a first group of Mexican migrants who work in the United States but who declare their permanent residence somewhere in Mexico. These represent the so-called “circular” migrants. The second group includes those who have been established in the United States for longer periods of time and who declared the United States to be their permanent country of residence. This is an important distinction because these two kinds of migrants have different impacts on both Mexican and US society as well as labour markets. Mexicans living in the United States demand more public services and labour benefits because they are with their families, while Mexicans who live between the United States and Mexico are younger migrants who represent a cost-benefit relation that is more favourable to the receiving country (Santibáñez, 1997). Interestingly, most Mexican migrants heading for the United States had some kind of job in Mexico prior to migrating; however, the EMIF reports that the percentage of Mexicans who did not work before migrating is still high and that this proportion significantly increased over the period 1993-95 from 26.7 to 45.6 per cent (considering the quarter from July to October). This apparent increase in emigration driven by unemployment may have been caused by the devaluation of the peso in December 1994 and the severe economic crisis of 1995. Based on EMIF data on the flow of migrants from north to south, we were able to compare the jobs held by Mexicans in the United States and the jobs they held in Mexico before migrating (Table 3). First, it can be seen that within the group of migrant workers who declared their residence in Mexico, those who had a job in the agricultural sector declined between 1993 and 1995 from 60 per cent to 50 per cent. Second, a similar decline in the share of employment in the agricultural sector was noted for the same group when working in the United States, falling from 57 per cent to 41 per cent over the same period. The share of employment increased in the commerce and services sectors, especially in the manufacturing sector where it rose from 20 per cent to 31 per cent during 1993-95. We should point out that despite this new trend, employment is still highest in the agricultural sector for migrant workers who declare their residence in Mexico. The structure of the sectoral employment of migrants who declare their residence in the United States is somewhat different (Table 3). A lower proportion is employed in the agricultural sector than is the case among migrants who declare their residence in Mexico. Migrants residing in the United States are also being increasingly employed in the commerce and services sectors. Another important group are undocumented migrant workers (Table 4). The proportion of undocumented workers who, prior to migrating, had a job in the agricultural sector rose between 1993 and 1995 from 41 per cent to 52 per cent. While in 1993 these workers were mainly employed in the agricultural sector in the United States, in 1995 undocumented workers were mainly absorbed by the manufacturing, commerce, and services sectors. 5.

Conclusions

We have tried to show in what ways the changing patterns of Mexican migration to the United States may be due to transformations within Mexican labour markets and to changes in the geographic distribution of the population in Mexico. The new patterns of internal migration flows in Mexico that characterised the 1980s and the early 1990s were an expression of the new era of urban transition and territorial mobility of the Mexican 159

population. Since then, big metropolitan areas have shown signs of exhausting their capacity to absorb the inflowing labour force, forcing large numbers of people to look for new job opportunities in other cities and also outside of the country. The redistribution of the population towards mediumsized cities of the country has, at the same time, brought new problems of its own. There is no doubt that the transformations and deterioration of employment within Mexican labour markets in recent years have directly affected Mexican labour migration. During the 1980s, the economic policies implemented by the Mexican government directly affected the level of employment and the salaries of workers. Both the agriculture and manufacturing sectors have shown an inability to create new jobs, the waging of the labour force has stopped, and a growing deterioration of employment has appeared in the regional economies of Mexico. While the maquiladora industry has grown significantly since the devaluation of the peso, particularly in the non-border states, the industry is still small within the greater context of the manufacturing industry in Mexico. Results from the EMIF show that the Mexican migrants who crossed the border to the United States were not generally attracted by the maquila before crossing. Diminishing employment opportunities in Mexico, as well as severe income disparities, seem directly linked to emigration to the United States. Although the great majority of Mexican migrants had a job before migrating, the proportion of unemployed Mexicans crossing the border has significantly increased, possibly due to the economic crisis of 1995. Once in the United States, the sectors in which Mexican migrants tend to be employed vary according to their migration status. Overall, migrants of all categories are increasingly participating in the manufacturing, commerce, and service sectors, while their participation in agriculture has decreased. With this change in sectoral composition, there has been a commensurate lengthening of the average duration of stay in the United States.

NOTE

1.

The EMIF began in March 1993 and provides data over the successive years for the months of March through December. This paper draws on data collected from March to December of 1993 and March to December of 1995.

160

BIBLIOGRAPHY

BEAN, F.D., PASSEL, J. and EDMONSTON, B. (1990), Undocumented Migration to the United States: IRCA and the Experience of the 1980, The Urban Institute Press, Washington, D.C. BEAN, F.D., DE LA GARZA, R.O., ROBERTS, B. and WEINTRAUB, S. (eds.) (1997), “Introduction”, in At the Crossroads: Mexican Migration and US Policy, Rowman & Littlefield Publishers, Inc., Lanham. BINATIONAL STUDY ON MIGRATION BETWEEN MEXICO AND THE UNITED STATES (1997), Commission on Immigration Reform, U.S.A. and Secretária de Relaciones Exteriores, Mexico. BOSWORTH, B., COLLINS, S.M. and LUSTIG, N.C. (eds.) (1997), “Introduction”, in Coming Together ? -- Mexico-US Relations, Brookings Institution Press, Washington, D.C. BROWNING, H. and CORONA, R. (1995), “La emigración inesperada de los chilangos”, in Demos, Carta demográfica sobre México -- 1995, Mexico City. BUSTAMANTE, J., SANTIBAÑEZ, J. and CORONA, R. (1996), “Migration and immigrants; research and policies, Mexico-United States labour migration flows: some theoretical and methodological innovations and research findings”, SOPEMI-MEXICO-OECD. BUSTAMANTE, J. (1994), “Migración de México a Estados Unidos: un Enfoque Sociológico”, in La Migración Laboral Mexicana a Estados Unidos de América: Una Perspectiva Bilateral desde México, Secretaría de Relaciones Exteriores, Mexico City. CORONA, R. (1993), “Migracion Permanente Interestatal e Internacional, 1950-1990”, in Comercio Exterior, Mexico City, No. 43(8), pp. 749-761. CORONA, R. and TUIRAN, R. (1994), “Migración hacia las ciudades de Tamaño intermedio. Profundas transformaciones regionales”, in Demos, Carta demográfica sobre México -- 1994, Mexico City. CRUZ PIÑEIRO, R. (1995), “Inestabilidad y Volatilidad en el empleo de la fuerza de trabajo fronteriza”, in Estudios Demográficos y Urbanos, Vol. 10, No. 3, El Colegio de México, Mexico City. GARCIA, B. (1996), “Las Implicaciones del nuevo modelo de desarrollo”, in Demos, Carta demográfica sobre México -- 1996, Mexico City.

161

KOPINAK, K. (1997), “The post-NAFTA Impact of Mexican Export Processing Industries on Migration”, forthcoming in Labour, Capita and Society, December. PARTIDA BUSH, V. (1994), Migración Interna, INEGI-EL COLMEX-ISSUNAM, Mexico City. RENDON, T. and SALAS, C. (1995), “Cambios sectoriales del empleo (1980-1993)”, in Demos, Carta demográfica sobre México -- 1995, Mexico City. ROBERTS, B.R. and ESCOBAR LATAPI, A. (1997), “Mexican Social and Economic policy and Emigration”, in Bean, F.D., de la Garza, R.O., Roberts, B. and Weintraub, S. (eds.), At the Crossroads: Mexican Migration and US Policy, Rowman & Littlefield Publishers, Inc., Lanham. SANTIBAÑEZ, R.J. (1997), Flujos Migratorios México -- Estados Unidos. Evolución Reciente y Características Básicas, El Colegio de la Frontera Norte, Tijuana, mimeo. SASSEN, S. (1990), The Mobility of Labor and Capital. A study in international investment and labor flow, Cambridge University Press, New York. SMITH, P.H. (1997), “NAFTA and Mexican immigration”, in Bean, F.D., de la Garza, R.O., Roberts, B. and Weintraub, S. (eds.), At the Crossroads: Mexican Migration and US Policy, Rowman & Littlefield Publishers, Inc., Lanham.

162

Table 1. Economically active population by industry in Mexico, 1950-1995 (Percentages) Industry

1950

1970

1979

1991

1995

Agriculture

58.3

40.8

28.9

26.8

24.7

Manufacture minery and energy

14.8

21.8

21.1

16.9

15.9

3.1

4.7

6.4

6.1

5.4

Construction Commerce Services Non-specified

8.8

10.8

13.8

15.9

18.5

14.9

21.8

29.3

33.6

35.1

-

-

0.5

0.6

0.4

100.0 8 345 000

Total (percentage) Total (numbers)

100.0 12 955 000

100.0 19 177 000

100.0 30 534 000

100.0 33 881 000

Sources: Mexican Population Census, INEGI, 1950 and 1970; Encuesta Continua de Ocupación, INEGI, 1979; Encuesta Nacional de Empleo, Secretaría del Trabajo y Previsión Social (STPS) e Instituto Nacional de Estadística, Geografía e Informática, INEGI, 1991 and 1995.

Table 2. Number of maquiladora plants and employment in Mexico, 1990-1997 Total Year

Plants

Northern border States Jobs

Plants

Jobs

Other States Plants

Jobs

1990

1 731

454 432

1 487

410 885

244

43 547

1991

1 948

476 534

1 631

422 058

317

54 476

1992

2 085

512 761

1 757

449 953

328

62 808

1993

2 126

545 252

1 765

465 016

361

80 236

1994

2 083

594 078

1 705

510 092

378

83 986

1995

2 145

653 775

1 695

549 909

450

103 866

1996

2 465

769 143

1 907

627 078

558

142 065

1997

2 758

920 989

2 122

741 130

636

179 859

Sources: Instituto Nacional de Estadística Geografía e Informática , INEGI.

163

Table 3. Distribution of Mexican labour migrants by economic sector, 1993-1995 (Percentages) Economic sector

Residents in Mexico Employment in Mexico

Residents in US

Employment in the US

Employment in the US

1993

1995

1993

1995

1993

1995

Agriculture

60.1

50.6

56.7

41.2

42.4

31.1

Manufacture

18.8

19.6

20.0

31.1

30.7

34.9

Commerce and services

19.7

27.8

22.8

27.7

25.5

33.5

Other

1.4

2.0

0.5

0.0

1.4

0.5

Total

100.0

100.0

100.0

100.0

100.0

100.0

Sources: Encuesta sobre Migración en la Frontera Norte de México (EMIF); Secretaría del Trabajo y Previsión Social (STPS); Consejo Nacional de Población (CONAPO); and El Colegio de la Norte (COLEF).

Table 4. Distribution of undocumented Mexican labour migrants by economic sector, 1993-1995 (Percentages) Economic sector

Employment in the US

Employment in Mexico

1993

1995

1993

1995

Agriculture

44.2

29.9

40.7

52.4

Manufacture

30.5

37.2

26.1

22.7

Commerce and services

23.6

32.1

30.4

24.7

Other

1.7

0.8

2.8

0.2

Total

100.0

100.0

100.0

100.0

Sources: Encuesta sobre Migración en la Frontera Norte de México (EMIF); Secretaría del Trabajo y Previsión Social (STPS); Consejo Nacional de Población (CONAPO); and El Colegio de la Norte (COLEF).

164

PART III

LINKS BETWEEN FREE TRADE, INVESTMENT AND MIGRATION

165

A. PERSPECTIVES ON THE NORTH AMERICAN FREE TRADE AGREEMENT

NAFTA, foreign direct investment and economic integration: a United States approach Sidney Weintraub NAFTA, foreign direct investment and economic integration: a Canadian approach John Kirton NAFTA, foreign direct investment and economic integration: a Mexican approach Fernando de Mateo Venturini

167

NAFTA, FOREIGN DIRECT INVESTMENT AND ECONOMIC INTEGRATION: A UNITED STATES APPROACH

by Sidney Weintraub University of Texas

1.

Introduction

The North American Free Trade Agreement (NAFTA), even though it receives mixed reviews from both the US and Mexican publics, will almost certainly endure, short of a major blow-up between the member countries. The vested interests in its continuation are by now quite powerful. Trade between the United States and the other two member countries, Canada and Mexico, was substantial before NAFTA, but there is considerable evidence that much of the additional increase in trade after NAFTA came into existence was due to the agreement (Gould, 1996; Weintraub and Gilbreath, 1996; Kouparitsas, 1997). This has been particularly true for two-way US-Mexico trade. As the charts that are part of this paper make evident, foreign direct investment (FDI) into Mexico also grew markedly once NAFTA was on the horizon and then after it came into existence on 1 January 1994. Conventional wisdom as recently as the early 1980s was that economic integration between Mexico and the United States was a pipe dream, a fantasy of those who did not understand either the economic situations or the political realities in the two countries. 1 The policy context changed when the United States and Canada signed their free-trade agreement (FTA) in 1988 because this meant that the United States had abandoned its earlier position of avoiding preferential trade agreements with significant trading partners; and the context changed in Mexico when then President Salinas de Gortari realised that he could expect little in the way of investment or improved market access from the European Union, which in the early 1990s was preoccupied with its own future and with the former Soviet Union and Eastern Europe. Mexico’s decision to seek free trade with the United States required not just opening its market, which Mexico began to do unilaterally after its economic collapse in 1982, but also cultivating a more co-operative political relationship with its neighbour to the north. It is fair to say, therefore, that NAFTA had many fathers: US disillusion with trade relations based solely on the multilateral system of the General Agreement on Tariffs and Trade (GATT); the Canadian decision to discard its deep concern over free trade with the United States; an economic crisis in Mexico which led to a change in its development paradigm and self-initiated market opening; and Mexico’s realisation that its best and only significant external market in the near term was the United States. Getting approval of NAFTA turned out to be more controversial in the United States than in either Canada or Mexico. The disquiet in the United States was mainly over competing with a low-wage 169

country, which for this very reason would attract new and runaway investment, plus a resurgence of nationalism, which proved to be a more anti-NAFTA factor in the United States than in Mexico, where it had been expected. Getting congressional approval of NAFTA was more arduous than for any other recent trade agreement, including the many rounds of multilateral negotiations in the GATT. The original disquiet exists to this day and explains the inability thus far of President Clinton to secure fast-track trade authority from the House of Representatives. 2 The remainder of this paper examines outcomes four years into NAFTA and then tries to answer the question of whether the original fears of NAFTA opponents in the United States were justified. In doing this, laying out factual information is not sufficient because perceptions and fears have lives of their own. 2.

Trade, investment, and production outcomes

This section will devote more attention to economic developments in US relations with Mexico than with Canada. The reason for this is that the Mexico connection is controversial in the United States whereas the Canada relationship is generally taken for granted. Some material on Canada will be included, however. Mexico’s overall economic performance has been less than spectacular since the early 1980s. Recovery from the economic decline brought on by the debt crisis at the end of 1982 began modestly about five years later and continued into the early 1990s. Figure 1 provides the data on GDP growth in Mexico since 1990. The main objective of macroeconomic policy in Mexico starting at the end of 1987, a year in which consumer prices rose by almost 160 per cent (December to December), was to gradually reduce inflation. Two principal techniques were employed to this end. The first was to use the exchange rate as the anchor to achieve this and the peso, while it was devalued modestly on a daily basis, did not depreciate by as much as inflation in the seven years after the anti-inflation policy was adopted. The second was a series of pactos, or agreements between government, labour, and business to co-operate in keeping down fiscal deficits, prices, and wage demands. Increasing the level of economic growth played a secondary role to inflation control based on the premise that sustained growth in GDP could be achieved only after inflation was exorcised. This, quite logically, had its impact on Mexico’s trade. Mexico, over these years, unilaterally lowered its border protection and this income effect had some influence on the level of imports. However, this aspect was overwhelmed by the overall demand in Mexico, best measured by changes in GDP (Weintraub and Leigh, 1992). GDP growth, as indicated, was modest, but it was positive until 1995. Figure 2 traces the relationship between GDP changes in Mexico and the level of US merchandise exports to that country. The relationship is almost uncanny in its correspondence. Market opening under NAFTA surely has had some influence on the level of Mexico’s merchandise imports, but this has been secondary compared with Mexico’s economic performance. Figure 3 provides information on US-Mexican merchandise trade. As can be seen, this rose reasonably steadily in both directions between 1990 and 1993 and jumped up sharply in both directions in 1994, the first year of NAFTA’s formal existence. Mexico suffered a sharp economic decline in 1995, and US exports there fell, although not dramatically, whereas Mexican exports to the United States soared. Trade in both directions resumed their growth path of about 20 per cent in each

170

direction, in current dollars, in 1996, as Mexico recovered from its depression. Mexico did less well in its trade with countries outside NAFTA. Indeed, the proportion of merchandise exports going to the two NAFTA countries, particularly the United States, climbed from about 65 per cent to more than 80 per cent after NAFTA came into the picture. In the years before NAFTA, Mexico’s response to a balance-of-payments crisis was to impose import controls. This is what was done after 1982 and US exports consequently plummeted. This time, in 1995, Mexico used macroeconomic measures, which slowed the economy, but did not impose restrictions on imports from countries with which it had free-trade agreements. This explains the smaller, in percentage terms, US export decline in 1995 compared with what happened in 1983-84. One of the objectives of NAFTA, as President Salinas made clear, was to lock in economic reforms; and, for better or worse, this worked in 1995, despite the biggest single-year drop in GDP in more than 60 years. Figure 3 also shows US merchandise trade with Canada in the 1990s. Canada is far and away the largest US trade partner and, as the graph shows, this trade has been growing steadily. One of the most vociferous complaints against NAFTA made by its opponents is that there was a small trade surplus with Mexico before NAFTA and this turned into a trade deficit after NAFTA.3 Canada, in fact, has had larger bilateral merchandise trade surpluses with the United States than Mexico, yet one hears few complaints about US trade relations with Canada. Moreover, Mexico for much of this period had what was probably an overvalued exchange rate in order to curtail inflation, whereas Canada in recent years has had what is surely an undervalued exchange rate with respect to the US dollar. Mexico is seen as being different from Canada mainly because it is a developing, low-wage country. Perceptions and fears carry much baggage in influencing US public opinion. Free trade with Canada is relatively non controversial, whereas free trade with Mexico is highly contentious -- and this disparity of sentiment has little to do with actual trade outcomes. Figure 4 shows total investment in Mexico, both direct and portfolio, since 1990. The precision of the data is suspect because the accounting method often confused planned from actual capital flows, but the direction and general magnitudes are more solid. Direct investment into Mexico clearly skyrocketed once NAFTA came into existence; the level actually began to rise in 1991, when the decision to negotiate NAFTA was taken. Portfolio investment shows the behaviour of volatile capital as the Mexican economy began to experience political and economic problems in 1994 and then during the collapse of the economy in 1995. As Mexico recovered in 1996 and since, investment recovered as well. It is interesting that in 1997, about half of foreign investment into Mexico has been direct, whereas in earlier years, 1991-93, portfolio capital was considerably more voluminous. Between 60 and 65 per cent of foreign direct investment in Mexico originates in the United States. Figure 5 provides data on direct investment between Canada and the United States since 1990. Table 1 is designed to give some indication of the growth of intra-industry trade in North America. The intellectual case for free trade is that it stimulates specialisation and, one hopes, economies of scale, within industries as investors seek optimal locations within the larger, integrated market. Much of the trade between the United States and the other two countries in North America takes place in intermediate products, and this facilitates specialisation. In 1996, according to Mexico’s ExportImport Bank, 80 per cent of the country’s merchandise imports were intermediate products then used for further production in Mexico. This type of trade is exemplified by the maquiladoras in Mexico, mostly along the border with the United States, under which partially completed products are shipped to Mexico, value is added there -- mostly labour -- and the final product is returned to the United

171

States. Maquiladoras are by no means unique in their co-production aspects. There is much USMexican co-production and intra-industry trade in non-maquiladora activities as well. The large amount of US FDI in Canada has stimulated much trade between affiliated companies in the two countries. One estimate is that 70 per cent of Canada’s trade is not transacted at arm’s length but rather that 40 per cent is intra-firm and another 30 per cent the result of licensing and other intercorporate contracts (Hart, 1994). About 50 per cent of Mexican manufactured exports to the United States are intra-firm (Weintraub, 1997). Table 1 provides some elementary indication of how much merchandise trade in both directions between the United States and each of the other two countries in North America takes place in variants of the same products. The table is based on three-digit Standard International Trade Classification (SITC) figures showing the leading 20 exports and imports of the United States in trade with Canada and Mexico. Of these, 14 are included in both US exports to and imports from Canada, and 12 of the top 20 are included in US trade in each direction with Mexico. The list is dominated by automobile and parts production, where economic integration has led to considerable specialisation and intra-firm trade, but also includes many other industrial products. 3.

The role of NAFTA in economic outcomes

At about the same time that the Canada-US FTA went into effect at the start of 1989, Canada found it necessary for domestic reasons to tighten monetary policy to deal with inflation pressures. The result was slow growth for a number of years, an appreciated Canadian dollar due to capital inflows taking advantage of higher interest rates, and, as Figure 3 shows, slow growth in Canada’s exports to the United States and a relatively small bilateral merchandise trade surplus. Unemployment in Canada rose and remained high, and is still much higher than in the United States. Opponents of the FTA in Canada were quick to say: "See, I told you so." Supporters of the FTA mocked this view by saying that every time a sparrow fell in the forest, blame was placed on the FTA. In any event, this period of retrenchment passed and the FTA -- NAFTA at this point -- is now less controversial in Canada than it is in either Mexico or the United States. The hope of the supporters of NAFTA in the United States was that once the agreement was functioning, it would be taken for granted, as it is in Canada. It has not, even though the objective situation in the United States has been different from what it was in Canada immediately following the FTA. The United States has had seven years of steady economic growth. Unemployment has not increased, as it did in Canada, but has dropped to 5 per cent or less. Canada had to deal with inflationary pressure in 1989, whereas US inflation has remained remarkably low despite the steady economic growth and declining unemployment. The factual situation makes it quite hard to comprehend the persistence of anti-NAFTA sentiment in the United States, but it is evident that the fears persist. I will try in the next section to explain why this is so; but this section will seek to analyse how much effect NAFTA has had on economic developments in Mexico in order to set the context for discussion of the US fears. The single most important development in Mexico since NAFTA was the crash of the economy in 1995. As in Canada, where temporary economic slowdown was blamed on the FTA, the crash was blamed on NAFTA. Post hoc, ergo propter hoc. Is this fair? It is hard to see that it is. The antecedents of the financial and economic debacle in Mexico in 1995 have been described over and over again. The elements that led to the calamity are reasonably clear.

172

They include an overvalued exchange rate, or at least the perception that this was the case; a large current account deficit approaching 8 per cent of GDP financed largely by volatile capital; rising US interest rates throughout 1994, complicating Mexico’s ability to attract foreign portfolio capital; the uprising in Chiapas; the loss of foreign reserves as a consequence of a plague of political assassinations; and the use of tesobonos, or dollar-indexed government bonds, mostly of a short-term nature. These developments all coincided in 1994. When something can go wrong, it does go wrong, and all at once -- that was the story in Mexico. Which of these developments might conceivably be attributed to NAFTA? Not the political assassinations; not the uprising in Chiapas, which had historical antecedents and was planned long before January 1, 1994, when it took place on the day NAFTA went into effect; not the steady rise in US interest rates that was part of the policy of the US Federal Reserve Open Market Committee to forestall inflation; not the use of tesobonos, which were in part stimulated by the US interest rate rise. The potential connection with NAFTA is stronger for the overvalued peso, the growing current account deficit, and its financing by volatile capital inflows. The gradual overvaluation of the peso was the result of deliberate anti-inflation policy begun at the end of 1987 when the first pacto (pact) was unveiled. The use of the exchange-rate anchor to combat inflation, therefore, long antedated not only NAFTA itself, but even the conception of the idea that there might be a free-trade agreement with the United States. The only argument that can be used to assert that the peso was not devalued earlier is that this would have complicated US congressional approval of NAFTA at the end of 1993. It probably would have had this effect if a devaluation had taken place at precisely that time. But there were plenty of other times that a peso devaluation could have occurred. It could have taken place well before the end of 1993, but was not then considered because the anti-inflation program had not yet run its course. (The rise in consumer prices in 1993 was 8 per cent.) It could have taken place many times in 1994 before 20 December, when it actually occurred, but did not because the antiinflation program still had not run its course. (Consumer prices rose 7 per cent in 1994.) It could have taken place just before Salinas turned over the presidency to Ernesto Zedillo on December 1, but we know from his subsequent statements that Salinas did not want to be known as a president who abruptly devalued the peso. 4 It is thus quite a stretch to argue that NAFTA led to overvaluation of the peso. There is a contrary argument that NAFTA opponents in the United States have used, that argument being that Mexico wanted to make itself a base for exports to the United States. This is not consistent with persistent overvaluation of the currency. When this is pointed out, the argument is then transformed that the December 1994 devaluation showed Mexico’s true colours to have an undervalued peso. This, to put it kindly, is fantasy. Can anyone really believe that sane political leaders would deliberately put their country through a horrible economic downturn, with damaging consequences for the continuation of the party in power (the Institutional Revolutionary Party), by a peso devaluation in order to promote exports -- something that could have been done much earlier, and more calmly and methodically, if that were their motive? The argument that the growing current account deficit was an artefact of NAFTA has somewhat more plausibility, but not much more. The current account deficit began its rise during the 1990s, well before NAFTA. In 1993, the year before NAFTA came into effect, it was 6.4 per cent of GDP; the year before that, in 1992, it was 7.4 per cent. A modest case can be made that the high deficit was due to Mexico’s unilateral reduction of import barriers, which antedated NAFTA, but even this is 173

hard to sustain. The evidence, as indicated earlier, is that the influx of imports in the 1990s was due more to GDP growth than to tariff reduction or elimination of border barriers. If one wants to ascribe causation for the continuing and rising current account deficits, there are two other plausible choices. The one favoured by the Bank of Mexico is that the instigating factor was the high capital inflows, whose counterpart is a current account deficit (Bank of Mexico, 1995). My own choice is the overvalued peso. Mexico in 1994 had a merchandise trade deficit of US$ 18 billion, of which only US$ 3 billion was with its NAFTA partners and the rest with Europe and Asia. The rest of the US$ 28 billion current account deficit was in the services account. In any event, the largest part of the deficit was not with NAFTA countries, but with others. This leaves only one other possible NAFTA-related explanation for Mexico’s economic debacle in 1994, namely, the heavy reliance on volatile capital inflows. The argument here is that without NAFTA, or the pending prospect of NAFTA, the large portfolio capital inflow into Mexico would not have occurred. The data are contained in Figure 4, which shows the steady rise in portfolio capital inflows into Mexico in the 1990s, peaking in 1993. The cause and effect -- NAFTA on the horizon and therefore higher capital inflows -- is almost surely correct, even though Mexico was not alone among emerging countries in attracting foreign portfolio capital. But condemning NAFTA because it instilled greater investor confidence in Mexico surely entails perverted logic. Is confidence-building an undesirable attribute for a country to cultivate? Mexico allowed its current account deficit to grow based on its own logic that there was no problem as long as the capital inflows originated independently and not from deliberate borrowing from banks, as in the early 1980s. Mexico had it in its power to take measures to reduce the current account deficit, such as by slowing the economy, but rejected this kind of action. The conclusion that results from careful examination of what got Mexico into trouble in 1994-95 is that this was the result of conscious, if misplaced, Mexican economic policies in the face of horrible political events and rising US interests rates complicating Mexico’s policy-making in 1994. It takes a leap without corroborating evidence to place the blame on NAFTA. NAFTA has helped in overcoming the financial and economic collapse in 1995. As Figure 1 shows, GDP has since recovered smartly. And Figure 3 shows that as the domestic market collapsed, the rise in exports to the United States cushioned the economic fall in 1995 and then served to stimulate GDP growth in 1996 and 1997. Mexico’s exports of goods and services now constitute upwards of 35 per cent of GDP, and about 80 per cent of these exports go to the United States and Canada. These exports to its NAFTA partners now make up between 25 and 30 per cent of Mexico’s GDP. For better or worse, Mexico is thoroughly integrated into North America. 4.

Fears about NAFTA in the United States

The main case made against NAFTA is that it has led to a loss of hundreds of thousands of US jobs. Even if accurate -- and the basis for the claim is never made fully clear -- the United States has created upwards of 10 million jobs in the four years of NAFTA’s existence. The US economy, by any reasonable calculation, is at full employment. Every time unemployment goes down, and this has been happening regularly over the past four years, the US Federal Reserve considers whether it must raise interest rates in order to dampen job creation lest inflation rise. For the country as a whole, therefore, there is no basis for concern over job loss. Policy is made in the belief that the United States is already below its non-accelerating inflation rate of unemployment 174

(NAIRA). Individuals obviously do lose jobs even when the economy is at full employment and some of these job losses can be traced to imports from Mexico and Canada and plant transplants into those countries, particularly Mexico. By the same token, jobs are created from growing exports to NAFTA partners and the US government estimate is that jobs created from this are modestly greater than jobs lost from trade with and investment in these countries. The big engine of job-creation is the large domestic economy of the United States. Most job losses and job changes, similarly, originate in the domestic economy. Yet, the main argument heard in the anti-NAFTA, anti-fast-track, debate is that the job loss from trade is excessive. The fear of free trade with Mexico arose in the original debate on entering into NAFTA. The basis for the concern was Mexico’s low wages, which allegedly would make its production more competitive in the US market and simultaneously would attract investment in industries that do not need skilled workers. This drumbeat of anti-NAFTA fears has been incessant, particularly by US labour unions, for the four years the agreement has been in existence. The shift in the bilateral trade balance (Figure 3) was touted as leading to more job losses, even as the national unemployment rate was declining. These fears about job insecurity have by now taken on a life of their own. The anti-NAFTA position also has been used to pit Main Street against Wall Street, to use the US shorthand. Here the logic is that big business, the multinational corporations, benefit from NAFTA, but at the expense of worker jobs and even wages. The large corporations are, in fact, the principal supporters of free trade, just as organised labour is the main opponent. It is also true that real wages of low-skilled workers have not increased much, although this phenomenon long antedates free trade with Mexico.5 NAFTA, in this manner of looking at the world, becomes part and parcel of the process of globalisation, and there is inchoate concern about what all of this means for Main Street. But surely US withdrawal from the world, the urge to protect instead of to compete, is impractical. 5.

Conclusions

Most anti-NAFTA arguments in the United States have little substance. No significant job loss can be attributed to NAFTA. Trade in all directions has boomed since NAFTA came into existence. Specialisation in production, which is what was sought, is taking place. Even the shift in the bilateral merchandise trade balance between Mexico and the United States, which is so decried by the mercantilistic opponents of NAFTA, served to compensate for the sharp economic decline in Mexico in 1995. NAFTA is uniting the economies of North America, and particularly, the economy of Mexico, the weakest of the three, with the other two. The agreement did not become part of the woodwork in the United States, as did the Canada-US FTA once it got over it rocky beginnings, because of the dramatic and traumatic events in Mexico in 1994 and the precipitous decline in that country’s economy in 1995. Fears still dominate US thinking about free trade, as has been demonstrated by the inability of the Clinton administration to secure fast-track authority for future trade negotiations.

175

NOTES

1.

I can attest to this from the reviews of my book Free Trade between Mexico and the United States (1984), Brookings Institution, Washington, D.C.

2.

For those uninformed about US congressional procedures, fast-track refers to the ability of the president to submit a trade agreement, once negotiated, to the Congress for approval or disapproval, but without amendments. The procedure commits the executive branch to keep congressional leaders informed about the progress of negotiations as they proceed. Fast-track was adopted in 1974 after other countries refused to enter into complex trade negotiations with the United States if this involved a second negotiation with 535 members of the Congress.

3.

I do not wish here to enter into the legitimacy of using a bilateral surplus or deficit as a measure of the success or failure of a trade agreement. My view is that this is mostly meaningless posturing for a number of reasons, such as the nature of the global trading system and the great influence of transitory developments, such as ups and downs in respective economic growth rates. The United States generally exports more services to Canada than it imports, whereas Mexico generally exports more services (particularly tourism) to the United States than it imports.

4.

The Mexico City newspaper Reforma carried an interview with Salinas on three successive days in February 1997 and in the February 5 edition, Salinas was quoted as stating that he had obtained a commitment from his finance secretary, Pedro Aspe, when he appointed him that there would be no abrupt devaluation of the peso. This was in 1988, well before NAFTA was a gleam in Salinas’s eye.

5.

There is much literature of recent vintage in the United States about the effect of trade on the wages of low-skilled workers. William R. Cline (1997), Trade and Income Distribution, Institute for International Economics, Washington, D.C., deals with this subject and contains a critical review of the literature. This theme is not developed in this paper because of lack of space and because it has figured less prominently than job loss in the political debate over NAFTA.

176

BIBLIOGRAPHY

BANK OF MEXICO (1995), Information Annual 1994, 1994 Annual Report. CLINE, W.R. (1997), Trade and Income Distribution, Institute for International Economics, Washington D.C. GOULD, D. (1996), “Distinguishing NAFTA from the Peso crisis”, Southwest Economy, Federal Reserve Bank of Dallas, September-October, pp. 6-10. HART, M. (1994), What’s Next: Canada, the Global Economy, and the New Trade Policy, Centre for Trade Policy and Law, Ottawa. KOUPARITSAS, M.A. (1997), “A dynamic macroeconomic analysis of NAFTA”, Economic Perspectives, Federal Reserve Bank of Chicago, Vol. 21, No. 1, January-February, pp. 14-35. WEINTRAUB, S. (1984), Free Trade between Mexico and the United States, Brookings Institution, Washington D.C. WEINTRAUB, S. (1997), NAFTA at Three: International Studies, Washington D.C.

A Progress Report, Center for Strategic and

WEINTRAUB, S. and GILBREATH J. (1996), “North American trade under NAFTA”, NAFTA Effects Working Paper No. 2, Commission for Environmental Cooperation. WEINTRAUB, S. and LEIGH, B.B. (1992), U.S.-Mexico Free Trade Agreement: Economic Impact on Texas, Lyndon B. Johnson School of Public Affairs, University of Texas, Austin.

177

Table 1. Overlap among the United States exports and imports of the main traded commodities, 1996 (by SITC code) SITC

With Canada

With Mexico

784

Auto parts

X

X

781

Motor cars

X

X

782

Motor vehicles for transport of goods

X

752

Automatic data processing

776

Thermionic cold cathode valves

X

X

713

Internal combustion piston engines

X

X

764

Telecommunications equipment

X

X

772

Electrical apparatus for switching

X

778

Electrical machinery

X

773

Equipment for distributing electricity

X

821

Furniture and parts

X

792

Aircraft and equipment

X

641

Paper and paperboard

X

771

Electric power machinery

931

Special transactions

X

759

Parts for office machinery

X

984

Low value import transactions

X

X

X

1

333

Crude oil

334

Oil, not crude

1

X

X

X

X

1. US imports of SITC 333 are among the top 20 from both Canada and Mexico and US exports of SITC 334 are among the top 20 to each country. Source: US Department of Commerce.

178

Figure 1. G D P grow th in M exico A nnual cha nge in pe rcentage s (1993 prices) 8 6 4 2 0 -2 -4 -6 -8 1990

1991

1992

1993

1994

1995

1996

* 1997: Provisional figures. Source: Bank of Mexico.

Figure 2: U.S. exports and Mexico’s economy US exports

GDP growth (1993=100)

US exports (Billions of US $)

GDP growth

110

70 65

105 60 55

100

50 95

45 40

90 35 30

85

25 80

20 1990 Q1

1991 Q1

1992 Q1

1993 Q1

Source: Bank of Mexico and U.S. Department of Commerce.

179

1994 Q1

1995 Q1

1996 Q1

1997*

Figure 3. United States: merchandise trade with Mexico and Canada, 1990-1997 Mexico

Canada

Exports Imports Balance

Exports Imports Balance Billions of US $

Billions of US $

150

150

100

100

50

50

0

0

-50

-50

-100

-100

-150

-150 1990 1991 1992 1993 1994 1995 1996 1997*

1990 1991 1992 1993 1994 1995 1996 1997*

* Provisional figures. Source: U.S. Department of Commerce.

Figure 4: Foreign investments in Mexico, 1990-1997 DirectPortfolio Billions of US $ 40

30

20

10

0

-10

-20 1990

1991

1992

1993

1994

1995

1996

1997*

* Provisional figures.

Source: Bank of Mexico; Mexican Investment Board; Bancomer, Economic Report, September 1997.

Figure 5. Direct investments between the United States and Canada, 1990-1996 US investments to Canada Canadian investments to the U.S. Billions of US $ 10

8

6

4

2

0

-2 1990 1991 1992 1993 Source: U.S. Department of Commerce.

180

1994

1995

1996

NAFTA, FOREIGN DIRECT INVESTMENT AND ECONOMIC INTEGRATION: A CANADIAN APPROACH

by John Kirton University of Toronto

1.

Introduction

The advent of NAFTA aroused little of the political controversy in Canada that arose with such prominence in the United States. Canadian policy and analytical energies had been, and remain, consumed by the 1989 Canada-US Free Trade Agreement (CUFTA) and its impacts on trade, investment, growth, employment, and income (Gaston and Trefler, 1997). Although NAFTA incorporated and expanded CUFTA, it was the latter agreement, governing the largest bilateral trading relationship in the world, that has had by far the largest effect on Canadian economic, political and social life. Yet NAFTA in general, and its investment provisions in particular, were of distinctive interest to Canadians for several reasons. As an investment agreement, NAFTA brought important advances to the CUFTA regime. Some of the more significant changes were the creation of an investment dispute settlement mechanism and reductions in Mexican investment restrictions in several sectors (Gestrin and Rugman, 1994; Rugman, 1994). Canada was host to high levels of US FDI and itself had high levels of FDI in the United States. This advanced level of economic integration, coupled with changes in rules and institutions promised to have a major, if indirect, long-term effect on the Canadian economy. The advent of NAFTA inspired several concerns among those Canadians focused on its investment implications. The first, a miniature replica of concerns in the United States, was fear that NAFTA would induce the migration of production through a switch in FDI from Canada to Mexico, in order to take advantage of the latter’s lower cost structure and weak regulatory environment. This would consequently place downward pressure on wages, input costs, as well as labour, environmental and other regulation in Canada. The second, a variant of the first, was that investment would flow, in a process of corporate rationalisation, from Canada to the United States. The United States would then become the hub from which production would be exported to both the now open, but “hollowed out”, Mexican and Canadian markets. In contrast, others argued that NAFTA would produce, in both trade and related investment, a broader process of North American integration and rationalisation. This would in turn create a more balanced increase in foreign, domestic, and offshore investment in all three NAFTA countries (Gestrin and Rugman, 1994). Such investment, and the technological and growth gains that came with it, could be expected to be most intense in the hitherto fragile Canadian-Mexican relationship, where, relative to 181

previous economic integration and intergovernmental arrangements, NAFTA would have its largest effect. Indeed, in a strong variant of this argument, some asserted “that NAFTA will have a negligible impact upon the inward and outward FDI positions of Canada and the United States. In addition, the Agreement will give rise to strong increases in inward FDI for Mexico” (Gestrin and Rugman, 1994). The evidence to date provides very little foundation to support the concerns of NAFTA’s critics. In the three full years since NAFTA took effect, flows and stocks of Canadian FDI to Mexico have indeed shown rapid increases. They have also risen for Canadian FDI in the United States. But these outflows have been accompanied by even greater investment in Canada by the United States, Mexico, overseas investors, and Canadians themselves. This pattern suggest that NAFTA has above all intensified a full regional rationalisation and integration, which has increased intra-firm and intracorporate trade and production. This has led to a diffusion of management expertise, enhanced technology transfer, and efficiency gains. This process has been most dramatic in the CanadaMexican relationship, where new FDI from Canada to Mexico has flowed in the first instance from sectors where each country had a traditional comparative advantage, but has extended rapidly to sectors, such as autos, where Canada-US and Mexican-US integration was already well advanced. The recent patterns of Canadian and US FDI in Mexico, and that of US FDI in Canada, suggest that NAFTA’s investment provisions are inducing a process of technological transfer into the two smaller member countries, although their diffusion throughout and benefits for the broader economy remain unclear. To identify these impacts, this paper first discusses Canada’s extensive pre-NAFTA integration with the United States and its distant relationship with Mexico, the major changes NAFTA brought to the rules and institutional arrangements governing investment in the region, and the Canadian government’s current overall policy and perspective on FDI. It then examines the changing flows and stocks of investment between each of the three countries and from extra-regional and domestic investors, at both the overall and (where possible) sectoral level. It finally considers the challenges for greater regional integration that these trends suggest. Before proceeding, two caveats are in order.1 First, as Mexican statistics on FDI rely on the number and value of approved investments, rather than actual investments undertaken, its figures do not compare with those of the United States and Canada. Thus US Department of Commerce figures are relied on for the Canadian-US and Mexican-US relationships, and those of Statistics Canada for the Canadian-Mexican relationship. Second, limitations of space preclude a detailed consideration of the important question of the way NAFTA-induced FDI has increased the flow of technology from Mexico and Canada to firms in, and the economy of, the United States and Canada. 2.

The new NAFTA regime

Background Canada and the United States, which maintain the world’s largest two-way trade and an intense investment relationship, have enjoyed high levels of economic integration, in part as a consequence of the Defence Production Sharing Agreements of the 1950s and 1960s, the Autopact of 1965, and the CUFTA that took formal effect in 1989. These arrangements led not only to very high levels of US FDI in Canada but to tight intracorporate and intrafirm production systems that ensured a high level of advanced technology flow to Canada, and its rapid diffusion throughout Canadian industry,

182

resulting in strong productivity gains for the Canadian economy. Thus, the original purpose of FDI in Canada, to jump the high tariff walls of the “National Policy” to produce for the protected domestic market, was transformed into serving as a production platform to serve the entire continent. By the 1980s, one third of Canada-US trade came from three multinational firms: General Motors, Ford, and Chrysler. And a full 70 per cent of Canadian exports to the United States were accounted for by only 50 firms, primarily US and Canadian-owned multinationals. During the 1980s in particular, Canadians began to invest heavily in the United States. By 1986, Canada, one-tenth the economic size of the US, had not one-tenth but 60 per cent of the level of US investment in Canada. By contrast, the Canada-Mexico investment relationship remained very fragile. NAFTA’s rules and institutions However, the advent of NAFTA, beginning with its anticipation in 1990 and following its conclusion in 1993, brought to the Canadian-US relationship a more extensive set of disciplines governing investment, parallel agreements, and institutions for environmental and labour co-operation, and added Mexico to Canada’s long-standing partnership with the United States. Of initial importance was the anticipation that NAFTA would provide a permanent, stable, transparent, rules-bound regime for investment. Amidst the many exemptions of key sectors, it was hoped that this regime would provide a further binding of already open sectors, specific openings in key sectors, general disciplines across all sectors, a dispute settlement mechanism to handle disagreements, and institutions to implement and extend the regime. The autonomous anticipatory effect on corporate strategy is revealed by a 1993 survey of major US firms, in which the “Prospect of NAFTA” was ranked first among nine factors as most likely to drive changes in firm strategy and structure in North America (Blank and Krajewski, 1995). In the Canadian case, NAFTA’s anticipatory effects took hold as early as 1991 when the CEO’s of the largest, domestically-owned Canadian and Mexican firms began to meet in a “retreat” to explore and forge business relationships -- in part to offset any tendency to be swamped by much larger US competitors in a hub and spoke arrangement. The firms involved embraced a wide range of sectors, including financial services, transportation equipment, energy, and food and beverages. The conclusion of NAFTA and its coming into force on 1 January 1994 did in fact bring new stability as a result of several specific rules (Gestrin and Rugman, 1994; Rugman, 1994). NAFTA expanded the CUFTA investment rules to apply national treatment to investors and their investments, to include minority shareholders and portfolio investment in its guarantees, and to limit performance requirements on foreign-owned firms. NAFTA also introduced strict rules of origin in several key sectors and thus created a strong incentive (in cases where the parties’ external tariffs were high) for investment to concentrate within North America, at the expense of firms outside. In a move that increased the transparency of the regime, NAFTA also provided that any exceptions to its investment disciplines be individually itemised in annexes for each country, rather than having all existing limitations grandfathered. It required state and provincial governments to specify their exceptions within two years. Although this time limit was not met, the process is well underway (Howse and Soloway, 1997). In overall terms, Canada had the fewest number of reservations (48, compared to 50 for the US, and 89 for Mexico), and focused its sectoral reservations on culture, natural resources (agriculture, energy, and fishing), social services, and transportation (Rugman and Gestrin, 1994).

183

The Mexican government accepted obligations to substantially open many sectors, provide compensation for expropriated investments, limit uncompetitive practices and protect intellectual property rights. Mexico’s NAFTA opening took place in two waves. The first stage occurred immediately with the coming into force of the agreement. The second takes place over the next ten years as the existing investment restrictions are phased out. Its extensive list of reservations embraced energy, social services, transportation, cultural industries, and, distinctively, many manufacturing sectors such as automobiles and parts, maquiladora industries, and basic petrochemicals (in all cases largely for a ten year period). Of particular importance were the rules contained in NAFTA’s Chapter 11 on investment, and the new investment dispute settlement mechanism it created. These rules allow individual investors (rather than governments) to initiate disputes against member states and have them arbitrated through one of two binding international arbitration processes. By late 1997, there was one case proceeding through this mechanism, brought by a US firm, Ethyl, which charged that the Canadian government had infringed its rights in banning, on environmental grounds, the interprovincial and international trade of a gasoline additive, MMT, of which it was the sole distributor in Canada. A second case currently pending involves a US waste disposal firm, Metalclad, that has been prevented from opening a new facility by a Mexican state government charging the facility has failed to meet local building and environmental regulations. More broadly, the fifty or so trilateral intergovernmental institutions created or catalysed by this “NAFTA regime” established new mechanisms for regional co-operation, and an impetus for the harmonisation of regional standards that had long created significant barriers for Canadian firms seeking access to the US and Mexican markets (Kirton and Fernandez de Castro, 1997). NAFTA’s professional mobility provisions were also of importance in allowing multinational firms to move to integrated production, by allowing designated professionals to work for extended periods in the other NAFTA countries. Canada’s Current FDI Policy and Perspective Canada’s willingness to open its own market to inward FDI in return for a major opening of the Mexican market and greater certainty in the US market reflected an evolving policy shift toward a regime of rapidly seeking more inward FDI from around the world, while supporting the growing outward FDI of Canada’s emerging multinationals. The Canadian government’s current investment strategy, adopted in June 1996, seeks primarily to substantially increase global inward FDI into Canada. It is based on a calculation that FDI in Canada directly and indirectly provides 30 per cent of Canada’s jobs, over 50 per cent of its overall exports, and 75 per cent of its manufacturing exports (DFAIT, 1997a). Every C$ 1 billion in new FDI in Canada is estimated to create 45 000 jobs in Canada over the ensuing five years. Yet Canada’s share of global FDI declined from 11 per cent in the early 1980s to 4.5 per cent in 1995, while Canada’s share of inward FDI dropped 50 per cent during the past decade, from 7.5 per cent in 1985 to 3.6 per cent by 1997. It is believed that “the substantial increases in investment in Canada attributable to NAFTA” were overwhelmed by the attraction of Asia and Europe as destinations for the burgeoning world stock of FDI, which quadrupled from 1985 to 1995 (DFAIT, 1997a). Canada’s strategy thus focused on the top five countries providing outward FDI (the US, Britain, Japan, Germany and France), and sought to attract inward FDI from them in eight priority sectors (information technology, life sciences, automotive, aerospace, agro-food, forest products, mining, and chemicals-petrochemicals). A core element of this strategy was to increase global consciousness of the benefits of locating and producing in Canada to serve the full NAFTA market. It was based on 184

analyses showing that among seven countries (Canada, the US, Britain, France, Germany, Italy, and Sweden), Canada has the lowest business costs, both overall and in each of the eight sectors studied. The US ranked fourth overall and in most of the individual sectors (KPMG, 1997). Underlying this policy was a desire to offset what appeared to be Canada’s diminishing attraction as a location within North America for inward FDI, as Mexican locations became more appealing, and as US locations appeared the best from which to service the entire NAFTA marketplace. Canada’s share of inward FDI stock in the North American region fell from 24 per cent in 1985 to 21 per cent in 1990 and to (an estimated) 15 per cent in 1996. The US share of inward FDI rose from 69 to 73 and 76 per cent over the same period, while that of Mexico also rose, from 7 to 6 and 9 per cent (Industry Canada, 1997). At the same time, however, Canadian and Mexican levels remained higher than their share of the regional economy as a whole. Moreover, Canada’s decline in the North American region was less significant than in its world share (which dropped from 9 to 7 to 4 per cent), and its G7 share (which dropped from 16 to 11 to 8 per cent). Both facts pointed to the independent positive effect of NAFTA as an “insurance policy” and its potential as a vehicle for offsetting otherwise disturbing trends. 3.

NAFTA’s Impact on Investment, Technology Transfer, and Host Economies

Although Canada’s overall FDI strategy focused on increasing inward FDI in Canada from the United States and the major overseas providers, the most sensitive and visible effect of NAFTA has been to increase Canadian outward FDI in Mexico. Simultaneously, the pre-NAFTA pace of Canadian outward FDI to the United States has slowed, while inward FDI from the United States to Canada has increased substantially. There has also been an increase in Mexican FDI in Canada, although the overall levels remain minuscule. The overall pattern suggests that NAFTA is fulfilling the objectives of the Canadian government’s overall investment strategy, despite the NAFTA’s strong tendency to pull US and Canadian investment into Mexico and away from Canada, especially when compared to the changes in investment between the United States and Mexico in the post-NAFTA period. Underlying this pattern is Canada’s strong cost advantages over the US as an investment location, and NAFTA’s overall impact in allowing for intensified integration and rationalisation of production systems. Canada’s FDI in Mexico Those Canadian critics of NAFTA who alleged it would induce a “sucking sound” southward to Mexico can find some support, at first glance, in the pattern of outward Canadian FDI to Mexico. For NAFTA does appear to have had a strong effect in dramatically stimulating Canadian investment in Mexico. The stock of Canadian investment in Mexico had remained at a constant low from 1988 to 1991, at a level of about only C$ 200 million. The anticipation of NAFTA prompted a sharp rise in 1992 and 1993 to a level of C$ 520 million. The completion of NAFTA led to a doubling in the single year of 1993 to 1994. After a small drop in 1995, the increase resumed, to reach a level of C$ 1 268 billion in 1996. There was thus a sixfold increase in the stock of Canadian FDI in Mexico in the five years from 1991 to 1996. This compares favourably with the increases in the stock of Canadian outward FDI globally, which increased only 164 per cent during the full decade 1986-96 (DFAIT, 1997b). An indication of NAFTA’s autonomous effect in inducing Canadian FDI into Mexico can be obtained by comparing the purely pre-NAFTA years of 1984 to 1990, with the NAFTA period of 1991 to 1994 185

(Ramirez de la O, 1996). During that time, the stock of outward Canadian FDI globally increased from an average annual level of C$ 56.2 billion to C$ 86.9 billion, for an increase of 55 per cent. Of that, investment flows into the United States rose from C$ 36.4 billion to C$ 48.5 billion, or 33 per cent, while those into Mexico rose from C$ 0.2 billion to 0.5 billion, for an increase of 150 per cent. However, from a Canadian perspective, the “sucking sound” represented by these increases was barely audible, in that the levels remained very low. They were a tiny fraction of total investment in the Canadian economy and of the increase in investment during this time. Moreover, Canadian FDI in Mexico represented under 1 per cent of the total C$ 74 billion of Canadian FDI abroad in 1995. To a substantial degree, these Canadian investments in Mexico were destined to serve the domestic Mexican market, or Latin American or global markets, rather than to produce for re-export to Canada or to the United States. In sectors such as food and beverages, the incentive was to take advantage of a demographically young Mexican market and thus diversify from the ageing Canadian market where consumption was static. In sectors such as mining, it was the availability of the raw material in Mexico (for example Placer Dome’s investment in Sonora) and production for the global market that provided the incentive. Yet in other cases, such as the automotive sector, its purpose was to become part of a North American regional production system. In few cases did it involve the closure of existing plants in Canada and the creation of new or expanded plants in Mexico to service customers previously supplied from the Canadian plant. From a Mexican perspective, inward FDI from Canada represented a large and welcome addition to Mexico’s rather low stock of FDI, a stock that remained heavily dominated by the United States. The importance of its NAFTA partners to Mexico was evident in the first eight months of 1994, when Canadian and US companies invested US$ 2.4 billion in Mexico, a figure which represented 55 per cent of Mexico’s total inward foreign investment (Ready and Sandver, 1995). These flows propelled Canada to the point where it commanded 9.4 per cent of total FDI in Mexico in the first half of 1995. Canada also was one of the most important sources of inflowing investment in 1996, helping lead Mexico’s recovery from the peso crisis of the previous year. This gave Canada a level of FDI in Mexico one-fifth as large as that of the United States -- a country ten times as large, much more proximate, and far more historically involved. The value of this investment to Mexico, and the technology flows it encouraged, can be assessed by examining its sectoral composition. Canada’s investment stock in Mexico at the end of 1994 was concentrated in mining, telecommunications, automotive parts, and beverages (Ramirez de la O, 1996). More broadly, as NAFTA took hold, Canadian FDI in Mexico came to be led by automobiles, computer software, real estate, services, mining, and telecommunications, with a major increase in banking. By the end of 1996, mining, banking, and telecommunications stood out from the rest (DFAIT, 1997b). Overall these sectors show the strong autonomous impact of NAFTA’s investment provisions, which helped liberalise the previously closed (to foreign ownership) sectors of mining, telecommunications, banking, and energy (Ramirez de la O, 1996). They also stand in partial contrast to the overall pattern of NAFTA-induced investment in Mexico from 1991-94, which concentrated in manufacturing (representing over half the increase), processed foods, textiles, metallic goods and machinery, restaurants and hotels, and professional services (Ramirez de la O, 1996). These three sectors are areas of Canadian comparative advantage globally, and ones where Canadian firms possess globally competitive technology. With the exception of mining, they suggest that Canadian firms (both home-based MNEs in the case of banking and transnational MNEs in the case of telecommunications) are primarily interested in producing in Mexico for the Mexican market and

186

using their firm-specific advantage of technology transfer to do so. In the case of banking, it was Canada’s leading experience with automatic teller machines and similar systems that provided a competitive edge. In the case of mining, Canada’s home-based MNEs appear to be exploiting their firm-specific advantage of technology and the location-specific advantage of raw material availability. The further potential, identified as existing in gas and energy, is similarly motivated (DFAIT, 1997b). Canadian FDI in Mexico has thus far focused on capital intensive sectors, although over time a spread into labour-intensive sectors should be expected, as Canadian firms increasingly take advantage of Mexico’s absolute advantage (Ramirez de la O, 1996). There is little evidence that this investment is thus far driven by a search for low wage or pollution havens in Mexico. The absence of any reports of major Canadian firms reducing or closing operation in Canada in order to relocate in Mexico, especially when not part of a major corporate rationalisation, reinforces this conclusion. Rather than looking at low standards or lack of enforcement as a competitive advantage or investment/relocation incentive, Canadian-owned subsidiaries in Mexico are concerned that they will respect Mexico’s generally high level of environmental regulations while their local competitors may not. Indeed, a recent study of the Canadian automotive parts industry indicates that environmental regulatory compliance was almost non-existent as a factor governing corporate strategy and production location (Eden, Husbands, and Molot, 1997). Generating intensive intrafirm trade and rapid technology shifts, Canadian investment has partly matched the high-engineering industries that were the focus of NAFTA’s investment provisions. It is here that investment brings modern technology, as Mexican operations are co-ordinated with those in Canada and the United States. These flows should increase as facilities gear up for export. Canadian investment is based (through both direct transfer and modern capitalisation) on modern technology in banking, auto parts, and telecommunications. High technology firms such as IBM Canada also value NAFTA’s role in protecting such technology through its provisions for intellectual property (patents, copyrights, trade secrets, and trademarks) (Ramirez de la O, 1996). However, such protections may, of course, inhibit the process of technological diffusion. Canada’s FDI in the US As Canadian outward FDI to Mexico surged, the pace of Canada’s growing FDI in the United States slackened. During the decade prior to NAFTA, the stock of Canadian FDI in the US had increased dramatically, from C$ 42 billion in 1986 to C$ 61 billion in 1992 (Table 1). During this time Canadian FDI in the US, as a ratio of US FDI in Canada had grown from 61 per cent to a peak of 69 per cent (Rugman, Kirton and Soloway, 1997). While Canadian FDI in the United States continued to increase to C$ 76 billion in 1995, the ratio declined to 68 per cent in that year. By another calculation, in 1996 the stock of Canadian FDI in the United States rose 7.5 per cent to reach C$ 93 billion (DFAIT, 1997b). This represents an historically low 54 per cent of Canadian global outward investment, a proportion which has been stable for the past four years. Canadian FDI flows into the United States rose steadily from US$ 1.3 billion in 1992 to $7.1 billion in 1995, but then dropped to $5.7 billion in 1996. These figures strongly suggest that NAFTA did not provide an incentive for Canadians to increase their FDI in the United States at the expense of overseas locations. There appears to be no investment diversion effect.

187

Table 1. Stocks of FDI between Canada and the United States, 1986-1995 Canadian FDI in the U.S. (C$ million) (1)

U.S. FDI in Canada (C$ million) (2)

Net position (C$ million) (1) - (2)

(1)/(2) in percent

1986

42 027

69 241

-27 214

60.7

1987

46 091

74 194

-28 103

62.1

1988

48 809

76 345

-27 536

63.8

1989

52 615

80 877

-28 262

65.1

1990

55 475

84 353

-28 878

65.8

1991

58 398

86 996

-28 598

67.1

1992

61 471

89 115

-27 644

69.0

1993

61 828

91 620

-29 792

67.5

1994

68 835

101 514

-32 679

67.8

1995

76 505

113 092

-36 587

67.6

Average rate of increase (%)

6.66

5.45

Source: Data are from Canada’s International Investment Position 1995 , Statistics Canada, Catalogue 67-202, March 1996, Ottawa.

US FDI in Canada At the same time, the post-NAFTA period witnessed a sharp increase in US and overseas FDI in Canada. Most generally, NAFTA’s primary effect appears to have been attracting investment flows towards the north, as FDI from around the world flooded into Canada. The trend is all the more remarkable given the stable (if low) Canadian dollar, slow growth in Canada, an October 1995 referendum on Quebec sovereignty, and a government not notably committed to investment flows (unlike its predecessors) until June 1996. Total FDI in Canada grew by 8.7 per cent in 1994, 9.3 per cent in 1995 (despite the October sovereignty referendum in Quebec), and 7.4 per cent in 1996, to reach a total of C$ 180 billion by the end of 1996 (DFAIT, 1997b). On a global scale, in 1996 Canada was the world’s third largest recipient of FDI by foreign MNEs, who provided C$ 12 billion of investment, in significant measure through reinvested profits. This strong increase of investment in Canada contributed to an increase in Canadian competitiveness. From 1990 to 1996, FDI in machinery and transport equipment rose 50 per cent. A Canadian government analysis concluded that this FDI increase had ‘an important effect on the renewal of plants and equipment” and “contributed to the higher productivity observed during this period” (DFAIT, 1997b). It was the United States, among major investors, that led this investment inflow into Canada. The stock of US FDI in Canada in 1996 rose for the fourth year in a row. The 1996 gain was a vibrant 9.1 per cent, compared to an overall figure of 7.4 per cent. The total stock of US FDI in Canada thus rose to C$ 122.7 billion, to account for 68 per cent of all FDI in Canada (DFAIT, 1997b). Strikingly, this 1996 figure for the growth of US FDI in Canada is substantially higher than the increase (7.5 per cent) of Canadian FDI in the United States that year. It may in part reflect a perception by US 188

investors of the prospects of a strongly growing post-referendum Canada, in contrast to those of a Mexico still recovering from the peso crisis of the previous year. On a sectoral level, the global investment gains in Canada during the post-NAFTA period were concentrated in financial services, transportation equipment, automobile equipment, chemicals, energy, communications, and food and beverages (DFAIT, 1997b). There is indirect evidence that NAFTA spurred productivity in Canada, through increased competition, lower-priced inputs, rationalisation, and specific investment guarantees. Mexican FDI in Canada NAFTA brought not only increased American investment to Canada, but Mexican investment, which had steadily decreased from 1988 to 1991, made a dramatic reversal. Mexican investment in Canada rose to C$ 60 million in 1992, and then more than doubled to reach C$ 154 million in 1993. It remained at this new level through 1994 and through the peso crisis year of 1995, then jumped again to C$ 239 million in 1996. This jump appears to foreshadow a sustained surge led by large Mexican firms. For example, in late 1997, Grupo Acerero del Norte SA announced the purchase of Nova Scotian-owned Sidney Steel for US$ 26 million, and its intention to inject a further US$ 100 million, along with advanced European technology into the plant (Gibbens, 1997). US-Mexican Investment Flows The policy concern about Canada receiving its “fair share” of US and overseas investment (against the new pull from Mexico), and the industry dynamic of region-wide rationalisation of production require a brief examination of Canadian investment in the context of the US-Mexican investment relationship under NAFTA. In the first year after NAFTA, the United States did indeed give a growing share of its outward FDI to Mexico and a diminishing share to Canada. Canada sharply increased its FDI stock in Mexico while holding steady in the United States. Mexican FDI in the United States soared, while remaining limited in Canada. However, this pull to the south soon ended. After the initial NAFTA adjustment, Mexico’s share of US FDI outflows, which had risen from 3.3 per cent (US$ 2.5 billion) in 1993 to 5.3 per cent ($3.7 billion) in 1994, dropped during the peso crisis year of 1995 to 3.4 per cent ($3.0 billion) and again to 3.1 per cent ($2.7 billion) in 1996. The stock of Mexican FDI in the US, which had risen from $1.0 billion in 1993 to $2.3 billion in 1994, fell to $2.0 billion in 1995, and again to $1.0 billion in 1996. On a sectoral level, NAFTA encouraged co-ordinated production in the automotive, telecommunications equipment, computer, electronic products, and textiles-apparel sectors, thereby increasing the exports of high value US component and services to firms in Mexico, and allowing Mexican firms to displace extra-NAFTA competitors. From 1993 to 1996, US FDI flows in Mexico were low in the automotive sector (with the Big Three investing $3 billion), steady in computers, household appliances and textiles-apparel, and negative in chemicals (where total US investment declined 47 per cent over the period) and printed products. In processed food and beverages, where more than 25 per cent of total US FDI in Mexico is concentrated, US FDI stock rose from $2.3 billion in 1993 to $2.8 billion in 1994 but declined to $2.3 billion in 1995. In contrast, Mexican FDI in the United States consists of a few strategic investments by large Mexican companies in cement (Cemex), glass, oil, refining, banking, and distribution (Ramirez de la O, 1996).

189

This pattern suggests that the technology transfer and host economy benefits flowing from NAFTA were experienced largely in Mexico in the initial year of adjustment to NAFTA. In key sectors such as automobiles, anticipation of NAFTA led to many US-Mexican joint ventures that brought Mexico technology, and modernised the sector as a whole (Studer, 1994). More recently, the United States has been the primary beneficiary, with investment from NAFTA and non-NAFTA countries concentrating in the centrally-located United States. Although Canada initially may have suffered from a diversion of US investment to Mexico, Canada’s intensified integration into a regional production system has brought new activity in sectors such as beef packing. Direct evidence of NAFTA-induced technology transfer through US MNEs in Mexico comes from the area of the environment. A 1995 survey by the American Chamber of Commerce in Mexico found that 85 per cent of the 125 companies (virtually all American-owned MNEs) increased their spending for pollution controls and environmental protection. Almost 60 per cent declared plans for further such expenditures, especially in water and air systems. Much of such technology and equipment is imported from the United States and up to the same state of the art standard used in the US parent. Interviews conducted in 1995 with selected firms in Mexico suggested that firms owned by other countries and Mexican-owned firms were also increasing their environmental expenditure through similar processes (Ramirez de la O, 1996). Consistent with this pattern is the sectoral flow of new US FDI into Mexico in the post-NAFTA period. It appears to be avoiding or declining in those sectors with a large ecological footprint (basic metals, industrial chemicals, and non-metal products) and concentrating in those with a smaller ecological footprint (textiles, metal products, food products). Such a pattern strongly suggests that US industry is not moving its heavily polluting production to Mexico. This is consistent with the pattern of initial post-NAFTA investment in Mexico (Ramirez de la O, 1996). With US FDI flows to Mexico in 1996 representing only 0.2 per cent of US gross private domestic fixed investment that year, there has been no general migration of US firms from “environmental sanctuaries” in the United States to alleged “pollution havens” in Mexico (USTR, 1997). Moreover, the most recent and well developed of several studies demonstrates a trend of US FDI into Mexico in industries characterised as lower polluting (Cole and Ensign, 1997). 4.

The Key Challenges for Greater Regional Integration

From a Canadian perspective, the experience of NAFTA to date has led to a broad judgement that NAFTA has been a striking success. Public opinion polls in Canada on the agreement show that the initial levels of opposition, in the majority, have steadily declined, to the point where NAFTA commands very high and broad levels of acceptance (McKenna, 1998). Canadians, based on their NAFTA experience, are also are strongly favourable to broader trade liberalisation, both within the hemisphere and across the Asian-Pacific region (Ready and Sandver, 1995). In order to secure the benefits of such expanded liberalisation, and to encourage the United States to fulfill its commitment, initial consultations began at the December 1994 Miami “Summit of the Americas” about how to expand NAFTA to include Chile. Canada concluded a NAFTA-compatible bilateral free trade agreement with Chile in 1996. As Canada rivals the United States as the largest direct foreign investor in Chile, the agreement’s value lies importantly in the extended guarantees for investment which it offers. Indeed, Canada preserved the right of a free flow of FDI by exempting FDI from the Chilean government’s demand that it have the right to prevent the repatriation of portfolio investment for a period after its initial entry into Chile.

190

Canada has also long favoured a deepening of NAFTA. It was a founder of the North American Financial Group, created at the ministerial level in the spring of 1994 to deal with exchange rate and other financial issues among the NAFTA members. This body has continued to meet, usually on the margins of broader multilateral gatherings. Canada also contributed C$ 1.5 billion to the support fund organised by the United States in the wake of the severe devaluation of the Mexican peso following 20 December 1994. The absence of sustained US and Canadian FDI inflows into Mexico during and after the economic crisis of 1995, when Mexican assets were especially attractively priced and available, points to the substantial costs of such financial crises, and the consequent need to strengthen the regime of the North American Financial Group. It has similarly encouraged the formation of other trilateral ministerial bodies in areas such as health, transportation, and (more embryonically) agriculture. Looking ahead, the intensification of region-wide integrated and rationalised trade suggests the need for improved standards harmonisation, both for products and for production processes. Although the diffusion of practices through affiliates or alliances, or through voluntary, industry-driven harmonisation may suffice in some cases, the need to guard against defection, and to build capacity in Mexico, suggests the need for a strong role for national governments and NAFTA institutions in this regard. There is a similar need to improve NAFTA’s capacity to guarantee strict and uniform enforcement of national and sub-federal regulations, both in the environmental and labour areas where the institutional foundation already exists, and in other regulatory areas where it does not. There is further need to expand the definition and ease the implementation of NAFTA’s provisions for the mobility of professional and business personnel. The free flow of such personnel is critical to a relationship in which trade is increasing in the services sector, and where such traded services are increasingly vital to maintain the production of foreign-owned subsidiaries that form part of a regional production network. They are also an essential component of the process of technology transfer and diffusion. The need to ensure and expand such mobility, and foster high level standards convergence underscores the immediate need to create a stronger NAFTA Free Trade Commission, with a new deputies institution and a single Secretariat with policy analysis capabilities. This is especially so as the process of NAFTA expansion into the hemisphere begins in earnest. Looking further ahead, it is likely that Canada’s strong support for deepening and broadening NAFTA will endure, in large part because NAFTA and the prospective FTAA represent a welcome diversification of Canada’s CUFTA, which generated deepening integration with the United States alone. This diversification instinct, and Canada’s continuing desire to protect key sectors of its own economy, such as culture and banking, from heavy foreign ownership (an approach that resonates with many other countries in the Americas) means Canada is well positioned to pioneer initiatives to both deepen and broaden the NAFTA regime. Some initiatives could be practical measures to lower transaction costs, such as the use of “smart cards” to speed the border transit of personnel admitted under NAFTA’s temporary entry provisions. Another would be strengthened provisions for government procurement, to ensure that Canadian companies in Mexico such as Bombardier, and indeed any company considering investing anywhere across national borders within North America, is not deterred by the suspicion that a perceived lack of full transparency can sometimes breed. Moreover, while Canadians generally welcome NAFTA’s rules of origin in the automotive and other sectors as a considerable advance over those prevailing under CUFTA, as experience accumulates and production techniques change, it is worth a review to ensure that quantitative levels and regulations fixed in 1994 do not exert an undue cumulative trade diversion effect in practice. More ambitiously, given the trend toward integrated transborder production systems, it is increasingly desirable for all NAFTA parties that NAFTA move more adventurously into the realm of competition policy, and 191

replace the time-consuming, expensive barrier of antidumping and countervailing procedures with a new regime based on region-wide rules and mechanisms against predatory pricing. Finally, one should highlight the importance of a well functioning and steadily improving NAFTA investment regime not only for the North American region and western hemisphere, but also as a stimulus and model for other regions and for the quest for multilateral investment regime building as well. The value of NAFTA’s pioneering provisions for investment dispute settlement under Chapter 11 and for protecting against pollution-haven-seeking investment relocations under Article 1114 acquire greater significance as resistance grows both to the rapid completion of a high standard Multilateral Investment Agreement under OECD and later WTO auspices, and to the ability of the WTO to incorporate improved investment provisions in a new round of multilateral trade liberalisation.

NOTE

1.

Unless otherwise indicated, the judgements in this paper are based on a programme of 59 semi-structured interviews conducted in the autumn of 1995 and the spring of 1996 with individuals, evenly balanced among the three countries, from business (17), Government (27), NGO’s (11) and the academic community (3). I am grateful to Denis Audet and Philip Hemmings for their comments on an earlier version of this report.

192

BIBLIOGRAPHY

BLANK, S. and KRAJEWSKI, S. (1995), “U.S. Firms in North America: Redefining Structure and Strategy”, North American Outlook, No. 5, February, pp. 5-57. COLE, E. and ENSIGN, P. (1997), “An Examination of the United States Foreign Direct Investment into Mexico and Its Relation to the North American Free Trade Agreement: Towards a Balanced Understanding of the Effects of Environmental Regulation and the Factor Endowments that Affect the Location Decision”, Paper presented at the Annual Meeting of the Academy of International Business, Monterrey, Mexico, 8-12 October. DEPARTMENT OF FOREIGN AFFAIRS AND INTERNATIONAL TRADE -- DFAIT (1997a), Asia Pacific Investment Strategy, , Ottawa. DEPARTMENT OF FOREIGN AFFAIRS AND INTERNATIONAL TRADE -- DFAIT (1997b), NAFTA: A Partnership at Work, , Ottawa. EDEN, L., HUSBANDS, K. and MOLOT, M.A. (1997), “Shocks and Responses: Canadian Auto Parts Suppliers Adjust to Free Trade and Lean Production”, Paper presented to the Annual Meeting of the Academy of International Business, Monterrey, Mexico, 8 -12 October. GASTON, N. and TREFLER, D. (1997), “The Labour Market Consequences of the Canada-US Free Trade Agreement”, Canadian Journal of Economics, No. 30, February 1997, pp. 18-41. GESTRIN, M. and RUGMAN, A. (1994), “The North American Free Trade Agreement and Foreign Direct Investment”, Transnational Corporations, No. 3, February, pp. 77-96. GIBBENS, R. (1997), “Mexican Firm Bids for Sysco”, Financial Post, 30 December, p. 4. HIRSHHORN, R. (1997), Industry Canada’s Foreign Investment Research: Messages and Policy Implications, Discussion Paper, No. 5, Industry Canada, Ottawa, October. HOWSE, R and SOLOWAY, J (1997), “Rethinking Foreign Ownership Restrictions as a Tool of Canadian Public Policy”, Working Paper, Centre for the Study of State and Market, Toronto, Forthcoming. INDUSTRY CANADA (1997), “Canada, Investment Rules and the World Economy”, Foreign Direct Investment: Canadian Trends, Micro-Economic Policy Analysis Branch, Ottawa, 4 November. KIRTON, J. and FERNANDEZ DE CASTRO, R. (1997), NAFTA’s Institutions: The Environmental Potential and Performance of the NAFTA Free Trade Commission and Related Bodies, Commission for Environmental Co-operation, Montreal.

193

KIRTON, J. et al. (1996), Building a Framework for Assessing NAFTA Environmental Effects, Commission for Environmental Co-operation, Montreal. KPMG (1997), The Competitive Alternative: A Comparison of Business Costs in Canada, Europe and the United States, Canada, October. McKENNA, B. (1998), “Canada, US Poles Apart on Free Trade”, Globe and Mail, 3 January, p. B1. RAMIREZ DE LA O, R. (1996), “North American Investment Under NAFTA”, Working Paper, No. 3, NAFTA Effects Working Paper Series, Commission for Environmental Co-operation, Montreal, April. READY, K. and SANDVER, M. (1995), “Policy Issues in Moving from NAFTA to the FTAA”, North American Outlook, No. 5, Winter, pp. 24-55. RUGMAN, A. (1994) (ed.), Foreign Investment and NAFTA, University of South Carolina Press, Columbia, S.C. RUGMAN, A., KIRTON, J. and SOLOWAY, J. (1997), “Canadian Corporate Strategy in a North American Region”, American Review of Canadian Studies, Summer, No. 27, pp. 199-218. STUDER (1994), “The Impact of NAFTA on the Mexican Auto Industry”, North American Outlook, No. 5, November, pp. 20-55. USTR (1997), Study on the Operation and Effects of the North American Free Trade Agreement, United States Trade Representative, Washington, D.C., July. WEINTRAUB, S. (1997), NAFTA at Three: A Progress Report, The Center for Strategic and International Studies, Washington, D.C.

194

NAFTA, FOREIGN DIRECT INVESTMENT AND ECONOMIC INTEGRATION: A MEXICAN APPROACH

by * Fernando de Mateo Venturini

1.

Introduction

Experts agree that trade and foreign direct investment are complementary. However, they also agree this has not necessarily always been the case. In fact, in the past firms invested abroad as a substitute for exports when faced with import restrictions in the host country,1 or did so as a means to access natural resources unavailable in the home country. Many developing countries used import restrictions as the cornerstone of their import-substitution development strategies. One of the by-products was the presence of foreign companies willing to take advantage of the domestic market of these countries. But foreign direct investment (FDI) also took place in many developed countries for the same reasons, either while there were high trade restrictions (e.g. Europe after World War II), or when it was perceived that markets would get closed (e.g. the perceived threat of Fortress Europe during the 1980s, or the threat perceived by Japanese companies with respect to the imposition of voluntary export restraints in developed countries, particularly in the automotive and electronic industries). Import substitution strategies in developing countries usually meant that as soon as the domestic production of a product started, or was about to start, protection by way of high import tariffs and/or import permits was introduced for the product in question. Foreign firms then faced the alternative of losing a market they had been providing from abroad or taking the initiative of establishing an affiliate before anyone else did in a market often too small to shelter more than a couple of producers. Import substitution strategies were usually carried out in different stages, starting with consumer goods, following with simple intermediate goods, and (when the domestic market was large enough) ending with capital and sophisticated intermediate goods. Traditionally, import substitution strategies have been analysed from the exclusive stand-point of merchandise. The CEPAL (Economic Commission for Latin America and the Caribbean) theoretical model developed in the late forties and early fifties was devoted to the goods sector. In practice, however, import substitution also covered services and, in fact, foreign direct investment. Given that *

Director General of International Negotiations in Services and Europe. Under-Secretariat of International Trade Negotiations. Secretariat of Trade and Industrial Development, Mexico. The views expressed in this paper are the author’s, and do not necessarily correspond with those of the institution for which he works. I want to thank Pierre Sauvé, Carlos García Fernandez, Patrick Low and Isabel de Mateo for their comments on an earlier version of this paper. I’d also like to thank Ricardo Lozada for his statistical help.

195

the most important mode of delivering services internationally is through establishment, import substitution in services implied substituting foreign for domestic ownership of the service producing firms. The same logic was also applied, albeit to a lesser extent, in the realm of merchandise, particularly in the non-renewable natural resources sector. Import substitution strategies in both goods and services could function as long as technologies were relatively stable and the product cycle was a long one. However, such strategies faltered when the scientific and technological revolution exploded in the mid-1970s, triggering a world-wide economic structural change. Rapidly changing soft technologies replaced hard technologies as the driving force of industrial development, shortening the product cycle. Financial markets around the world grew at 2 an incredible speed as a result of both deregulation and new technologies. A services revolution was also set in motion, causing this sector, and producer services in particular, to become information-, knowledge- and organisation-intensive, as well as a fundamental input in the production process of 3 manufactured goods and all other tradables. Transnational corporations (TNCs) and producers of goods and services changed their strategies and behaviour accordingly, particularly through a global approach to cost reduction and economies of scale and scope. On the one hand, their integration strategies became increasingly complex, with the “decomposition of the production process into discrete inputs and functions that can be located 4 anywhere in the world”, thus with TNCs choosing what they believed to be the most efficient locations for the fulfilment of a specific function. On the other hand, service providers followed their clients abroad at a higher rate. Therefore, though globalisation is a product of liberalisation, it also 5 put in motion forces that have contributed to accelerating the latter. A number of developing countries became aware that the continuation of inward-looking development strategies would be tantamount to becoming permanently marginalised from world economic development. The question was not whether they should insert themselves into the globalisation process, but how deeply they should do it and by what means. The bottom-line was how “contestable” their markets should be allowed to become for both foreign and domestic economic 6 agents. In this respect, without liberalised trade, foreign-owned companies would not be in a position to import the necessary inputs to pursue global strategies. Likewise, the mere liberalisation of trade without a corresponding FDI liberalisation would mean that an important part of today’s world trade, 7 i.e. intrafirm trade, would be left out. Moreover, free trade in goods (or its substantial liberalisation) without corresponding changes for services implies negative rates of effective protection to the former.8 Bearing in mind that establishment is the main mode of delivery of services, it is also true that free trade in goods (or its substantial liberalisation) not accompanied by FDI liberalisation also implies negative rates of effective protection. Therefore, the efficient insertion of a country in the globalisation process implies the liberalisation of trade in goods, of trade in services, and of FDI. There is a long history of the liberalisation of trade in goods, both at the multilateral and regional levels. The same is not the case for trade in services, although during the Uruguay Round the GATS 9 was established and a number of sectors were liberalised by participant countries, and certain 10 regional deep integration agreements have been negotiated over the past few years. In investment -- notwithstanding the investment rules proposed under the Havana Charter and the OECD codes on capital movements and invisible transactions established in the early 1960s11 -- the experience is even

196

more restricted, for it is concentrated in regional integration agreements, bilateral investment 12 13 treaties, and the Trade Related Investment Measures Agreement of the Uruguay Round (TRIMs). There seems to be agreement among experts on a number of hard-core conditions to make markets 14 more fully “contestable” as far as FDI is concerned. The fulfilment of such conditions will promote a more efficient use of resources and, therefore, foster welfare and economic growth. They include the following: full national treatment (including full market access to, or right of establishment in, the host country); the right not to establish (i.e. the supply of a good or a service should not be subjected to the obligation of establishment in the host country); the elimination of performance requirements (i.e. approval of FDI conditional on the attainment of certain objectives, such as a given level of employment or exports); the granting of MFN status (i.e. no discrimination between countries); the elimination of subsidies to FDI; and finally, the protection of investors and investments through a workable dispute settlement mechanism. However, to attain fully contestable markets, free trade in goods and services and free FDI have to be complemented by other important elements, such as the liberalisation of government procurement, full protection to intellectual property rights, and neutral competition policies (i.e. not geared to interfere in the granting of full national treatment). 2.

FDI import substitution in Mexico and first steps towards liberalisation

The import substitution process of goods in Mexico really started during World War II and for all practical purposes concluded in the mid-1970s. The process started with non-durable consumer goods. In the 1950s it shifted towards the production of some durables and non-sophisticated intermediate goods. The 1960s witnessed the import-substitution of durables such as automobiles and complex household appliances (e.g. televisions), as well as increasingly sophisticated intermediate goods. Simple capital goods were also produced. In the early 1970s import substitution of capital and increasingly complex intermediate goods were promoted. By the mid-1970s the import substitution development strategy was clearly faltering, so a leading-sector (oil) development strategy was adopted, based on high international oil prices. This strategy was, in turn, replaced by a more outward-oriented one as a result of the crisis that began in 1982. 15

FDI “import substitution” clearly began early this century. In 1908, 58 per cent of the railroad 16 system was nationalised. After the 1910-21 Revolution, a number of FDI-related measures were taken. Telegraphic and radiotelegraphic services were de jure nationalised. In 1925 the Petroleum Law was approved, regulating an industry with an important FDI component. This law was the juridical support for the 1938 expropriation of the industry -- this is still one of the most important symbols of Mexico’s national sovereignty. In 1934 a new law was passed requiring that banks invest their holdings in Mexican securities. In protest, all foreign banks, except for one, left the country. Similar requirements were imposed on insurance companies, producing similar results. No new foreign affiliates of foreign banks or insurance companies were allowed in Mexico during the next six decades. In 1936 foreign participation was barred from all broadcasting activities (extended in the 1960s to cable television). The nationalisation of the railroad was almost complete by 1937. Since the mid-1940s the construction sector has been reserved to Mexican nationals. FDI in road transportation (both cargo and passengers) has been prohibited since the early 1930s. In 1947 foreign participation in the rubber industry and in some other industries of lesser importance was limited to 49 per cent. As a general rule, FDI could acquire no more than 49 per cent of existing companies (and, of course, only in those sectors where FDI was allowed). In 1958 the monopolistic 197

telephone company was “mexicanised”. The similarly monopolistic electric power company was also nationalised. In 1962 foreign participation in Mexican maritime transportation companies was prohibited. In 1970 the participation of FDI in the steel, cement, glass, fertilisers, cellulose and aluminium industries was limited to 49 per cent, and in the automotive parts industry it was restricted to 40 per cent. In mining, foreign participation was to 34 to 49 per cent. The height of “import substitution” in FDI was reached in the 1973 Law to Promote Mexican Investment and Regulate Foreign Investment (FIL). The FIL codified all the economic activities reserved to the State, as well as those where FDI was excluded. It established a 49 per cent limit on FDI in all industries not specifically subject to lower participation rates. It did, however, provide for the possibility of a higher percentage of foreign ownership in an undertaking if it was considered to be in the public interest, but then subject to the fulfilment of a number of performance requirements. Notwithstanding these unfavourable conditions, the stock of FDI continued to grow during the 1973-76 period, albeit at a slow rate. The momentum of FDI “import-substitution” was considerably reduced during the leading-sector development strategy, though no major effort was undertaken to liberalise it. While a high-growth, oil-rich economy attracted increasing amounts of FDI -- at an annual rate averaging nearly 20 per cent -- FDI flows suffered a significant reduction with the collapse of international oil prices and of the development strategy predicated on them (Table 1). After the 1981-82 external shock and the ensuing debt crisis, the approach changed from regulatory and restrictive to supportive of FDI, and a number of important measures in the application of the FIL 17 were introduced. Perhaps more important from the foreign investors’ perspective was the Mexican government’s commitment to trade liberalisation. This commitment began in 1983, and in 1986 Mexico acceded to the GATT. By 1987 the government had on its own undertaken a substantial reduction of tariffs and an elimination of most quantitative restrictions. The weighted-by-trade average tariff was cut by more than half (or to 10 per cent) and prior import permits (covering 100 per cent of imports in 1982) were eliminated for all but a handful of products. The result of these reforms was a substantial increase in the growth rate of FDI flows into the country, outstanding even by international standards (Tables 1 and 2). In 1989 new regulations of FDI were introduced along with other structural reforms, including the public external debt re-negotiation. Though FDI liberalisation was important, it did not touch a long list of activities restricting FDI participation. 3.

The rules for FDI liberalisation -- the maquiladora export industry and NAFTA

The maquiladora export industry The Mexican government first realised the close relationship between free trade and free foreign 18 In the mid-1940s the US government established the Bracero investment in the mid 1960s. Programme, allowing for the temporary entry of Mexican farm workers into the US during the crop season. When it was terminated in 1964, the Maquiladora Programme was established in order to counter the loss of employment on the Mexican side of the border. The first maquiladora plants were devoted to performing very simple assembly and/or packaging operations, but as time went on they became the site of labour-intensive activity requiring skilled 198

workers. The work involved increasingly sophisticated, high-tech production processes. At the present time, there is evidence of the “widespread use of modern managerial techniques” in these plants. “Just-in-time inventories, statistical product control, zero-defect techniques and work teams 19 are now common practices.” While in 1987 there were only eight export manufacturing categories (textiles, furniture, electrical equipment, shoes, transportation equipment, toys and sporting goods, food products, and services), in 1996 these also included sectors such as chemicals, machinery, 20 electronics, telecommunications equipment, and automotive goods. Since its inception, this industry has provided for free trade of both inputs and final products, and for FDI without restrictions. It fulfils almost all the hard core conditions experts consider characteristic of free FDI that were mentioned earlier: national treatment (including right of establishment), MFN status, no obligation of local establishment for the supply of a good or a service to the maquiladora companies, no performance requirements attached to the establishment of foreign companies, transparency in the regulations, and others. The only important condition not met is the presence of a 21 state-to-state and/or investor-to-state international dispute settlement mechanism. Both as value-added generator and as job provider, the maquiladora industry has been a great success. In the 1980-96 period, value added in the maquiladora industry increased by 13.8 per cent annually. Likewise, between 1980 and April 1997, employment grew at an annual average rate of 12.2 per cent (Table 3), with chemical products (32.4 per cent per year), and transportation equipment and accessories (20.1 per cent per year) showing the highest rates.22 In 1996 the maquiladoras exported 37 billion dollars, representing 40 per cent of Mexico’s total exports and 47 per cent of its total manufacturing exports (Table 4). In 1997 about half of the maquiladora plants were owned in full or in part by US investors, 44 per cent were Mexican, 4 per cent were Asian, 1 per cent European, and 1 per cent Latin American. The US market is by far the most important destination for maquiladora production. Though 80 per cent of the value added in the maquiladora industry is generated in the northern border states of the country, more than 50 per cent of new maquiladora enterprises (those established since 1994) are located outside the border region, as far as the Yucatán Peninsula and the southern state of Oaxaca. NAFTA The GATT, even in the then on-going Uruguay Round, did not fulfil the conditions necessary for Mexico to fully participate in the globalisation process. There were three main reasons for this: first, after the far-reaching unilateral trade liberalisation of 1987 the country’s different producing sectors were not ready to further open the market without full reciprocity, particularly, but not exclusively, on the part of Mexico’s main trading partners. This reciprocity was not obtained during the Uruguay Round. Second, the Uruguay Round coverage was rather limited, particularly in areas other than merchandise trade liberalisation, whereas real-life or de facto integration with the United States was already extensive in that and other areas. Finally, though it was subject to negotiation in the Uruguay Round, the GATT’s dispute settlement mechanism was not very reliable, for it was not binding for the offending party. For Mexico to further open its economy, a workable dispute settlement mechanism was needed by Mexican exporters. Therefore, it seemed preferable to adopt a regional liberalisation process encompassing Mexico’s main trading partners in the form of a GATT-compatible free trade agreement.

199

The maquiladora industry’s long history demonstrates two things with respect to NAFTA. On the one hand, the maquiladora export industry strongly contributed to the so-called silent economic 23 integration of Mexico and the United States in the 1980s. On the other, it showed that NAFTA 24 should be a deep integration agreement rather than a traditional, merchandises-only free trade area. As a deep integration scheme, NAFTA provides for free trade in all goods for a period no longer than 25 26 ten years after its implementation, state-of-the-art provisions on intellectual property rights, and the 27 liberalisation of government procurement. NAFTA also provides for a liberalised regime that 28 facilitates the temporary entry of business people, and for the liberalisation of cross-border trade in 29 services. It further includes a workable dispute settlement mechanism. As far as investment is concerned, the negotiations resulted in a liberal regime in all three countries. The Agreement includes: − A definition of investment covering virtually every investment, not only direct investment (i.e. it covers all equity, debt security, business loans, real estate or other property investment, and other commitments of capital -- including, for instance, turnkey contracts). It is the broadest definition existing so far under any bilateral, regional, or multilateral agreement. − A national treatment concept (Articles 1102, 1202 and 1405), which includes market access (i.e. establishment), a concept treated separately in other international 30 agreements, such as the GATS, the OECD Code of Liberalisation of Capital movements, and the draft APEC investment code. − A right not to establish (Article 1205), as a complement to national treatment, which does not require the establishment of the provider when services may be supplied crossborder. − Minimum standard of treatment (Article 1105), which also goes beyond national treatment, obliging the parties to maintain minimum internationally accepted standards for the treatment of investors. − MFN status (Articles 1103, 1203 and 1406), supplemented by the obligation to provide the better of MFN or national treatments (Articles 1104 and 1204). − Performance requirements (Article 1106) that go well beyond those achieved in the TRIMS Agreement of the Uruguay Round. They are applicable to both goods and services, and include prohibitions on export requirements; domestic content requirements; the restriction of sales in the home country by relating them to exports, technology or proprietary knowledge transfer; and the obligation of market segmentation. It also prohibits performance requirements linked with the conferral of advantages, relating allowable imports to exports performance, the achievement of a given amount of domestic content, and relating exports to domestic sales. These prohibitions apply to all investments, not only to those from NAFTA investors. − The standstill and rollback procedure, or the “top-down” or “negative list” approach, that requires all remaining measures inconsistent with NAFTA be inscribed as reservations to the application of the relevant principles. Both the rollback or standstill of a measure decided during the negotiations was complemented by the “ratchet clause” -- if for whatever reason an investment measure is further liberalised, the new level of liberalisation is bound under NAFTA. 31

200

− A provision for payments and transfers to be freely made, without delay, and in a freely usable currency. The only exceptions to this refer to balance of payments difficulties and to the good faith application of domestic laws in cases relating to bankruptcy, criminal offences, and similar situations. − A dispute settlement mechanism that goes beyond that of any other deep integration agreement. It contains provisions related to both state-to-state and investor-to-state disputes. In cases where negotiations fail to resolve a dispute, an investor may file a complaint against the government of the host country in its domestic courts or seek mandatory arbitration of the dispute according to established international procedures, like the World Bank’s International Centre for the Settlement of Investment Disputes or the United Nations Commission on International Trade Law. Therefore, NAFTA covers two fundamental conditions for being a state-of-the-art regional agreement: taken together, its provisions make it a deep integration scheme, and, with respect to foreign direct investment, it covers practically all the hard-core conditions that experts consider necessary in an international agreement of this nature. This is why NAFTA has been extensively used as a reference during the negotiations of a Multilateral Investment Agreement in the OECD. 32

According to some authors, NAFTA’s main shortcomings are a weak chapter on competition -- not even subjected to the dispute settlement mechanism -- and the exclusion of national treatment on 33 subsidies and “incentives”. Moreover, strict rules of origin either give rise to trade and investment 34 diversion or reduce competition. To the above shortcomings it may be added that the elimination of quantitative restrictions in services is not a principle, but a negotiating objective; these restrictions may also affect investment decisions. There are a number of reservations (exceptions) with respect to 35 NAFTA principles for both investment and services in important (albeit, few) sectors. 4.

FDI liberalisation -- NAFTA and beyond

The FDI liberalisation introduced by NAFTA and the almost equally important liberalisation further undertaken by the Mexican government (remember the NAFTA “ratchet” clause) has resulted in a very liberal FDI regime in Mexico.36 The following section offers a brief description of this process in the NAFTA countries. However, it must be borne in mind that, with a few exceptions, such a 37 liberalised regime is applied erga omnes to the rest of the world. Subject until the year 2000 to specific market quotas for different financial services, FDI was liberalised with respect to Canadian and US suppliers, including extra-regional providers with affiliates in North America. The insurance industry was also fully liberalised for green-field investment, though subjected as well to temporary market quotas. However, for mergers and acquisitions, FDI is limited to 49 per cent. There is a cap of 10 per cent for foreign investment in co-operative production societies. The cap is set at 25 per cent for air transportation companies, and at 49 per cent for the production and commerce of explosives and firearms, newspapers, agriculture, fishing, ports, cabotage maritime transportation, and supply of fuel to maritime vessels, aircraft, and railroad locomotives and equipment. With prior authorisation of the National Commission on Foreign Investment (NCFI), 100 per cent FDI is now allowed for high-seas maritime transportation. FDI in international cargo road transportation is being liberalised according to a time-table agreed upon during the NAFTA

201

38 negotiations. The same liberalisation schedule applies to passenger services. The railroad system is being privatised, allowing 100 per cent FDI, upon prior approval of the NCFI.

There are no restrictions for value-added telecommunication services,39 nor in the construction sector, subject to the prior approval of the NCFI. Foreign direct investment of up to 100 per cent is also permitted in the automotive parts industry as well as in mining. No restriction is imposed on FDI in 40 professional services. The Mexican satellite telecommunications system is being privatised, with a 49 per cent FDI cap. This cap is also applicable to most of the other basic telecommunication services, except for resale (100 per cent FDI) and cellular telephony (100 per cent FDI with prior approval of the NCFI). Cable television and DTH are considered basic telecommunications services, and are thus subject to a 49 per cent restriction on FDI. In 1991 the telephone monopoly was privatised, allowing up to 49 per cent FDI. It remained a monopoly until January 1997, when full competition started. Today, more than 30 foreign companies are providing basic (including non-telephone) telecommunication services in the country. Mexico made a head-start in the liberalisation of this sector compared to most developed countries, including its NAFTA partners. Basic telecommunications were not negotiated under NAFTA, but Mexico bound its liberal regime in the WTO negotiations for the sector, concluded in February 1997. To sum up, FDI is restricted in only a handful of activities. Some sectors are reserved to the state by the Constitution. Among them, energy is the most conspicuous. However, as already mentioned, it is one of the symbols of Mexican sovereignty, ranking as high or even higher than water resources or cultural industries for the Canadians, or maritime transportation for the US. For historical reasons that have little economic importance today, telegraphy and radiotelegraphy are also state monopolies. Mail too is a state-run activity; however, there is no FDI limitation on the courier and small packages industries. As in most countries in the world, central banking services are reserved to the state. The number of activities reserved to Mexican nationals is also very small. The most important among them is broadcasting (excluding cable television and DTH services, where 49 per cent of FDI is allowed). For cultural and other reasons, FDI in this sector is restricted in many countries around the world, including Mexico’s NAFTA partners. All other activities reserved to Mexican nationals have, except for domestic cargo road transportation, very limited economic importance: gasoline retail services, distribution of liquefied gas, credit unions, and development banking. The recently enacted FIL also contains two important mechanisms that allow for increased FDI in activities where it is limited (except, of course, in the case of state-run companies). The first involves the issuing of non-voting shares. In the case of basic telecommunications where FDI is limited to 49 per cent, for instance, through non-voting shares foreigners may own as much as 80 per cent of the total capital of a company. The second mechanism involves the redefining of a Mexican-controlled company as 100 per cent Mexican when, through a joint-venture, a third company controlled by Mexicans is created. In the same way NAFTA and the ensuing unilateral measures taken by Mexico produced an open, liberal FDI regime, the liberalisation of cross-border trade in services has been equally important. The only restriction left was the requirement of establishment for the provision of a certain number of 41 services. However, at the moment most of these services are not technologically apt to be provided across the frontiers, though with future technological advances, the restriction may develop teeth. In

202

contrast with FDI measures, Mexican states have a limited number of measures that may restrict trade in services. The NAFTA “ratchet” clause also applies to subnational governments. All remaining activities, which account for roughly 80 per cent of GDP, are completely open to FDI, 42 and most of them without the need for prior approval by the NCFI. It should be recalled that in Mexico, FDI measures are of the exclusive prerogative of the Federal government. 5.

Recent behaviour of FDI in Mexico

Mexico is, after China, the most important recipient of FDI among the developing countries (Table 6). At the same time, Mexico’s own participation in in flows into Latin America represented about one-third of its total participation in in flows during the first half of the 1990s. Mexico’s FDI to developing countries represented about 9 per cent of its total FDI in the first half of the 1990s. (Table 5). 43 There was a substantial increase in FDI flows into Mexico during 1994, the year NAFTA was implemented (Table 1). The rate of growth of these flows slowed down in 1995 and 1996, but remained remarkably high considering that during that time Mexico faced its worst economic crisis since 1932, with GDP falling by 6.2 per cent, inflation soaring above 50 per cent, and unemployment at its highest levels in recent history.

Beside the fact that most investment taking place in 1995 was in the pipeline since 1994, the explanation lies to a great extent in the fact that most foreign companies, and an ever-increasing number of Mexican-owned companies, are now geared to the export market, or to both the domestic and foreign markets. More interesting, though, is the fact that they now possess a high capability to swiftly switch from one market to the other. From 1993 to 1996, the number of companies exporting from Mexico grew from 21 447 to 31 860; many of these are small and medium-sized firms. In fact, from the mid-1980s to 1996 total exports have grown at an annual average rate of 12.3 per cent. This rate accelerated to 17.6 per cent between 1991 and 1996, and to 22.8 per cent since NAFTA’s implementation. This export performance has converted Mexico into the tenth world trade power in just a few years. At the same time, while in the early 1980s around 80 per cent of Mexico’s exports were petroleum products, today 80 per cent are manufactured goods. Exports to the United States have steadily augmented since the mid-1980s, at an annual rate of 12.6 per cent. Between 1991 and 1996 the annual growth was of 19 per cent. In the latter year, exports were almost two-fold the figure registered when NAFTA entered into force, amounting to a yearly increase of 23.2 per cent. At an annual rate of 14 per cent between 1993 and 1996, exports to Canada have also grown rapidly. As far as total imports are concerned, between 1985 and 1996 they grew at a yearly average of 15.4 per cent, and at a slower pace from 1991 (12.3 per cent) and from 1993 (11 per cent). This, of course, is the result of the fall registered in 1995 as a consequence of the economic crisis. The figures with respect to the United States are not very different, though between 1993 and 1996 it fared better in the case of total imports (13.2 per cent). Imports from Canada increased at even faster rates: 21 per cent between 1991 and 1996 and 14.4 per cent from 1993. The insertion of Mexico in the globalisation process through foreign-controlled companies can be illustrated through the changes experienced in what the United Nations Conference on Trade and 203

Development (UNCTAD) calls the “export propensity” coefficient (exports as percent of total sales) of US majority-owned foreign affiliates in manufacturing (Table 7). Corresponding to a “pure” import-substitution development strategy, the figure for Mexico in the mid-1960s is half that for Latin America and almost one-third of the figure corresponding to all developing countries. By the midseventies and early eighties, the export propensity of US affiliates in Mexico was half that shown for the developing countries as a whole. From the mid-eighties until 1993, the coefficient was very similar to that of the developing countries, and substantially above the coefficient for Latin America. No doubt the coefficients shown in Table 7 have substantially increased during the last few years in Mexico, particularly during the 1995 crisis and in 1996. This trend can be seen in the case of automobiles manufactured by US affiliates in Mexico (Ford, Chrysler, and General Motors). During the whole import-substitution period export propensity was nil (Table 8). In fact, by the early eighties effective protection for the sector was negative and 44 45 smaller than 100 per cent, and with a strong anti-export bias. However, by the mid-1980s more than one-third of total production was exported, rapidly growing during the 1990s to 63.6 per cent by 1994.46 In 1995, domestic sales of the three US-owned companies floundered by 66 per cent. Notwithstanding the above, their total output -- 88 per cent of which was exported -- was reduced by 47 only 2.3 per cent. In 1996, 83 per cent of total production was sent abroad. The automobile industry is just an illuminating example of the rapid response of exports to the dramatic contraction in domestic demand in 1995. It is also an example of Mexico’s insertion in the globalisation process and of the role foreign-owned companies are playing in such insertion. The 33 per cent increase in total exports registered in 1995 and its very dynamic behaviour in 1996 and 1997 is in part the result of NAFTA’s tariff and non-tariff barriers elimination in Mexico’s trading partners, and of its transparent rules providing certainty to all economic agents. In fact, the existence of a workable dispute settlement mechanism may have discouraged the resort to protectionist measures in Mexico’s main export markets. It may be argued that without this mechanism, the protectionist lobby, particularly in the United States, would have been much more effective than it actually proved to be. NAFTA’s rules provided ammunition to the US government to resist domestic protectionist pressures which had been exacerbated by the change in sign of the trade balance between the two countries. The flip side of this export surge is of course NAFTA’s very positive contribution to making Mexican markets (goods, services, investment, government procurement) “contestable”. It has been the engine behind export efficiency and increased productivity of domestic production in both goods and 48 services. In other words, NAFTA has not only opened the markets of Mexico’s main trading partners, but also provided, through the liberalisation of the domestic markets, the means to supply those markets. This is reflected in the increasing participation of foreign companies in Mexico, for many of them were attracted by NAFTA itself. According to the United Nations Conference on Trade and Development (UNCTAD), intrafirm trade between US majority-owned foreign affiliates in Mexico and their parents in the United States rose from 24 per cent of total trade between the two countries in 1989 to 27 per cent in 1993. In contrast, intrafirm trade between US majority-owned foreign affiliates in Latin America (excluding Mexico) and the Caribbean and their US parents increased 49 from 17 to 20 per cent of total trade between the two regions over the same period.

204

Had it not been for the very positive behaviour of Mexico’s total exports during 1995, the decrease in 50 GDP would had been of around 11 per cent instead of the 6.2 per cent recorded. Taken together, the above-mentioned facts strongly suggest that NAFTA played a very important role in avoiding an even deeper recession than the one actually experienced. By the same token, the very positive contribution of exports in the impressive economic recovery shown by Mexico, with GDP growing 5 per cent in 1996 and 7 per cent during the first semester of 1997, may also be at least in part attributed to 51 NAFTA. It is clear that even without NAFTA the structural reforms introduced in Mexico would have determined a more rapid recovery from the 1995 crisis than if the country had followed an inwardlooking development strategy. But it may also be argued that without NAFTA the reforms would have been less profound and would have taken longer. In any case, without NAFTA exports would have increased at a lower pace. Therefore, the recession would have been deeper and the recovery would have taken longer, including, of course, the recovery of imports. Actually, during the first five months of 1997, imports increased at an annual rate of above 25 per cent. Foreign direct investment coming from NAFTA countries is just above 60 per cent of the total FDI received (Table 9). Around 20 per cent originates in the EU member states, and the rest in Japan, Korea, and Switzerland, among others. Much of the non-NAFTA investment has been attracted to Mexico with the whole North American market in mind. Most of it is geared either exclusively towards foreign markets (as is the case of a number of Japanese, German and Korean maquiladora companies) or to satisfy both the domestic and foreign markets. Foreign-owned companies exclusively dedicated to supplying the domestic market seem to be the minority. The main destination of FDI in Mexico is the manufacturing sector with more than half of the total (Table 10). The services sector receives most of the remaining balance. Mexico’s insertion in the global market has also taken place through Mexican companies investing abroad. Though there are no reliable statistics for the yearly amounts of Mexican outward direct investment, it seems to be substantial. Six of the ten largest Latin American TNCs with assets abroad are of Mexican origin (Table 11). The remaining four are Brazilian. Two of the Mexican corporations are among the top ten TNCs based in developing countries, ranked by foreign assets (Table 12). It is interesting to note that the majority of Mexico’s foreign investment abroad is directed to the US. However, substantial amounts of Mexican investment have been committed in Latin American countries and, to a lesser extent, in European countries.

205

NOTES

1.

Of course, protection was not the only condition required for attracting firms into the protected industries of home countries. Cf. Bhagwati (1978).

2.

Today, one trillion dollars moves daily from country to country looking for the highest possible over-night returns.

3.

During the 1980s and early 1990s an extensive literature on the subject appeared, addressing everything from sectorial issues to the characteristics of the so-called services revolution. The main reason for such a comprehensive literature was the then on-going GATS and other regional trade negotiations that included services. A classic on the subject is UNCTAD (1988); see also De Mateo (forthcoming).

4.

Cf. Aranda et al. (1995).

5.

Cf. UNCTAD (1996b).

6.

There is not doubt that “globalisation” and all it involves was the primary reason that many developing countries changed their development strategies. However, many other considerations, both economic (e.g. the need for fresh foreign capital after the debt crisis), and political (e.g. the fall of the Berlin Wall, symbolising the failure of centrally planned economies), were also very important. But as in the past, protectionism was not a sufficient condition to attract FDI; liberalisation per se is not enough to create a favourable climate for foreign investment in the context of increasing globalisation.

7.

Intra-firm trade is a prominent component of foreign trade in today’s world. According to UNCTAD, in the late 1980s at least 80 per cent of US international trade was related to TNCs; that is, US parent firms and their affiliates abroad, and US affiliates of foreign TNCs (cf. Aranda et al., 1995). Perhaps more importantly, as a proportion of intra-firm exports to foreign affiliates, those to the US rose from 37 per cent in 1977 to 53 per cent in 1983, and to 60 per cent in 1993 (cf. UNCTAD, 1996b).

8.

Effective protection is defined as the percentage difference between value-added in the production of a product at domestic prices and value-added at world prices. For a demonstration that trade liberalisation of goods requires an accompanying trade liberalisation of services, see De Mateo and Carner (1996).

9.

For an analysis of the shortcomings of both the GATS and the results of the sectorial negotiations, see Sauvé (1995), and De Mateo and Carner (1994).

10.

The expression “deep integration agreements” is understood in this paper as encompassing not only free trade in merchandise within a regional agreement (as in EFTA; or in the different FTAs between the European Union on the one hand, and other European, Middle East, and North African countries on the other; or as in agreements between and among a number of developing countries), but also the liberalisation of, at least, services and foreign investment, with workable dispute settlement procedures.

11.

For a history on international foreign investment negotiations, see Hart (1996).

12.

There are around 1 300 BITs, 75 per cent of which have been concluded since 1990.

13.

The TRIMS Agreement “... remains the odds-on favourite for Trade Policy Hall of Fame’s Lifetime Underachievement Award ...” (Sauvé, 1995).

14.

As Pierre Sauvé correctly pointed out to me, full market contestability is more an objective than reality, but getting there progressively through the more coherent pursuit of competition-friendly policies is what the trade, investment, and regulatory conduct agendas are increasingly about.

206

15.

For a more detailed description, see De Mateo and Carner (1996), Carner (1990), and Amigo (1990).

16.

The term Mexicanisation will be used to indicate that foreign participation in a sector is no longer allowed, while nationalisation indicates the state’s assumption of the ownership of, and exclusive rights for, the production of a good or supply of a service.

17.

The reforms to the FIL management were part of a much larger set of measures directed at establishing a new development strategy. These included, among other things, a stabilisation programme that was put in place in 1983. A far reaching privatisation scheme was also implemented, reducing the number of stateowned enterprises from 1 214 to 468 between 1983 and 1988. Other measures conducive to FDI were also introduced, such as strengthening the Law of Inventions and Trademarks which made Mexico’s intellectual property regulations similar to those of its principal trading partners.

18.

This may seem contradictory in a country where, at that moment, a “hard-core” import-substitution model was being followed. However, it was realised that free trade for both inputs and final goods required free investment as well in order for the maquiladora programme to work properly.

19.

Calderón et al. (1994).

20.

“While the devaluations of the 1980s made Mexican assembly operations much more convenient for TNCs operating in the US market, the early low wage advantages seem to have given way to more longrange considerations for corporate strategies in terms of international competitiveness in modern activities” (Calderón et al., 1994).

21.

This is no longer true with respect to al the countries with which Mexico has concluded free trade agreements (Costa Rica, Colombia, Venezuela, and Bolivia in addition to the NAFTA countries), and in relation to the countries with which Mexico has in place BITs.

22.

Jobs in the garment industry received a big boost as a result of trade liberalisation of this sector in the NAFTA. They increased around 160 per cent between 1993 and April 1997.

23.

Though it may be argued that the main engine for such integration was the trade liberalisation programme followed by Mexico during that decade.

24.

The 1988 bilateral FTA between Canada and the United States (CUSFTA), which explicitly recognised the high integration their economies had achieved through the years, provided an immediate precedent of a deep integration agreement. However, in most areas the NAFTA went beyond the CUFTA. Foreign direct investment was one of them.

25.

There are minor exceptions to this general rule; for certain products accounting for just 1 per cent of Mexico’s total trade with its NAFTA partners complete liberalisation will take up to fifteen years.

26.

Besides subjecting intellectual property disputes to the NAFTA’s dispute settlement mechanisms, its provisions also cover, among other things, products in the “pipeline”, i.e. in the process of being developed.

27.

Not covering sub-federal measures, but with very low thresholds above which the procurement is to be subjected to international bidding.

28.

Covering business visitors, traders and investors, intra-company transferees, and professionals.

29.

National treatment, MFN and the right not to establish being the main principles. As in investment, a topdown approach was adopted.

30.

At least four out of six kind of measures typified in the GATS as market access restrictions are in fact quantitative restrictions applicable to both domestic and foreign firms.

31.

Rugman and Gestrin (1996), state that “One main reason for the decision to use negative lists was that Mexico had so many existing measures running counter to the NAFTA that grand-fathering was not a practical option”. In fact, it was Mexico who proposed to use the top-down approach, for the state and provincial measures in the US and Canada, respectively, were unknown quantities. As it turned out, the (albeit, small) number of reservations of any of the three countries was not very different from the other two.

32.

Cf. Low (1995), Graham and Sauvé, and Rugman and Gestrin, op. cit.

207

33.

The most conspicuous example being the SEMATECH, a consortium of US manufacturers of computer chips receiving government subsidies and excluding foreign participation.

34.

Cf. Rugman and Gestrin (1996). Rules of origin “will serve as an impediment to competition as nonNorth American producers will be constrained in their ability to source from familiar supply networks ...”.

35.

Notwithstanding these shortcomings, NAFTA still qualifies as a deep integration scheme, albeit an incomplete one.

36.

Both through autonomous liberalisation and by the way of multilateral negotiations in the GATS, as in the case of basic telecommunications. In December 1996 a new, more liberal Foreign Investment Law substituted the one in force since December 1993. For a description, cf. Novoa (1997).

37.

Domestic road transportation passenger services, and banking and other financial services. NAFTA treatment is provided to the latter if the investment originates in a foreign affiliate established in the United States and Canada.

38.

As from December 1995 there is a 49 per cap, increasing to 51 per cent in 2001, and to 100 per cent in 2004. Cabotage services remain reserved to Mexican nationals.

39.

It is interesting to note that if the pre-NAFTA regulations on both FDI and cross-border services were applied today, consumers in Mexico would not be allowed to make use of the Internet services. This is just an example of the problems producers of goods and services faced in using one of the most dynamic sectors, indispensable for the insertion of a country into the globalisation process: the telecommunications value-added services.

40.

FDI in Law services requires prior approval of the NCFI to reach 100 per cent. Public notaries and commercial public notaries have to be Mexican-born citizens, and are not allowed to associate with any other person for the provision of their services.

41.

Cf. De Mateo and Carner (1996).

42.

It most also be recalled that there is a market quota -albeit a growing one- with respect to FDI in the financial sector. However, the market quota have not been filled up in any of the different, individual services. Such quotas will be eliminated by year 2000.

43.

Figures in the period 1993-1996 are not fully comparable with previous years because of a change of methodology. However, independently of the methodology utilised, FDI inflows doubled between 1993 and 1994.

44.

A negative effective protection with a value smaller than 100 per cent means that, at world prices, value added in the production of a tradable would be negative. Cf. Instituto Mexicano de Comercio Exterior (1984).

45.

Id.

46.

In terms of productivity and quality, the new production facilities in Mexico may have even surpassed the benchmarks established by the US automobile industry. Cf. Calderón et al. (1994).

47.

Besides the three US automakers, Volkswagen and Nissan have had a presence in Mexico since the 1950s and 1960s, respectively, (Renault stopped assembling cars in the mid-1980s, but continues to produce motors in Mexico). During the 1990s, Mercedes Benz, Honda, BMW, and British Leyland have production facilities in the country.

48.

The lack of reliable statistics makes it difficult to ascertain Mexican exports in sectors different from travel and transportation. However, there is evidence of important exports, in the four modes of delivery, in a number of sectors such as professional services, construction, telecommunications value-added services, basic telecommunications, wholesale and retail trade, maintenance and repairs, restaurants and hotels, among others.

49.

UNCTAD (1995).

50.

This estimation was done by both the Balassa and the Chenery models. These models decompose GDP growth in the behaviour of domestic demand, exports and import substitution.

51.

In 1996 exports was the single most important factor explaining the 5.1 per cent growth in GDP. Cf. Banco de Mexico (1997).

208

BIBLIOGRAPHY

AMIGO, J. (1990), “The Legal Framework of Direct Foreign Investment in Mexico”, in Direct Foreign Investment, Banamex. ARANDA, V., ECONOMOU, P. and SAUVANT, K. (1995), “Trends and Policies in Market Presence”, in OECD, New Dimensions in Market Access in a Globalising World Economy, Paris. BANCO DE MEXICO (1997), The Mexican Economy. BHAGWATI, J.N. (1978), Foreign Exchange Regimes and Economic Development: Anatomy and Consequences of Exchange Control Regimes, Ballinger, Cambridge. CALDERÓN, A., MORTIMORE, M. and PERES, W. (1994), “Mexico’s Integration into the North American Economy: The Role of Foreign Investment”, Foreign Direct Investment in Developing Countries: The case of Latin America, IRELA and the European Commission, Madrid. CARNER, F. (1990), “Foreign Investment in Services in Mexico”, in Direct Foreign Investment, Banamex. DE MATEO, F. and CARNER, F. (1994), “Los servicios en la Ronda Uruguay” (“Services in the Uruguay Round”), in Comercio Exterior, December. DE MATEO, F. and CARNER, F. (1996), “Los servicios en el sector externo” (“The role of services in the external sector”), in Comercio Exterior, February. DE MATEO, F. (forthcoming), “El sector servicios en la apertura económica y el Tratado de Libre Comercio” (“The services sector in economic liberalisation and NAFTA”). GRAHAM, E.M. (1996), “Toward a Rules-Based Regime for Investment: Issues and Challenges”, in Sauvé, P. and Schwanen, D. (eds.), Investment Rules for the Global Economy, C.D. Howe Institute, Toronto. GRAHAM, E.M. and SAUVÉ, P. (1996), “Toward a Rules-Based Regime for Investment: Issues and Challenges”, in Sauvé, P. and Schwanen, D. (eds.), Investment Rules for the Global Economy, C.D. Howe Institute, Toronto. HART, M. (1996), “A Multilateral Agreement on Foreign Direct Investment”, in Sauvé, P. and Schwanen, D. (eds.), Investment Rules for the Global Economy, C.D. Howe Institute, Toronto.

209

INSTITUTO MEXICANO DE COMERCIO EXTERIOR (1984), La protección efectiva en México: 1979-1983 (Effective protection in Mexico: 1979-1983), México. LOW, P. (1995), “Market Access Through Market Presence: a Look at the Issues”, in OECD, New Dimensions in Market Access in a Globalising World Economy, Paris. NOVOA, C. (1997), “Reformas a la ley de inversión extranjera” (“Reforms to the foreign investment law”), in Revista mexicana de derecho internacional privado, México, April. RUGMAN, A. and GESTRIN, M. (1996), “A Framework for Multilateral Agreement on Investment”, in Sauvé, P. and Schwanen, D. (eds.), Investment Rules for the Global Economy, C.D. Howe Institute, Toronto. SAUVÉ, P. (1995), “The General Agreement on Trade in Services: Half Full or Half Empty?”, Journal of World Trade, August. SAUVÉ, P. (1996), “Market Access through Market Presence”, in Sauvé, P. and Schwanen, D. (eds.), Investment Rules for the Global Economy, C.D. Howe Institute, Toronto. UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT) (1988), World Trade and Development Report -- 1988, United Nations, New York. UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT (1995), World Investment Report -- 1995, United Nations, New York. UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT (1996a), El desarrollo frente a dos corrientes poderosas: la globalización y la liberalización (Development vis à vis globalisation and liberalisation), United Nations, New York. UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT (1996b), World Investment Report -- 1996, United Nations, New York.

210

Table 1. Foreign Direct Investment in Mexico 1 (Millions of dollars) Year

New investment

Change (%)

FDI stock

Change (%)

1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993

287.3 362.2 295.0 299.1 327.1 383.3 810.0 1 622.6 1 701.1 626.5 683.7 1 442.2 1 871.0 2 424.2 3 877.2 3 157.1 2 499.7 3 722.4 3 565.0 3 599.6 4 900.7

26.1 -18.6 1.4 9.4 17.2 111.3 100.3 4.8 -63.2 9.1 110.9 29.7 29.6 59.9 -18.6 -20.8 48.9 -4.2 1.0 36.1

4 359.5 4 721.7 5 016.7 5 315.8 5 642.9 6 026.2 6 836.2 8 458.8 10 159.9 10 786.4 11 470.1 12 899.9 14 628.9 17 053.1 20 930.3 24 087.4 26 587.1 30 309.5 33 874.5 37 474.1 42 374.8

8.3 6.2 6.0 6.2 6.8 13.4 23.7 20.1 6.2 6.3 12.5 13.4 16.6 22.7 15.1 10.4 14.0 11.8 10.6 13.1

1994 1995 1996

10 158.8 7 613.3 5 646.1

107.3 -25.1 -25.8

52 533.6 61 146.9 65 793.0

24.0 14.5 9.4

1. As from 1994 figures include FDI flows notified to the foreign Investment National Register and imports or fixed assets by the maquiladora industry. They do not include profit reinvestment or transfer among companies. Source: Dirección General de Inversión Extranjera, La Secretaría de Comercio y Fomento Industrial (SECOFI).

Table 2. Foreign investment in selected countries (Stocks) Mexico Year 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989

million $ 8 459 10 160 10 786 11 470 12 900 14 628 17 053 20 930 24 087 26 587

Brazil

change %

23.7 20.1 6.1 6.3 12.5 13.4 16.6 22.7 15.1 10.4

million $ 16 846 18 440 21 041 23 302 25 045 26 380 27 053 29 171 31 381 37 792

Canada

Taiwan

change %

billion $

change %

9.5 14.1 11.0 7.3 5.6 2.7 8.1 7.5 8.3

129.0 134.0 171.9 184.7 209.4 217.3 267.9 256.3 260.2 287.3

3.9 28.2 7.4 13.4 3.8 9.5 7.7 1.5 10.4

million $ 1 050 2 251 3 501 4 698 5 899 7 002 8 201 9 620 10 800 12 310

change %

114.3 55.5 34.2 25.5 18.7 17.1 17.3 12.3 14.0

Source: León Opalín and Laura Iturbide, Direct Foreign Investment: An International Comparative Analysis, Direct Foreign Investment, Banamex, 1990.

211

Table 3. Maquiladora industry: main indicators, 1980-1997 Number of plants Year 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 19973

Value added 1

Employment

In all of Mexico

In Border States2

In all of Mexico

In Border States

In all of Mexico

In Border States

578 605 585 600 672 760 891 1 125 1 396 1 655 1 703 1 914 2 075 2 114 2 085 2 130 2 411 2 607

533 564 546 564 631 714 830 1 017 1 253 1 461 1 462 1 614 1 746 1 764 1 718 1 692 1 874 2 020

119 546 130 973 127 049 150 867 199 684 211 968 249 833 305 253 369 489 429 725 446 436 467 352 505 698 542 074 583 044 648 263 754 858 846 078

112 548 123 142 120 384 143 623 189 850 202 140 237 000 287 091 343 406 395 727 402 426 413 628 443 204 465 575 500 633 545 441 619 505 687 063

17 729.0 19 089.2 23 573.6 25 252.7 30 204.4 32 486.0 44 081.8 50 752.7 59 928.3 75 715.5 81 584.1 87 070.3 92 084.6 99 885.0 111 104.1 127 563.0 139 867.5 57 136.5

15 995.0 17 435.9 21 485.6 23 150.5 27 780.1 30 311.0 40 809.4 46 446.4 54 232.0 67 503.3 71 336.6 74 597.1 77 866.0 84 393.5 94 031.4 105 056.3 113 198.3 46 029.1

1. Thousand of 1980 pesos. 2. Border States: Baja California Norte, Sonora, Chihuahua, Coahuila y Tamaulipas. 3. Figures for 1997 first quarter. Source: Instituto Nacional de Estadística Geografía e Informática (INEGI).

Table 4. Mexico’s Trade Balance, 1983-1997 (Millions of dollars) Exports Year 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1 1997

Total

Imports

Maquila

25 953 29 100 26 757 21 804 27 600 30 691 35 171 40 711 42 688 46 196 51 886 60 882 79 542 96 000 43 000

3 641 4 904 5 093 5 646 7 105 10 146 12 329 13 873 15 833 18 680 21 853 26 269 31 103 36 920 16 735

1. January - May 1997. Source: Banco de México.

212

Total 11 848 15 916 18 359 16 784 18 812 28 082 34 766 41 593 49 967 62 129 65 367 79 346 72 453 89 469 41 216

Maquila 2 823 3 749 3 826 4 351 5 507 7 808 9 328 10 321 11 782 13 937 16 443 20 466 26 179 30 505 13 549

Table 5. Mexico’s participation in FDI inflows, 1990-1995 (Percentages) Year

Latin America

Developing coutries

1990 1991 1992 1993 1994 1995

34.2 38.7 30.9 28.3 38.4 1 28.2

11.0 8.7 7.1 6.7 11.7 7.6

1. Estimate. Source: Economic Commission for Latin America and the Caribbean (CEPAL), “Inversion extranjera directa e intrarregional en América Latina y el Caribe: Una propuesta para su seguimiento”, LC/R. 1996, 27 December 1996 ; UNCTAD, 1996 b, p. 4.

Table 6. FDI flows to developing countries, 1994-1996 (Billions of dollars) China Mexico Malaysia Brazil Indonesia Poland Hungary Thailand Chile Argentina

112.1 31.6 16.3 13.5 12.2 9.8 7.3 6.4 5.7 3.9

Source: La Secretaría de Comercio y Fomento Industrial (SECOFI) and World Bank.

213

Table 7. Export propensities of United States majority-owned foreign affiliates in manufacturing, 1966-19931 (Percentages) Economy

1966

1977

1982

1986

1989

1993

Developed economies Developing economies Latin America and the Caribbean Mexico India Malaysia Hong Kong Korea Singapore Taiwan Province of China

20.4 8.4 6.2 3.2 6.9 .. .. .. .. ..

33.1 18.1 9.7 10.4 3.6 76.2 80.5 68.4 93.2 71.4

36.6 22.0 11.9 10.8 .. 81.5 77.4 .. 91.8 59.4

39.3 32.5 20.0 34.5 4.1 ≥ 83.7 71.8 58.0 89.7 63.7

38.0 2 36.7 22.0 33.7 .. 74.7 68.6 38.5 87.2 46.4

40.6 2 38.7 22.2 32.1 .. 84.9 55.0 27.9 85.9 38.8

Developed and Developing economies

18.6

30.8

33.9

38.4

37.8

40.3

2

2

1. Exports (total sales minus local sales) as percent of total sales. 2. Exports by manufacturing affiliates in Africa and Korea in 1982, Africa and Middle East in 1986 and Israel and New Zealand in 1989 and 1993, included in these figures, are estimates. Source: United States, Department of Commerce, various years. Taken from UNCTAD, World Investment Report, 1996b.

Table 8. Mexico: Export propensities of United States majority-owned foreign 1 affiliates in the automobile industry (Percentages) 1978-82

1983-87

1991

1992

1993

1994

1995

1996

0

38.4

44.4

45.4

55.0

63.6

88.9

82.9

1. Exports as percent of total production in Mexico expressed in number of automobiles (Chrysler, Ford and General Motors). Source: 1991-1996, estimates on the basis of data provided by AMIA; 1978-1987, estimates on the basis of data appearing in Calderón et al. (1994).

214

Table 9. Mexico: FDI flows by country of origin, 1981-1996 (Millions of dollars) 1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1

304.4

96.4

282.9

217.7

137.2

737.4

637.6

1 309.0

270.0

946.9

870.6

744.6

France

10.3

6.8

110.0

8.7

10.7

316.9

31.2

152.4

16.5

181.0

500.5

68.9

146.3

39.9

110.0

152.5

55.4

218.6

46.9

136.7

84.7

288.2

84.7

84.9

EEC/EU

Germany Italy Spain U.K. 2

Netherlands

1993

1994

1995

1996

608.7

1 890.4

1 748.8

76.9

89.3

116.4

69.8

111.4

304.8

543.1

162.7

753.3

5.1

1.9

1.0

0.5

0.6

4.0

2.8

0.0

6.6

4.6

1.9

7.5

4.6

2.4

10.4

17.6

101.8

40.4

12.7

38.7

-13.0

93.7

125.8

34.1

44.0

10.9

43.8

37.2

63.5

141.5

39.2

40.4

40.9

7.4

49.2

44.3

56.4

104.3

430.9

767.6

44.7

114.4

74.2

426.8

189.2

589.2

206.9

59.2

8.6

6.3

13.1

17.8

22.4

10.2

30.3

71.1

47.8

126.1

119.5

83.1

88.2

727.9

690.7

325.0

Sweden

15.3

-2.0

29.1

61.1

5.5

24.6

36.7

32.5

6.9

13.3

13.9

2.0

2.4

9.3

61.1

60.2

Switzerland

74.9

23.1

16.2

59.8

141.2

34.1

95.2

86.3

194.4

148.0

68.0

315.2

101.7

53.6

200.1

66.8

1 077.3

434.2

558.6

647.6

1 388.7

1 247.0

2 688.9

1 275.5

1 851.2

2 364.1

2 460.3

1 740.1

3 577.8

5 260.8

4 977.7

4 217.1

1 072.1

426.1

536.5

615.1

1 353.8

1 206.4

2 669.6

1 241.6

1 813.8

2 308.0

2 386.1

1 651.7

3 503.6

4 525.3

4 817.4

3 723.9

5.2

8.1

22.1

32.5

34.9

40.6

19.3

33.9

37.4

56.1

74.2

88.4

74.2

735.5

160.3

493.2

Japan

212.1

65.4

3.8

35.6

79.3

145.7

132.3

145.8

15.7

120.8

73.5

86.9

73.6

628.3

159.4

94.7

Other

8.5

3.1

49.9

93.3

-18.2

225.3

255.7

455.4

187.2

351.0

124.7

747.0

611.4

2 351.7

608.6

532.6

Total

1 701.1

626.5

953.6

1 159.9

1 729.0

2 424.4

3 876.7

3 157.4

2 499.7

3 722.4

3 565.0

3 599.6

4 900.7

10 158.8

7 613.6

5 646.1

North America

215

United States Canada

1. 1981-1985: 10 countries; 1986-1994: 12 countries; 1995-1997: 15 countries. 2. 1981-1988: Netherlands; Belgium and Luxembourg; 1989-1997: Netherlands. Source: Dirección General de Inversión Extranjera, La Secretaría de Comercio y Fomento Industrial (SECOFI).

Table 10. Sectorial distribution of FDI, Mexico, 1994-1997 (Millions of dollars) 1994 Sector Agriculture Mining Manufacturing of which maquiladora Electricity and water Construction Commerce Financial services2 Transport and communic. Other services Total

Value

1995 %

Value

Acum. 1994-19971

1996 %

Value

%

Value

%

7.9

0.1

8.9

0.1

22.9

0.4

40.0

0.2

102.9

1.0

79.0

1.0

65.5

1.2

249.0

1.0

5 878.9 777.0

57.9 7.6

4 294.0 1 098.6

56.4 14.4

3 196.9 947.6

56.6 16.8

13 911.2 3 230.0

57.6 13.4

15.2

0.1

2.1

0.0

2.3

0.0

20.5

0.1

258.0

2.5

24.0

0.3

25.5

0.5

317.7

1.3

1 239.6

12.2

923.5

12.1

526.4

9.3

2 712.5

11.2

938.4

9.2

1 052.2

13.8

1 124.7

19.9

3 188.8

13.2

591.9

5.8

860.0

11.3

347.8

6.2

1 852.6

7.7

1 126.0

11.1

369.6

4.9

334.1

5.9

1 842.6

7.6

10 158.8

100.0

7 613.3

100.0

5 646.1

100.0

24 134.9

100.0

1. 1994-May 1997. 2. Including real estate. Source: Dirección General de Inversión Extranjera , La Secretaría de Comercio y Fomento Industrial (SECOFI).

216

Table 11. The largest TNCs in Latin America, ranked by foreign assets, 1994 (Millions of dollars and number of employees) Assets Foreign

Sales Total

Foreign

Employment

217

Corporation

Country

Industry

Total

Foreign

Total

Cemex, S.A.

Mexico

Cement

2 847

7 893

744

2 101

8 073

20 997

Grupo Televisa, S.A. de C.V.

Mexico

Media

1 371

3 260

286

1 288

..

21 600

Souza Cruz, S.A.

Brazil

Tobacco

935

1 246

316

3 784

63

11 387

Petrolero Brasilerio, S/A - Petrobras

Brazil

Petroleum refining

715

30 162

2 316

26 396

24

50 295

Empresas Ica Sociedad

Mexico

Construction

321

3 264

95

1 386

2 136

25 267

Controladora S.A.Sadia Concordia S/A

Brazil

Food

313

1 405

567

2 784

57

32 357

Industria e Comercio Desc S.A. de C.V.

Mexico

Diversified

313

1 902

313

1 633

3 431

19 288

Grupo Industrial Bimbo, S.A. de C.V

Mexico

Food

..

1 221

252

1 252

..

42 463

Companhia Cervejaria Brahma

Brazil

Food

187

1 755

80

1 249

476

9 606

Grupo Sidek

Mexico

Tourism

114

2 831

25

575

10 438

10 774

Source: UNCTAD, World Investment Report, 1996b.

Table 12. The top 10 TNCs based in developing economies, ranked by foreign assets, 1994 (Millions of dollars and number of employees) Assets Foreign assets

Index1

1

11

Corporation

Economy 2

Daewoo

Korea

Sales

Employment

Industry

Foreign

Total

Foreign

Total

Foreign

Total

Index1

Electronics

..

3

33 000

16 000

40 000

100 000

200 000

33.0

3

34.4

2

10

Hutchison Whampoa Ldt.

Hong Kong

Diversified

..

52 192

12 500

30 168

15 086

26 855

3

8

Cemex S.A.

Mexico

Cement

2 847

7 893

744

2 101

8 073

20 997

36.6

4

5

Jardine Matheson

Hong Kong

Construction

2 539

6 350

6 463

9 559

50 000

220 000

43.4

5

..

Holdings Ltd. China State China construction

Engineering Corp.

Construction

2 189

..

5

1 010

..

..

..

..

6

..

China Chemicals Imports and Exports

China

Trading

1 915

..5

7 914

..5

..5

..5

..5

7

20

Samsung Co., Ltd.

Korea2

Electronics

..3

38 000

21 440

67 000

42 235

195 429

19.5

2

..3

25 000

8 600

43 000

29 061

59 200

25.1

4

5

5

5

5

218

8

17

LG Group

Korea

Electronics

9

19

Grupo Televisa, S.A. de C.V.

Mexico

Media

1 371

3 260

286

1 288

..6

21 600

22.2

10

34

Hyundai

Korea2

Diversified

1 293

9 657

1 610

13 081

814

44 835

9.2

1. The index of transnationality is calculated as the average of foreign assets to total assets, foreign sales to total sales and foreign employment. The ranking shown is based on a ranking of the top 50 TNCs bases in developing countries. 2. The accounting standards of Korea do not require the publication of consolidated financial statements including both domestic and foreign affiliates. The figures provided here are estimates of consolidated financial statements as provided by the companies in response to a survey by the United Nations Conference on Trade and Development (UNCTAD). Depending on the availability of the data on foreign components, the data for business group totals are used. 3. Data on foreign assets are either supressed to avoid disclosure or they are not available. In the case of non-availability, they are estimated on the basis of the ratio of foreign total employment, foreign to total sales and similar ratios for the transnationality index. 4. A subsidiary of Jardine Matheson Holdings of Bermuda. 5. Data are not available. 6. Data on foreign employments are supressed to avoid disclosure or not available. In the case of non-availability of the data, they are estimated on the basis of other foreign component ratios for the transnationality index. Source: UNCTAD, World Investment Report , 1996b, p. 34.

B. MODELLING RESULTS: NAFTA AND NORTH AMERICAN MIGRATION

Applied general equilibrium models: the Mexican experience of NAFTA Horacio Sobarzo Fimbres Economic effects of NAFTA: employment and migration modelling results Raúl Hinojosa-Ojeda, Robert McCleery and Fernando De Paolis

219

APPLIED GENERAL EQUILIBRIUM MODELS: THE MEXICAN EXPERIENCE OF NAFTA

by Horacio Sobarzo Fimbres Centro de Estudios Económicos, El Colegio de México

1.

Introduction

This paper presents some reflections on the use of the applied general equilibrium (AGE) approach to evaluate economic integration, taking as a point of reference the Mexican experience prior to the approval of NAFTA. Although some reference to specific results will be made, the intention is to offer general comments about the main characteristics of AGE models, emphasising both their weaknesses as well as their merits. The document is structured as follows: the first section outlines the basic principles of these models and their resolution; the second very briefly discusses the main applications as well as the main limitations of the approach and compares them with other models, such as macroeconometric models; the third outlines the main results of the models elaborated around the NAFTA experience; and finally, the fourth presents the main conclusions and some implications for the modelling of economic integration. 2.

Applied general equilibrium models

AGE models can be traced back to Leontief’s work based on fixed input-output coefficients. Unlike input-output models, however, AGE models allow for substitution both on the consumption and the production sides. The most well known analytical application in a two sector model was developed by Harberger (1962) to analyse issues related with fiscal incidence. In 1967 Scarf (1967) developed a computer algorithm which subsequently permitted a detailed disaggregation of these models. The work by Shoven and Whalley (1972) is perhaps the first example of a highly disaggregated AGE model dealing with tax reform. During the 1970s and 1980s, a large number of models were developed for different countries, most 1 of them focusing on tax and trade reforms. In particular, numerous AGE models were built to analyse the effects of trade and tax reforms in developing countries, introducing great detail into the analysis 2 of the structure of income distribution and rigidities presumed characteristic to these countries. More 3 recently, this kind of model has been used to explore the effects of ecological taxes. The structure of AGE models is well documented elsewhere and therefore we will not go into the details of specific models. For the purposes of this document it will suffice to outline their basic underlying principles, regardless of the many different ways in which AGE models can be specified.

221

Essentially, these models simulate together both supply and demand for every market; the solution for the model is reached when a vector of prices that clears all markets is found. Both supply and demand are modelled using behavioural assumptions for producers and consumers. Producers are normally modelled using a constant returns scale technology, sometimes allowing for some input substitution, mostly between domestic and imported commodities. Consumers have initial endowment and demand functions for each commodity. These demands are a function of prices and income, and satisfy Walras’ Law. Finally, producers are assumed to maximise profits whereas consumers are assumed to maximise utility. Equilibrium is characterised by a set of prices and production levels such that demand equals supply in every market. It is therefore necessary to construct an initial or base equilibrium which is done by calibrating, for any year, a database from national accounts, which normally include an input-output matrix describing the inter-industry transactions. To put all of the database into a consistent framework requires a good deal of effort, particularly if the model is to incorporate a detailed production and consumption structure. Once the model reproducing the base equilibrium is calibrated, experiments can be undertaken by changing the value of certain exogenous variables (for instance tariff rates), and finding the new vector of prices that clears every market. The resolution of the model is an iterative process where both producers and consumers react to prices until both find a vector in which profit and utility levels are maximised, given the technological and income restrictions. The description provided above is of course very simplistic. In practice the results depend on the model’s behavioural assumptions (such as functional forms) as well as on the parameters (elasticities) chosen. For instance, critical behavioural assumptions include the choice of the so-called closure rule, whether or not factors are mobile between sectors, and whether factors are fully employed or not. The decisions concerning these behavioural assumptions can lead to very different, even opposite results. It is therefore important to bear in mind that the results of AGE models should be interpreted carefully, for their reliability is ultimately dependent on the choice of a reasonable set of assumptions. Two final points should be mentioned. First, most models explicitly incorporate the following economic agents: producers, consumers, a government, and a foreign sector (often referred to as the ‘rest of the world’). It is commonly assumed that governments collect taxes and then redistribute the resulting revenue in a lump-sum fashion. This is due to the difficulty of modelling public goods. Second, while most models built in the 1970s and 1980s were static and assumed perfect competition, models in recent years have begun incorporating forms of imperfect competition and/or dynamic behaviour. It has to be said, however, that whereas static models are in some sense standard, when it comes to introducing dynamics and imperfect competition, a great number of behavioural assumptions are added, making it more difficult to achieve a consensus with regards to the results. Although static models may be useful, they do not answer such important questions as the mechanics of transition between equilibria, which are essential to understanding how labour markets adjust in a trade liberalisation process, for instance. 3.

Main applications and limitations

As mentioned earlier, AGE models have traditionally been used to analyse trade and tax reforms, and more recently ecological tax reforms. It is also common to add some elements typical of a trade liberalisation process, such as capital flows. In general, however, AGE models address issues such as

222

fiscal incidence, or more generally, incidence on real incomes. This is, in fact, the main attractiveness of the model: given that they incorporate dynamic effects, resulting welfare changes are a measure of the changes in the real income of economic agents. In other words, to the extent that these models explicitly incorporate the production and consumption structure, and therefore the interaction of economic agents, the solution equilibrium -- unlike that of partial equilibrium models -incorporates all effects. A simple illustration of this might be tax reform. In a partial equilibrium model one normally has to assume, a priori, whether a particular tax is shifted forwards or backwards. This choice will crucially determine the results of the model. In AGE models, rather than being an assumption, this is a result. While this characteristic is among the model’s main assets, it is perhaps also their main weakness: the results are heavily dependent on the assumptions adopted by the modeller, beyond the problems of choosing a specific value for parameters that normally are not empirically estimated. Yet partial equilibrium models implicitly assume certain economic behaviours; the famous “ceteris paribus” clause ultimately means that the modeller has made a decision on behavioural assumptions, although these assumptions are not explicit. In that sense, the results of AGE models are very interesting, as they are the outcome of a complete and explicit set of assumptions. Other weaknesses include difficulties in choosing appropriate elasticities and other parameter values; in incorporating detailed micro-data sets; and in accounting for financial frameworks, which often treat elements like savings instruments in the aggregate. Finally, in models that incorporate dynamics and/or imperfect competition, their resolution very quickly becomes complicated, resulting in much less disaggregated models. More importantly, there is far less agreement as to how sensible are the assumptions adopted in these models. As a result, the economic processes of intertemporal behaviour, for instance, are still areas where economic theory provides insufficient explanations. The best example is perhaps endogenous growth theory which, although very appealing, still has a long way to go in explaining the mechanisms at work, not to mention the difficulty in measuring them. Ultimately, there is the issue of what these models are good for. The question is important because in the policy-making debate the results of this kind of model are sometimes wrongly seen as forecasts. While AGE models are indeed useful for policy analysis, they should not be used for forecasting. Although they incorporate the whole structure of the economy they do not -- unlike macroeconomic forecasting models -- incorporate lagged processes of behaviour capturing expectations. In AGE models the focus is on the structure of the economy and the mechanisms at work. In other words, AGE are simulation models and, therefore, their main attractiveness derives from their ability to provide a detailed description of adjustments resulting from economic changes. In the case of models designed to analyse trade liberalisation, for example, the results should not be seen as forecasts of trade deficits, but instead as indicators of how resources would have to be allocated in order to reach certain levels of trade deficit. More generally, AGE models are good for describing adjustment processes resulting from determined economic changes. As such, they provide the policy maker with a general idea as to the feasibility of certain processes. While there seems to be general agreement on this point, in an interesting exercise Kehoe, Polo, and Sancho (1995) compared the results generated by an AGE model with the changes that actually occurred in Spain during the period 1985-86 and concluded that the model performed well in predicting the changes that actually occurred. They argue that AGE models can be used to make conditional forecasts with some accuracy. Perhaps this argument should be interpreted as saying that, provided AGE models are built with realistic assumptions, their simulations are also realistic. 223

Therefore, it should not be surprising that they may be useful in making conditional forecasts. To be certain of the accuracy of the forecasts, however, one needs to be certain of the assumptions that it adopts. This is not always easy. In summary, despite the weaknesses of AGE models and the questionable realism of their underlying behavioural assumptions, whether they are right or wrong should not be at issue. Instead, they should be seen as an additional tool for policy makers who need to evaluate policy options. 4.

The Mexican experience with NAFTA

This section briefly summarises the experimental use of AGE models to analyse the impacts of NAFTA, in particular as they concern Mexico. The models and their results have already been 4 analysed and surveyed in great detail elsewhere. The following discussion will focus on their findings and suggest how these simulations might be of help in the debate on economic integration. Prior to the approval of NAFTA, several models were built to evaluate its possible economic impacts. All of them were AGE models although each model focused on different aspects. Tables 1, 2, and 3 5 describe the main characteristics of these models, as reported by Francois and Shiells (1994). As can 6 be seen, five models are multi-sector, two of which are multi-country, whereas the other two refer to Mexico and Canada. Another four models are sector focused, and two more are dynamic models. Of the four multi-sector models, three of them incorporate some form of imperfect competition. All of the models predict welfare gains for the three countries. Furthermore, all of the models predict that Mexico will be the country that benefits the most. This last result should not be surprising for at least two reasons. First, of the three countries Mexico is the smallest and therefore opening up to trade means that Mexico benefits disproportionally from access to the common market. Second, according to these models the removal of distortions leads to more efficiency -- prior to NAFTA, of the three countries Mexico was the most protected economy, especially if we consider that Canada and the US already had a free trade agreement. In numbers, the welfare gains for Mexico range from 1 to 5 per cent of GDP whereas the United States would gain, according to these models, no more that 0.1 per cent of GDP. Moreover, according to the results of these models, labour disruption in the United States does not seem to be important. According to the models that incorporate some form of imperfect competition, welfare gains to the three countries are greater than those predicted by models assuming perfect competition. Again, this result assumes that countries -- especially Mexico -- can exploit scale economies. This contrasts with the results of traditional models where welfare gains from trade liberalisation are normally inferior to 1 per cent of GDP, even in countries like Mexico. Turning to the sector focused models, the principal suggestion of the models focusing on agriculture is that the effects on labour migration may be great, which would suggest to policy makers that some form of adjustment program may be necessary. The study of the auto sector suggests that the Mexican industry may be in need of thorough restructuring. Finally, the model on textiles indicates that the United States could lose market share unless it relaxes its quotas. The results of the dynamic models indicate that welfare gains could be larger than those predicted by static models. These gains could come from a liberalisation of the capital goods sector, which in turn would increase investment. Moreover, while the empirical evidence is much weaker, Kehoe argues

224

that if countries learn in the process of doing, welfare gains could be substantially larger than those predicted by conventional static AGE models. Beyond the quantitative results of the AGE models, which depend on the assumptions adopted by the modellers, there appears to be agreement on the fact that the Mexican economy would benefit the most from the rationalisation process that NAFTA would in the long term produce. This consensus should be taken seriously by policy makers because it implies that the Mexican economy will experience a profound process of adjustment, particularly in the labour market. The lesson to be learned from this modelling experience is not so much that Mexico will automatically obtain welfare gains from 1 to 5 per cent of GDP as a result of NAFTA. Rather, the results suggest that if Mexico wants to have a substantial welfare gain from NAFTA, it has to go through a process of labour market reform, among other reallocation processes. Whether or not these processes will occur is something that the models do not address. Two final points are in order. First, it is very often argued in the political debate that the AGE models were too optimistic, especially in light of Mexico’s present economic situation. Such a line of argument is, however, “unfair” in my opinion because the models were built to answer to a very specific scenario. Moreover, the results of these models should be seen as taking place in the long run, which would be at least as long as the duration of the implementation of NAFTA -- between 10 to 15 years. In this respect, it is appropriate to recall a point made by Francois and Shiells (1994): Canadian public perception of CAFTA suggests strongly that NAFTA will be blamed, at least in some circles, for most -- if not all -- negative economic shocks experienced during or following implementation. Because this ‘scapegoat’ effect is not necessarily linked to real effects or developments, governments in all three countries need to be prepared to deal with pressure for adjustment assistance and related trade remedies, not only for NAFTAinduced structural adjustment, but also for those blamed erroneously on NAFTA. To support this point, one only needs to look to Canada, where mainstream parties have blamed the global recession on CAFTA. In a similar vein, structural adjustments in the United States associated with micro-imbalances are often attributed to ‘unfair trade’ in individual sectors, rather than to the broader macro-environment. The second point that should be mentioned is that although the models suggest strong adjustment processes in the Mexican labour market, none of them explicitly modelled the possible adjustment paths. To do that would require a fully dynamic model. Indeed, as Kehoe and Kehoe (1995) suggested, much of the future research work on AGE modelling should take this line of research as central. 5.

Conclusions

Despite their weaknesses, AGE models are a very useful instrument for policy makers, to the extent that they provide them with an extra tool with which to evaluate policy options. This tool should not be seen as an alternative to macroeconometric models, but rather as a useful complement. The purpose of macroeconometric models is to make forecasts based on lagged processes of behaviour, whereas AGE models try to describe the adjustment of the economy resulting from a particular change (or changes). To the extent that AGE models are reasonably built, using realistic assumptions, their results can be seen as merely conditional forecasts. Only in this very narrow sense do both approaches coincide. 225

The recent development of numerical algorithms and the rapid increase in computer power now permit the resolution of large and sophisticated models. Yet, when it comes to modelling dynamic behaviour, a very wide range of modelling possibilities are opened up, fuelling an ongoing discussion as to which modelling strategy is best. This situation can perhaps be explained by the fact that economic theory itself is still relatively undeveloped in this area.

NOTES

1.

See Shoven and Whalley (1984) for a detailed survey on these models.

2.

Dervis, de Melo and Robinson (1982) provide a rich discussion on the use of AGE models in developing countries.

3.

See, for instance, Whalley and Wigle (1991).

4.

See Francois and Shiells (1994) and Kehoe and Kehoe (1995).

5.

For a model focused on agricultural policies and migration see Hinojosa and Robinson (1991). More generally, a good description of modelling results on labour market effects can be found in Hinojosa and Robinson (1993).

6.

The number of sectors is from 12 to 27.

226

BIBLIOGRAPHY

DERVIS, K., De MELO, J. and ROBINSON, S. (1982), General Equilibrium Models for Development Policy, World Bank. FRANCOIS, J. and SHIELLS, C. (1994) (Eds.), Modeling Trade Policy: Applied General Equilibrium Assessments in North American Free Trade, Cambridge University Press, Cambridge. HARBERGER, A. (1962), “The Incidence of the Corporation Income Tax”, Journal of Political Economy, Vol. 70, No. 3, pp. 215-240. HINOJOSA, R. and ROBINSON, S. (1991), “Alternative Scenarios of US Mexico Integration: A Computable General Equilibrium Approach”, Working Paper, No. 609, Berkeley, April. HINOJOSA, R. and ROBINSON, S. (1993), “Cuestiones laborales en una zona norteamericana de libre comercio”, Economia Mexicana, Centro de Investigacion y Docencia Economicas, México, June. KEHOE, P. and KEHOE, T. (1995), Modeling North American Economic Integration, Kluwer Academic Publishers. KEHOE, T., POLO, C. and SANCHO, F. (1995), “An Evaluation of the Performance of an Applied General Equilibrium Model of the Spanish Economy”, Economic Theory, Vol. 6, pp. 115-141. SCARF, H. (1967), “On the Computation of Equilibrium Prices”, in Fellner, W.J. (ed.), Ten Economic Studies in the Tradition of Irving Fisher, New York. SHOVEN, J. and WHALLEY, J. (1972), “A General Equilibrium Calculation of the Effects of Differential Taxation of Income from Capital in the US”, Journal of Public Economics, Vol. 1, Nos. 3-4, pp. 281-321. SHOVEN, J. and WHALLEY, J. (1984), “Applied General Equilibrium Models of Taxation and International Trade: An Introduction and Survey”, Journal of Economic Literature, Vol. 22, pp. 1007-1051. WHALLEY, J. and WIGLE, R. (1991), “The International Incidence of Carbon Taxes”, Global Warming: Economic Policy Responses, MIT Press, Cambridge, Mass.

227

Table 1. Summary of demand and production structure Model

Countries

Multisector models

Demand side Demand functions

Production side Disaggregation

Production functions

Disaggregation

Brown

Home and Foreign

2-level utility functions 1-Cobb-Douglas between composite goods 2-CES between firm varieties.

1 consumer in each country

CES function of capital and labour for fixed and variable inputs

2 composite goods

Cox

Canada

Domestic final demand: 2-level utility functions 1-Cobb-Douglas between composite goods 2-CES between national varieties.

1 consumer in each region (Canada, US, Mexico, and ROW) buys goods from each region

Fixed costs in imperfectly competitive industries are linear functions of labour and capital input prices.

19 production sectors in Canada

Foreign final demand: Derived from CES utility functions between national varieties

Unit variable costs are Cobb- Douglas functions of labour, capital, and intermedite inputs

Canada Mexico US

2-level utility functions 1-LES between composite goods 2-CES between national varieties

1 consumer in each country

CES value-added functions of labour and capital inputs; gross output is a CET function of domestic sales and exports to each region; fixed coefficient use of intermediate goods

26 production sectors in each country

Sobarzo

Mexico

2-level utility functions 1-Cobb-Douglas between composite goods 2-CES between national varieties (Mexico, other North America, and ROW)

1 consumer

Cobb-Douglas value-added functions of labour and capital inputs; fixed coefficient use of intermediate goods; fixed coefficients between intermediate inputs and value added

27 production sectors in Mexico

Young and Romero

Mexico

Cobb-Douglas preferences between consumption goods sectors; imports and domestic products are perfect substitutes

1 consumer

2-level cost functions 1-traslog functions of prices of labour and composite intermediate goods and rental fee on composite capital 2-composite capital price is a translog function of capital goods prices; composite intermediate price is a Cobb-Douglas function of intermediate goods prices

9 consumption goods and 3 capital goods sectors

228

Roland-Holst, Reinert and Shiells

Table 1. Summary of demand and production structure (cont.) Model

Countries

Sector-focused models

Demand side

Production side

Demand functions

Disaggregation

Production functions

Disaggregation

Burfisher, Robinson and Thierfelder

Mexico US

2-level utility functions 1-Cobb-Douglas between composite goods 2-Almost ideal demand system between national varieties (Mexico, US, and ROW)

1 consumer in each country

CES value-added functions of primary factors (4 labour types, capital, and agricultural land); domestic outputs is a CET function of domestic sales and exports to each region; fixed coefficient use of intermediate goods

11 production sectors (4 farm and 1 food processing) in each country

Levy and van Wijnbergen

Mexico

3-level utility functions 1-Cobb-Douglas between 3 goods (industry, services, and a composite agricultural good) 2-CES between 5 rural goods 3-CES between raw corn and tortillas.

6 household types

Cobb-Douglas value-added functions of primary factors (2 labour types, 2 capital types, and 3 land types); mostly fixed coefficient use of intermediate goods; exogenous rate of Hicks- neutral technical change

7 production sectors (5 agricultural)

1 consumer in each region (Canada, Mexico, US, and ROW)

Final assembly of autos: Multinational producers; fixed cost plus constant marginal cost.

4 production sectors (final assembly, parts, engines, and other goods) in each of 4 regions (Canada, Mexico, US, and ROW)

Imports and domestic goods are perfect substitutes

229

Lopez-de- Silanes, Markusen, and Rutherford

Canada Mexico US ROW

3-level utility functions 1-Cobb-Douglas between composite autos and a composite of other goods 2-CES between composites of autos produced by different firm types (US and ROW ownership) 3-CES between varieties of autos produced by different firms of each type.

Auto parts: 2-level CES 1-CES production function between national composites 2-CES production function between firm varieties; fixed cost plus constant marginal cost. Engines: One engine per car; engines are produced using capital and labour at constant cost.

Table 1. Summary of demand and production structure (cont.) Model

Countries

Sector-focused models Trela and Whalley: Steel

Trela and Whalley: Textiles and Apparel

Canada Mexico US ROA

Canada Mexico US ROA

Demand side

Production side

230

Demand Functions

Disaggregation

Production functions

Disaggregation

CES utility function between 3 goods (steel producing sector, steel consuming sector, and all other goods); domestic production and imports from each of 3 regions (Canada, Mexico, and ROA, a 22-country aggregate of other non-VRA countries) are perfect substitutes.

1 consumer for each region (Canada, Mexico, US and ROA)

Domestic production: CES value-added functions of labour and capital inputs; fixed coeffficients between intermediate inputs and value added

3 production sectors (steel producing, steel consuming, and other) in each region.

3-level utility functions 1-CES between a composite of all textile and apparel products and a composite of all other goods 2-CES between a composite of all textile goods and a composite of all apparel goods 3-CES between MFA-restricted and unrestricted goods.

1 consumer for each region (Canada, Mexico, US, and ROA)

Foreign production: CET function between output of 3 production sectors 3-level production possibilities frontiers 1-CET between a composite of all textile and apparel products and a composite of all other goods 2-CET between a composite of all textile goods and a compsite of all apparel goods 3-CET between MFA-restricted and unrestricted goods.

Domestic production and imports from each of 3 regions (Canada, Mexico, and ROA, a 33-country aggregate of other supplying MFA countries) are perfect substitutes.

Source: Joseph P. Francois and Clinton R. Shiells, Modeling Trade Policy , Cambridge University Press, 1994, pp. 16-19, Table 1.1.

4 production sectors in each region

Table 2. Summary of market structure and model closure Model

Multisector models

Market structure Firm behavior

Model closure Entry and exit

Balance of payments

Capital market

Labour market

231

Brown

Monopolistic competition with Berthand pricing

Free

Walras’ Law implies the current account is in equilibrum

Fixed aggregate capital stock: perfectly mobile between sectors; immobile internationally

Fixed aggregate labour supply; perfectly mobile between sectors; immobile internationally

Cox

Imperfectly competitive industries use a weighted average of Eastman-Stykolt and monopolisticallcompetitive pricing

Free

Walras’ Law implies the current account is in equilibrum

Fixed rental rate on capital services; perfectly mobile between sectors and internationally

Fixed aggregate labour supply; perfectly mobile between sectors; immobile internationally

RolandHolst,Reinert and Shiells

Perfect competition, average cost pricing due to constable markets, or Cournot markup pricing

Free under Cournot

Trade balance is equal to net foreign borrowing

Fixed aggregate capital stock; perfectly mobile between sectors; immobile internationally

Fixed wage rate; perfectly mobile between sectors; immobile internationally

Sobarzo

Imperfectly competitive industries use monopolisticall ycompetitive pricing

Free

Variable trade balance and fixed real exchange rate

Fixed world rental rate on capital services: perfectly mobile between sectors and internationally

Fixed aggregate labour supply; perfectly mobile between sectors; immobile internationally

Young and Romero

Perfect competition

Free

Walras’ Law implies the current account is in equilibrium

Exogenous world prices of machines and vehicles, use transferable between sectors and internationally; exogenous supply of building, use transferable between sectors but not internationally

Exogenous aggregate labour supply; mobile between sectors within specified bounds; immobile internationally

Table 2. Summary of market structure and model closure (cont.) Model

Sector-focused models

Market structure Firm behavior

Model closure Entry and exit

Balance of payments

Capital market

Labour market

Burfisher, Robinson, and Thierfelder

Perfect competition

Free

Current account balance is equal to net foreign savings

Fixed aggregate capital stock; perfectly mobile between sectors; immobile internationally

Fixed aggregate supply of each labour type; perfectly mobile between sectors; labour markets are segmented and linked through migration

Levy and van Wijnbergen

Perfect competition

Free

Walras’ law implies that trade is balanced in equilibrium

Exogenous rate of growth of sectorspecific capital; immobile internationally

Exogenous rate of growth of rural and urban labour types; labour markets are segmented and linked through migration

Lopez-deSilanes, Markusssen and Rutherford

Final assembly: Multinational firms employ a Cournot markup rule

Free

Walras’ law implies that trade is balanced in equilibrium

Capital used in parts and other goods differs from capital used in auto assembly and engine production

“Labour” is a composite of all factors that are perfectly mobile between sectors and immobile internationally

232 Trela and Whalley:

Parts: Monopolistic competition with Berthand pricing

Capital is perfectly mobile between parts and other goods but immobile internationally

Engines: Intra-firm production under constant cost

Capital used in auto assembly and engines has some mobility between sectors and internationally

Perfect competition

Free

Trade balance is Zero and steel exports to the US are quota- constrained

Fixed aggregate capital stock; perfectly mobile between sectors; immobile internationally

Fixed aggregate labour supply; partially mobile between sectors, with external adjustment costs of moving between sectors; immobile internationally

Perfect competition

Free

Trade balance is zero and some categories of textile and apparel trade are quotaconstrained

Capital market is implicit only, due to use of the CET production possibilities frontier

Labour market is implicit only, due to use of the CET production possibilities frontier

Steel

Trela and Whalley: Textiles and Apparel

Source: Joseph P. Francois and Clinton R. Shiells, Modeling Trade Policy , Cambridge University Press, 1994, pp. 22-24, Table 1.2.

Table 3. Sources of data and elasticities in the models Model

Year(s) replicated

Extraneous use of elasticities

Production data

Demand data

Trade data

Multisector model

233

Cox

1981

Elasticities of substitution between national varieties (4 specifications for sensitivity analysis), and inverse scale elasticities

Statistics Canada input- output tables, Statistics Canada data on number of firms, capital stocks, and non-capital variable costs

Input-output tables

Input- output tables

Roland-Holst, Reinert, and Shiells

1988

Elasticities of substitution between national varieties (own estimates), elasticities of substitution between capital and labour, elasticities of transformation between domestic supply and exports, and cost disadvantage ratios

Social accounting matrix (own construction)

Social accounting matrix

Social accounting matrix

Sobarzo

1985

Elasticities of substitution between national varieties, export demand elasticities, and inverse scale elasticities

Social accounting matrix (own construction)

Social accounting matrix

Social accounting matrix

Young and Romero

1992-2002

Translog cost function parameters (own estimates)

Instituto Nacional de Estadística, Geografía e Informática (INEGI) National Accounts and input-output tables, Banco de Mexico capital stocks

National Accounts

National Accounts

Burfisher, Robinson, and Thierfelder

1988 (Mexico)

Almost ideal import demand system calibrated based on expenditure and substitution elasticities, elasticities of transformation between domestic supply and exports, export demand elasticities for the rest of the world

Social Accounting matrices (own construction)

Social Accounting matrices

Social Accounting matrices

Levy and van Wijnbergen

1991-2000

Elasticity of substitution between corn and basic grain in livestock production, aggregate supply elasticity for maize, elasticities of substitution between 5 rural goods in household demand, elasticities of substitution between raw corn and tortillas in household demand, rural-urban migration elasticities, inter-temporal substitution elasticity

Social Accounting matrix (own construction)

1984 IncomeExpenditure Survey and social accounting matrix

Social accounting matrix

Sector-focused models

1987 (US)

Table 3. Sources of Data and Elasticities in the Models (cont.) Model

Year(s) replicated

Extraneous use of elasticities

Production data

Demand data

Trade data

Sector-focused models 1989

Elasticities of labour supply to firms, elasticities of scale (from engineering studies)

Labour shares of value-added, production of autos, engines, and parts by region, and GDP

Consumer auto price indices

Matrix of trade flows between regions in autos, engines, and parts

Trela and Whalley: Steel

1986

Elasticities of substitution between labour and capital are based on a literature search, CES and CET parameters are set to 1 (6 specifications for sensitivity analysis)

UN, US Department of Commerce, UNIDO, USITC

Consumption is determined as a residual, from production and net trade data

US Department of Commerce, USITC

Trela and Whaley: Textiles and Apparel

1986

Elasticities of substitution and transformation at each of 3 levels (7 specifications for sensitivity analysis)

UN, US Department of Commerce, World Bank

Consumption is determined as a residual, from production and net trade data

US Department of Commerce and Canadian Department of External Affairs

234

Lopez-deSilanes, Markussen, and Rutherford

Note: Brown’s model is not included in this table because it is not calibrated to the actual world economy. Source: Joseph P. Francois and Clinton R. Shiells, Modeling Trade Policy , Cambridge University Press, 1994, pp. 25-27, Table 1.3.

ECONOMIC EFFECTS OF NAFTA: EMPLOYMENT AND MIGRATION MODELLING 1 RESULTS

by Raúl Hinojosa-Ojeda, NAID Center, UCLA, Robert McCleery, Claremont McKenna College and Fernando De Paolis, NAID Center, UCLA

1.

Introduction

Computable general equilibrium (CGE) models are used to highlight the complex interactions among economic variables. A CGE model can capture the upstream and downstream effects of a change in production levels or prices. For instance, a change in trade policy that makes the production of light manufactures in Mexico more profitable (raising the price Mexican producers receive on shipments to the United States, for instance, by cutting tariffs) will generate demand for goods used intensively as intermediates in such production. Conversely, as the price of light manufactures rises in response to the higher US demand, producers in Mexico that use light manufactures as inputs into further production will face higher costs. CGE models can also capture changes in the demand and returns to factors of production. As Mexican production of light manufactures expands, labour demand rises. The increase in demand causes a rise in wages, and as more labour (particularly unskilled labour) relative to capital is bid away from other sectors, the return to capital may fall. The change in the allocation of income between labour and capital may have further ripple effects through the economy if the structure of demand is different for the different groups. Thus a CGE model presents a comprehensive and consistent answer to the question: What happens to the economy or economies in question if a certain policy change is implemented? The CALNAFTA model places an emphasis on the trade links and concomitant flows of labour (in the form of migration) between Mexico and the United States, with California treated as a separate region. Simulations are used to demonstrate: i) the effects of free-trade and its relative importance compared with migration; ii) the implications of restrictive immigration policy, such as California’s 2 Proposition 187 and iii) the effects of increased foreign investment in Mexico and the effect of adopting greater equalisation of labour rights for immigrants.

235

2.

Model description

The CALNAFTA model is a revised version of the model used in Hinojosa-Ojeda et al., 1992 (see Appendix for a complete description). There are four regions: California, Mexico, rest of United States, and rest of world. Production is accomplished with capital, land, unskilled labour, and skilled labour, and allocated to one of eleven productive sectors. Unskilled labour is further segmented into rural unskilled (used in the production of corn and feed grains, other program crops, fruits and vegetables, and other agricultural products), and urban unskilled (used in producing output in the other seven sectors of each economy: oil, services, food processing, other light manufacturing, intermediate goods, consumer durables, and capital goods). Skilled labour is subdivided into unionised and non-union labour. Land represents agricultural land, and is used only in the production of agricultural output. Capital is used in production in all sectors. As in most CGE models, factors flow freely within countries in response to any differential in rates of return, such that in equilibrium the rates of return are equal. Clearly, migration (whether domestic or international) is not a costless process. We thus build equilibrium differentials into the model, and migration only takes place when the benefits of migration, in terms of higher wages, exceed the costs. In an equilibrium without migration, the US urban unskilled wage would exceed the US rural unskilled wage, which would in turn dwarf the Mexican urban unskilled wage, which itself exceeds the Mexican rural unskilled wage. Thus the original source of all migration can be thought of as rural Mexico. In that case, one might ask why all migrants do not head for the US urban unskilled labour market? In fact, given that wages in California are generally higher than in the United States as a whole, why don’t all migrants go to California? The lure of the greater wage differential between Mexico and the United States is partially offset by the greater cost of cross-border migration and/or of the perceived social costs to having to live as an undocumented migrant. These relative benefit and cost functions are used in the model to test various assumptions of how immigrants might behave towards different policy approaches, such as Proposition 187 in California and increased wage growth in Mexico. In short, the model is parameterised on the initial data, with the initial levels of immigration and initial wage differentials taken to represent an equilibrium. This equilibrium differential may represent a number of different variables, including the cost (in both money and income foregone) of migration, the preference of Mexicans for Mexico (in other words working in the United States must confer a substantial wage premium over Mexico to induce migration) and discrimination based on undocumented status.3 Another unique feature of this model is the detailed and explicit treatment of not just tariffs and other trade barriers but the myriad of complex farm support programs in each country. Mexico, for example, used direct price supports for corn and program crops, in addition to credit subsidies, input (fertiliser, etc.) subsidies, and consumption subsidies, often with significant variation across crops. The United States has various export subsidies, initial deficit payments, and numerous other price, quantity, or land use controls. The cost and impact of each are calculated and incorporated into the model structure, and each is maintained as a separate policy variable. Trade is motivated by the Almost Ideal Demand System (AIDS), a more robust functional form than traditional formulations. Along with the level of tariff protection, other direct and indirect tax rates are given as policy variables.

236

Scenarios and results The CGE modelling framework outlined above is designed to analyse the impacts of alternative scenarios of trade, migration and capital flows between Mexico, California and the rest of the United States. The scenarios which were generated for this paper are summarised in Table 1, and the impacts on real GDP, net migration, and factor returns of each scenario compared with various baseline scenarios are shown in Tables 2, 3 and 4. Below, a brief outline of the scenarios is provided and is followed by discussion of the results. Scenario 1 sets all tariffs and subsidies on trade to zero and is compared to a baseline scenario which assumes the system of tariffs and quotas existing in the late 1980s and net migration based on current flows. The scenario therefore represents the hypothetical scenario of free-trade and can therefore be seen as approximating the impact of NAFTA and Mexican liberalisation policies. − Scenario 1a calculates the impact of NAFTA and liberalisation whilst maintaining current migration levels, while scenario 1b calculates the same effect allowing net migration to change from current levels in response to the imposition of free-trade. In a sense, scenario 1b can be regarded as the current status quo and is used as the baseline in subsequent scenarios. Scenario 2 simulates alternative assumptions about the full implementation of anti-immigration legislation (e.g., Proposition 187) and is compared to a baseline scenario of NAFTA, liberalisation and migration (i.e. scenario 1b). − Scenario 2a) calculates the impact of the departure of all undocumented workers from California. − Scenario 2b calculates the impact of the departure of all undocumented workers from the entire United States. − Scenario 2c reports the possible impact if anti-immigration legislation does not result in the departure of undocumented immigrants but rather in a worsening of their employment conditions. Scenarios 3 and 4 report impacts of increased investment and increases in the reservation wage of immigrant workers. As in scenario 2, the results in Tables 2, 3 and 4 are reported as changes in relation to scenario 1b. − Scenarios 3a, 3b and 3c present the impacts of NAFTA combined with increases in restof-world investments in Mexico of 5 per cent, 10 per cent and 15 per cent respectively. − Scenarios 4a, 4b and 4c are the same as scenarios 3a, 3b and 3c except that migrants reservation wage has been increased. The motivation for this is discussed below. Scenario 1 illustrates that, compared with trade, migration is empirically more important in the economic link between the United States and Mexico. Scenario 1a shows the hypothetical effect of NAFTA in a world without immigration. By contrasting these results with those of scenario 1b, NAFTA with migration, one can clearly see the differential impact of the flow of labour in comparison with the flow of goods. Table 2 shows the impact on real GDP in the rest of the United States is more than two hundred times greater and nearly one hundred times greater for California.4 237

The quantity of additional migration occurring in this scenario is fairly large, with nearly 400 thousand migrants moving from rural Mexico to the rest of the United States, nearly 150 thousand coming to California, and more than 300 thousand migrating from rural to urban Mexico (Table 3). In both cases of international migration, migrants enter the rural and urban unskilled sectors in roughly the same proportions as the initial labour force, meaning that the vast majority enter the urban labour market. Without migration, changes in factor returns in California due to free-trade are relatively minor. The main benefits accrue to agriculture in the rest of the United States, with a rise of nearly 1 per cent in the returns to rural labour (i.e. agricultural unskilled wages) and of nearly 0.5 per cent in the return to land (i.e. farmers incomes). Mexico, by contrast, would experience a sharp contraction in returns to these same two categories of about 5 per cent, combined with a modest increase in the return to capital of 0.5 per cent. With the migration option, wages of rural labour in Mexico rise slightly instead of declining by nearly 5 per cent, which would occur without migration. The inflow of workers depresses wages of unskilled workers in California and the rest of the United States. Scenarios 2a and 2b simulate effective migration restrictions in California and the entire United States respectively, on top of the impact of NAFTA in scenario 1b. One way to interpret scenario 2a is to assume that the proponents of Proposition 187 get what they say they want and all undocumented migrants go back to Mexico and no more come to California. Thus not only does NAFTA not generate the additional 150 000 migrants we posit in scenario 1b, but the entire stock of undocumented migrants currently in California, estimated at 200 000 rural and 550 000 urban workers, return to Mexico. This turns out to hurt the economy of California, the tax base, and thus the fiscal health of the state. CGE models of course assume instantaneous and costless adjustments, with unskilled wages rapidly rising (43 per cent increase in returns for rural labour and a more modest 6 per cent for the unskilled in urban areas) to induce natives to take some of the vacated immigrant jobs. Even without adjustment costs, the wages required by US residents would still mean large price rises and a collapse in State GDP of 2.5 per cent (about $25 billion). This economic crisis would hardly improve the fiscal status of the state, swelling deficits and/or reducing services to natives. Given the distribution of wealth, the fall in returns to capital of nearly 2 per cent would most likely affect the incomes of the more highly skilled and farmers would see falls in the return to land of 11.5 per cent. The shocks are more muted for the rest of the United States, which keeps its migrants but loses some labour to California in response to the higher wages there. Perfect capital mobility between California and the rest of the United States ensures that capital income falls by the same 1.8 per cent across the nation. Strongly linked and highly elastic rural labour causes a similar increase in factor returns to rural labour. However less integrated urban unskilled labour markets see a much smaller wage rise of just over 2 per cent. Mexico experiences significant adjustments as well, with wages of urban elites (skilled workers and capital owners) all rising by about 2.5 per cent, and both rural and urban unskilled wages falling by more than 4 per cent. Scenario 2b is similar to 2a, except that not only are all migrants expelled from California, a similar quantity (another three-quarters of a million workers) are deported from the rest of the United States as well. In this scenario we would expect large contractions in output and wage adjustments in all three regions. That is precisely what we see. Rural wages rise even more, up about 85 per cent over the base. Urban unskilled wages rise by an additional 6 per cent in California and 2 per cent in the rest of the United States over scenario 2a, for total increases of 11.6 per cent and 4 per cent over the

238

base. With the return to agricultural land tumbling by over 23 per cent, land values in rural California would collapse, with serious implications for the tax base and state and local government revenues. It goes without saying that Mexico would face an enormous adjustment problem if an additional 1.5 million workers suddenly returned to the cities and the campo. Over 80 per cent of these workers (1.2 million) end up in the cities, to the delight of capitalists (earnings up 4.5 per cent), union labour (up 4.5 per cent) and profession labour (up 4.7 per cent). Wages for unskilled urban workers are driven down drastically (5.5 per cent), even more than rural wages (down 5.1 per cent). Thus, if Proposition 187 is potentially as effective as some believe, California would be locked into a permanently lower path of output and income growth, especially if one considers the additional influence of declining population growth rate in the United States. Mexican real wages for unskilled workers, in both the poor rural sectors and politically important cities, would suffer a significant decline and acutely exacerbate social and political tensions in Mexico. Scenario 2c simulates the possibility that Proposition 187 would not deter undocumented immigration and that the undocumented population in California would become increasingly marginalised. There is general agreement among researchers that this will be the more likely scenario since immigrants are not drawn to the United States due to a presumed availability of social services (Alarcón, 1995; Hinojosa and Schey, 1996). The employers of undocumented workers, traditionally strong supporters of Governor Pete Wilson, must also believe this to be the outcome or else they would have vocally opposed Proposition 187 in the same way they have opposed efforts to limit their access to undocumented labour in the past. Scenario 2c simulates that a possible impact of anti-immigrant legislation like Proposition 187 would be that wages for undocumented immigrants would continue to fall, as they did after the passage of a weak employer sanctions law in 1986 (IRCA). The results in Table 2 show that this scenario would have a positive impact on California GDP. Table 4 shows that scenario 2c would also benefit the rate of profit of employers as well as the consumers of low wage goods. The ironic result is that since wages for undocumented workers fall, the demand for them will rise, resulting in even greater dependence on low wage labour by California employers. In fact, as Table 3 shows, immigration would rise in this scenario in contradiction to the aims of Proposition 187. While this may seem a positive short term scenario for some sections of Californian society, there are some negative implications of Proposition 187 which are not captured by the CGE model. First, implementation of Proposition 187 may lead to increased, rather than reduced health-care costs in the long-run. Under Proposition 187 pregnant women will be denied prenatal care and children who have been abused or neglected will not receive social services. Immigrant adults and children with serious illnesses will be denied needed health care. Others, fearing deportation, will be deterred from seeking care when they need it. Illnesses susceptible to early treatment will deteriorate into life-threatening emergencies and communicable diseases will spread, endangering everyone’s health. Medical experts point out that the cost of health care may skyrocket as health workers are mandated to provide expensive emergency services that could have been avoided by far less costly early intervention. However, Proposition 187 would in effect deny treatment when it is least costly and most effective. As a result, the Chief of Staff of the Los Angeles County Medical Center, the largest public hospital in California, has stated that while implementing Proposition 187 may save the state about $9 million annually, the medical costs in emergency care and treatment of US citizens with communicable diseases will rise by $47 million, leaving net costs of over $38 million.

239

Second, Proposition 187 could lead to lower levels of human capital and additional public expenditure in the longer-term. One section of Proposition 187 bars undocumented children from attending public schools. Testimonials from teachers and educators throughout the country speak to the irreparable harm caused to children banned from attending school. As most undocumented children eventually become lawful permanent residents, and later US citizens, the long-term costs of implementing Proposition 187 on California are enormous. They will be largely unemployable, rely extensively on public support programs, and be far more likely to tax the criminal justice system. People who could have produced goods and services in California, adding to the production of wealth, will instead have fewer skills and contributions to offer. Finally, there are implications for discrimination. If fully implemented, Proposition 187 mandates a vast state network of immigration informants comprising employees of public schools, medical clinics, social service agencies, and state and local law enforcement agencies, all of whom will be required to identify suspected deportable aliens and report them not only to the INS, but to high state officials as well. Those required to make decisions of “suspected” undocumented immigration status have no training to make such determinations. There can be no doubt but that race and national origin will often play a role in the classification of persons suspected of not possessing lawful immigration status, resulting in greater discrimination and higher social tension between those who “look like immigrants” and those who “look like they belong here.” 3.

Scenarios for alternative policies approaches

It thus appears that full implementation of Proposition 187 may in fact lead to increased migration, and although short-run effects may be beneficial for some, there are negative implications in the longer-run. Given this somewhat contradictory outcome, it is essential to present an alternative vision for a more economically sound and humane immigration policy capable of mobilising broad political support on both sides of the border. The same bi-national framework which exposes the contradictions of Proposition 187 can be a guide for constructing a policy approach that can improve wages, working conditions, human rights, and economic growth in both the United States and Mexico. Scenarios 3 and 4 report the possible impacts of different elements of an alternative policy approach to integration and immigration that would combine NAFTA with increased capital investments flows to Mexico and the introduction of a more liberal immigration regime that would guarantee equal labour rights to Mexican immigrants in the United States. Scenario 3 presents the possibility that, as a result of NAFTA in combination with other regional mechanisms to increase investment, Mexico could experience a boom in capital inflows. What impact might substantial capital inflows have on migration flows? Will the “immigration problems” go away as capital inflows fuel more rapid growth of employment and incomes in Mexico? To some extent, that appears to be true. Even five per cent increase of total investment reverses the flow of new immigration due to NAFTA, from a net inflow of nearly 650 000 to a net outflow of almost 165 000, from the United States as a whole (Table 3). This modest outflow of migrants has a small impact on GDP, reversing the 0.1 per cent gains from NAFTA in scenario 1b to small losses. As the amount of investment is increased to 10 per cent of total investment in scenario 4b, the impacts on GDP and wages are more significant. GDP losses for California and the rest of the United States are about 0.25 per cent each. Although unskilled wages in the United States rise by nearly 2 per cent,

240

wages for the two categories of skilled labour fall slightly. The large flows of returning migrants cause the same problems for Mexico as in scenarios 2a and 2b, but in greater magnitudes. 5 Scenario 4 extends this scenario to include changes in the regime of labour rights of Mexican immigrants in the United States. The scenario we are attempting to simulate here is one where all immigrants in the United States, regardless of their immigration status, would be allowed the same ability to claim the full basic labour rights accorded to US citizens, including the right to organise and strike, protected legal standing to complain about working conditions, minimum wage, etc. One possible mechanism for this equalisation of immigrant/non-immigrant labour market status could be a special North American Work Visa. We assume that under such a new regime of transnationally protected worker rights, Mexican immigrants would be in a position to demand better compensation from their employment or seek new employment, thereby closing the “10-15 per cent” difference in wages in the same way that recently legalised immigrants were able to achieve after the Immigration Reform and Control Act (IRCA) . In scenario 4, we simulate this new immigrant rights regime by increasing the “reservation wage” that immigrants would use to compare their wages in Mexico with a wage in the United States in order to determine whether to migrate or not. The effect is to require the clearing of the US labour market for immigrants at a higher wage level in order to induce migration, an effect which would not only increase wages in the United States but reduce the demand for Mexican immigrants. The results of scenario 4 suggest some interesting dynamics which point to a viable argument as to why such a regime of more liberal, yet labour-rights-based migration could be politically acceptable in the United States. First, Mexican migration would actually fall in the rights-based scenario compared to the status quo, even without any capital transfers to Mexico. Migrants would become more expensive and less of them would be used by capital. Second, if this new immigration regime were accompanied by an increase of capital transfers to Mexico as in scenario 3, reductions in migration would actually be less expensive to achieve. Less investment capital would be required to raise Mexican wages to a level where potential migrants would decide to remain in Mexico. Third, income inequality would improve in both the United States and, to a lesser degree, Mexico under this scenario compared to all others. Finally, scenario 4 could generate much greater reduction in levels of migration compared to the Proposition 187 approach without dramatic reduction in US economic activity and without a deterioration of a regime of rights in the United States, rather than their significant improvement. 4.

Conclusion

The results of these CGE simulations suggest a number of stylised facts. First, migration appears to be empirically far more important compared with trade in the economic link between the United States and Mexico, suggesting that the implications of immigration policy carry far more weight than policies directed at trade. Second, the simulations suggest that free-trade by itself is not sufficient to generate the type of development needed to close the income inequality gaps in North America and that more significant impacts are to be found in increased foreign investment. This suggests a degree of impotence in the effects of trade-liberalisation and the need for it to be combined with greater depth in agreements which create greater freedom of capital movement. Finally, the results of the CGE model and consideration of longer-term effects suggests that policy measures of the type included in Proposition 187 have the potential of being counterproductive both in terms of reducing undocumented migration and in terms of achieving economic growth.

241

Further simulations have shown that more successful outcomes are generated by a policy option centred on furthering long-term agreements on trade, capital, and labour migration relations, as well as creating a shared North American agreement on labour and human rights, all crafted with the clear goal of sustainable and equitable development. As the results of this CGE model suggest, if moderately high growth in Mexican investment has a positive impact on labour market outcomes for workers, the trend towards North American wage convergence would be powerful enough that potential migrants would begin to choose to stay in Mexico, even if US wages for immigrants also rise. Sustained growth in Mexico would also generate a large net increase in high wage US export jobs, as well as help close the gap between high and low wage labour in the United States. The static nature of this CGE model inevitably leaves some issues unexplored. In particular, demographic trends raise important issues. Since the Mexican birth rate declined precipitously over the last 25 years, the number of new entrants into the labour force will also fall over the next decade. This suggests an increased importance for closing the poverty gap in the migrant sending regions of Mexico, thus preventing further loss of growth in the stock of human capital. Meanwhile, the US population is ageing and perhaps prompting the need for new immigrants to sustain growth and care for an older population, particularly in California (David Hayes-Bautista, 1992). In conclusion, the top priority is thus not stopping US immigration but rather to improve the human capital and social conditions of those people who will have to carry California and the rest of the United States into the next century, exactly the opposite of what measures such as Proposition 187 will do. Economic and demographic realities between the United States and Mexico suggest that North America has the opportunity to adopt a new form of economic integration which could, if managed properly, produce a pattern of upward growth and convergence in productivity and income levels across North America.

242

Appendix THE CALNAFTA CGE MODEL

The CALNAFTA CGE model is an 11-sector, four-country, computable general equilibrium (CGE) model composed of three single-country CGE models (of Mexico, California, and the rest of the United States), linked through trade and migration flows, plus a set of export-demand and import-supply equations to represent the rest of the world. The model is an extension of an earlier CGE model of the United States developed at the Department of Agriculture (USDA). A three-country (United States, Mexico, and rest of world) extension of the USDA model was initially developed by Hinojosa and Robinson (1991), who also introduced the use of domestic and international migration functions. That model was extended by Robinson et al. (1991) to include an explicit modelling of domestic farm programs in both the United States and Mexico. The CALNAFTA model is very close in structure to the model by Robinson et al., but adds an additional ‘country’, California. The core model follows the standard theoretical specification of trade-focused CGE models. Each sector (four of which cover agricultural products) produces a composite commodity that can be transformed according to a constant elasticity of transformation (CET) function into a commodity sold on the domestic market or into an export. Output is produced according to a CES production function in primary factors, and fixed input-output coefficients for intermediate inputs. The model simulates a market economy, with prices and quantities assumed to adjust to clear markets. Supply in the market for each sector is generated by six primary factors of production: four types of labour, capital, and agricultural land. The aggregate quantity of each is determined exogenously. The four labour types are rural, urban unskilled, urban skilled, and professional. In factor markets, full employment for all labour categories is assumed. The model can incorporate different assumptions about factor mobility, and various labour markets are linked by migration flows. The results should be seen as reflecting adjustment in the long-run, with capital able to switch between sectors. Domestic demand is comprised of consumption, intermediate demand, government expenditure, and investment. Consumer demand is based on Cobb-Douglas utility functions, implying fixed expenditure shares. Households pay income taxes to the government and save a fixed proportion of their income. Intermediate demand is given by fixed input-output coefficients. Real government demand and real investment are fixed exogenously. Imports and exports are incorporated through the use of sectoral export supply and import demand functions for each country, and solves for a set of world prices that achieves equilibrium in world commodity markets. At the sectoral level, in each country, demanders differentiate goods by country of origin and exporters differentiate goods by country of destination. In the commodity markets, California is treated as a separate country, with relative prices that may differ from the rest of the United States. Three key macro balances are maintained in each country model: the government deficit, aggregate investment and savings, and the balance of trade. Government saving is the difference between revenue and spending, with real spending fixed exogenously but revenue depending on a variety of tax instruments. The government deficit is therefore determined endogenously. Real investment is set exogenously, and aggregate private savings is 6 determined residually to achieve the nominal savings-investment balance. The balance of trade for each country (and hence foreign savings) is set exogenously, valued in world prices. In the case of California, the balance of

243

trade for California and the rest of the Untied States together is assumed fixed, but it is not fixed for each separately. In determining equilibrium prices and quantities, each country model solves for relative domestic prices and factor returns which clear the factor and product markets, and for an equilibrium real exchange rate given the exogenous aggregate balance of trade in each country. The model determines two equilibrium real exchange rates, one each for the United States and Mexico, which are measured with respect to the rest of the world. The cross rate (United States to Mexico) is implicitly determined by an arbitrage condition. Of course, the United States-California exchange rate is fixed at one, so the California-Mexico exchange rate is the same as the United States-Mexico rate. The GDP deflator defines the numeraire in each country model, and the currency of the rest of the world defines the international numeraire. Four types of elasticity parameters are used in the model. The production specification requires sectoral elasticities of substitution among primary factors. The CET export supply functions require elasticities of transformation between goods sold on the home and export markets. The Almost Ideal Demand System import demand functions require sectoral income elasticities and substitution elasticities for home goods and for goods from each import source. We have drawn on estimates and guesstimates from various studies, including Hinojosa and Robinson (1991). In lieu of econometric estimation, sensitivity analysis was carried out to check for the robustness of the model results using alternative elasticity parameters. The fully parameterised model is applied to data relating to the late 1980s. The base year for Mexico is mostly 7 1988. The United States and California use a 1987 base year because of the severe contraction of agricultural output following the 1988 drought. All bilateral trade flows are from 1988. Tariffs and tariff equivalents of quotas are 1988 trade-weighted rates. Therefore where simulations incorporate tariffs and quotas, they are essentially describing a pre-NAFTA scenario.

1.

Import demand equations

The standard approach in trade-focused CGE models is to assume that domestic and imported goods are 8 imperfect substitutes and to specify a constant elasticity of substitution (CES) import aggregation function. In the case of a multi-country model, the function aggregates imports from all countries of origin. In the simplest case, the CES function is extended to include goods from many countries, with the substitution elasticity assumed 9 to be the same for all pairwise comparisons of goods by country of origin. The first-order conditions define import demand as a function of relative prices and the elasticity parameter. The use of CES functions in multi-country Armington trade models has led to empirical problems due to the restrictive nature of the CES functions. Instead of the CES import aggregation function, we use import demand equations based on the Almost Ideal Demand System (or AIDS). The AIDS function is a flexible functional form in that it can generate arbitrary values of substitution elasticities at a given set of prices, and also allows expenditure elasticities different from one. Various restrictions on the parameters of the model are required to have the system satisfy standard properties of expenditure functions such as symmetry, homogeneity, adding up, and local concavity. We calibrated the parameters for the CALNAFTA model by starting from a set of expenditure elasticities and substitution elasticities for each sector in each country. Higher substitution elasticities for goods traded between the United States and California are specified, and lower elasticities are 10 specified for trade with Mexico and the rest of the world. Because each sector in the model consists of 11 composite goods aggregate prices need to be used. These were generated by a translog index.

2.

Migration

The CALNAFTA model specifies a number of labour migration flows: rural Mexico to rural United States and rural California; urban unskilled Mexico to urban unskilled United States and California; internal migration

244

within Mexico from rural to unskilled urban labour markets; and migration between rural and urban unskilled between the United States and California. Migration is assumed to be a function of wage differentials between the linked labour markets. In equilibrium, migration levels are determined which maintain a specified ratio of real wages for each labour category in the countries, measured in a common currency, and a specified ratio of real wages between the rural and unskilled urban markets in Mexico. An implication of this specification of Mexican-US-California migration flows is that real wages measured in a common currency are equated, but they can grow at different rates measured in the domestic currency. It is therefore possible to observe migrants between Mexico and the United States/California moving from a labour market where real wages are rising to one in which they are falling in domestic currency terms. The issue is in the specification of what motivates migrants. For example, if they are motivated by the desire to accumulate savings which they intend to repatriate, then migration will be sensitive to the exchange rate. On the other hand, if they are motivated by observations on relative changes within the two economies then migration could be expected to be insensitive to the exchange rate. The model probably overstates the sensitivity of migration to the exchange rate, generating a backward flow of migrants into Mexico when the Mexican peso appreciates. Migration flows generated by the CALNAFTA model refer to changes in migration from a base of zero. They should be seen as additional migration flows due to the policy change, adding to current flows. Note also that these additional flows are those generated by economic incentive (i.e. workers or heads of households) and therefore do not include associated flows of dependants.

3.

Agricultural programs

Five of the sectors in the model are agricultural, four are groups of primary products, and there is one intermediate food-processing sector. Particular attention has been given to the complex set of trade policies and domestic agricultural programs, as these policies distort production, consumption and trade, and require significant fiscal expenditures in both Mexico and the United States. Mexican agricultural program expenditure in 1988, totalling $1.6 billion, represented over one-half of total national subsidy expenditure, and equalled 12 almost one per cent of GDP. In the United States, deficiency payments and expenditures on the export enhancement program (EEP) in 1987 totalled $11.5 billion, or one per cent of government spending and 10 per cent of the fiscal deficit. Note that the policies described in the model relate to those existing at the end of the 1980s. The agricultural policies are modelled either as price wedges, which affect output decisions, or lump-sum income transfers. The wedges and transfers are either specified exogenously or determined endogenously, based on the 13 institutional characteristics of the program being modelled. California is assumed to have the same agricultural policies as the rest of the United States. Note that because the model assumes there is imperfect substitution between domestic and foreign goods, the link between trade and domestic prices is weaker than a model where all goods are perfect substitutes. Six Mexican policies are modelled. In the four agricultural sectors, these are input subsidies, tariffs, and quotas. In the food processing sector, we model direct subsidies and price subsidies, in addition to tariffs and quotas. The sixth Mexican policy is the low income, or tortilla, subsidy. Note that one of the main motivations for the systems of tariffs and quotas is the maintenance of targets for domestic farm prices and controlled retail prices for almost all basic foods. Two US farm programs are included in the model, the ‘deficiency payments program’ and the Export Enhancing Program (EEP). The deficiency payments program provides payments to farmers who participate in feed grain, wheat, rice, or cotton programs. The EEP program is intended to counter competitors’ subsidies and other unfair trade practices in targeted US agricultural export markets, and to develop, expand, or maintain foreign markets.

245

NOTES

1.

Limitations of space dictate that this paper be presented here in its shorter form..

2.

Proposition 187 is a piece of Californian legislation comprising a series of measures aimed at illegal aliens. The Proposition was approved by public vote in November 1986 and consists of five major sections, each of which can be implemented separately: −

barring of illegal aliens from public education systems and requires public education institutions to verify the legal status of both students and their parents;

−

providers of all publicly-paid, non-emergency health-care services must verify the legal status of persons seeking services in order to be reimbursed by the state of California. In this regard, the implication that prenatal care may be denied to illegal alien women has been particularly controversial;

−

requires that all persons seeking cash assistance and other benefits must verify their status before receiving benefits;

−

all service providers must report suspected illegal aliens to California’s Attorney General and to the Immigration and Naturalisation Service (INS) and police must determine the legal status of persons arrested;

−

the making and use of false documents is a state felony.

In fact, the core provisions of Proposition 187 have not been implemented as a result of a series of successful court challenges. The latest of these was in March 1998 when US District Judge, Mariana Pfaelzer, declared the core provisions of the Proposition as unconstitutional on the grounds that the US constitution gives the federal government exclusive power to regulate immigration. 3.

For a full description of these effects in a slightly different modelling context, see Hinojosa and McCleery (1992).

4.

Note that these are GDP numbers, not welfare gains. In particular, the higher US GDP due to NAFTA and immigration may lead to a lower GDP/capita figure, given the relatively low skill level of the immigrants. Is the United States really better off? A side calculation indicates that prior US residents clearly are better off. If one takes the increase in US GDP of nearly $5 billion (0.1272 per cent of 3 875.3 billion) and subtracts net wages paid to new immigrants (approximately 400 000 new immigrants at $10 000 each), there is a welfare gain of a billion dollars. This dwarfs the gain of $20 million (an increase of 0.0005 per cent over the base rest of US GDP of $3875.3 billion) from trade alone.

5.

At the higher levels of investment, the impact on migration, and thus on wages in both countries, appears to be exaggerated. We need to carefully analyse and perhaps revise the implicit migration elasticities in the model.

6.

Enterprise savings rates are assumed to adjust to achieve the necessary level of aggregate savings in each country. In the CGE literature, this treatment is known as Johansen macro closure, after Lief Johansen, who used it in the earliest single-country CGE model.

246

7.

Some of the Mexican agricultural support data refer to 1989.

8.

The properties of single-country CGE models incorporating CES import aggregation functions have been extensively studied.

9.

Other generalisations of the CES function could allow different, but fixed, elasticities of substitution between goods from different countries. It is also common to use nested CES functions, with a twogood CES function specifying substitution between domestically produced goods and a composite of imports, which is itself a CES function of goods from various countries of origin.

10.

We drew on work at the International Trade Commission for estimates of the various elasticities.

11.

The geometric price index is usually called a Stone index.

12.

This total represents agricultural expenditures for 1989 and subsidies to food processing for 1988.

13.

The policies are modelled as in the US-Mexico model Robinson et al. (1991), where the policies are described in detail.

247

BIBLIOGRAPHY

ABOWD, J.M. and FREEMAN R.B. (1991), Immigration, Trade and the Labour Market, University of Chicago Press for the National Bureau of Economic Research, Chicago. ALARCÓN, R. (1992), “Norteñización: Self-Perpetuating Migration from a Mexican Town”, in Jorge Bustamante, Clark Reynolds, and Raúl Hinojosa-Ojeda (eds.), U.S. Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. ALARCÓN, R. (1995), Immigrants or Transnational Workers? Proposition 187 and The Settlement Process among Mexicans in Rural California, Report for the California Institute for Rural Studies, Davis, CA. ALBA, F. (1992), “Migrant Labour Supply and Demand in México and the United States: A Global Perspective”, in Bustamante, J., Reynolds, C. and Hinojosa-Ojeda, R. (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. BLANCHARD, O., DORNBUSCH, R., KRUGMAN, P., LAYARD, R. and SUMMERS L. (1991), Reform in Eastern Europe, The MIT Press, Cambridge, Massachusetts. BLISS, C. and DE MACEDO, J.B. (1990) (eds.), Unity with Diversity in the European Economy: The Community’s Southern Frontier, Centre for Economy Policy Research, Cambridge University Press. BUSTAMANTE, J., REYNOLDS, C. and HINOJOSA-OJEDA, R. (1992) (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. CALIFORNIA DEPARTMENT OF FINANCE (1991), California’s Growing Taxpayer Squeeze, Finance Report. CALVA, J.L. (1992), Probables Efectos de un Tratado de Libre Comercio en el Campo México, Fontamara, México. CLINTON, B. (1992), “Expanding Trade and Creating American Jobs”, Speech at North Carolina State University, Raleigh, N.C., October 1992. CRAIG, A. (1990), “Legal Constraints and Mobilization Strategies in the Countryside”, in Foweraker and Craig (eds.), Popular Movements and Political Change in México, Lynne Rienner Publishers, Boulder, Colorado

248

CORNELIUS, W. (1990), “Labour Migration to the United States: Development Outcomes and Alternatives in Mexican Sending Communities”, in Diaz-Briquets, S. and Weintraub, S. (eds.), Regional and Sectoral Development as Alternatives to Migration, Westview Press, Boulder, Colorado. CORNELIUS, W. (1992a), “From Sojourners to Settlers: The Changing Profile of Mexican Immigration to the United States”, in Bustamante, J., Reynolds, C. and Hinojosa-Ojeda, R. (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. CORNELIUS, W. (1992b), “The Politics and Economics of Reforming the Ejido Sector in México: An Overview and Research Agenda”, LASA Forum, Vol. XXIII, No. 3. DONATO, K., MASSEY, D. and DURNAD, J. (1992), “Stemming the Tide”, Demography 29. ESPENSHADE (1992), “Responsiveness of the México-U.S. Flow of Undocumented Migrants to Relative Improvements in the Mexican Economy”. FISHLOW, A., ROBINSON, S. and HINOJOSA-OJEDA, R. (1991), “Proposal for a North American Regional Development Bank and Adjustment Fund”, Federal Bank of Dallas, North American Free Trade: Proceedings of a Conference, Federal Reserve Bank of Dallas, Dallas, TX. FIX, M. (1991) (ed.), Employer Sanctions -- Implementation, Impact, and Reform, Urban Institute Press, Washington, D.C. FIX, M. and PASSEL, J.S. (1994), Immigration and Immigrants, The Urban Institute, Washington, D.C., May 1994. FOX, J. (1992), “Democratic Rural Development: Leadership Accountability in Regional Peasant Organizations”, Development and Change, April 1992, Vol. 23, No. 2, pp. 1-36. FOX, J. and GORDILLO, G. (1989), “Between State and Market: The Prospects for Autonomous Grassroots Development in Rural México”, in Cornelius, P. et al. (eds.), México’s Alternative Political Futures, University of California, Center for US Mexican Studies, La Jolla, CA. FRIEDMANN, J. (1992), Empowerment: Cambridge.

The Politics of Alternative Development, Blackwell,

GARCIA Y GRIEGO, M. (1989), “The Mexican Labour Supply, 1990-2010,”, in Cornelius, W. and Bustamante, J. (eds.), Mexican Migration to the United States: Origins, Consequences and Policy Options, UCSD Center for US-Mexican Studies, La Jolla, CA. GARTEN, J.E. (1992), A Cold Peace: American, Japan, Germany and the Struggle for Supremacy, Times Books/Twentieth Century Fund, New York. GEPHARDT, R.A. (1992), “Address on the Status of the North American Free Trade Agreement”, Institute for International Economics, 27 July.

249

GOLDEN, T. (1992), “Mexican President Seeks to Address Clinton’s Concerns: Salinas Calls for Efforts to Ease Curbs on Mexicans’ Moving Across Border into U.S”, New York Times, No. 21, A1, November . GOLDRING, L. (1991), “Development and Migration? A Comparative Analysis of Two Mexican Migrant Circuits”, Paper prepared for the Commission for the Study of International Migration and Cooperative Economic Development. GRINDLE, M. (1988), Searching for Rural Development, Cornell University Press, Ithaca, New York. HAYES-BAUTISTA, D., SCHINK, W. and CHAPA, J. (1992), “The Young Latino Population in an Ageing American Society: Policy Issues Evoked by the Emergence of an Age-Race Stratified Society”, in Bustamante, J., Reynolds, C. and Hinojosa-Ojeda, R. (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. HERNANDEZ, L. (1991), Autonomia y desarrollo: La Lucha en el Campo en la Hora de la Concertacion, Cuadernos desarrollo de base, ADN Editores, México, D.F. HERNANDEZ, L. and BRAY, D. (1990), Privatizing the Peasantry in México: INMECAFE and the National Coffee Farmers Movement. HINOJOSA-OJEDA, R. (1993), “The Political Economy of Latino Income and Employment in the United States”, in Weaver, T. (ed.), Handbook of Hispanic Cultures in the United States, Vol. I: The Anthropology of Hispanic Groups in the US, Arte Publico Press/University of Houston Press, Houston, forthcoming, Spanish version to be published by Instituto Cooperativo Ibero Americano, Madrid. HINOJOSA-OJEDA, R. (1994a), “The North American Development Bank Forging in New Directions in Regional Integration Policy”, Journal of the American Planning Association, No. 60(3), pp. 301-304. HINOJOSA-OJEDA, R. (1994b), “The North American Free Trade Agreement and Migration”, Migration and Development of New Partnerships for Cooperation, OECD, Paris, pp. 229-239. HINOJOSA-OJEDA, R. and McCLEERY, R. (1992), “A U.S.-México Interdependence, Social Pacts and Policy Perspectives: A Computable General Equilibrium Approach”, in Bustamante, J., Reynolds, C. and Hinojosa-Ojeda, R. (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. HINOJOSA-OJEDA, R. and MORALES, R. (1992), “International Restructuring and Labour Market Interdependence: The Automobile Industry in México and the United States”, in Bustamante, J., Reynolds, C. and Hinojosa-Ojeda, R. (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. HINOJOSA-OJEDA, R. and ROBINSON, S. (1991), “Alternative Scenarios of US México Integration: A Computable General Equilibrium Approach”, Working Paper No. 609, Department of Agricultural and Resource Economics, University of California, Berkeley.

250

HINOJOSA-OJEDA, R. and ROBINSON, S. (1992), “Labour Issues in a North American Free Trade Area”, in Lustig, N., Bosworth, B. and Lawrence, R. (eds.), North American Free Trade: Assessing the Impact, The Brookings Institution, Washington, D.C. HINOJOSA-OJEDA, R. and SCHEY, P. (1995), “The Faulty Logic of the Anti-Immigrant Rhetoric”, NACLA Report on the Americas, Nov/Dec 1995, Vol. 29, No. 3, pp. 18-24. HINOJOSA-OJEDA, R. and SCHEY P. (1996), “Proposition 187 and the Future of North American Labour and Human Rights: Manifest Destiny’s Last Stand?”, Report presented to National Academy of Sciences, Committee on Population & Committee on National Statistics, Panel on Demographic and Economic Impacts of Immigration, UCLA North American Integration and Development Center, Los Angeles. HINOJOSA-OJEDA, R., CARNOY, M. and DALEY, H. (1991), “An Even Greater ‘U-Turn’: Latinos and the New Inequality”, in Melendez, E. et al. (eds.), Hispanics in the Labour Force, Plenum Press, New York. HINOJOSA-OJEDA, R., ROBINSON, S. and WOLFF, G. (1992), The Impact of a North American Free Trade Agreement on California: A Summary of Key Research Findings, Lewis Center for Regional Policy Studies Working Paper No. 3, University of California, Los Angeles, September. HINOJOSA-OJEDA, R., RUNSTEN, D., WOLFF, C., DE PAOLIS, F., WOLFF, G., ROBINSON, S., DOWDS, C. and McCLEERY, R. (1997), “US Employment Impacts of NAFTA at Three Years”, Paper presented at the American Economic Association -- Allied Social Science Associations Annual Meetings, Chicago, 3 January 1998. HUDDLE, D. (1993), “The Net National Costs of Immigration”, Carrying Capacity Network, July. HUDDLE, D. (1994), “The Net National Costs of Immigration in 1993”, Department of Economics, Rice University, Houston, TX, 27 June. INTERNAL SERVICES DIVISION (1992), “Impact of Undocumented Persons and Other Immigrants on Costs, Revenues and Services in Los Angeles County”, Report prepared for Los Angeles County Board of Supervisors, November. JOHNSON, W.B. (1987), “Workforce 2000: Work and Workers for the 21st Century”, Report to the US Department of Labour, Hudson Institute, Indianapolis, Indiana. LAWRENCE, R.Z. (1991), “Emerging Regional Arrangements: Building Blocks or Stumbling Blocks”, in O’Brien, R. (ed.), Finance and the International Economy, the AMEX Bank Review Prize Essays, Oxford University Press, Oxford. LAYARD, R., BLANCHARD, O., DORNBUSCH, R. and KRUGMAN, P. (1992), East West Migration: The Alternative, MIT Press, Cambridge. LEVY, S. and VAN WIJNBERGEN, S. (1991a), “Maize and the México-United States Free Trade Agreement”, unpublished manuscript, World Bank, February.

251

LEVY, S. and VAN WIJNBERGEN, S. (1991b), “Agriculture in the México-USA Free Trade Agreement”, unpublished manuscript, World Bank, revised version, April. LEVY, S. and VAN WIJNBERGEN, S. (1991c), “Labour Markets, Migration and Welfare: Agriculture in the México-USA Free Trade Agreement”, unpublished manuscript, World Bank, June. LEVY, S. and VAN WIJNBERGEN, S. (1991d), “Transition Problems in Economic Reform: Agriculture in the México-USA Free Trade Agreement”, unpublished manuscript, World Bank, December. LIPTON, M. (1980), “Migration from Rural Areas of Poor Countries: The Impact on Rural Development and Income Distribution”, World Development, No. 8, pp. 1-25. LOPEZ, R. (1992), “UCEFO”, unpublished manuscript, Oaxaca, México. LOZANO ASCENCIO, F. (1992a), Bringing it Back Home: Remittances to México from Migrant Workers in the United States, UCSD Center for U.S.-Mexican Studies, La Jolla, CA. LOZANO ASCENCIO, F. (1992b), “Las Remesas Enviadas por los Trabajadores Mexicanos en Estados Unidos en 1990: Reflexiones Sobre su Monto y Significado Economico y Social”, México-Estados Unidos 1990, El Colegio de México, México. MARTIN, P. (1991), The Unfinished Story: Turkish Migration to the EC, ILO, Geneva. MARTIN, P. and TAYLOR, J.E. (1992), “NAFTA and Mexican Migration to the US”, Paper presented at a conference on North American Free Trade Agreement: Implications for California Agriculture, sponsored by the UC Agricultural Issues Center and the Guanine Foundation of Agricultural Economics, Los Angeles, CA, 5 March. MASSEY, D.S. (1988), “Economic Development and International Migration in Comparative Perspective”, Population and Development Review, Vol. 14, No. 3, pp. 383-413, September. MASSEY, D. and DURANCE, J. (1992), “Mexican Migration to the United States”, Latin American Research Review, Spring. MASSEY, D. and SINGER, A. (1995), A new Estimates of Undocumented Mexican Migration and the Probability of Apprehension, Demography, forthcoming. MASSEY, D., ALARCÓN, R., DURANCE, J. and GONZALES, H. (1987), Return to Aztlan: The Social Process of International Migration from Western México, University of California Press, Berkeley, CA. McCLEERY, R. and HINOJOSA-OJEDA, R. (1993), “Asia and NAFTA: Possible Extensions of and Reactions to NAFTA in the Context of the Global Trading System”, Paper prepared for presentation at the Institute for Developing Economies Conference on Regional Integration and its Impact on Developing Countries, Tokyo, 3 and 4 February. OHMAE, K. (1985), Triad Power: New York.

The Coming Shape of Global Competition, Macmillan,

252

PASOS, L. (1992), “La Empresa Social Forestal: Program Pasos, México, D.F.

Tercer Analisis de Experiencias Forestales”,

PASSEL, J. and WOODROW, K.A. (1984), “Geographic Distribution of Undocumented Immigrants: Estimates of Undocumented Aliens Counted in the 1980 Census by State”, International Migration Review, No. 18, pp. 642-671. PORTER, M.E. (1990), The Competitive Advantage of Nations, The Free Press, New York. PORTES, A. and RUMBAUT, R. (1991), Immigrant America, University of California Press, Berkeley, CA. PRZEWORSKI, A. (1990), The State and Economy Under Capitalism, Harwood Academic Publishers, New York. PRZEWORSKI, A. (1991), Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America, Cambridge University Press, Cambridge. REICH, R.B. (1991), The Work of Nations, Vintage Books, New York. REYNOLDS, C. (1992), “Will a Free trade Agreement Lead to Wage Convergence? Implications for México and the United States”, in Bustamante, J., Reynolds, C. and Hinojosa-Ojeda, R. (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. ROBINSON, S.E., BURFISHER, M., HINOJOSA-OJEDA, R. and THIERFELDER, K.E. (1991), “Agricultural Policies and Migration in a US-México Free Trade Area: A Computable General Equilibrium Analysis”, Working Paper No. 617, Department of Agricultural and Resource Economics, University of California, Berkeley. ROSENBERG, J.M. (1992), The New American Community: A Response to the European and Asian Economic Challenges, Praeger, New York. RUNSTEN, D. and MARSH, R. (1992), “Peasant Participation in Fresh Fruit and Vegetable Production”, unpublished grant proposal. RUNSTEN, D. and WILCOX, L. (1992), “Demand for Labour, Wages, and Productivity in Mexican Fruits and Vegetables: Preliminary Estimates and Implications for NAFTA”, Paper presented at the Latin American Studies Associations Meetings, Los Angeles, September. SCHEY, P.A. (1992), Fare Trade and the Human Rights of Migrant Workers, The Immigration Law Bulletin, Center for Human Rights and Constitutional Law, Los Angeles, Vol. X, No. 1, pp. 1-7. SHAIKEN, H. (1990), México in the Global Economy: High Technology and Work Organization in Export Industries, Center for US Mexican Studies, University of California, San Diego, CA. SIROLA, P. (1992), “Beyond Survival: Latino Immigrant Street Vendors in the Los Angeles Informal Sector”, Paper presented at the XVII International Congress of the Latin American Studies Association, Los Angeles, California, September. 253

SMITH, R. (1992) “Los Ausentes y Los Presentes: The Imagining, Making and Politics of a Transnational Community Between New York and Ticuani, Puebla”, Paper presented at the XVII International Congress of the Latin American Studies Association, Los Angeles, California, September. STUART, J. and KEARNEY, M. (1981), “Causes and Effects of Agricultural Labour Migration from the Mixteca of Oaxaca to California”, Working Papers, US Mexican Studies, No. 30. TAYLOR, J.E. (1987), “Undocumented México-US migration and the returns to households in rural México”, American Journal of Agricultural Economics, p. 66. TREJO REYES, S. (1992), “Mexican-American Employment Relations: The Mexican Context”, in Bustamante, J., Reynolds, C. and Hinojosa-Ojeda, R. (eds.), US Mexican Relations: Labour Market Interdependence, Stanford University Press, Stanford, CA. THUROW, L. (1992), Head to Head: The Coming Economic Battle Among Japan, Europe, and America, William Morrow, New York. US DEPARTMENT OF LABOUR (1991), Handbook of Labour Statistics, US Department of Labour, Washington, D.C. VERNEZ, G. and McCARTHY, K. (1995), “The Fiscal Costs of Immigration: Analytical and Policy Issues”, Paper prepared for the Irvine Foundation by the Center for Research on Immigration Policy, February. WALLACE, S.P. (1988), “Central American and Mexican Immigrant Characteristics and Economic Incorporation in California”, International Migration Review, No. 22, pp. 664-665. WARREN REPORT (1994), “Estimates of the Unauthorized Immigrant Population Residing in the United States, by Country of Origin and State of Residence: October 1992”, US Immigration and Naturalization Service, April. WELLS, M. (1990), “Mexican Farm Workers Become Strawberry Farmers”, Human Organization, Vol. 49, No. 2. WILLIAMSON, O.E. (1985), The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting, The Free Press, New York. ZABIN, C. (1985), “The Impact of Circular Migration on Economic Activity in Indigenous Communities in México: A Case Study of Laxopa in the Sierra Juarez, Oaxaca, México”, Unpublished dissertation prospectus. ZABIN, C. (1992), “Binational Labour Markets and Segmentation by Gender and Ethnicity: the case of fresh fruit and vegetable production in Baja California and California”, Proceedings, Women’s Economic Policy Conference, American University.

254

Table 1. Scenario description Scenario

Description

1a 1b 2a 2b 2c 3a 3b 3c 4a 4b 4c

NAFTA - All tarifs and subsidies set to Zero. No migration 1a plus migration Migrants expelled from California Migrants expelled from the United States and California Migrants allowed to stay in California but wages reduced Mexican total investment increased by 5 per cent Mexican total investment increased by 10 per cent Mexican total investment increased by 15 per cent 3a plus increase in migrants’ reservation wage 3b plus increase in migrants’ reservation wage 3c plus increase in migrants’ reservation wage

Table 2. Real GDP (Percentage change from base) 1a

1b

2a

2b

2c

3a

3b

3c

4a

4b

4c

United States

0.0005

0.1272

-0.3878

-0.7755

0.4080

-0.0393

-0.2580

-0.4461

-0.2642

-0.4386

-0.6191

California

0.0014

0.1104

-1.2056

-2.4926

0.3518

0.0339

-0.2222

-0.3845

-0.2270

-0.3759

-0.5325

Mexico

0.2284

0.0648

0.5385

0.7786

0.2635

0.1097

0.1239

0.1205

0.3979

0.4090

0.3714

Source: NAID CalNAFTA CGE Model.

255

Table 3. Net change in migration (Thousands) 1a

1b

2a

2b

2c

3a

3b

3c

4a

4b

4c

256

USMIGAG Migrants to US rural sector

0.00

18.40

-200.22

-400.00

-57.75

6.16

-15.16

-30.25

-24.38

-35.10

-45.91

USMIGURB Migrants to US urban sector

0.00

370.03

-550.00

-1 100.00

183.61

-125.63

-773.30

-1 330.23

-785.26

-1 303.61

-1 841.61

CAMIGAG Migrants to CA rural sector

0.00

7.25

-200.00

-200.00

64.60

4.05

-2.99

-7.39

-6.19

-9.24

-12.28

CAMIGURB Migrants to CA urban sector

0.00

140.90

-550.00

-550.00

977.72

-49.01

-293.44

-504.36

-298.18

-494.94

-698.55

MXMIG Migrants in MX urban sector

0.00

305.90

-9.62

-273.29

267.81

176.58

-74.78

-245.59

-256.27

-420.61

Source: NAID CalNAFTA CGE Model.

Table 4. Return to factors (Percentage change from base)

1a

1b

2a

2b

2c

3a

3b

3c

4a

4b

4c

-0.277

United States Capital

-0.040

0.057

-1.790

-3.602

0.181

-0.045

-0.119

-0.193

-0.130

-0.215

0.860

-0.917

42.993

85.566

-2.877

0.277

1.907

3.292

2.715

3.644

4.607

Urban unskilled labour

-0.050

-0.917

2.041

4.040

-2.877

0.277

1.907

3.291

1.919

3.210

4.555

Union labour

-0.041

0.054

-0.526

-1.097

0.198

-0.050

-0.122

-0.202

-0.142

-0.218

-0.286

Professional

-0.044

0.048

-0.482

-1.011

0.181

-0.050

-0.113

-0.187

-0.135

-0.202

-0.266

0.425

0.585

6.805

13.477

0.744

0.585

0.346

0.266

0.239

0.133

0.053

Rural labour

Land California Capital

-0.030

0.053

-1.789

-3.601

0.189

-0.047

-0.124

-0.201

-0.136

-0.218

-0.283

0.006

-0.917

43.013

85.606

-2.877

0.277

1.907

3.291

2.715

3.644

4.607

Urban unskilled labour

-0.020

-0.917

5.699

11.566

-2.877

0.277

1.907

3.291

1.919

3.210

4.554

Union labour

-0.014

0.054

-0.535

-1.113

0.198

-0.022

-0.122

-0.202

-0.142

-0.218

-0.286

Professional

-0.012

0.048

0.143

0.375

0.181

-0.012

-0.113

-0.187

-0.134

-0.202

-0.266

0.013

0.188

-11.527

-23.283

0.705

-0.235

-0.651

-1.054

-0.678

-1.034

-1.390

Rural labour

257

Land Mexico Capital

0.515

0.315

2.307

4.538

0.444

-0.642

-1.757

-2.843

-0.397

-1.460

-2.555

-4.611

0.981

-4.264

-5.146

-0.922

-4.347

-10.645

-16.510

-11.465

-16.925

-22.680

Urban unskilled labour

0.284

0.982

-4.423

-5.466

-0.923

-4.347

-10.645

-16.510

-7.391

-12.850

-18.606

Union labour

0.463

0.298

2.398

4.475

0.397

-0.752

-2.003

-3.164

-0.585

-1.745

-2.912

Professional

0.435

0.264

2.522

4.752

0.364

-0.704

-1.851

-2.917

-0.516

-1.585

-2.657

-5.724

-6.761

-2.617

5.467

-6.271

-10.501

-12.822

-16.324

-8.431

-11.719

-15.463

Rural labour

Land

Source: NAID CalNAFTA CGE Model.

PART IV

BROADER PERSPECTIVES ON MIGRATION AND REGIONAL INTEGRATION

259

MIGRATION POLICIES IN A FREE TRADE AREA: THE ISSUE OF CONVERGENCE WITH THE ECONOMIC INTEGRATION PROCESS

by Francisco Alba (Colegio de México), Jean-Pierre Garson (OECD) and El Mouhoub Mouhoud (University of Evry)

1.

Introduction

The impact of regional economic integration on international migration flows is currently an issue in the debate on ways to regulate more effectively, or reduce, flows from developing to developed countries. By liberalising trade in goods and services and increasing direct investment flows, it is hoped that more jobs can be created in countries with high migration potential, so reducing the incentive to emigrate over the longer term. Comparing the formation of regional blocs, chiefly in Europe and North America, shows various levels of regional integration. Regionalisation is progressing against a background of greatly accelerating economic globalisation in both physical and financial terms, accompanied by the polarisation of trade in goods and movements of capital and technology among the developed countries. Host countries hope that trade liberalisation will speed up development in countries of emigration and that this will help regulate migration flows. Countries of emigration, whose sole advantages in many cases lie in natural resources or abundant labour, hope that free trade agreements will give them more of a share in international trade. The aim of reducing flows of emigrants, as such, is not one of their prime concerns. Differing objectives of developed and less developed countries in some policy areas complicate the debate on the convergence of migration policies in free trade areas with the process of economic integration. This paper will first attempt to provide a framework for analysing the effects of free trade on economic development and migration by identifying a number of levels of integration and analysing the underlying adjustment mechanisms at each level. This framework is then applied to various specific migration casebooks, based on the experience of European Union and NAFTA countries. The key question is whether liberalisation of migration should proceed concurrently with liberalisation of trade and economic convergence, or whether it could only come into play after the effects of regional integration have resulted in closer economic convergence. In other words, can the free movement of persons proceed in parallel with the stages of regional economic integration or should it remain a priority objective, but one that would be implemented only when economic convergence has reached an adequate level?

261

2.

The different levels of regional integration and their effect on economic development and international migration

The world economy is experiencing two quite distinct patterns of integration: globalisation, reflected in the growing liberalisation of trade in goods and services and capital movements, and the formation 1 of regional blocs in which there may be a greater or lesser extent of institutional integration. Although these two trends are interrelated, their effects on less developed countries’ economic growth and ability to catch up differ, depending on time-scale and the degree of initial asymmetry in the level of development of the countries concerned. Globalisation tends to change the role and position of these countries in the international economic order. Although essentially trans-national in nature, it is also evident that regionalisation encourages the economic integration of countries that are not necessarily at the same level of economic development. In this paper, we focus more closely on regional economic integration and the different forms it can take, rather than the broader trend of globalisation. In this regard, the degree of economic convergence and the degree of mobility of factors of production are taken as fundamental indicators of the level of integration of the economic area concerned. The formation of regional blocs Regional integration can be defined as a process whereby trade between countries belonging to the same free trade area, customs union or economic and monetary union, intensifies at the expense of extra-regional trade. It is important to distinguish between two types of regionalisation. The first, de jure regionalisation, is the outcome of institutional and political decisions by countries with the aim of building a more or less integrated regional area (European Union, NAFTA, MERCOSUR, ASEAN, etc.). Alongside these institutionalised forms of regional integration are de facto forms (also known as “deep” forms of integration) which may or may not coincide with de jure processes. Geographical and cultural proximity, history and migration are the main bases for regional integration preceding the political decision to form a regional bloc. De jure regional integration may speed up, or sometimes slow down, de facto processes. De jure and de facto regionalisation can both accommodate differences between countries in the economic area concerned. If the institutional decision is taken to have countries at different stages of development participate in a regional area, trade based on differing factor endowments may take place, in accordance with the principle of comparative advantage. The short-term effects of regional integration can be summed up, as in the classic analysis developed by Viner in the 1950s, as trade creation and trade diversion, i.e. the intensification of intra-regional trade at the expense of extraregional trade. The effect of trade diversion is regarded as benefiting member countries even when non-member countries have comparative advantages. However, the classical approach does not analyse the long-term effects (economic growth, capital accumulation and technological change) of regional integration processes on economic development and migration. There are now endogenous 2 growth models for open economies that take these effects into account. Regional integration and economic convergence A new and less developed entrant to an economic region benefits from a promising environment that can halt the threat of cumulative divergence that globalisation poses to less developed countries. Empirical tests of the convergence hypothesis, along with some endogenous growth models3 for open

262

economies, show the risks of cumulative divergence between countries at different stages of technological development. Contrary to classical theory, endogenous growth models indicate that opening up to international trade is not necessarily beneficial for all concerned. What happens is that increasing returns produce asymmetries between sectors. Specialisation will promote growth if it occurs in sectors where returns are increasing, but not if it occurs in sectors where they are diminishing (Lucas, 1988). Specialisations become self-reinforcing over time, through cumulative mechanisms. With this in view, L. Rivera-Batiz and Romer (1991a) divided the effects of international trade into three categories: the integration effect, the redundancy effect, and the allocation effect. The first two have a positive impact on growth. The third, the allocation effect, may have a negative impact if resources are transferred from the research sector to the industry sector. When the resources allocated to research are reduced, the growth rate declines in the country concerned. Where countries are at roughly the same stage of technological development, the allocation effect is limited and the first two effects are predominant. Where countries are at different stages of technological development, the allocation effect may be sufficiently strong to become predominant. A country whose comparative advantage prompts a reallocation of resources towards the industry sector at the expense of the research sector will see its growth rate decline. Where that country already has a smaller technological endowment, the gap between it and the most advanced countries is likely to widen. This said, the above theories underestimate other effects of regional integration. Regionalisation may be accompanied by an economic, social and institutional environment which also favours economic growth. The economic factors are principally infrastructure, transport and telecommunications. The social factors are partly those that determine human capital accumulation, particularly the standard of education. Among the institutional factors, legislation on the protection of intellectual property plays a decisive role in encouraging innovation and promoting technology transfer mechanisms. In addition, at regional level, economies of scale, particularly in the use of human capital, make it easier to catch-up on technology. Information on technology circulates freely in the region, increasing the stock of available knowledge for countries belonging to that area, and consequently 4 increasing the growth rate of economies that are less technologically advanced. The cumulative effects should therefore tend more towards convergence than divergence. Moreover, direct investment is more likely to be channelled to less developed countries in the regional area than to other LDCs. The entry of Portugal and Spain to the European Community is a case in point. In practice, the announcement of this type of integration influences firms’ expectations and decisions on the location of production units or on joint-venture operations, so as to reap the benefits of being the 5 first on the local markets. From that point on, technological catch-up and economic convergence may accelerate at a rate that is strictly dependent on the depth of regional integration. These catch-up mechanisms, which derive from the beneficial effects of creating a favourable environment, may play a smaller role in the case of regional integration between countries of the South (i.e. between developing countries) because human capital is initially too small and inefficient. In addition, the risk of redundant and duplicate international specialisations is higher for economies whose specialisations (most often in the primary sector) are largely identical. The key issue that then arises for many less developed countries whose sole advantages lie in natural resources and plentiful labour is the pertinence of forming a regional grouping with countries whose international specialisations and level of development are complementary.

263

Level of regional integration and impact on economic development, employment and migration. Three levels of regional integration can be identified whose dynamic effects in terms of economic catch-up, foreign direct investment, job creation and international migration may differ. At the first level we have (more or less comprehensive) free trade agreements without full liberalisation of factor flows (NAFTA, for example, or EU free trade agreements with Mediterranean countries). At the second we have agreements to enlarge existing regional blocs to peripheral countries. These expressly provide for budget transfers to new members but keep barriers to the mobility of labour (the enlargement of the European Union to CEECs, for example). The third level is full economic and monetary integration of member countries (economic and monetary union as with the next stage of the European Union, for example) including freedom of movement and establishment within the area. Free trade agreements without full liberalisation of factor flows At the first level, free trade agreements do not provide for significant budget transfers. The effects expected vary with the size and level of development of the signatory countries. These agreements may or may not be comprehensive: they may cover all goods and services, or be confined to certain sectors. In the case of comprehensive agreements, well-documented effects such as improved welfare related to better resource allocation are found, provided that the trade-diversion effect does not outweigh the new trade-creating effect. Any effects on migration will only be seen over the longer term, via the tendency for factor income between trading partners to level out, according to the 6 standard theory of international trade. The agreements recently concluded between the European Union and the countries of the southern Mediterranean exclude one of the main areas of comparative advantage for the latter (agricultural produce). In this case, the risk that the southern Mediterranean countries will experience unwanted effects from trade liberalisation is high. The elasticity of national supply in exports of agricultural goods is high, but this export sector has not been liberalised. Given the low elasticity of supply in industrial products, the liberalisation of trade in manufactures will not be of much help to countries whose main advantages lie in natural resources and abundant labour. In addition, these countries are likely to have higher imports of industrial goods and lower fiscal revenues as a result of the abolition of customs duties. In the short term, problems in effecting the transition to a fiscal structure that is primarily based on direct domestic taxation may well reduce 7 public and private investment. This could result in a deterioration of public infrastructure despite the fact that it is a key factor in attracting FDI. Overall, increased external indebtedness and weak growth or a decline in per capita GDP may trigger a deterioration in the macroeconomic factors attracting FDI. Even if new signatory countries retain a wage-cost advantage, it will not be sufficient to attract FDI, particularly if positive externalities (efficient transport and telecommunications infrastructure, for example) are absent. The issue of FDI is examined in a recent computable general equilibrium model (Cogneau et al., 8 1998) where direct investment is treated as endogenous -- in the form of a lower FDI risk premium (most studies on the effects of free trade to date regard FDI as exogenous, lumped together with public transfers). In this model, the effect of a free trade agreement itself on the industrial export growth rate determines the requisite level of FDI. Industrial exports, themselves dependent on the elasticity of national industrial supply on the one hand, and the price-elasticity of demand in the

264

regional bloc on the other, will have some impact on GDP growth rates and employment, and therefore on the incentive to emigrate. Foreign firms’ expectations about the size of the market (measured by GDP growth rates and per capita GDP), wage cost differentials and public infrastructure determine, in principal, the future amounts and sectoral distribution of direct investment flows to initially less developed signatory countries. However, in the case of partial free trade agreements (those excluding agriculture, for example) these mechanisms are unable to operate and the manufacturing sector will be unable to create enough jobs, through trade liberalisation, to reduce the incentive to emigrate in the longer term. Agreements enlarging existing regional blocs In the case of agreements enlarging a regional bloc to peripheral countries (for example, the enlargement of the European Union to the CEECs) that provide for budget transfers but maintain barriers to labour mobility, the transfers may offset any crowding out effect on public expenditure and domestic investment. They may also increase the import capacity of the new entrants. In this case, the absence of restrictions on the liberalisation of trade would increase the potential volume of export goods and increase export income. This would have a positive effect on GDP growth rates and on per capita GDP. Foreign firms could expect the market in new entrant countries to grow, and their public infrastructure to improve. The maintenance of a wage-cost advantage coupled with improved public 9 externalities and the expected growth in market size would then help to attract FDI. The latter would have a positive effect on employment and in the long run produce a change in the structure of international specialisation in the least advanced countries integrated into the regional bloc. An increase in the number of skilled jobs and a boost to local economic development are two factors that would help significantly decrease the incentive to emigrate over the longer run. Economic and monetary integration The third level, economic and monetary integration, is the most radical form of integration. Where there are major structural differences (particularly in technology) between advanced and less developed countries in a monetary area, the exchange rate cannot be used as an adjustment tool and cumulative divergence is a real risk for the least developed countries. The inability to make nominal adjustments makes real or structural adjustment necessary, which may prove costly in the event of 10 asymmetric shocks (productivity, demand shocks etc.). Should such shocks occur, the weakest member countries would see a deterioration in the competitiveness of their export sectors (to the advantage of production of non-tradable goods), and job losses there as well. If labour mobility in the area is high, migration flows between member countries would increase. Conversely, if labour mobility within the area tends not to be high, unemployment could well increase in countries that are not converging at the real and structural level. If the labour market is flexible, then adjustment through nominal salaries could partially offset real adjustments. Another adjustment mechanism lies in the scope some countries will have to recruit cheaper foreign labour (new legal entrants, or illegals) rather than lowering the wages of residents (nationals and legally resident foreigners). In this particular case, even though workers of member countries are allowed to circulate freely within the area, the movements may well include foreign labour from countries outside the area.

265

In the three levels of regional integration outlined above, the effects on migration of the structured adjustment and economic convergence processes differ. Likewise, the degree of mobility of factors of production differs. Economic and monetary union explicitly provides for the free movement of persons and therefore of labour. However, labour’s mobility within the area may tend to remain low, or at any rate lower than the mobility of capital. The first two levels of regional integration (see above) do not allow free movement of persons, prohibiting labour mobility. The tendency towards economic convergence is not necessarily paralleled by any liberalisation of migration. Can trade liberalisation policies be pursued and achieve greater regional economic integration while migration policy remains restrictive? In other words, is further integration at regional level possible without liberalising migration? 3.

Migration policy and the development of regional integration

Migration policy has three main facets. The first has to do with flow regulation and the movement of persons, the second with the status of immigrants and measures to assist their integration. The third facet places migration policy in the framework of development co-operation between host countries and countries of origin; for instance, the mobility of skilled people (including students and trainees) contributes to the development of economic ties and transfers of knowledge, and may promote flows of direct investment from host countries to countries of origin. Economic convergence and liberalisation of migration: the case of the European Union In theory, moves towards free trade and regional integration should over time reduce the incentive to emigrate. In the case of partial or comprehensive free trade agreements, however, migration is likely to continue until -- in the medium term -- there is more marked convergence between economies at differing degrees of development. Free trade areas and relatively advanced regional integration develop in most cases with barriers to the free movement of persons, and workers, still in place. Are these restrictions on the mobility of workers an impediment to deeper integration? This question is much more than a matter of international trade theory. It is an item, whether explicit or not, in the ongoing discussions on agreements to liberalise trade within regional groupings of countries at differing degrees of development. The accession of Greece, in 1981, and Portugal and Spain, in 1986, to the European Economic Community is an interesting precedent in this respect. All three countries had close and long-standing ties with other Western European countries, in particular through migration. In the case of Spain’s emigration (see Chart 1 regarding foreign trade and the migration flow between Spain and France, one of the principal receiving countries), one observes that the flows started to reverse towards the middle of the 1970s, i.e. before the EEC accession agreements had been concluded. Even so, at the time of signature a six-year period of adaptation was proposed, before free movement came into full effect. The incentive to emigrate was substantially reduced because, as with Greece and Portugal, economic catch-up and inflows of foreign investment boosted job creation and narrowed wage differentials. Greece, Portugal and Spain have today all become countries of immigration. In the European Economic Community, migration preceded the liberalisation of trade. Conversely, the free movement of persons was institutionalised only following a period of more intense and more liberal trade and capital regimes (ten years, for example, for the six founder members of the EEC). 266

The impact of regional economic integration on migration flows in this case shows that movements of EEC nationals fell back, reflecting the approximation of standards of living and the lesser incentive to emigrate. But immigration from non-member countries, less developed and with high migration potential, increased even though the EEC countries decided to close their borders to would-be immigrant workers in the mid-1970s. The dynamics of migration thus show that more intense economic and human links (deep integration, see above) will outweigh institutional factors, such as the institutional application of free movement for the EEC countries, or the closure of borders for non-member countries (Tapinos, 1994). That means that legal or institutional barriers cannot be regarded as the main inhibitor of migration flows. Conversely, free movement of persons within an economic area does not necessarily result in continuing or increased flows between member countries. What is more, the closure of borders did not prevent continued immigration flows from non-member countries. The case of the European Union shows that political determination to establish, by stages, an area for free movement of goods, services and capital assisted the economic integration of the less advanced countries. Free movement of persons was simply the final stage of integration before a single currency. It was introduced against the background of reversed flows in the EU’s traditional countries of emigration. Economic development, fostered by regional integration, lessened the incentive to emigrate. Accordingly, the decline in migration flows appears to be the result less of institutional decisions (NAFTA, Euro-Mediterranean agreements) to liberalise trade in order to replace migration by flows of goods and capital, than the outcome of economic development and technological catch-up fostered (or not) by regional integration. Free movement of persons and economic convergence The presence, in some OECD countries, of relatively large numbers of non-OECD nationals raises, from the standpoint of regional integration, the questions of their rights, status and terms of movement within the OECD area. Without any provision for free movement, there have been for instance in Western Europe large-scale inflows of people from Turkey, the Maghreb countries, and more recently ex-Yugoslavia (see Table 1). As has been shown in the case of the European Union (see above), free movement is not in itself a forerunner or trigger of migration. The free trade agreements which the European Union has recently signed with various countries along the southern and eastern shorelines of the Mediterranean do not touch on the issue of free movement of persons. A chapter of the NAFTA agreement proposes measures to facilitate the movement of specific categories of personnel (business people, investors and staff on secondment) within the North American area (Alba, 1993). However, this movement of persons is seen as complementary to the investment flows. At the same time, it is something of a puzzle that Mexico does not fill its current quota of 5 000 slots for Mexican professionals to enter the United States under NAFTA regulations. The references to migration in the association agreements between the European Union and various countries around the Mediterranean and in Central and Eastern Europe largely concern the integration of immigrants in host countries and equal treatment for immigrants and host country nationals as workers. Alongside these agreements, host countries have retained varying scope to accept immigrants from non-member countries for family reunion reasons or under movements to cover cyclical or sectoral labour shortages.

267

In addition, bilateral agreements provide for the admission of new immigrant workers under fixedterm contracts or as seasonal, trainee or frontier workers. These movements occasionally, and semiofficially, have a background of historical, geopolitical or cultural preferences. Host countries can also relax their own rules on the admission of immigrants: specified categories of workers may, for instance, be given special status on a temporary basis. Both the United States and Canada continue each year to accept very large numbers of new permanent immigrants, mainly from Mexico in the case of the United States (OECD, 1997). Accordingly, the absence of free movement is not a barrier to migration from one country in a regional bloc to another. At the same time, free movement is among the sensitive issues to be considered in the framework of deepening NAFTA, and particularly so in the process of accession to the European Union of the first group of countries listed at the Luxembourg Summit in December 1997 (the Czech Republic, Cyprus, Estonia, Hungary, Poland and Slovenia). As much is at stake for host countries as for countries of origin. The economic circumstances of Western European countries at present are not those which prevailed in the 1960s and 1970s, when Greece, Portugal and Spain set out to join the EEC. Growth rates are lower today, unemployment is higher and the established immigrant population is extremely vulnerable to unemployment. In addition, the three countries which joined the European Union in 1995, Sweden, Finland and Austria, are still familiarising themselves with the legal and institutional procedures flowing from Community rules applied to foreign residents in the other EU countries. Discussions about free movement of persons and free establishment of workers from would-be entrant countries cover only one of the facets of migration policy. They are likely to overshadow discussion of appropriate measures, first, to move from plain and sometimes incomplete liberalisation of trade to genuine economic integration (enlargement agreements including budget transfers) and, second, to improve the position of immigrants already established in host countries and to give new impetus to co-operation regarding migration and country-of-origin development. The example of the Southern European countries shows that their integration in the European Union was accompanied by reform of economic and financial structures, development of infrastructure and enhancement of human capital, all significant factors for economic development in countries with high migration potential. For the European Union, if structural convergence in would-be entrant countries (enlargement from 15 to 21) proves insufficient, introducing free movement could have an adverse impact on their development. From the standpoint of economic efficiency, and given that international movements of labour increasingly involve migration by skilled workers, the short or medium-term prospects of extending free movement may very well result in a marked rise in emigration by skilled and highly skilled workers from the new member countries. These workers would be lost to their countries of origin at the very time when the latter are entering a decisive phase of insertion in international trade 11 and engaging in a process of regional economic integration. While free movement, as we have seen earlier, is not in itself a precondition for intensive migration between countries at differing levels of development, is it nonetheless an essential factor in accelerating economic convergence and integration in a free trade area? The answer depends on the scale of migration in the period leading up to the establishment and development of the free trade area. In the case of NAFTA, for example, the key feature of movements between member countries was (and still is) the predominance of flows from Mexico to 268

the United States. In 1950, workers of Mexican origin represented 0.5 per cent of the labour force in the United States. The figure was around 4 per cent in 1990, and in some geographical areas or particular economic sectors it is much higher. The asymmetric character of these movements demonstrates the mutual dependence of two of the three NAFTA countries with regard to migration within the area. In this particular case, it may be posited that migration has already helped regional economic integration to develop, though it has not had sufficient impact to accelerate economic convergence. The emphasis placed, in the NAFTA context, on the liberalisation of trade and capital movements, without reference to free movement of workers, can be interpreted in two ways. On the one hand, liberalisation reflects the signatory countries’ determination to see that the sensitive issue of migration does not inhibit the deepening of economic and financial integration, in particular between the United States and Mexico. On the other hand, it indicates that one of the implicit objectives of trade liberalisation under NAFTA is to lessen the incentive to emigrate from Mexico to the other two countries, and chiefly the United States. 4.

Conclusion

Until recently, migration and its linkages with economic integration were viewed very differently in Europe and North America. Free movement was set out in the original proposals for the EEC, and the notion of establishing a trade area in order to limit potential flows of labour is not reflected in the Treaty of Rome. The introduction of free movement within the Community was not seen as an immediate step, however. For various reasons, a transitional period was set. In North America, inclusion of Mexico in the free trade area previously established by the United States and Canada stemmed partly from the wish to regulate immigrant flows from Mexico; the issue of migration was thus excluded from the negotiations leading to the three-country accord. The free trade agreements which the EU countries have signed in the course of the 1990s with various countries in the Mediterranean basin or Central and Eastern Europe now take the same approach: preventive substitution of trade flows for migration movements. Free trade is seen as an alternative to migration, as an instrument which may make flow regulation and control policies more credible. There must be some doubt, however, about the anticipated short and long-term effects of regional integration, on the one hand, and about the specific determinants of international migration on the other. In the short term, the effects of free trade on less advanced economies and the outlook for job creation both depend on the depth of regional integration being achieved between countries at differing stages of development. The anticipated effects of substituting trade flows for factor flows are restricted by the unwanted short-term side-effects of free trade on the less advanced economies. Domestic supply there cannot fully respond to demand in the regional market on account of their dependence on commodities or on goods making intensive use of unskilled labour. Foreign direct investment, relied on to offset fiscal shortfalls due to revenue forgone by dismantling tariff barriers, will increase appreciably only if the economic environment (public infrastructure, financial system, substantial sub-regional market, prospects of sustained growth in demand, and hence for rising income) is substantially enhanced. Unless employment can be stabilised in countries of origin, emigration is likely to increase over the short term. The positive effects anticipated over the longer term from the establishment of regional blocs (increase in human capital, development of public social and educational infrastructure, modernised transport and telecommunications, job creation in sectors with increasing returns) depend on the degree of regional integration proposed under the agreements establishing them. In the absence of

269

deepening regional integration (for instance, enlargement agreements between the European Union and the CEECs), the unwanted effects of free trade may well not be offset by transfer dynamics that will set off the modernisation of the economic and industrial structures of associate countries (countries of emigration): reallocation of resources to sectors with increasing returns, employment of skilled labour in those sectors. The determinants of international migration are partly immune to the substitution effects predicted by international trade theory. The turnaround of migration flows in some countries (Greece and Spain, for example) took place before the barriers to trade had been completely dismantled. For example, in the case of the relationship between the United States and Mexico on the one hand, and between Germany and Poland on the other, we observe in Charts 2 and 3 respectively that the intensity of foreign trade is high but that the reversal of migration flows has yet to clearly begin. Economic development and catch-up, the introduction of democratic political structures, manpower training and up-skilling, and a massive inflow of foreign direct investment, including investment by emigrants in their countries of origin, will all play a prominent role in this reversal. In the case of Poland, the prospect of joining the European Union fosters the catch-up process. In itself, the absence of the free movement of persons is not an obstacle to regional economic integration. The liberalisation of migration could remain an objective for the longer term, once economic convergence has gone further, and may proceed in transitional stages over a fairly long time-scale and for clearly specified categories of labour. Conversely, a free trade agreement cannot form a credible alternative to a proper migration policy. The latter could begin with the objective of discouraging unauthorised immigration, with emphasis on better regulation of flows within the free trade area and on monitoring that authorised immigrant workers have the same status as host country nationals. That objective entails active participation by the authorities in both host countries and countries of origin in combating illicit channels of immigration and the employment of unauthorised foreigners. Concerted action of this kind is particularly necessary when the member countries of a given free trade area have differing levels of development. In that connection, another facet of migration policy has still to be applied, co-operation between host countries and countries of origin. When such co-operation is first set in place, it often omits to involve those principally concerned, the migrants themselves.

270

NOTES

1.

See Mouhoud (1996) and Oman (1997).

2.

See principally Rivera-Batiz and Romer (1991b); Krugman and Venables (1990).

3.

Most authors who study the links between growth and international trade using new models in which technical progress is endogenous conclude that the economic integration of a country into a large area increases that country’s growth and therefore its ability to catch up on others (Krugman, 1991; Rivera-Batiz and Romer 1991a and 1991b inter al.).

4.

If two identical countries are integrated, the stock of human capital is doubled; the growth rate more than doubles since the efficiency of human capital increases with size (Rivera-Batiz and Romer, 1991a).

5.

For example, the announcement in the early 1980s that Portugal and Spain would be entering the European Community had a positive influence on investors’ expectations and they redeployed and increased their direct investment in both countries even before 1986, the actual year of entry. The 1997 announcement of the enlargement of the European Community to include certain Central and Eastern European countries may also produce positive expectations for those countries. See Mouhoud (1997) on this subject.

6.

For an analysis of the links between trade policies and migration in the short and long term, see Schiff (1995).

7.

On the European-Mediterranean free trade agreement, see Cogneau and Tapinos (1995). On the links between regional integration and foreign direct investment, see Mouhoud (1998).

8.

Rodrik (1991) links the lowering of the FDI risk premium to the irreversible character attributed to free trade agreements.

9.

For an analysis devoted to such effects in Central and Eastern Europe, Africa and Latin America, see Schiff (1996).

10.

For an analysis in terms of Optimal Monetary Areas, see Mundell (1957). See also Loufir, Haunaut and Mouhoud (1997) on measuring nominal, real and structural convergence processes in 15 EU member states and adjustment problems between countries at different stages of development in the context of monetary union.

11.

In order to analyse the impact of trade liberalisation on certain developing countries, Lopez and Schiff (1995) establish a distinction between two groups of countries: those which have effectuated their demographic transition and those which are still experiencing rapid population growth. The results differ only slightly. According to Lopez and Schiff, free-trade will lead to an increase in the migration flow of unskilled workers from central and eastern Europe and the former Soviet Union (countries from the first group) but will have no effects on the migration flow of skilled workers. Conversely, they form the opinion that trade liberalisation will facilitate the reduction in the flow of skilled workers from developing countries with rapidly growing populations. This kind of analysis fails to take into account sufficiently the factors determining the demand for foreign labour in developed countries and accords excessive importance to demographic variables.

271

BIBLIOGRAPHY

ALBA, F. (1993), “La emigración mexicana a Estados Unidos y la iniciativa del Tratado de Libre Comercio”, in Vega Canovas, G. (ed.), Liberación económica y libre comercio en America del Norte: consideraciones políticas, sociales y culturales, El Colegio de México, México. COGNEAU, D. and TAPINOS, G. (1995), “Libre-échange, répartition du revenu et migration au Maroc”, Revue d’Economie du Développement, No. 1. COGNEAU, D., DUMONT, J.-C. and IZZO, P. (1998), “Intégration régionale, investissements directs et migrations dans l’espace euro-méditerranéen: enseignement d’un modèle d’Equilibre général calculable”, in OECD, Migration, Free Trade and Regional integration in the Mediterranean Basin (forthcoming), Paris. KRUGMAN, P. (1991), Geography and Trade, MIT Press, Cambridge Mass. KRUGMAN, P. and VENABLES, A. (1990), “Integration and the competitiveness of peripheral industry”, in Bliss, C. and Braga Macedo, K. (eds), Unity with Diversity in the European Community, Cambridge University Press, Cambridge. LOPEZ, R. and SCHIFF, M. (1995), “Migration and the skill composition of the labour force - The impact of trade liberalisation in developing countries”, Working Paper, World Bank Policy Research. LOUFIR, R., HANAUT, H. and MOUHOUD, E.M. (1997), Convergence des économies et intégration européenne, OFCE-EPEE-Evry report for the French Commissariat général du plan, June . LUCAS, R.E. (1988), “On the mechanics of economic development”, Journal of Monetary Economics, No. 22, pp. 3-42. MOUHOUD, E.M. (1996), “Régionalisation, globalisation et polarisation de l’économie mondiale: quelle place pour les pays en développement ?”, Région et Développement, No. 2. MOUHOUD, E.M (1997), “The links between migration, free trade and regional integration: specific characteristics of the CEECs”, in OECD/WIFO, Migration, Free Trade and Regional Integration in Central and Eastern Europe, Paris. MOUHOUD, E.M. (1998), “Investissements directs étrangers, migrations et intégration régionale”, in OCDE, Migration, Free Trade and Regional Integration in the Mediterranean Basin (forthcoming), Paris.

272

MUNDELL, R.A. (1957), “International trade and factor mobility”, American Economic Review, No. 47, pp. 321-335. OECD (1997), Trends in International Migration, Annual Report 1996, Paris. OMAN, C. (1997), “The policy challenges of globalisation and regionalisation”, in OECD/WIFO, Migration, Free Trade and Regional Integration in Central and Eastern Europe, Paris. RIVERA-BATIZ, L. and ROMER, P. (1991a), “International trade with endogenous technical change”, European Economic Review, No. 35, pp. 971-1004. RIVERA-BATIZ, L. and ROMER, P. (1991b), “Economic integration and endogenous growth”, Quarterly Journal of Economics, No. 106, pp. 531-556. RODRIK, D. (1991), “Policy uncertainty and private investment in developing countries”, Journal of Development Economics, No. 36, pp. 229-242. SCHIFF, M. (1995), “Trade policy and international migration in the short and the long run”, in Revue d’économie du développement, pp. 3-25. SCHIFF, M. (1996), “South-North migration and trade - A survey”, World Bank Policy Research Working Paper. TAPINOS, G. (1994), “Regional Economic Integration and its effetcs on employment and Migration”, in Migration and Development, New partnerships for Co-operation, OECD, Paris. VINER, J. (1950), The customs union issue, Carnegie Endowment for International Peace, New York. VINER, J. (1955), Studies in the Theory of International Trade, George Allen & Unwin, London.

273

Table 1. Maghrebian, Turkish and former Yugoslavians residents in selected European OECD countries, 1996 (thousands, percentage of total foreign population and percentage of foreign labour force) Foreign population1 Total foreign

Of which:

population

Algeria

Belgium

911.9

9.2

Denmark

237.7

..

France

3 596.6

614.2

Germany

7 314.0

17.2

Italy

1 095.6

..

Netherlands

679.9

1.1

Norway

160.8

..

Spain

539.0

..

Sweden

526.6

..

1 337.6

..

1 972

..

Switzerland United Kingdom

%

%

Morocco

1.0 .. 17.1 0.2

138.3

.. 0.2 .. .. .. .. ..

119.5

3.4 572.7 82.9

138.7 1.6 77.2 .. .. ..

15.2 1.4 15.9 1.1 10.9 20.4 1.0 14.3 .. .. ..

Tunisia 5.1 .. 206.3 25.7 44.8 1.6 .. .. .. .. ..

%

Turkey

0.6 .. 5.7 0.4 4.1 0.2 .. .. .. .. ..

78.5 36.8 197.7 2 049.1 .. 127.0 4.4 .. 18.9 79.4 42

% 8.6 15.5 5.5 28.0 .. 18.7 2.8 .. 3.6 5.9 2.1

Former Yugoslavia

% 0.9 13.5 1.5 10.3

8.1 32.2 52.5 754.32

4.0 4.8 10.9 .. 7.0 22.8 ..

44.3 32.8 17.6 .. 36.6 305.0 ..

3

Foreign labour force

Austria4

Total foreign

Of which:

labour force

Algeria 257.2

..

Belgium

335.3

3.1

France5

1 573.3

245.6

Germany6

2 559.3

..

Netherlands Spain7 Sweden8 Switzerland9 United Kingdom

218

..

161.9

3.1

218

..

728.7

..

1 031.7

0.8

% .. 0.9 15.6 .. .. 1.9 .. .. 0.1

%

Morocco .. 42.9 197.5 .. 32 59.2 .. .. 1.8

.. 12.8 12.6 .. 14.7 36.5 .. .. 0.2

Tunisia .. 1.9 81.0 .. .. .. .. .. 1.9

% .. 0.6 5.1 .. .. .. .. .. 0.2

Turkey 52.2 21.0 66.4 759.1 33 .. 7 35.6 13.7

% 20.3 6.3 4.2 29.7 15.1 .. 3.2 4.9 1.3

Former Yugoslavia 141.5 3.0 32.3 454.52 .. .. 23 134.6 8.5

% 55.0 0.9 2.1 17.8 .. .. 10.6 18.5 0.8

1. Data are from population registers for all countries except for France (Census), Italy and Spain (residence permits) and the United Kingdom (Labour Force Survey). 2. Data are for Serbia, Montenegro and Bosnia-Herzegovina. 3. Figures include unemployed except for Austria and the Netherlands. Data for Belgium and the United Kingdom are from the Labour Force Survey (Eurostat, 1995); for the other countries see the introduction to the tables of the Statistical Annex. 4. Annual average of valid work permits. Unemployed and self-employed are not included. 5. Data are derived from the Labour Force Survey (1995). 6. Data (as of 30 September) refer to salaried workers only. Figures cover only western Germany. 7. Valid work permits. Workers from the European Union are not included in total. 8. Annual average. Data are from the annual Labour Force Survey. 9. Data are counts of the number of foreigners with an annual residence permit or a settlement permit (permanent permit), who engage in gainful activity. Data are for 1995. Source: Trends in International Migration , OECD, 1997, Paris.

274

Chart 1. France and Spain: Relative intensity of foreign trade and migration flows, 1961-1995

60

3

50

Intensity of trade

2.5

40 2 30 1.5 20 1

10

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

1967

1965

1963

0 1961

0.5

Inflows of Spanish citizens to France (% of total)

Intensity of Spanish exports (1) to France Intensity of Spanish imports (2) from France Inflows of Spanish citizens to France (3) (% of total)

1. For exports (X) from i to j, the indicator is: (Xij/Xi.)/(M.j/M..) 2. For imports (M) to i from j, the indicator is: (Mij/Mi.)/(X.j/X..) The indicator equals 1 when the bilateral trade flows are absolutely proportional to the place occupied by both partners in world trade rankings. An indicator equal to 2 signifies that the trade flows are twice the volume that would normally be expected given the trading partners’ share of world trade. The share of France in world trade is calculated on the basis of data on foreign trade for OECD countries only; this would lead to underestimate the relative intensity indicators. 3. Inflows of Spanish permanent workers as a percentage of total inflow. Sources: OECD Foreign trade database; Trends in International Migration, OECD, 1997.

275

Chart 2. United States and Mexico: Relative intensity of foreign trade and migration flows, 1980-1995 Intensity of Mexican imports (1) from the United States Intensity of Mexican exports (2) to the United States

Inflows of Mexican-born persons to the US (%)

1995

1994

0 1993

0 1992

10

1991

1

1990*

20

1989

2

1988

30

1987

3

1986

40

1985

4

1984

50

1983

5

1982

60

1981

6

1980

Intensity of trade

Inflows of Mexican-born persons to the United States (3) (% of total)

* Change in series: before 1990, data on foreign trade are based on the CHELEM database (CEPII). From 1990 on, they are issued from the OECD Foreign trade database. 1. For imports (M) to i from j, the indicator is: (Mij/Mi.)/(X.j/X..). 2. For exports (X) from i to j, the indicator is: (Xij/Xi.)/(M.j/M..). The indicator equal 1 when the bilateral trade flows are absolutely proportional to the place occupied by both partners in world trade rankings. An indicator equal to 2 signifies that the trade flows are twice the volume that would normally be expected given the trading partners’ share of world trade. The share of the United States in world trade is calculated on the basis of data on foreign trade for OECD countries only; this would lead to underestimate the relative intensity indicators. 3. Data refer to fiscal years (October to September of the given year). The lower dotted line is related to the percentage of Mexican-born immigrants excluding persons who benefited from the IRCA regularisation programme. From 1989 to 1996, approximately 2.7 millions people benefited from this program (Of which 1.1 million during the 1991 fiscal year). Sources : OECD Foreign trade database; Trends in International Migration, OECD, 1997.

276

Chart 3. Germany and Poland: Relative intensity of foreign trade and migration flows, 1980-1995 Intensity of Polish imports from Germany (1) Intensity of Polish exports to Germany (1)

5

3

4.5

Intensity of trade

2.5

4 3.5

2

3 2.5

1.5

2 1

1.5 1

0.5

0.5 1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

0 1980

0

Polish citizens in Germany (% of total foreigners)

Polish citizens in Germany (% of total foreigners)

Note : see notes 1 et 2 of Chart 1 for the definitions of the relative intensity indicators. 1. Data on bilateral foreign trade and on German imports and exports are based on the CHELEM database (CEPII). Imports and exports of Poland are issued from International Financial Statistics Yearbook , IMF, 1996. The share of Germany in world trade is calculated on the basis of data on foreign trade for OECD countries only; this would lead to underestimate the relative intensity indicators. Sources: OECD Foreign trade database; CHELEM database (CEPII); Trends in International Migration, OECD Trends in International Migration , OECD, 1997.

277

NEW DIRECTIONS FOR MANAGING US-MEXICAN MIGRATION

by Demetrios Papademetriou Carnegie Endowment for International Peace, Washington

1.

Introduction

The political passions that mushroomed during the battles leading to the ratification by the United States of the North American Free Trade Agreement (NAFTA) led to increasingly extraordinary claims about its aims that tended to confuse, rather than inform the discussion about it. That confusion has made the ongoing assessment of the Agreement’s effects more complicated than needs be, while simultaneously obscuring the importance of some of its policy consequences that are not 1 directly trade/economic related. As the US debate moved from the understandable hyperbole that is involved in the “selling” of politically contentious issues, to irresponsible rhetoric, and then to apocalyptic “... if NAFTA fails” scenarios, it became increasingly evident that discussions about the Agreement would be more about the effects of trade liberalisation on less skilled workers (and less directly about free trade’s contributions to income inequality) than about NAFTA itself. This has become even more obvious with the Clinton Administration’s current difficulties in obtaining Congressional approval for “fasttrack” trade negotiating authority despite a sympathetic Republic Congressional leadership. And as the global trading regime has been put on public trial, it has become even more apparent that the fight 2 is really about two sharply contrasting visions of America’s strengths and future direction. In its ideal form, the first view is very much anchored in the past -- the vision of an economy relying to a substantial degree on well-paying manufacturing jobs in which often unionised workers with relatively modest amounts of education produce products to be sold to US consumers without challenge from foreign competition. In addition to not having existed for a quarter of a century and not being likely to reoccur, this is also a vision clouded by frustration and a pronounced sense of betrayal (see below). The second vision is that of an activist, forward-looking America that perceives the present as it is and faces the future with confidence, actively embracing it and energetically, even aggressively, trying to shape it in ways that will best serve its long-term interests. This clash of visions has troubled the United States for much of the time since the 1970s, as the country has gone through a difficult period of economic restructuring. The fact that the ratification of NAFTA suddenly became the issue on which the two would wage an all out war may have had less to do with any “unique” attributes of NAFTA than with the economic and political circumstances of the moment: i) a stagnant economy and high relative unemployment that affected managers and professionals, segments that until then had been largely spared the effects of restructuring; and ii) the

279

ability of the Democrats to exploit those circumstances to recapture the White House (the dominant political slogan of the time was “it’s the economy, stupid”). This report is not meant to take sides on whether those that opposed NAFTA had a legitimate grievance. Rather, it is intended to emphasise that the core of that grievance may not have been with the Agreement per se. It was (and in many ways it continues to be) a grievance with official Washington and many of its high priests in the private sector who advise it. At the heart of this grievance was a deep anger about the unilateral near-abrogation of the implicit social contract between the government and the governed -- a contract on which the prosperous post-Depression America was built -- and about the unequal burdens and rewards of restructuring and globalisation for different social and economic classes. It is this still unresolved anger that turned the debate about a highly esoteric and technical document with very little effect on the United States into an antiWashington, anti-insider, anti-expert crusade, whose repercussions threatened to stir up a difficult-tochannel social protest movement. Echoes of that crusade are still resonant and have chastened the government in a number of areas of social and economic policy, ranging from minimum wage legislation to a variety of health and related initiatives. At the risk of further burdening NAFTA, the Agreement’s effect on unauthorised immigration -- reviewed four years after taking effect and its projected effects in the longer-term -- is inescapably political and hence a policy question of the first order. Furthermore, it touches on other policy issues that do not specifically pertain to economic integration, such as the broader US/Mexican relationship in the fight against drug trafficking, etc. These more ancillary issues seem to have displaced the trade/economic relationship as the axis about which the success of “integration” is measured to the point where during the first three years of NAFTA, the US/Mexican relationship appeared to have been reduced to border and unauthorised immigration issues, and more recently to drug trafficking first and unwanted immigration second. Remarkably, immigration had played a minor role in US calculations about whether to enter into trade negotiations with Mexico and had provided little of the energy driving the Agreement within the Administration during the negotiations -- except perhaps in a “defensive” way. Specifically, the administration had to anticipate concerns that opening the US labour market “too much” to Mexican business persons could entangle the Agreement in the gathering storm about immigration and sidetrack, or even derail it. (It was not until the negotiations were nearly complete that the wisdom of this precaution became obvious, as immigration became a critical and, for a period, a nearly uncontainable political issue in the United States.) As a result of this calculation, the Agreement (Chapter 16) does not facilitate entry into the United States for Mexican business persons and professionals, treating 3 them unequally with Canadians in this respect. 2.

NAFTA and unauthorised immigration pressures

One of the most important non-trade related issues might have a significant enough positive economic impact on Mexico (in terms of creating jobs, raising the standard of living of Mexicans from sectors and regions that contribute the most US-bound immigrants, and increasing Mexican national income) so as to reduce pressures for unauthorised Mexican migration measurably. Claims by some NAFTA boosters had painted the agreement as a virtual panacea in the long-standing but intermittently pursued US effort to contain unauthorised immigration and had created correspondingly unrealistic expectations. On the other side of the philosophical spectrum, counterclaims that NAFTA would

280

increase such immigration substantially had put both NAFTA supporters and the community of immigration experts on the defensive. This latter view was based on the common sense judgement that NAFTA would likely augment pressures for emigration in the short to medium-term, but that after a “transition” period of undetermined duration such pressures would likely abate as a result of NAFTA-assisted 4 development. It merits noting, however, that this judgement was primarily of the nature of “informed speculation” and that estimating the temporal threshold when migration pressures would abate as a result of development was (and is) impossible. One can safely assume, that even under the most favourable NAFTA scenario, it will take Mexico considerable time to bring its employment problems under control, although the recently released Binational Study on Migration (1997) estimates that Mexico may be able to create enough jobs by the end of the next decade to accommodate its first-time labour market entrants. However, since Mexico’s overall job-creation deficit is expected to continue for the foreseeable future, it is indeed likely that some of the Mexicans who enter the migration stream as a direct or indirect result of NAFTA will seek to enter the United States illegally and obtain unauthorised employment there. What is equally clear, however, is that in the absence of NAFTA, the economic transition period (i.e. the period during which Mexico would continue to be in the grasp of development forces that create and aggravate external migration pressures) would have been drawn out enormously, causing 5 more unauthorised immigration implications for the United States. Of course, during this transition period, and until Mexico produces enough additional economic activity to make the uncertain trek to the United States less attractive to Mexican nationals, the United States will have to make certain that it improves its border controls and that it more stringently enforces its labour market control laws -- actions incumbent upon a state regardless of regional trade accords. 3.

Managing the US-Mexico relationship in the shadow of unauthorised immigration

Unauthorised migration is a challenge that confronts all OECD countries. Successfully managing it -- let alone controlling it -- cannot be achieved unless substantial gains are made in parallel efforts to reduce the more egregious instances of economic scarcity and to redress the politics and economics of exclusion, which are at the heart of most migration flows. Without such improvements, even draconian control measures to eradicate unauthorised immigration would be only partially successful. Yet, the US-Mexico migration relationship may be simultaneously less and more complex than has just been suggested. It may be less complex because it is primarily economic improvement, rather than overt social and political oppression and conflict, that is the principal motive for northward migration. It may be more complex because the ready availability of US jobs for poorly educated but hard working Mexicans, and the existence of mature migration networks nurture and reinforce this movement. This is what many Mexican analysts and politicians mean when they repeat with mantralike conviction that Mexicans just respond to US employer demand. The reality is that what has developed is a near natural symbiosis between the goals of Mexican migrant workers and the interests and expectations of employers in increasing numbers of US economic sectors. Yet, despite a migration relationship more than a hundred years old, which has developed its own social rules and economic logic, US policy discussions about border management and unauthorised immigration controls continue to focus almost exclusively on unilateral actions. This occurs despite

281

the fact that it is becoming obvious that such actions are likely to fall far short of the mark and must thus be augmented by thoughtful bilateral and regional initiatives. While several vehicles exist for pursuing migration policy, and other policy objectives, none has had a chance to mature enough to evaluate its adequacy. Not surprisingly, most meaningful US-Mexican co-operation has occurred since the beginning of the courtship surrounding NAFTA -- making the agreement the most promising instrument in recent memory, from the perspective of the United States, for pursuing bilateral co-operation on a number of key issues that it addresses. For instance, the US State and Justice Departments have been engaged in a dialogue with their institutional counterparts in Mexico on a large number of specific issues related to immigration as part of the US-Mexico binational relationship (Binational Commissioner), which is now more than fifteen years 6 old. And an evolving Department of Labor relationship with Mexico’s Ministry of Labour, both enhanced and complicated enormously by the North American Agreement on Labor Cooperation 7 (NAALC), provides a potentially crucial, yet still very much immature, additional vehicle for such discussions. In addition to these formal bilateral contacts, one must also emphasise the opportunities created by the numerous informal and mostly private sector led initiatives. Perhaps most notable among them has been an innovative, quasi-official joint US-Mexican study of migration (the “Binational Study”) that has created a unique “knowledge baseline” on what has until now been highly contested terrain, both theoretically and empirically. Other important initiatives include a number of other binational studies on narrower issues; numerous “dialogues” that engage, in different combinations, representatives from the business community, the academy, civil society institutions, and federal, state, and local government officials (often in their private capacities); conversations with official and quasi-official officials at the border that emphasise the resolution of local problems; and a myriad of small and large business and government contacts across an array of spheres. Thus the NAFTA may well have been an essential prerequisite for beginning a serious dialogue with Mexico on a variety of mutually sensitive issues -- with immigration having the dubious distinction of being at the very centre of discussion during the Agreement’s first years. It is important to appreciate, however, that further meaningful co-operation and concrete results from such a dialogue hinge on the ability not just to sustain the dialogue but to engage it in a manner that gives roughly 8 proportional weight to actual priorities and respects the sensitivities of both sides. 4.

Building upon the relationship

Two issues in the bilateral relationship have the longest history and carry the largest social and institutional “baggage”, and thus help demonstrate both the complexity of the relation but also the available room for co-operation: US border controls and the human and labour rights of Mexican nationals in the United States. Eliciting Mexico’s co-operation in US southern border management efforts The starting point of any promising conversation about controls must be border “management”, a concept far broader and with a different dynamic than border “enforcement.” Border management implies a comprehensive effort that seeks to simultaneously facilitate legal access while better controlling unauthorised entry into a country. These are in some ways competing functions. Specific measures to achieve the joint goals of border management include developing visas that are more 282

resistant to fraud but more easily read, continuously improving dual use border physical infrastructure, and expanding human and technological resources. While the United States has made strides towards better understanding what effective border management implies, and can point to significant improvements in both facilitation and control, most of the control measures have been unilateral policing operations. The various iterations of “Hold the Line” -- the bold El Paso, Texas initiative that inaugurated both a new approach and a new commitment to border control and which exemplifies the emphasis of US approaches to date -- have convincingly demonstrated that a targeted saturation enforcement effort concentrating on a small segment of the border can indeed control unauthorised entries through that segment. A number of other things are, however, much less clear. − Whether this control-through-deterrence strategy reduces overall unauthorised immigration. The research literature is highly sceptical about the relationship between saturation policing in a few high traffic areas and overall reductions in unauthorised entries, pointing to several unintended but predictable outcomes of tighter enforcement. These include changes in crossing patterns and routes by unauthorised immigrants and the converting of many unauthorised border crossers from sojourners to long-term immigrants. If this is indeed the case, upping the enforcement ante becomes a never ending process, for otherwise saturation policing of some segments will only have limited effect. − Whether the United States Government and US taxpayers will be willing to continue to make the even higher financial commitments a saturation enforcement strategy requires over time. − Whether the political establishment, affected ethnic communities, civil society institutions, and popular opinion will speak with a single voice about interior controls -- especially with regard to their implications for civil rights and liberties and the acceptable reach of government. − Whether Mexico will be willing and able to continue to live with the domestic political fallout from a US approach to immigration enforcement that is increasing militarised 9 and police-oriented. Responding to Mexico’s human and labour rights concerns There is no doubt that experimentation in border controls is important if the United States is to develop better methods of assuming its legitimate sovereign responsibility of controlling unauthorised entry into the country. What muddies the waters, however, is the imbalance between the diligence of that effort and the relative indifference to Mexico’s long-standing concerns about the treatment of its nationals by US employers and, since 1996, the fact that Mexican nationals have been “targeted” both by several US legislative initiatives (ranging from the removal of social safety nets for legal immigrants to new income requirements for sponsorship -- both of which will affect the poorest immigrants most and, hence, disproportionately persons of Mexican origin) and by the Administration’s increasing efforts to remove unauthorised immigrants (most of whom are again Mexican).

283

Part of the explanation for this relative indifference to each other’s sensitivities lies simply in the “dialogue of the deaf” character of many aspects of the relationship on migration. Consistent with this image, each side focuses entirely on its agenda and domestic political requirements and fundamentally treats the other side’s interests as something akin to background noise (if often 10 sympathetically). 5.

Where are we now with immigration controls?

Despite extraordinary investments in physical and human infrastructure at the border (the Border Patrol grew by about three fifths between 1993 and 1997; it is slated to add an additional 1 000 positions per year till the year 2001; and it is equipped with better technology than any other border control effort anywhere), it is not clear that unauthorised immigration has been affected beyond the specific places where these investments have been made (United States General Accounting Office, 1997). In fact, entries have become more difficult and costlier (both in terms of the dollar value of crossing and the consequences for miscalculation, which can at times be counted increasingly in death tolls). These additional costs -- the pecuniary ones borne primarily by the migrants’ sponsors in the United States, the non-pecuniary ones more directly by the migrants and their family -- probably deter those not fully committed to crossing the border. They have, in any event, diverted the flow to more distant and less hospitable routes. It is also possible that modifications in the perceptions about the border’s vulnerability may eventually affect the behaviour of new migrants. Moreover, because the new controls make unassisted entry much more difficult, they may increase reliance on professional smugglers. For instance, the increasing reliance on professionals for crossing is thought to have kept the probablility of being apprehended fairly steady despite vast increases in 11 Furthermore, profits accruing to smugglers and forgers makes these activities enforcement. attractive to organised crime and also allows them to become more technologically sophisticated in their methods. While this development may increase the probability of successfully penetrating the border, in some ways, it also creates a greater incentive and a more familiar target for governments to attempt to infiltrate, thus offering the potential for greater regulation. Finally, increased costs decrease circular migration significantly, so now and in the near future at least, the total size of net unauthorised migration is likely to continue to rise despite additional border measures. If the new status quo is having so many perverse effects, is there still a value in continuing to invest in it? The answer to this question must be a qualified “yes,” even if data about the effort’s success are ambiguous, even if the approach generates greater incentives for smuggling, forging, and corrupting public officials, and despite the fact that it probably curbs unauthorised migration primarily at the margin. A key reason is that the appearance of a more orderly border is a precondition to a more mature and even-handed bilateral conversation about immigration in the future. As Andreas (1998) has argued, symbolism -- in the form of managing perceptions about the border -- matters a great deal. The success and adeptness of the Administration in conveying the appearance of effective border control matters as much now as California’s ability to convey an image of a border totally out of control mattered in 1993-94. One key measure of the political effectiveness of each initiative is that those that oppose the message find it extremely difficult to be heard -- let alone to be heeded -almost regardless of the soundness of their position. Put differently, the anger of Californians in 1993-94, which infected much of the country, has now been replaced with a sense of reassurance (at least in the rest of the country) that significant progress against unauthorised immigration is indeed being made, that orderliness has by and large returned to the border, and that the Administration, 284

despite continued attacks by some Republican legislators, has managed the issue of border controls effectively. But are these efforts consonant with the new relationship exemplified by NAFTA? In other words, is saturation policing of the border compatible with the spirit of NAFTA? These are complicated questions that may require nothing less than a re-conceptualization of borders. Suffice it to suggest at this time that the openness in commercial, economic, social and cultural exchanges brought about by regional integration need not necessarily imply less emphasis on border regulation (although the nature of the challenge to virtually all borders is no longer maintaining their integrity from possible military aggression) (Andreas, 1998). It implies instead equal parts of facilitating legitimate transactions while keeping out terrorists, new types of unsafe products and illicit substances, and smuggled people). This new emphasis on inspection is more in the nature of rebuilding borders, than in dismantling them, and if done properly, can allow the somewhat paradoxical coexistence of what Andreas calls the “borderless economy”, which economic integration demands, with the “barricaded border” that the new challenges to the nation state appear to require. 6.

What should we be doing next?

There are at least two ways to think about the next three-to-five years of North American integration and its effect on unauthorised immigration. The first is probably best captured by a recent paper by Sidney Weintraub (1997) that argues for doing more of what is already being done while waiting for development to take hold. That includes more border inspectors of every kind, better technologies and detection methodologies, more emphasis on enforcement at the workplace, etc. This is certainly the pattern that the US Government appears committed to following despite evidence of meagre results and increasing Mexican frustration. The second way argues for less patience but recognises that neither side can afford to walk away from the dialogue. Still, the apotheosis of dialogue has natural limits. These can be reached when the prospects that the dialogue will become meaningful -- and thus lead to concrete results -- for both parties are no longer realistic. The US/Mexico dialogue on immigration may not be at that point yet. However, it seems unlikely (if not inconceivable) that the United States can continue to count on Mexico’s remarkable understanding and co-operation as it has done so far. A third way of looking at the issues starts with acknowledging the political imperative of continuing with a robust border presence but then moves beyond unilateral controls to engage Mexico in a serious (and hence relatively open-ended) conversation about what it might take to gain its active cooperation in looking at borders and controls within a truly regional framework. This conversation would include institutional actors other than just immigration and border control authorities from both sides and would engage head on the issues of economic and socio-political development whose resolution would gradually eliminate the continuing unauthorised immigration crisis. Only by opening up a conversation about the future will the many smaller measures that each partner now performs half-heartedly continue to be possible -- from Mexico’s anti-loitering and anti-smuggling measures through Grupo Beta forces, to US attention to the human rights and welfare of Mexican nationals, to co-operation on drugs and even limited labour mobility issues. And only by making progress in these areas can a worthwhile North American vision be possible -- a vision in which each country takes responsibility for its people and their actions. This vision includes a border that works,

285

communities in both countries that are viable, a real commitment to rights, an active civil society, and 12 an emphasis on policies that improve peoples’ lives throughout the region.

7.

Conclusion

Responding responsibly to the challenge of international migration is an immense test that requires the thoughtful and sustained attention of all members of the world community. Being successful in meeting the challenge, so as to continue to benefit from the opportunities migration also offers, requires that we properly understand both the structural causes of migration and the personal ambitions and aspirations of those who move, and that we are willing to push the proper policy levers (including trade-offs), in the proper sequence. Whether the issue is the most appropriate border or labour market controls, creating or expanding channels of regular immigration access, safeguarding the human and labour market rights of all workers (regardless of immigration status), affirming protections for bona fide refugees, or working co-operatively to achieve measurable progress in the development of the “South,” the only viable long-term solution lies in international co-operation. As the US Commission for the Study of International Migration and Cooperative Economic Development (1990) pointed out, many less developed countries face a “stark dilemma” between “exporting goods and services to create jobs at home and exporting people”. With NAFTA now four years old, has the cause of co-operation spilled over to other realms? Procedurally, and in terms of building the institutional infrastructure for a deeper relationship, the answer is a qualified yes. In a substantive sense on immigration, however, NAFTA has so far been little more than a small downpayment on the long-term investment in a culture of co-operation among the three North American partners. Of course, investing in a long-term co-operative relationship (with the promise such an investment holds for becoming the foundation for fair, equitable, and responsible ways of resolving differences) is preferable to the “viewpoint”, which until the late 1980s had resulted in Mexico’s being treated -- and subsequently being induced to act -- like a troubled neighbour who is mostly the source of problems. Moreover, NAFTA should not be seen as the last word on either the bilateral or the trilateral relationship. Far from it. To loosely borrow from Churchill’s views about a united Europe, broad relationships between and among neighbours are living things that grow and adapt in response to shifting conditions on the ground. NAFTA can indeed make an important contribution to the growth of a more successful “living” relations. A final point also needs emphasising. The pursuit of a deeper relationship must be motivated as much by self-interest as by altruism or philanthropy: without sustained development, and given the force of demographic momentum (that leads the absolute size of each of several subsequent generations to be greater than the previous one despite declines in fertility), many less developed societies will be unable to cope with their rates of population growth. Nothing short of the social and political stability of entire regions may be at stake. As we debate the benefits of working together, we might revisit the seminal 1992 study by the Congressional Office of Technology Assessment (OTA), entitled “US-Mexico Trade: Pulling Together or Pulling Apart”, which provided both succour and the analytical ammunition for many of the most thoughtful arguments in opposition to NAFTA. In his letter of the study’s transmittal, John

286

Gibbons, the OTA Director, noted that neither the most dismal nor the most optimistic NAFTA scenario is likely, and chose to emphasise the importance of institutions in managing change. [T]he key to success in managing the social and economic transformations of the coming decades lies with the institutions that frame public and private choices -- decisions made by employers, by workers, by government officials. Gibbons probably had public institutions primarily in mind and more formal public/private partnerships than the broader forms of “civil society” that are now in vogue. The spirit of his message, however, is both crystal clear and equally valid today: the change implied both by NAFTA and globalisation in general needs to be managed successfully or it will “pull-us apart.”

287

NOTES

1.

Among the specific issues were the Agreement’s effect on job gains and loses, the environment, sovereignty, worker’s rights in Mexico, labour standards in the United States, immigration, and the longterm relationship with Mexico.

2.

The “vision” concept is borrowed from P. Behr, “NAFTA Fight Is a Duel over Visions: Accord’s Friends and Foes Have Different Views of US Economic Strengths”, Washington Post, 15 September 1993, pp. D1,2,3.

3.

The United States sought to pattern the immigration aspects of the Agreement after the US legal and regulatory regimes -- much as it had done with the free trade agreement with Canada a few years earlier. The overall goal was to obtain a single set of stable and transparent commitments with which it could live both substantively and politically. This created two sets of challenges. First, the Mexicans had to be persuaded to accept disparate (in effect, inferior) treatment in certain of the terms of their nationals’ access to the United States when compared to Canada’s access. (The United States cited the fact that it had tightened the relevant immigration provisions since the agreement with Canada had been negotiated and its concern about possible Mexican visa fraud -- about one-third of visa applications from Mexicans are rejected as fraudulent.) Second, the United States had to negotiate an instrument whereby entry by US business persons and professionals into Mexico would be fully transparent and guaranteed in what was otherwise an immigration regime allowing for an extraordinary amount of discretion, even arbitrariness. In both instances, Mexico’s focus on the real “prize” -- securing a proper trade agreement --and, to a lesser degree, the explicit prior understanding that “immigration” would not be on the negotiating table, enabled the United States to meet both challenges successfully.

4.

This position was anchored on the 1990 final report of The Commission for the Study of International Migration and Cooperative Economic Development that had noted that “[t]he transformations intrinsic to the development process are at first destabilising. They initially promote rather than impede migration. Better communications and transportation and other improvements in the quality of life of people working hard to make a living raise expectations and enhance their ability to migrate (1990:34; emphasis added).”

5.

The behaviour of networks and how that behaviour might affect the NAFTA’s short-to-medium-term immigration effects has been one of many areas where the growing business of modelling this and similar agreements’ effects has provided little useful guidance. Networks complicate most calculations about immigration enormously. For instance, take the issue of the enhanced legal access to the United States brought about by US immigration legislation in 1986 and 1990. The research literature suggests strongly that the legalisation programs of the former, and certain “family unity” provisions of the latter, by increasing substantially “main gate” (that is, legal) immigration opportunities for Mexicans and their immediate family members, also contributed to additional unauthorised migration. Since immigrants travel along pathways established and being kept active by immigrant networks, and since a large share of unauthorised immigration is made up of immigrants who reunify with their families de facto, (i.e. outside of the immigration laws) and/or are otherwise “plugged-in” into specific networks—expanding legal immigration opportunities may in fact also contribute to unauthorised immigration precisely by that very group of Mexicans for whom illegal immigration and employment in the United States has become part of its economic survival strategy, i.e. by those coming from communities in Mexico which have by now historical immigration relationships with their US “sister” communities. (Massey et al., 1993, 1994; Massey and Espinosa, 1997; Kritz, 1997). That is the reason that so many researchers studying Mexican migration consider networks as near independent variables in immigration to the United States. By extension, once a community’s migration contacts with the US mature, migration from it becomes nearly impervious to regulation. In other words, that body of literature suggests that the overall NAFTA effect on unauthorised migration from Mexico for the foreseeable future will be likely small.

6.

These same two Departments, occasionally with representation from the Department of Labor, are also routinely involved in similar discussions with Canada on a number of immigration-related issues, including a now stalled but nonetheless nearly-completed conversation regarding the designation of each other as “safe” countries for the purposes of allocating responsibility for the adjudication of asylum

288

claims. Intermittent trilateral discussions with Mexico fashioned after the US-Canadian ones have also occurred, but Mexico is far less inclined to (and far less capable of) entering into and effectively discharging similar co-operative obligations. 7.

The NAALC evolved directly as a result of the NAFTA negotiations. The US Department of Labor sought and completed a Memorandum of Understanding (MOU) launching the relationship in 1991. The impetus for seeking the MOU was US concerns that without engaging the Mexicans on these issues, support for the entire agreement might be jeopardised. The courting of the US labour movement’s support (or, more accurately, the hoped-for attenuation of some of its opposition) also played a role in pursuing the MOU but was never thought to be a realistic objective. Following the Clinton election in 1992, the United States vastly expanded the MOU into the NAALC, over a reluctant Mexico and a highly sceptical Republican Party. The hard-won NAALC failed to change organised labour’s opposition to and has been at best marginally valuable to the US labour movement since then. Furthermore, it has again become a major all around point of contention in the Administration’s continuing attempt to obtain fast track trade negotiating authority.

8.

It is important to appreciate in this regard that the US is striving to develop an ever-closer relationship with a sovereign nation. Such relationships can only be successful in the long term if each party to the discussion is treated as an equal with respect to its sovereign responsibilities. Common approaches in any one-policy area are likely to be reached only if such approaches are demonstrably to the benefit of all partners -- or concessions are made to “compensate” for benefit imbalances. No negotiating course can sour relations among old or new friends faster than trying to impose unilaterally one’s views on the other in areas of extreme sensitivity.

9.

By and large, senior Mexican officials have been remarkably restrained both in their public statements and during their official contacts with US officials about these US initiatives. (The US rescue of the Mexican peso has certainly played a large role in that behaviour.) Privately, however, they complain bitterly and are indignant about the treatment of Mexico, while appearing to be at a loss about their options. Their level of indignation varies with the incident. For instance, their reaction to the initial El Paso actions, which they likened to the US blockade of Haiti, was one of shock; they in fact saw the initiative as treatment more appropriate toward an “outlaw” state. Subsequent incidents were seen variably as insensitive and unnecessarily provocative. What seems to make Mexican frustration even greater, however, is that despite their close co-operation with the United States on a number of sensitive issues (ranging from the interdiction of Chinese immigrants to working closely on anti-smuggling operations) they seem to be getting little consideration for their efforts.

10.

In one sense, there is nothing new to this dynamic. In fact, some analysts have questioned whether any bilateral “relationship” on immigration-related issues at the official level either really exists or is likely in the near term.

11.

Although past studies have consistently found that the probability of apprehension hovered at approximately 30 per cent (Espenshade and Acevedo, 1995; Singer and Massey, 1998), no recent studies have been published that determine whether new strategies are effective in either deterring immigrants from attempting an initial unauthorised crossing or in changing the probability of apprehension (GAO, 1997).

12.

Unlike NAFTA, the European Community (EC, originally, EEC) explicitly included a freedom of movement of labour component in its constitutive agreement, the 1957 Treaty of Rome. A major reason for this provision was because the Community’s core members needed workers -- a situation at considerable variance with US and Canadian objectives in the NAFTA or realities on the ground. Furthermore, and because of the EC’s unusual political and philosophical aims (the Europeans sought to avert future Continental wars through the deep integration of the economies of key states), that agreement’s basic architecture is very different than that of the NAFTA. Finally, and again unlike with the NAFTA, Europe opted for the approach of vast income transfer programs for agricultural subsidies and for large-scale Community investments in the improvement of the physical and social infrastructure of the less developed regions of the Community. In other words, North America and Europe set out to create vastly different living organisms. It should thus not be surprising that the two outcomes are not at this time really comparable. And although some may see in the NAFTA the first stage of a process leading to a far “deeper” integration than the current trade relationship sets up, support of that line of thinking at this time requires the making of vast inferential leaps. In fact, whether further integration is even conceivable in the near term may rest squarely on the successful management of the emerging relationship in precisely such non-trade related matters as unauthorised immigration, drugs, and similar issues.

289

BIBLIOGRAPHY

ANDREAS, P. (1998), “The U.S. Immigration Control Offensive: Constructing an Image of Order on the Southwest Border”, in Marcelo Suarez-Orozco (ed.), Crossings: Mexican Immigration in Interdisciplinary Perspective, Cambridge, Harvard University Press. BEHR, P. (1993), “NAFTA Fight Is a Duel Over Visions: Accord’s Friends and Foes Have Different Views of U.S. Economic Strengths”, Washington Post, p. D1. BINATIONAL STUDY ON MIGRATION (1997), Binational Study on Migratio Between Mexico and The United States, Mexican Ministry of Foreign Affairs & The US Commission on Immigration Reform, Mexico. COMMISSION FOR THE STUDY OF INTERNATIONAL MIGRATION AND COOPERATIVE ECONOMIC DEVELOPMENT (1990), Unauthorized Migration: A Development Response, Government Printing Office, Washington, DC. ESPENSHADE, T.J. and ACEVEDO, D. (1995), “Migration Cohort Size, Enforcement Effort, and The Apprehension of Undocumented Aliens”, Population Research and Policy Review, Vol. 14, pp. 145-172. KRITZ, M. (1997), “Population Growth and International Migration: Is there a Link?”, in A. Zolberg & P. Benda (eds.), Global Migrants, Global Refugees: Problems and Solutions, Berghahn Books, Providence, R.I. MASSEY, D.S. and. ESPINOSA, K.E. (1997), “What’s Driving Mexico-U.S. Migration? A Theoretical, Empirical, and Policy Analysis”, American Journal of Sociology, Vol. 102, No. 4, January, pp. 939-999. SINGER, A. and. MASSEY, D.S. (1998), “The Social Process of Undocumented Border Crossing Among Mexican Migrants”, International Migration Review, Vol. 32, No. 4 (Fall), pp. 561-592. UNITED STATES COMMISSION FOR THE STUDY OF INTERNATIONAL MIGRATION AND COOPERATIVE ECONOMIC DEVELOPMENT (1990), Washington, DC. UNITED STATES CONGRESS, OFFICE OF TECHNOLOGY ASSESSMENT, (1992), U.S. Mexico Trade: Pulling Together or Pulling Apart, US Government Printing Office, Washington, DC. UNITED STATES GENERAL ACCOUNTING OFFICE (1997), Illegal Immigration, Southwest Border Strategy Results Inconclusive: More Evaluation Needed, GAO/GGD-98-21, Washington, DC. WEINTRAUB, S. (1997), “Ways to Ease Migration Tensions Between Mexico and the United States”, Discussion paper prepared for the meeting of the United States-Mexico Consultative Group of the Carnegie Endowment for International Peace, 16 January, Los Angeles, CA.

290

ANNEX SELECTED COMMENTS

291

Comments on the Paper by Cécile Thoreau and Tania Paracini, “Demographic Situation, Employment and Economic Performance in North America” by Pablo S. Reyes Pruneda Undersecretary of Labour, Secretaria del Trabajo y Prevision Social, Mexico

Recent economic indicators point to a return to levels comparable to those before the crisis −

Taking 1994 as the base year, during the first three quarters of 1997 GDP grew at an annualised average rate of 5.2 per cent. In 1997, GDP is estimated to have increased by 7 per cent on 1996.

−

Even though the inflation rate for the second quarter of 1997 was relatively high (6.49 per cent) compared to the 1994 rate for the same quarter (3.61 per cent), it is declining and is at a low level.

−

In August 1996, Mexico re-attained the level of employment of December 1994. Employment has since grown by 10.9 per cent.

−

The equivalent to almost 668 000 full-time jobs, (i.e. 939 000 posts) has been created each year in Mexico over the course of the past two years.

An examination of developments in job creation, unemployment, real wages and social conflict reveals that the labour market situation has improved (in a more or less substantial manner) −

Employment in the formal sector grew by 7.4 per cent between December 1996 and December 1997.

−

In November 1997, the open unemployment rate stood at 3.3 per cent. This is virtually the same as the rate for December 1994 (3.2 per cent) and lower than that in November 1994 (3.9 per cent). Since August 1997, the rate has fluctuated within the range of 3.0 and 3.5 per cent.

−

As the inflation rate has been lower than expected, contract-based salaries have increased in real terms. In 1997 for example, nominal wages increased by 18 per cent while recorded inflation was no higher than 15.7 per cent

−

Employment creation has increased whilst at the same time the number of posts suppressed has decreased. In 1995, posts suppressed were equal to 1.25 per cent of the posts created. This percentage fell to 0.67 in 1996 and to 0.48 in 1997. This indicator bears witness to the improvement in the employment situation.

293

The document states that employment creation has taken place principally in the informal sector and in the manufacturing industry (Maquiladoras) −

From December 1994 to October 1997, 967 465 jobs were created in the formal sector; 348 073 (i.e. 35.9 per cent of this total) concerned the Maquiladoras (export sector). If one chooses as the point of comparison the lowest level of employment in the formal sector, July 1995, the corresponding figure is 312 540 jobs created in the Maquiladoras, i.e. 21.4 per cent of the total.

−

Looking at the number of workers regularly making social security contributions, 1 412 148 posts were created between December 1994 and October 1997, 1 823 853 between August 1995 and October 1997. The Maquiladoras accounted for respectively 24.6 and 17.1 per cent of the posts created.

Migration: the document states that the 1994 crisis is, among other explanatory factors, at the origin of the increase in migration flows towards the United States −

According to the results of the inquiry on international migration across the northern border (EMIF) and contrary to that which has most frequently been put forward by the media, the devaluation of the peso led to a reduction in the emigration flow.

−

Regarding the relationship between unemployment and migration, the inquiry revealed that the proportion of the unemployed in the migration flows had increased: prior to 1994, they accounted for less than 30 per cent; data for the second half of 1996 suggest that between 40 and 60 per cent of the migrants had not undertaken any form of paid activity during the 30 days immediately prior to emigration.

−

The document suggests that the Maquiladoras have served as a trampoline for the migrants. Yet, as the EMIF inquiry shows, migrants are predominantly male whereas the majority of the employees in the Maquiladoras are female (migration towards the northern border is dominated by young women, unlike migration towards the United States).

−

Moreover, less than a quarter of the migrants originate from the northern States. Almost half originate from central Mexico.

−

In a similar vein, the document mentions, correctly, that the determining factor for migration is the demand for labour in the United States. The inquiry showed that the proportion of unmarried Mexicans has increased, which confirms the existence of plentiful job opportunities.

−

Finally, almost three quarters of migrants (again, according to the inquiry) had already spent periods in the United States with the intention of working there. Almost two thirds had already been employed in the United States and almost 70 per cent had relatives in the country. Hence, the character of migration is essentially structural with variations according to the state of the economy.

−

Only nine per cent of the money earned by Mexicans in the United States is transferred to the origin country. The remainder is absorbed by the United States economy in consumption, taxes, business start-ups, etc.

294

Concerning the Maquiladoras −

This industry has its own dynamic, with marked variations in employment and real wages (respectively 19.4 and 2.6 per cent between 1996 and 1997, if one compares the average level over the period January-October for each of these years).

The document suggests that recent export growth is due to the 1994-1995 devaluation and to the fall in domestic demand; yet, Mexico has always conducted a policy of openness, especially with regard to export promotion −

In 1997, the total value of trade was in the region of one half of GDP. Mexico is the world’s tenth largest exporter and the most important in Latin America. This is due principally to its membership in NAFTA.

Other precisions −

The increase in income inequalities is not due simply to increased wage flexibility; it is also a consequence of the devaluation and of inflation.

−

In Mexico there does not of course exist employment insurance as such. On the other hand, there is a system of indemnities for those made redundant. Moreover, even though the minimum wage is very low, to affirm that workers are prepared to accept any job simply in order to survive is excessive.

−

Regarding the working-age population, the following remarks could be added: as a proportion of the total population it has increased considerably over the course of the last 25 years, passing from 50 to 60 per cent (in absolute terms from 24 to 56 million). Due to the demographic repercussions of past developments this number will continue to increase, and should reach 77 million people in 2010.

−

Although female participation rate is very low, be it in comparison to Mexican males as much as to American or Canadian women, it should however be noted that it doubled between 1970 and 1995, passing from 16.7 per cent to 32.1 per cent (the increase was at its strongest between 1980 and 1991).

−

According to the 1996 employment survey, the agriculture and fishing sector accounts for 22.5 per cent of total employment (i.e. 7 928 000 persons).

−

Employment in the manufacturing sector has indeed registered a sharp decline since the 1994 crisis. However, a revival has been underway since August 1995 and since April 1996 the annual growth rates (i.e. compared to the corresponding month of the preceding year) have been positive. In October 1997, recorded employment attained a level equal to that of December 1995.

295

−

If it is true that there exists some mismatch between job vacancies and the skills of the unemployed, the active labour policies (PROBCAT, CIMO, SNE) constitute a fundamental instrument as much for the unemployed in their search for work as for the employees who hope to improve their situation.

−

Without underestimating the importance of FDI, domestic saving is the driving force of economic growth and macro-economic stability. Current economic policy places a high priority on promoting this internal saving.

296

Comments on the Paper by Cécile Thoreau and Tania Paracini “Demographic Situation, Employment and Economic Performance in North America” by Andrew Sharpe Centre for the Study of Living Standard, Ottawa

I have a number of comments on the paper pertaining to all three countries, and several caveats on the data used in the paper.

Points related to the United States The paper states that “the United States is focused mainly on services”. I think this is somewhat misleading. Even though the share of total employment (and to a lesser degree current dollar output) in services is continually rising, this does not mean that the goods sector is becoming less important as a source of economic dynamism. For example, exports of goods such as airplanes or office equipment remain extremely important for the health of the US economy. It is stated that the US unemployment rate at 5.5 per cent is below its natural level. Inflation is supposed to rise when the actual unemployment rate is below the natural rate. Yet in 1997 the US unemployment rate averaged around 5 per cent and inflation actually fell. This suggests that the natural unemployment rate may have fallen. In another point related to the US unemployment rate, I am not sure it is correct to state that “unskilled workers in the United States face difficulty finding jobs”. There currently appears to be evidence of labour shortages in most parts of the United States, even for unskilled workers.

Points related to Mexico The paper notes that the low official unemployment rate in Mexico (3.8 per cent in 1996) does not have the same meaning as that in Canada and the United States. If this is the case one wonders why the OECD publishes this rate in the OECD Economic Outlook. It would be extremely useful if the paper attempted to provide an estimate of the Mexican unemployment rate based on Canadian and US norms and definitions.

Points related to Canada Use of the 1984-94 period (Table 5) for analysis of developments in the Canadian economy is somewhat misleading given the great difference in economic performance between the second half of the 1980s and the first half of the 1990s. A better periodisation for Canada-US comparisons would be the peak-to-peak 1981-89 period and the 1989-1996 or 1997 period. The years 1981 and 1989 were cyclical peaks in both countries. Such a periodisation shows that since 1989 Canada has experienced a markedly inferior economic performance relative to the United States.

297

I am not sure of the meaning of the statement that “Canadian finance houses are less stretched than their US counterparts and should be able therefore to profit more effectively from liberalisation of the Mexican banking sector”. Canadian banks complain that they are not big enough to be world players. Possibly the statement is referring to the high profitability and return on equity of the Canadian banking sector. In any case, the statement needs clarification. The paper fails to recognise a number of significant differences between Canada and the United States. Admittedly, these differences may be considered minor in the context of the developmental gap between Canada and the US on the one hand and Mexico on the other. But at least from the point of view of Canadians, they are important. Canada and the United States are still two separate countries and economies. Since 1989, Canada has experienced much higher unemployment and lower labour force participation than the United States. Another important difference is that income is much more equally distributed in Canada. Also, because of Canada’s more generous social safety net, the country did not experience any increase in inequality for total income, at least up to 1995, in contrast to major increases in inequality in the United States.

Data caveats There are a number of data issues that are relevant to the statistical material presented in the paper. I believe the paper would be strengthened by recognition of these issues, at least in footnotes. The OECD purchasing power parity for the Canadian dollar vis-a-vis the US dollar ($0.82 US per Canadian $) appears to have been used in the calculation of the 1996 Canadian level of per capita GDP. Some observers, myself included, feel this rate is too low. For example, Summers and Heston in the most recent Penn PPP tables estimate the PPP at 0.88 US per Canadian dollar. This significantly increases Canada’s standard of living relative to that in the United States. It is noted that productivity gains have been greatest in the United States in high technology industries. But this situation may reflect the use of hedonic price indexes for computers. This has resulted in a massive decline in the deflator for office equipment (primarily computers) and hence very rapid growth in real (constant dollar) output in this sector. If the computer price deflators are inappropriate, as some observers suggest, the rapid productivity growth of the high technology industries may be misleading. The paper points out that real wages in the United States fell 3 per cent between 1985 and 1995. I assume this number is derived by dividing nominal wages by the Consumer Price Index (CPI). The Boskin Commission reported that the US CPI has an average upward bias of 1.1 per cent per year. If this is so, then instead of a 3 per cent fall over the 10 year 1985-95 period, there has been an 8 per cent increase. The uncertainty associated with CPI bias should at least be mentioned.

Key characteristics and trends in the Canadian economy In a short paper describing economic developments in three countries it is not possible to do justice to any one country. This section presents what I consider key characteristics and trends in the Canadian economy that received inadequate or no treatment in the paper: −

Economic growth in Canada in the 1990s has been dismal, meriting the phrase “the great Canadian slump”. Between 1989 and 1996 growth of real GDP averaged 1.2 per cent per year, down from 3.2 per cent per year over the 1981-89 period. This has lead to a large gap between actual and potential output, with estimates of the size of the current gap varying from 2 to 10 per cent of GDP.

298

−

The rise in Canada-US unemployment rate gap from 2 percentage points in the 1980s to an average 4 points in the 1990s largely reflected weaker economic growth in Canada compared to the United States (1.9 per cent per year).

−

Because of the weak macroeconomic environment, the 1990s has been the lost decade for personal income growth for Canadians, with 1996 per capita real personal income 5.1 per cent below the 1989 level and per capita real disposable personal income 8.8 per cent below the 1989 level.

−

A key explanation of this weak output growth in Canada in the 1990s is the very weak demand growth experienced in the early 1990s due to the Bank of Canada’s restrictive monetary policy aimed at achieving price stability. This policy has produced lower inflation in Canada that in the United States since 1991.

−

The fiscal position of all levels of government in Canada has improved dramatically recently due to deep cuts in spending and a strong economy in 1997. The federal government will run a surplus in the 1998-99 fiscal year.

−

There is growing disparity in unemployment within Canada. In Quebec and the Atlantic provinces the unemployment rate remains above 10 per cent, while in the Prairie provinces the unemployment rate has fallen to the 5-6 per cent range.

−

The cost competitiveness of the Canadian economy is at an all-time high, at least vis-a-vis the United States, due to the greatly undervalued Canadian currency and low wage growth. This situation accounts for much of the merchandise trade surplus with the United States.

299

Comments on the Paper by B. Lindsay Lowell “Migration Trends and Policies in the United States” by Eva Haagensen Department of Immigrant And Refugee Affairs Ministry of Local Government and Regional Development, Oslo, Norway

B. Lindsay Lowell’s paper provides a useful description of migration trends and policies in the United States, with particular focus on the Immigration Act of 1990.

Migration legislation The author recalls that immigration and its regulation divide into historical phases, each with distinct characteristics. If immigration was initially unregulated, which is not likely to occur again in the foreseeable future, it is now regulated by a sophisticated and complex regime. Lowell also points out that the immigration acts of 1965, 1986 and 1990 have had unexpected consequences for the volume of migration flows into the United States, and that the failure to develop effective strategies to control unlawful immigration has blurred the public’s perception of the distinction between legal and illegal immigrants. This perception problem was also mentioned in Roderic Beaujot’s paper on Canada, and touches on an issue that is also of concern to European countries. According to Lowell, the Immigration Act of 1990 attempted to balance competing interests by: −

establishing annual numerical targets for total legal immigration;

−

creating a guaranteed minimum number of visas for close family members;

−

addressing present and future labour market needs by increasing the number of persons admitted for employment reasons, with higher priority given to professionals and highly skilled persons; and

−

creating a diversity class of admissions for individuals from nations that have not recently sent many immigrants to the United States.

As part of the intended consequences of the Act, admissions numbers have been higher than in the 1980s, and it is expected that future admissions will continue to grow. The unexpected consequences include backlogs in processing applications and a surge in naturalisations applications as of 1995. Naturalisations have not, however, resulted in increased applications for admissions in the family preference categories from the relatives of new US citizens. This may be due to the welfare reforms (1996) that placed additional requirements on sponsors.

Demographics and population growth Lowell’s paper also provides information on the general characteristics of contemporary US immigration. These can be summarised as follows:

300

−

Mexican and other Latin Americans make up 40 per cent of the current immigrant population in the United States, and Asians make up another 40 per cent.

−

Most immigrants settle in just a few states, meaning that they have disproportionate impacts on them: California, New York, Texas, Florida and New Jersey.

−

Immigrants are younger than the US population but not as young as they used to be.

−

Female immigrants are now in the majority.

−

Employment based immigrants report (self-reporting) the greatest English ability.

−

Immigrants differ significantly in their educational attainment by class of admission.

−

The humanitarian classes of admission are less educated than US natives or the employmentbased, but are better educated than family admissions.

−

Family and humanitarian immigrants are primarily blue-collar workers, and employment and diversity immigrants primarily white-collar workers.

Lowell also discusses the current debate regarding the impact of immigrants on the employment and earnings of the native-born. He emphasises that there is evidence showing that immigration has had a largely positive impact on the national level, but that it has tangible costs to certain sectors of the labour market and to certain communities. For lesser skilled workers, there is evidence of wage competition, and communities with large foreign-born populations, especially with many school-age children, have high costs associated with foreign-born households. He adds that undocumented immigrants generate more in public costs than they contributed in revenues to governments. It is estimated that international migration makes up somewhere between one-fourth and one-third of the net annual US population increase. There are approximately 800 000 legal admissions and an estimated 200 000 to 300 000 unauthorised admissions yearly, yielding a net annual increase in the foreign-born population of about 700 000. The paper also draws attention to the effects of changes in welfare legislation which have reduced the eligibility of certain immigrants to welfare benefits. Evidence suggests that these changes have been partially bypassed through increased naturalisations. Difficulties in enforcing this welfare legislation are expected.

Concluding comments States, or governments, define migration policy and consequently how migration questions are framed. Lowell’s paper, but also the paper of Roderic Beaujot, seem to indicate that solving the problem of illegal immigration is a high priority in Canadian and US immigration policy. It is, moreover, an issue that also concerns European nations. The stabilisation and economic growth of the regions concerned, promises to make a significant contribution to the reduction of undocumented migration. Economic convergence would decrease the incentive to migrate and bring flows to more manageable proportions. Obviously, this would only be possible in the longer term. In the interim, it seems that the United States and Mexico would benefit from a temporal seasonal worker regime, particularly for the agricultural sector. This would enable people to legally work for a limited time in response to labour market demands and then return to their country of origin at the end of the season. Current border control measures may have the unintended effect of simply prolonging the stay of otherwise temporary migrant workers.

301

Comments on the Paper by Don J. Devoretz and Samuel A. Laryea “Migration and the Labour Market: Sectoral and Regional Effects in Canada” by Helena Gaytan-Fregoso University of Essex

Don DeVoretz and Samuel Laryea address the impact of migration on Canada’s labour market over the last 30 years. Canada has always been a country of immigration, yet the number and characteristics of these migrations have changed over time. According to the authors, fluctuations in the number of immigrants arriving in Canada follow changes in the country’s macroeconomic conditions and, in particular, changes in the trade regimes, employment rates, and other political considerations. According to traditional views on early twentieth-century immigration flows, which the authors use as a reference point for their analysis, immigration creates job displacement pressures and compresses wages, thus promoting the emigration of Canadian nationals to the United States. At the same time, if the immigration flows were accompanied by complementary capital, then immigration might result in higher employment rates, higher wages, and higher returns to capitalists. The authors, therefore, ask which of these two views applies to contemporary Canada and which preconditions coincide with those present at the beginning of this century. Modern Canadian immigration is characterised by a very high concentration in urban areas. In addition, the advent of the Canada-United States Free Trade Agreement and of NAFTA brought about a series of reforms to old migration regulations affecting the pattern of migration. Unlike the previous immigration experience, current immigrants possess substantial financial and human capital, thus labour markets in some regions and sectors may have received gains from immigration, whereas others may have suffered wage reductions and displaced native labour. The migration policy regime, which under the 1910 Immigration Act essentially only excluded some unhealthy and criminal elements, now involves a number of admissions categories and a points system for independent admissions under the 1976 Act. The paper is primarily concerned with determining what labour market effects have been produced as a result of the recent changes in immigration policy, in other words, as a result of the current demographic and skill characteristics of contemporary immigrants. To do so, the authors assume a binding quota of immigrants that increases the labour supply, and also, that wages and unemployment changes depend on the domestic elasticities of labour supply, and demand on the magnitude of the immigration quota. The more inelastic are the supply and demand for labour, the greater will be the reduction of domestic wages due to immigration. In addition, the job displacement effect would be greater the more elastic is the labour supply and the less elastic is the labour demand. These elasticities are tested for 125 industries that employ about the 95 per cent of the total labour force. The paper shows that there are no displacement effects considering all industries in the aggregate. However, estimation by sector demonstrates that in those industries with a higher concentration of foreign-born labour, there is greater substitution of native labour. Taking the case of 59 industries characterised by a high concentration of foreign born labour at skilled and unskilled levels, the elasticities show that immigrants are significant substitutes for Canadian-born labour, whereas capital is not yet a significant complementary input to immigrants. The result also reports that this lack of complementarity between immigrant labour and capital is the result of a high level of human capital represented by arriving immigrants.

302

The authors then attempt to determine how immigration of skilled labour may increase natives’ wages and employment due to the human capital brought by the immigrants. Analogously, they explain why earnings degradation may occur if the immigrants have little complementary human capital. To that end, a simulation exercise is presented that illustrates the effects of a 20 per cent increase in the number of recent immigrants. This is meant to establish the wage impact on native Canadian labour and other immigrants. The economy-wide impact proves to be minimal and to have no adverse impacts on the Canadian labour market. The wage decrease is of no more than 1 per cent, with all variations depending on the age cohort studied. The same analysis by sectors, shows that wage depression occurred in some primary sub-sectors and services. These sub-sectors are usually the entry points of immigrants into the labour market. Finally, DeVoretz and Laryea show whether there was a labour market adjustment mechanism open to Canadians who felt the impact of immigration, namely emigration to the United Sates. Statistical information shows that throughout the 1980s and 1990s there has been a considerable increase of emigration, especially among the highly skilled or trained. In conclusion, the paper shows that for a limited number of sectors and types of human capital some negative aspects of the job displacement and wage compression thesis hold. Moreover, emigration to the United States in some highly trained professions has reappeared in the midst of large scale immigration flows. It is unclear in the paper, however, whether the emigration of Canadians had its origin in the same 59 sectors where immigration and a high concentration of foreign-born labour proved to have displaced Canadian labour or in those sectors where wage degradation occurred. It should be added that since immigration to Canada from Mexico is negligible, there may be little effect on the Canadian labour market as a result of trade liberalisation through NAFTA.

303

Comments on the Paper by Rodolfo Cruz Piñeiro “Migration and the Labour Market: Sectoral and Regional Effects in Mexico” by Helena Gaytan-Fregoso University of Essex

Rodolfo Cruz Piñeiro summarises the literature indicating that Mexico’s participation in NAFTA has not yet affected the migration of Mexicans to the United States. He also describes the internal factors that have affected economic development, employment structure, and the territorial distribution of the population. The paper shows the existence of contradictory forces in the Mexican labour market, which have had an ambiguous effect on Mexico’s employment structure and migration, both internal and external. One of the influential factors is the accelerated process of internal migration from rural to urban areas. A saturation of urban job markets, as a result of internal migration, triggers emigration to the United States. However, the development of medium-sized cities in the 1980s appeared as a new force that redirected migration flows in Mexico. The author presents data from the Mexican census and a survey of the country’s northern border to illustrate that the new distribution of population in Mexico affects the pattern of migration to the United States. The economic dynamics and economic problems experienced in Mexico have transformed the employment structure. Influential factors include the greater opening to foreign trade and the austerity measures applied at the end of the 1980s. Transformations in the productive structure have affected the sectoral composition and the regional structure of employment. While these are certainly factors that one would expect to affect migration, Cruz did not supply sufficient data to show this link. As the paper argues, it is clear that current international migrations are largely due to urban problems in metropolitan areas and to the transformation of the productive sector. Nevertheless, in explaining the transformations suffered in the productive sector, the author omitted the important changes operated in the Mexican agricultural sector which undoubtedly played a significant role in the redirection of migration flows (see for instance, Young and Romero, 1993; Levy and Vanwijnbergen, 1994, 1995). In explaining the case of the maquiladora industry, the author confirms the view that this industry has attracted labour migration to the cities where they are established. The paper shows that according to data from labour surveys, while this type of industry is highly concentrated in border cities only a third of its workers expressed a willingness to migrate to the United States. Yet, the author does not point out the fact that the maquiladora industry is characterised by a high personnel turnover and that contracts are made on the temporary basis. Finally, the paper summarises some of the results obtained from the Binational Study of Migration in order to show the transformation in occupations and sector distribution of migrants to the United States. According to those results, the author agrees with Piore’s (1979) hypothesis that suggests that migration to the United States may be interpreted as a part of the configuration of the labour market at the binational level. That is, a market where the supply of labour from Mexico is as real as the demand determined by the current conditions of the US economy.

304

The paper is a remarkable summary of hypothetical causes affecting migration patterns in Mexico. However, the author does not give formal measures on how these factors determined the transformation of the labour market and promoted changes in the pattern of migration. More accurate measures of these hypothetical influences are therefore needed in order to make policy proposals.

REFERENCES

LEVY, S. and VanWIJNBERGEN, S. (1994), “Labour market, migration and welfare: agriculture in the North American Free Trade Agreement”, Journal of Development Economics, Vol. 43, No. 2, pp. 738-754. LEVY, S. and VanWIJNBERGEN, S. (1995), “Transition problems in economic reform: agriculture in the North American Free Trade Agreement”, American Economic Review, Vol. 85, No. 4. PIORE, M.J. (1979), Birds of Passage: Migrant Labor and Industrial Societies, Cambridge University Press, Cambridge and New York. YOUNG, L. and ROMERO, J. (1993), “Constant growth and transition to a dual-dynamic model of the North American Free Trade Agreement”, El Trimestre Economico, Vol. 60, pp. 353-370.

305

Comments on the Papers by Sidney Weintraub,by John Kirton and by Fernando de Mateo Venturini “NAFTA, Foreign Direct Investment and Economic Integration: US, Canadian and Mexican Approaches” by Denis Audet Trade Directorate, OECD

We have three positive assessments of NAFTA from representatives of the three NAFTA countries. John Kirton notes that the stock of Canadian FDI in Mexico increased six fold between 1991-96, with a majority going in sectors previously closed to FDI. He further adds that Canada nevertheless remains a host of FDI, with increased inflows from both NAFTA and non-NAFTA countries. Syndey Weintraub stressed that “10 million jobs had been created in the USA in the four years of NAFTA’s existence”; that “trade in all directions has boomed”; and that “NAFTA is uniting the economies of North America”. Fernando de Mateo, for his part, noted that “Mexico was the second largest recipient of FDI among developing countries” between 1994-96; and that the export propensity of foreign affiliates in Mexico has increased significantly, which is a sign of integration into the world economy. Overall, we have a consensus that NAFTA’s investment provisions have been instrumental in the closer integration of the three economies. Technology transfers, with important spill-over effects, took place through investment. Employees in all three countries have enjoyed better job opportunities than before NAFTA. The Agreement has also brought its members politically closer, in the sense that North American stability and growth are perceived as shared objectives. When one member suffered from a financial crisis, assistance quickly came from the other members. Weintraub and de Mateo have highlighted the importance of NAFTA as a determining factor in minimising the adverse effects of the Mexican crisis, both in terms of depth and duration. “For better or worse”, as Weintraub noted, Mexico is thoroughly integrated into North America.

Investment chapter I would like to briefly talk about the role of investment as a medium of integration, the innovations brought about by NAFTA’s investment rules and its contributions to the policy dialogue for improving investment disciplines at the multilateral level. Investment is the principal means of delivering goods and services to foreign markets and the leading factor in organising international production. The sales of foreign affiliates of multinational corporations have, in recent years, far exceeded the value of world trade in goods and services. In 1996, they reached about US$ 6.5 trillion. Both de Mateo and Kirton have alluded to the need for a deep integration arrangement to complement investment rules, with liberalisation in other policy domains. These policy areas include: trade in services, the protection of intellectual property rights, government procurement of goods and services, border facilitation with respect to temporary entry of business people and their tools of trade, and a work programme on standards harmonisation and on mutual recognition arrangements for the certification and licensing of professional-service providers. To a large extent, the success of NAFTA is attributable to the establishment of a coherent and comprehensive set of doing-business rules that treat trade and investment as inseparable forces of market integration.

306

Looking at the contribution of NAFTA to the policy-making process, there are a number of innovations that are worth noting. For example: a wide definition of investment that includes non-equity interests (this is an improvement from the Canada-United States Free Trade Agreement), the absence of a general grandfathering of non-conforming measures (the non-conforming measures must be listed in an annex rather than establishing a list of measures which are subject to the new disciplines), and an investor-state dispute settlement procedure. Overall NAFTA’s investment rules are resulting in stronger and deeper commitments than those embodied in the WTO Agreements on Trade-Related Investment Measures (TRIMs) and on Services (GATS). NAFTA’s investment rules have also provided the blueprint for the negotiation of the OECD Multilateral Agreement on Investment (MAI). It would be interesting to have responses from the authors to the following questions: What are the policy lessons to be drawn from NAFTA’s investment rules? What is the unfinished agenda, or incomplete liberalisation in investment and investment-related fields, that would maximise the economic gains of the regional integration? Are there still significant investment liberalisation measures that are not applied on a MFN basis for all WTO members?

Comparison NAFTA, as a regional integration arrangement, is very different from other regional arrangements. Many regional arrangements were launched during the 1960s and failed due to their limited economic and political commitments. Members of the European Economic Community and now the European Union have progressively moved to establish a comprehensive common market, with the establishment of the four freedoms, as stated in the objectives of the Treaty of Rome: free internal movements of goods, services, labour and capital. More recently, the Single Market or EC-1992 programme led to the enforcement of EU-wide competition policy, disciplines on state aids, public procurement, and wide application of mutual recognition of standards and certification procedures. The Single Market is one form of deep market integration which is the result of a long and sustained process of gradual integration. I would like to focus on two aspects to illustrate the differences between NAFTA and the EU, more precisely on the movement of labour and competition policy. Chapter sixteen of NAFTA concerning the temporary entry of business people is very useful, as it allows citizens of each country to pursue business opportunities without meeting a labour-market test. Four categories of business people are identified: business visitors, traders and investors, intra-company transferees, and professionals. Each country retains its rights to protect the permanent base of its domestic workers. The United States has, however, reserved the right to restrict the entry of one category, professionals from Mexico, to 5 500 a year during the first ten years of NAFTA. There are no quotas for Canadian professionals entering the United States or Mexico for professional purposes. Within the EU, workers can freely look for and take-up employment in any of the member countries. The political and historical context in Europe is indeed far different from the situation in North America. Granting full labour mobility, or indeed establishing a common labour market, was not among the objectives stated in NAFTA. With regard to competition policy, NAFTA established a Working Party on trade and competition with the mandate to address competition matters in the integrated market. So far, the three national competition authorities still work under the purview of distinct national markets. Anti-dumping measures, for example, can still be applied on products originating within NAFTA. In other regional arrangements, the EU, the AustraliaNew Zealand and the recent Canada-Chile agreement, anti-dumping measures are no longer (or will soon cease to be) an acceptable instrument to address competition concerns for products originating within the regional arrangement.

307

What kind of policy lessons can we draw from NAFTA and other formats of regional integration arrangements with respect to labour mobility and competition policy? Are there additional efficiency gains to be achieved by pursuing the integration process further?

Apparent paradox In light of the three papers, it appears that there is an apparent paradox about free trade. While both Canada and Mexico suffered economically in the 1990s: “Canada’s lost decade” (as noted by Andrew Sharp), and the Mexican financial crisis in 1994-95, Canada and Mexico have nevertheless entered into additional free trade agreements with other countries. Canada and Chile now have a free trade agreement in force, and in December 1997 Mexico signed a Memorandum of Understanding with the European Union for negotiating a free trade agreement. Within NAFTA, both Canada and Mexico are actually seeking to adopt a broader concept of competition policy and deeper integration commitments. Meanwhile, the United States, which has probably benefited the most from the rationalisation process brought about by NAFTA, is the country that is most reluctant to further pursue integration within or outside NAFTA. It is paradoxical that with a unanimous assessment of the benefits of NAFTA, the prospect of extending the free trade area to cover the whole of the Americas is currently deadlocked because of opposition in the US Congress. In closing, there are several questions that merit further discussion: What are the obstacles for the true beginning of the Free Trade Area of the Americas (FTAA) negotiations? What are the factors that explain the fears of free trade initiatives in the United States, as indicated by Weintraub? What is envisaged in terms of labour movement within the upcoming negotiations between Mexico and the EU for their free trade agreement?

308

Comments on the Paper by Horacio Sobarzo Fimbres “Applied General Equilibrium Models: The Mexican Experience of NAFTA” by Jean-Christophe Dumont Dial-Orstom, Paris

Horacio Sobarzo’s paper provides the opportunity to examine assumptions of Computable General Equilibrium models (CGE) in a very concrete context: the North American Free Trade Agreement (NAFTA). The first section of the paper recalls the technical principles which drive CGE modelling, the second addresses some of their limitations, and the last presents the main results that have been obtained in the NAFTA context.

CGE models: assumptions and limitations As the author points out, CGE models are not new; the first trials can be traced back to the 1960s. Today, CGE models are well known, and so, I will simply touch on the fundamentals as they are outlined in the paper. CGE models have the following characteristics: −

They are macroeconomic models with microeconomic foundations.

−

They are grounded on the neo-classical framework. This means that in their standard version the prices clear all markets, but some types of quantitative adjustments, with unemployment or underutilisation of capital, can be introduced.

−

Supply and demand are endogenous. They result from the maximisation of profit and utility.

−

CGE models are traditionally static and real but there are numerous examples of dynamic exercises and of models which incorporate a financial framework.

From an empirical point of view, data are needed for a base year, which is usually synthesised in a Social Accounting Matrix (SAM), and a full set of elasticities. The fundamental hypothesis lies in: −

the functional specification of the production and the demand functions;

−

the macro closures;

−

the degree of aggregation and the structural specifications.

As regards the functional specification, econometric estimations should ideally be used, but as this is rarely the case, they are often replaced with a priori mathematical formulations and the corresponding parameters. For that matter, it is not unusual to import the values from other models and other countries, i.e. guesstimates. These specifications are nevertheless important. In terms of production, they reflect the feasibility of the adjustment of the production process: the possibility to substitute different types of inputs (capital and labour). So, if significant factor reallocation is expected to accompany trade liberalisation, this becomes an important point to be dealt with. However, the author points out that alternative specifications can be assessed to some extent by a sensitivity analysis.

309

Whatever their weaknesses, these modelling exercises are undoubtedly useful to assess the global effects of trade liberalisation. They allow the reallocation of resources and welfare gains to be described. They also provide information about the challenges the concerned economies are likely to face. At the same time, it is also very important to recognise what models do not tell. The author concludes the second section of his paper with a discussion of the use of CGE modelling, arguing that “they are structural models useful for policy analysis but that they should not be used for forecasting”. I concur on this point. CGE models can only provide counterfactual analysis. They help us to understand the main mechanisms at work and to quantify them under a specific set of assumptions.

Comments on CGE results The last section proposes a brief review of the experimental use of CGE models in analysing the impact of NAFTA, as it concerns Mexico. These efforts have yielded four common outcomes: −

Whatever the type of exercise, welfare gains are expected in all three countries involved: Canada, Mexico and the United States.

−

Mexico, being the smallest economy, will benefit disproportionally from the Agreement.

−

In the models that incorporate some form of imperfect competition, the welfare gains are greater than in those assuming perfect competition.

−

Dynamic models predict even greater efficiency gains than the static ones.

Sobarzo’s summary of the modelling results briefly addresses two questions that nonetheless merit further discussion. First, I would point out that although the models indicate that every country gains from trade liberalisation, the estimated gains are very small: welfare gains for Mexico range from 1 to 5 per cent of GDP over the whole period of 10 to 15 years. This represents almost nothing on a yearly basis. Furthermore, if we take into account the frictional cost, which could be very high considering the expected adjustment, gains might well be negative in the short term and null in the long run. In addition, as the author stresses, anticipated outcomes would be different under imperfect competition or in an endogenous growth framework. For instance, if there are increasing returns in some tradable sectors, an exportled growth industrialisation process might be expected. In the same vein, with trade liberalisation, increased Foreign Direct Investment (FDI) flows would be expected, opening new economic opportunities and providing cumulative factor productivity gains. So, it would appear that a dynamic model, as proposed by the author in his conclusion, could be built in order to assess issues that really matter for Mexico. Second, it is not clear that NAFTA will reduce existing inequalities. As Denis Cogneau pointed out in his comments on the paper of Raul Hinojosa, the immobility of factors, such as land, would lead to sectoral problems. In such a case, land earnings would decline and rural labour would suffer as a result of trade liberalisation.

310

Comments on the Paper by Raúl Hinojosa-Ojeda “Economic Effects of NAFTA: Employment and Migration Modelling Results”

by Denis Cogneau DIAL-ORSTOM, Paris

The paper of Raúl Hinojosa and his co-authors is very rich. I would identify four main topics that make their paper particularly interesting. The first one is statistical and historical evidence for the high level of asymmetric interdependence between the United States and Mexico, with a particular focus on labour market issues and the status of Latino immigrants in the United States. The second is a survey of previous studies about the impact of NAFTA on both economies, making use of various modelling techniques, again with a particular focus on the consequences of NAFTA on migration flows and the welfare of Latino immigrants in the United States. The third is a critique on logical and empirical grounds of what the authors call “neo-nativist” political propositions. These propositions rely on the argument that undocumented migration is one of the major causes of California’s economic and fiscal problems, further contending that illegal immigration can be stopped by prohibiting the provisions of public services to undocumented immigrants and by raising the intensity of internal and border controls. The fourth topic is an original set of policy experiments using a CGE model designed by the authors, called CALNAFTA. Given the originality of the modelling experiments, I will focus my comments on the fourth topic, but not before making a few remarks about the preceding ones. Productive comparisons can be drawn between the situation of US-Mexican integration and that of the European Union and Southern Mediterranean Countries.

Trade liberalisation and migration flows As for the first and the second topics, the authors argue that there is a historical interdependence between Mexico and United States, in the form of trade and investment flows, but that this interdependence remains highly asymmetrical. Mexico’s total economic weight is roughly equivalent to that of Los Angeles county, meaning that economic events in Mexico have only a marginal impact on the US economy as a whole, though some sectors and segments are more directly concerned. In all these respects, there is great similarity between this relationship and the one existing between European and Southern-Mediterranean countries. At the moment, NAFTA has not modified this state of affairs, and, even if NAFTA represents the most ambitious step ever attempted towards contractual formal integration, its results are not qualitatively distinct from long-term historical trends. Indeed, most counterfactual experiments agree that NAFTA will only have a small effect on aggregate Mexican GDP and on the average welfare of Mexican citizens. Moreover, the distributive and migration consequences of the Agreement appear to be far more problematic than a revealed by modelling experiments based on standard trade theory. If we were to apply a simple Hecksher-Ohlin-Samuelson model with only two factors (capital and labour) and two countries, we would predict that NAFTA would raise the earnings of the most abundant factor in each country; hence, unskilled labour in the case of Mexico, which should produce a slowdown in migration flows. Trade flows would substitute for factor flows. Everything would change, however, if immobile or sector-specific

311

factors, like land, were introduced. In such a case, there would be a decline of land earnings in non-competitive land-intensive sectors when these sectors are highly protected, as with the Mexican corn and grain sectors. Unskilled rural labour would then possibly suffer from trade liberalisation. The removal of tariff and non-tariff barriers on agricultural products seems to lead to a collapse of Mexican corn and grain production, with the rise in fruits and vegetables exports being not high enough to absorb the rural labour released from corn production. As a consequence, rural unskilled labour migrates to the cities or emigrates from the country. In the EuroMediterranean case, however, decreased output in non-competitive agricultural sectors like grain, milk, and meat is offset by the rise in the fruits and vegetables sectors. Yet (and perhaps unsurprisingly), the agreements between the European Union and Mediterranean countries do not, for the moment, include agriculture. At the same time, it seems that, in itself, the free trade agreement between the EU and Mediterranean countries would not lead to a substantial decline in migration pressure, given the weak impact on the welfare of potential migrants. Thus, in both cases, free trade has disappointing effects on welfare if one puts these effects in balance with stabilisation and reallocation costs. I agree with the authors that the implementation of trade liberalisation should be accompanied by structural reforms of the labour market, strong investments in human capital, and, I would add, public expenditures on infrastructure. I must say, for my part, that I do not see how the developing countries could implement such reforms without stronger financial support from their northern partners. Despite the “neo-nativist” ideas to which the authors allude, most studies conclude that migrants cost less than they bring to the country to which they immigrate. In this respect, the analysis they present appears to me sound and convincing.

CGE modelling Turning to the original model built by the authors, it is standard in the field of CGE models, in line with the Walras-Arrow-Debreu framework: perfect flexibility of prices and full-employment of factors, maximising atomistic agents, and constant returns in production. Its only peculiarity is the exogeneity of real investment, whereas in more traditional neo-classical models investment is driven by savings. One might ask the authors why they chose the so-called “Johansen’s closure rule”, and what the consequence of this choice are on their results? Geographically speaking, the model distinguishes four regions: Mexico, California, the rest of the United States, and the rest of the world. Those four regions are linked by trade flows, and the United States and Mexico by labour migration flows. International capital movements are not modelled. As for migration, Mexican immigrants are perfect substitutes of the relevant category of native US labour: unskilled rural or unskilled urban. Migration behaviour is perfectly elastic so that the ratio of wages of migrants to the wage of the corresponding labour type in Mexico, converted at the prevailing real exchange rate, is constant. In this area, three hypotheses of the model can be questioned: − − −

Is migrant labour a perfect substitute for native US labour? Are migration flows perfectly elastic to wage differentials, inside Mexico and internationally? Is migration responsive to the exchange rate? The more permanent migration is, the less it should be responsive to the exchange rate, since permanent migrants compare the purchasing power of both wages in the currency of the country of settlement.

Moving on to the simulations, there are four types presented in the paper. The scenarios labelled “1” are the pure NAFTA scenarios. I will not comment on them, as their results are in line with previous results of constant returns models: the welfare effects are very small and the distributive effects are negative in the case of Mexico. The scenarios labelled “2” try to model the impact of a restrictive migration policy. When all undocumented Mexican migrants are deported from California or the United States as a whole, the negative shock on the labour supply in the United States is so large that it is not surprising to find that the US economy would experience a dramatic downturn. Besides being unrealistic, these policies are clearly to be rejected on the grounds of

312

economic efficiency. The third type of scenarios examines the impact of an increase of foreign direct investment (FDI) towards Mexico and of an increase of domestic investment. In the context of a static framework, I do not understand the goal of these simulations. In such a context, a rise in investment cannot have any impact on the capital stock or on the productive capacity of the sectors, and hence cannot have any supply effects, only demand effects. Therefore, from that perspective, FDI plays the part of any foreign transfer, contributing to ease the balance of payment constraint. Alternately, one could have modelled an exogenous increase in Mexican capital stock, as it is said “falling from the sky”, to get supply effects. Anyway, in both cases, given the scope of the shocks, there is again no surprise in finding strong positive effects on incomes and a strong slowdown of migration. The results of the model confirm that FDI, at least for its demand effects, is a strategic variable in the regional integration process. I will comment on the third simulation of type “2”, labelled “2c”, together with the fourth group of scenarios. The last scenarios labelled “4” combine the positive investment shocks of the third-type scenarios with an increase of the reservation wages of migrants. This increase is meant to represent the effects of a “new regime of transnationally protected worker rights”, allowing migrants in the United States to get better compensation for their employment. While I subscribe to the policy objectives put forward by the authors, I must say that I am not convinced by the way the implementation of the policies is modelled here. Where is the problem? The problem lies in the fact that the model does not make the distinction, on the labour market, between migrant labour (whether undocumented or not) and native US labour. This means that the wage gap between these two categories, migrants and natives, is not explicitly modelled, while being in fact the critical variables of the migration policies they are trying to assess. Instead of playing on this wage gap, the authors have chosen to manipulate the reservation wage. But one should realise that this reservation wage plays the same role as migration costs. So, an increase in the reservation wage may as well represent, in the model, the impact of a more restrictive policy. It is therefore not a surprise to find a decrease of labour migration in the scenarios of type “4” and an increase in scenario 2c. A simple solution would be to disagreggate unskilled labour in the United States between undocumented migrants, documented migrants, and natives, and to introduce the prevailing wage gaps between these three categories, whatever their origin: cost of illegality, discrimination, etc. A decrease in the wage gap would mean increasing costs for US enterprise, lowering the demand for migrant labour, while the ex-ante increase of the migrant wages would raise the supply of migrant labour. The equilibrium wages would depend on the elasticity of the demand for migrant labour. Since the resulting migration flow is unknown (its sign being even ambiguous), it would be a more interesting situation to study with the help of a quantitative model. And even then, as the model cannot make clear the distinction between restrictive migration policies and the upgrading of the labour conditions of migrants, I am afraid that there would still be research to carry out before being able to properly simulate those policy options.

313

Comments on the Paper by Demetrios Papademetriou, “New Directions for Managing US-Mexican Migration” by Antonio Spilimbergo International Monetary Fund

My comments are focused on two issues raised by Demetrios Papademetriou: i) that NAFTA should be considered not just on the basis of trade integration but also for its effects on immigration, ii) how the MexicanUS border should be managed. In the debate prior to NAFTA, politicians and economists suggested that NAFTA would decrease migration by raising the wages of unskilled workers in Mexico. Contrary to these expectations, migration increased substantially after the Agreement’s implementation. The reasons for this surge in immigration are the crisis that followed the devaluation of the peso in December 1994 and the slowness of the mechanisms through which trade affects migration. I will concentrate on the second structural reason. The way in which trade is supposed to affect migration is based on two hypotheses: first, US and Mexican labour markets are integrated so that migration is sensitive to relative wages; second, trade should improve relative employment opportunities in Mexico, especially for less skilled workers. The first hypothesis, that labour markets are integrated, seems valid even before NAFTA considering evidence both from wages and migration flows. Concerning wages, Raymond Roberts has shown that prior to NAFTA, the Mexican labour markets along the northern border were more integrated with the United States than with the Mexican interior. Concerning migration flows, Gordon Hanson and I (1996) have found that a decline of 10 per cent in Mexican wages relative to US wages is associated with an increase of 8 per cent in the people apprehended while trying to illegally cross the US-Mexican border -- this is true when starting from the same month and controlling for changes in the enforcement policy -- which means that the devaluation of the peso in December 1994 can explain as much as 320 000 new apprehensions in 1995. It can be concluded that the Mexican and the US labour markets for unskilled workers are tightly connected, even in the short run. The second hypothesis, that trade should improve the wages of unskilled workers, is more difficult to verify. In the debate leading up to NAFTA, many politicians and economists took for granted that NAFTA would rapidly create many new jobs in Mexico. There was consensus on this point among both critics and supporters of NAFTA. The former were eager to point to the “giant sucking sound” and the transfer of jobs from the United States to Mexico, and the latter (mostly economists) stressed that migration and trade are broadly speaking substitutable. The fact that both groups, although divided on many other issues, supported this view created very optimistic expectations for job creation in Mexico. The post-NAFTA reality has not matched these expectations but has been more similar to what previous experience had shown. Indeed, trade liberalisation creates new job opportunities, but with some lag, and for more qualified workers. Here, the example of the assembly industry in Mexico (maquiladora) is interesting. Although maquiladora have existed since 1965, they employed just 119 546 workers in 1980; the industry grew to 446 436 workers in 1990, and to 692 142 in 1996. This shows that trade liberalisation and capital mobility have created jobs, but with a considerable lag (see G. Hanson and A. Spilimbergo, 1998). Moreover, trade liberalisation and the accompanying technological progress have considerably increased the premium for higher education, which in turn has increased the incentive to obtain higher education in Mexico, the final result of

314

which may well be less migration in the long term. In summary, there is evidence that trade creates jobs, but this happens with a considerable delay and not within the short political horizon of political debate. If we put together the observations that the US and Mexican labour markets are already tightly connected and the fact that trade creates new jobs only with time, it follows that NAFTA will not stop migration in the next few years, but could actually prompt more migration by inducing local shocks and structural adjustments. This leads me to a final point: if (illegal) migration will happen, what is the best way of managing the border? G. Hanson and I (1998) have considered several policy options to deal with the management of the border and illegal migration. A convenient way to classify the policy options is in terms of the trade-off between short and long-run effects. The first option, to increase the border patrol, can only be effective in the short term because the border is too long to be effectively patrolled and because this does not affect the wage differential, which is the primary cause of migration. The second option, to increase sanctions on employers of illegal immigrants, is more efficient because sanctions reduce the available jobs in the United States, but such measures are politically difficult to implement. The third option, the legalisation of illegal immigrants, would create expectations of future periodic legalisations and so would send an ambiguous signal to potential immigrants. The forth option, an effective collaboration with the Mexican authorities on migration policy, is the only way to directly affect the primary cause of (illegal) migration, i.e. the persistent differential in wages and social conditions. Recent experience (e.g. in El Salvador in the 1980s) has shown that social unrest can cause large migration flows that persist even after the social situation improves. A common policy would avoid the deterioration of local conditions in critically poor Mexican regions. More generally, international experience has shown that it is almost impossible to close a long border when there are incentives to migrate; on the other hand, an international collaboration on migration, trade, and labour policies can significantly diminish the social costs of migration for both sides.

REFERENCES

HANSON, G. and SPILIMBERGO, A. (1996), “Illegal immigration, border enforcement, and relative wages: Evidence from apprehensions at the U.S.-Mexico Border”, NBER Working Paper No. 5592, May. HANSON, G. and SPILIMBERGO, A. (1998), “Mexican migration and U.S. policy options”, in Immigration in U.S.-Mexican Relations, Brooking Institutions and Inter-American Dialogue, January. ROBERTS, R. (1997) “Wage shocks and North American labor market integration”, Discussion Paper No. 84, Syracuse University.

315

OECD PUBLICATIONS, 2, rue Andre-Pascal, ´ 75775 PARIS CEDEX 16 PRINTED IN FRANCE (81 98 12 1 P) ISBN 92-64-16188-0 – No. 50363 1998