MIGRATION AND DEVELOPMENT

IN NORTH AND CENTRAL AMERICA:

A SYNTHETIC VIEW

International Organization for Migration (OIM)

Economic Commission for Latin America and the Caribbean

(ECLAC/CELADE>

This text is based on the document Migración y desarrollo en Centro y Norteamérica: elementos para una discusión (1998) prepared for the IOM and ECLAC/CELADE by the consultant señor Agustín Escobar Latapí.

MIGRATION AND DEVELOPMENT IN NORTH AND CENTRAL AMERICA:

A SYNTHETIC VIEW [1]

International Organization for Migration (IOM) Economic Commission for Latin America and the Caribbean (ECLACICELADE)

Introduction.

The relationships between international migration and development are manifold, and involve reciprocal links of causality. Just as the movement of people across international frontiers is driven by economic, social, political, cultural and environmental factors that form part of the development process, this movement in turn affects the development patterns of the areas of origin and destination, as well as altering the living conditions of the individuals and families that are directly involved in this migration.

Of course, the interactive links between migration and development differ according to historical circumstances and vary between countries. In the modern world, with the growing interdependence between countries that characterizes it, these links have become very complex. Together with changes in the spheres of production and trade, technological progress in communications has given people greater access to information and to the lifestyles, consumption patterns and cultural norms prevailing in the more highly developed nations, and has thus made them more aware of the large discrepancies that exist between countries in terms of their stage of development and ability to meet the needs of their populations. The information that people have about such inequalities, and the way these are perceived, act as an incentive to migration. Again, sociopolitical upheavals, which are not absent from the changing international scene, have been of decisive importance in increasing the volume and diversity of migratory movements. At the same time as the number of migrants has risen, migration has become more diverse in terms of its characteristics and duration.

Within this varied panorama of population exchanges, a large number of countries have been rapidly changing from recipients of population into exporters, a few have become more attractive to migrants, while others may be regarded as countries of transit or "transmigration" The region comprising the countries of North and Central America provides a specific example of what happens when different patterns of development and migration intersect. In this region, the historical variability of the relationships linking the two processes, whose peculiarities are highlighted by the diversity of the economic, social, political and cultural conditions existing in the different countries, is particularly apparent. In this text, an attempt will be made to explore some of these relationships to provide the basis for thinking that can contribute to the evaluation of policy options for the near future.

I. Development and international migration processes

1. Aspects of development and international migration. The different aspects of the development process have an influence on migration, affecting its trends (alteration or consolidation of patterns of population interchange), its characteristics (composition, characteristics and duration) and its scale (both absolute and relative). As has been noted, however, this influence is reciprocal, since the trends, characteristics and scale of migration also exercise an influence on the different aspects of development. Within this field of interactions it is worth concentrating on certain linkages which are bound up with demographic trends, economic developments, the sociopolitical situation, environmental conditions, reforms of an institutional nature and the relations between societies.

(b) The forms of political organization and participation in decision-making processes that exist in different societies are closely linked with the degree of equity obtaining there. If socio-economic inequalities are acute, vast sectors of the population will find that the aspiration of exercising their rights as citizens is a virtually unattainable one, and some of these may seek to realize it in other social contexts. Exacerbation of tensions resulting from sociopolitical exclusion tends to lead to various forms of instability and violence, which generally result in forced movements of population.

(c) The emergence of production and consumption patterns that are detrimental to ecosystems, such as the speculative exploitation of natural resources leading to these being exhausted and the generation of volumes of waste that exceed disposal capacity, combined with the persistence of certain. traditional practices, has given rise to situations that involve severe damage to the environment. This type of damage, which is found both in highly populated areas and in recently settled ones, is detrimental to the economic and social sustainability of a number of districts, and consequently tends to cause migration. As the scale and pace of environmental change have tended to increase, the effects of such change on migration appear to have grown.

(d) Over the last decade, many countries have implemented institutional reforms which have changed the traditional role of the State as a generator of job s, provider of services and regulator of markets. Application of the criteria of efficiency, cost recovery and privatization has led to changes which are affecting people's living conditions and the way they work and reproduce. In particular, the move towards flexible labour markets, which has been accompanied by internationalization of the more dynamic production activities, appears to have been a factor in encouraging people to migrate.

(e) Over recent decades, globalization has become more far-reaching and has taken a firmer hold. In macrosocial terms, the greater interdependence of world markets for capital, goods and services has led to the sphere of national autonomy being reduced in relative terms and to many decision-making processes being internationalized. At the microsocial level, with the cultural globalization that has also taken place, the social networks among households and communities have been strengthened, and this has led to a reduction in the friction effects produced by geographical and cultural distance. Both of these circumstances have helped to weaken many of the obstacles which previously stood in the way of migration.

2. Development and migration situations in Mexico and Central America. Although there are differences between them, the Central American countries and Mexico constitute a distinctive sub region within the North and Central America area as a whole. During the last thirty-five years this sub region has undergone profound changes in its development pattern, and these have been accompanied by fluctuations in international migration trends. Using a somewhat schematic approach, these changes may be divided into three stages, as set out below.

(a) In the first of these stages, covering the 1960s and the beginning of the 1970s, the development model that predominated in the sub region attached particular importance to import substitution. This strategy attained its highest development in Mexico, where the industrialization process reached an advanced stage, helping to bring about greater urbanization and a marked concentration of population. The Central American economies, while displaying differences ranging from a social administration approach in Costa Rica to an enclave style in other countries, concentrated mainly on agricultural export activities; nonetheless, growing numbers of people who were not absorbed by these activities moved to the cities and found employment, often of a marginal kind, in the tertiary sector.

During this stage, when the economy appeared to be able to generate a quantity of jobs that more or less tallied with the size of the available labour force, international migration from the sub region stood at a fairly low level. In the case of Central America, most international movements took place between countries that shared borders, a situation that has deep historical roots and is linked with developments in agricultural export activities and the drive to occupy new areas. In the case of Mexico, migration was basically towards the United States, as has been the case since the beginning of the twentieth century; the country's historical links with the south-eastern part of the United States, and the use of various mechanisms to hire labour, stimulated a continuous flow of migrant Mexican workers, giving "rise to the existence of a de facto labour market between the two countries" (Bustamante, 1997, p. 129). This market has been subject to the ebbs and flows that periods of economic boom and contraction are liable to produce, and these have contributed to changes in patterns of job creation between different sectors of the economy (Cornelius, 1989; Fernández, 1983; Vernez and Ronfeldt, 1991).

(b) Towards the middle of the 1970s there began a second stage during which a number of problems associated with the development approaches then in favour became more acute and widespread. Both enclave type exporting economies and import substitution industrialization had come up against the structural limits of their growth potential, and the inability of these economies to generate employment of sufficient quality or in sufficient quantities became increasingly plain.

Again, a number of countries suffered crises of political exclusion, exacerbated by deep social inequities which, besides jeopardizing respect for human rights, limited the scope for improving human resources. The rigidities of these economies, reproduced in the sphere of employment (with underutilization of labour being manifested in numerous ways) and combined with growing political instability, gave rise to escalating violence which was to break out into armed conflicts.

During this second stage, which included the end of the 1970s and the 1980s, all the countries in the subregion went through a profound economic crisis. Against a background of severe inadequacies in their development levels, most of these countries experienced negative growth in gross domestic product, increasing unemployment, falling incomes from work and rising indices of poverty. At the same time, sociopolitical instability led to widespread violence, with armed conflicts that reached their greatest intensity in El Salvador, Nicaragua and Guatemala, but were felt around the whole subregion.

The highly unfavourable conditions that obtained during this stage prepared the ground for emigration, and social violence proved to be a precipitating factor. Large-scale migratory movements were set off; great numbers of people were forced to migrate within Central America and towards Mexico, and many of them ended up in the United States and Canada. These movements were different from previous ones in that their composition was very varied, comprising refugees, displaced persons, illegal immigrants, families and professionals, and for this very reason they proved to be extremely difficult to handle. In this increasingly complex situation, with the number of migrants growing at an unprecedented rate, international migration gradually shifted from the "south-south" to the "south-north" axis. This meant that these movements were more visible, which turned them into a source of growing concern, particularly in the destination societies.

(c) The third stage, which began in the early 1 990s, holds out some promise, at least if it is compared with the previous ones. What we are seeing in fact is a gradual renewal of economic growth and the re-establishment of peacefull coexistence. Nonetheless, there can be no doubt that in the 1990s numerous socio-economic problems which have affected the subregion for a long time are still to be found, among them: the productive structure it has inherited, which has very little diversification and is extremely dependent on international demand; a highly unequal income distribution model; high levels of unemployment and underemployment; a manifest lack of social equity; a high incidence of poverty; and severe environmental damage in certain territories. All these problems, taken together, represent burdensome legacies from previous decades and pose complex challenges which will have to be addressed in the future; the persistence of these problems gives grounds for supposing that conditions will continue to be favourable to emigration.

What is peculiar to this third stage is the development of institutional reform processes redefining the role of the State and that of the markets. This peculiarity is given heightened prominence by the fact that it comes against a background of new forms of relationship between these countries and the outside world, which are laying the groundwork for regional and global economic integration. The restoration of civil order, a gradual recovery in economic growth, the institutional reforms being implemented and the changes taking place in the international environment all open up opportunities for the countries of the subregion, but they do not obviate the need to address the major challenges involved in carrying through the far-reaching transformations in the structure of production that will be required if a higher degree of social equity is to be achieved and the development process is to be given greater sociopolitical and environmental sustainability.

During this third stage, some expatriates have returned and many of those who entered other countries as refugees have been given permanent status. These processes have not been without difficulties, as is illustrated by instances where populations have been resettled in areas that are extremely run-down or that have been reoccupied by other groups, and by the fact that a large number of migrants still have no official status. Of course, a large proportion of those who emigrated during the 1970s and 1980s have not returned to their countries of origin. Furthermore, the conditions for emigration have not disappeared, and in certain countries it is still taking place on a large scale.

3. The current development and migration situation. Given the characteristics of the current development situation, including the processes of institutional reform and globalization, we need to ask what the repercussions of these may be for migration trends in the subregion. Since the groundwork for structural economic adjustments has already been put in place in most of the countries in the subregion, the current situation may be described as one of transition towards a new period of economic growth. Insofar as this supposition may be correct, there is reason to think that the development process is still a cause of migration.

Although it is too early to make judgements about a situation that is only now beginning to take shape, and despite the lack of solid and up-to-date information about migration in the subregion, it is highly likely that newly open markets and new forms of participation in the international economy, including integration schemes and open regionalism, will cause migration to continue. What these changes tend to do is rapidly modify and reposition economies and labour markets both within countries and internationally, undermining existing forms of subsistence. These repercussions come on top of the expectations that people have gradually been forming in a cultural context that leans towards individualism; these expectations, which take on a visible character in the sphere of consumption, cannot always be satisfied in people's societies of origin, and so act as incentives to emigration. The processes of trade liberalization and integration, together with the powerful effects of new technologies, have also led to communications and transportation becoming easier, and these advances, combined with the instability of employment and with the social networks that were created or consolidated during the 1980s, mean that larger and larger segments of the population are now in a position to respond rapidly to far-off opportunities and information.

A number of recent analyses, although not yet crystallized into a coherent theory, give a picture of how international migration works in the context of economic globalization and the new international division of labour (Castells, 1989; Lim, 1993; Portes and Walton, 1981; Sassen, 1988). It is argued that segments of developed country labour markets that incorporate immigrants are becoming consolidated, and that these niches tend to be reinforced by the operation of networks of migrants and, in many cases, of those who recruit this immigrant labour.

This analysis emphasizes the growing hegemony of multinational corporations and the paradox that in many developing countries investments made in agriculture and industrial plant with external markets in view are instrumental in increasing, rather than reducing, migratory tendencies. In other words, the way in which economic development takes place against a background of global interdependence tends to disrupt traditional local economies and lead to acuter inequality and higher unemployment. This context of globalization, with the stronger economic ties (due to improved communications and transport systems) and intenser political, social and cultural relationships that it entails, provides a strong stimulus to international migration, which is facilitated in practice by the existence of social networks.

Again, the subregion formed by the United States and Canada also passed through deep structural crises and reforms in the 1 980s. Nonetheless, these countries differ very profoundly from their southern neighbors in terms of the degree of development and prosperity they have attained. Among the developments of most importance to international migration that have occurred in Anglo-America since the 1980s, particular mention must be made of changes in the sphere of employment. The demand for labour has undergone a structural change, while at the same time the use of flexible types of labour contract has become widespread, something that appears to be associated with unstable forms of employment; likewise, the relative size of economic sectors has altered, with services expanding (Levy and Murnane, 1992; Sassen, 1997).

In the United States, not only has the demand for labour been strong in recent years, but changes in the labour market have helped to enhance its traditional attractions for migrants. This has given rise to a tension between the demand for labour and the application of regulations placing restrictions on migration, although trends in the latter seem to show that, in practice, these regulations have been overtaken by events. Awareness of migration trends has given rise to an upsurge in xenophobic attitudes, with immigrants being seen as the source of economic circumstances that in fact derive from processes which are essentially endogenous. From a different point of view, efforts have been made in Canada to coordinate migration policy with changes in the labour market, an approach that provides a broad conceptual and operational framework for the regulation of migration processes.

Viewing the situation realistically, then, what is proposed here, in the context of the relationships between migration and development, is that the governments of the region, rather than preparing for a slackening off of migration, should be planning ways of imposing some order on future flows and adapting them to national and regional development needs. This task may be made easier by a growing political willingness to establish bilateral and multilateral agreements, a shared desire to see greater social equity, and the consolidation of open regionalism. If migration appeared to be out of hand in the 1980s, this was due not just to its actual scale and composition, but also to the failure by many governments to pay sufficient attention to it, because of the urgent priorities that the sociopolitical context created. The present situation is objectively different: with greater sociopolitical and economic stability now obtaining, migration can be treated as an issue of the highest importance, and can be dealt with by coordinated policies. Given the changes that are taking place in the different aspects of development in the countries of the region, attempts to keep populations immobile seem to be unjustified, and indeed may entail very high economic, social and political costs.

II. Regional migration over the last three decades

1. Information about patterns of migration. If we are to examine the relationships between development and international migration, we need to be aware of trends in the latter. This means discovering its scale (in absolute numbers and relative frequencies), directions and characteristics (of both a demographic and a socio-economic nature) and the different ways in which population movements manifest themselves~ This task is a particularly complex one in the case of the North and Central America region due to the -considerable heterogeneity displayed by population movements. Existing sources of data collected in accordance with common criteria provide only a partial picture of this very diverse situation. The lack of appropriate, relevant and up-to-date information hinders efforts to examine migratory trends and behavior in a rigorous way, makes it hard to predict future changes, impedes evaluation of the consequences of international migration and increases the difficulty of designing realistic policies in this field.

Although most of the countries in the region keep records of those entering and leaving through international ports, these records are known to be seriously deficient 2. Other ongoing statistics (such as records of passports, resident foreigners, visas and work permits) are affected by even greater problems. These limitations have led to consideration being given to the potential of national population censuses as a source for the study of international migration. The relevant. data covers only "stocks" of migrants from abroad to be found in each country, or the number of surviving migrants that has accumulated as of the date of each census, which makes it difficult to form an understanding of migration as a process 3. Despite this limitation, these data do make it possible to trace the broad outlines of the international migration situation in the North and Central America region from 1970 up to the beginning of the 1990s 4.

The information available, although incomplete due to the irregularity of census operations, particularly during periods of economic crisis and violence, does enable us to identify three major international migration flows within the North and Central America region. One of these follows a "south-south" direction, taking place within the countries of the subregion comprising Mexico and the Central American states. A second flow takes -a "south-north" course, being the sum of the movements linking that subregion with the more developed countries of North America. Finally, a third flow is formed by movements of a "north-north" type, including the United States and Canada alone. As has already been suggested, as development tendencies have shifted, so the incidence of these patterns has changed over time. Examination of this information is supplemented by a description of some of the sociodemographic characteristics of migrants into the United States.

2. Growth and changing destinations of migratory stocks in the countries of the region.

Since the 1 970s, there has been vigorous growth in the number of migrants, while their movement patterns within the territory of the region have shifted. The stocks of migrants listed in national censuses (tables 1 to 3), notwithstanding some gaps due to a lack of data, make this increase apparent. The total number of intraregional migrants (natives of the countries of the region recorded in censuses in the other countries of North and Central America) rose from some two million four hundred thousand around 1970 to four million around 1980 and about seven million in 1990. Examination of the figures shows that people originating from Mexico and the Central American countries have accounted for a larger and larger proportion of intraregional migrants: in 1970, at just over a million people, they represented 40% of the regional stock of migrants; by 1990 they numbered more than five and a half million people, accounting for a share of over 80%. As well as growing in scale, emigration from Central America in the 1970s and 1980s also changed its destination. Thus, the 1970 round of censuses showed that somewhat over half (140 thousand) of all those born in the countries of Central America who were then living elsewhere in the region (268 thousand) were still within Central America. Ten years later, in 1980, that proportion was down to just over a fifth, with around three quarters (331 thousand) of all emigrants from Central America being in the United States. In other words, it appears that during the 1970s the geographical centre of gravity for migration from Central America underwent a shift. Nonetheless, this shift in destination should not be exaggerated, as there was also a very large increase in movements between the countries of Central America itself In Costa Rica, for example, immigrants from other Central American countries doubled in number between the 1973 and 1984 censuses (31 thousand and 62 thousand people respectively). Furthermore, it needs to be added that the data on stocks of migrants recorded in the national population censuses did not include the effects of large-scale movements of refugees and displaced persons which took place mainly within the subregion comprising Central America and Mexico. Somewhat conjectural estimates which have been arrived at for the middle of the 1980s put the total number of displaced persons (between and within countries) at around two million (ECLAC, 1993); a decade later, these figures had fallen considerably due to return programmes and to the change in status of refugees (ACNUR, 1997).

Notwithstanding the above, there are grounds for maintaining that Central American migration of the more "permanent" kind tended to go to destinations outside of the subregion, thus following the same pattern as that traditionally shown by migration of Mexican origin, be this permanent, temporary or cyclical. Indicative of this is the fact that over 80% (one million people) of the total stock of Central American emigrants that had accumulated by the end of the 1980s was registered in the United States census for 1990. Central American emigration to Canada, which had hitherto remained within fairly narrow limits (with less than 5 thousand people in 1981) multiplied tenfold during that decade (reaching 48 thousand people in 1991).

If figures for migration within the region are broken down by country of destination, the share of the United States is found to be particularly high. According to the 1970 census, that country then contained around one million eight hundred thousand immigrants originating from the rest of North and Central America, half of whom were natives of Canada. In other words, three quarters of the total of intraregional migrants that had accumulated as of 1970 had the United States as their destination. Although the number of people in the United States who were natives of Canada fell slightly during the 1970s, the stock of immigrants from the region that was reported in the United States census of 1980 was double the 1970 figure, standing at three million four hundred thousand people. This increase is explained by a tripling of the number of migrants from Mexico and Central America. In 1990, the figure for immigrants from the rest of the region recorded by the United States census stood at more than six million people, and 88% of these were natives of Mexico and Central America. Although the bulk of the increase in the stock of immigrants from the region living in the United States was accounted for by natives of Mexico, for whom their northern neighbor has traditionally been the preferred destination, the increase in the number of Central American immigrants was particularly pronounced, with virtually a tenfold rise between 1970 and 1990.

It needs to be added that the rate of increase in the numbers of people emigrating from the Mexico and Central America subregion to the United States was not constant between 1970 and 1990. This increase was at its peak in the 1970s, when the annual average rate of growth stood at around 10%; this rate fell to a little over 7% in the 1 980s (table 4). This decrease is also found to hold for most of the flows originating from the individual countries of the subregion, although in certain cases, Nicaragua and Honduras being prime examples, the 1 980s saw an increase. In any event, it needs to be stressed that the highest indices of relative growth between 1980 and 1990 were among immigrants originating in countries that do not border the United States, such as El Salvador, Guatemala and Nicaragua. The data provided by the United States Current Population Survey for 1996 suggest that the rate of increase in regional immigration has tended to remain within the limits that were found to hold in the 1 980s. Extrapolations from this survey indicate that the stock of regional immigrants stood at nine million people, 90% of them being natives of Mexico and Central America.

The importance of the United States as a destination for intraregional migration should not obscure the fact that other nations have also been in receipt of substantial flows. This is true in the case of Costa Rica, which has historically proved attractive to the population of neighboring countries, Nicaragua in particular; the number of Central American immigrants given by the Costa Rican census of 1984, most of them from Nicaragua and El Salvador, was more than double the stock that existed in 1973 (tables 1 and 2). Having remained almost totally immune from the sociopolitical and economic upheavals that affected the other countries of Central America, Costa Rica played a vital role in taking in refugees and displaced persons, many of them without any official status, during the 1 980s. According to official figures, in 1997 or thereabouts there was a total of 100 thousand foreigners in Costa Rica, and three quarters of these were Nicaraguans (MEIC, 1998); this figure is net of displaced persons who have been repatriated and of more than fifty thousand refugees who have been integrated into Costa Rican society.

Canada, despite having traditionally had a net outflow of migrants to the United States (which has been declining), has been another country of destination for intraregional migration. Although in absolute terms the figures for immigration into Canada from Mexico and Central America are substantially lower than those for the United States, the rate at which these flows have grown, particularly in the case of immigration from El Salvador, has been higher if anything, picking up in the 1980s and declining during the first half of the 1990s (table 5).

Although there are no comparative figures for Honduras and Belize, there are clear signs that both countries were important destinations for migration from other countries in Central America. During the 1980s Honduras took in a large number of displaced persons from neighboring countries, especially Nicaragua. Belize, for its part, is an example of large-scale "transmigration", as emigration of the native population has been combined with in flows of migrants from El Salvador and Guatemala; since this is the least populated country in the region, these movements have wrought profound changes which are reflected in the ethnic composition of the inhabitants and their distribution across the territory: the mixed-race Latin population increased from one third of all inhabitants in 1980 to more than 40% in 1991 (overtaking the Afro-Belizean population), while at the same time the population showed a tendency to become more rural (Woods and others, 1997). Finally, alongside the increase in emigration to the United States, Mexico took in a large number of displaced persons and refugees from the countries of Central America. As can be seen from the figures available, the stock of Central American migrants recorded in Mexican censuses virtually quadrupled between 1980 and 1990 (fourteen thousand and forty-nine thousand people, respectively); the bulk of these migrants originated in Guatemala.

To sum up, the large increase in the number of intraregional migrants and the change in destinations are indicative of the effects both of economic restrictions and of the sociopolitical instability experienced by the countries of Central America between the middle of the 1970s and the end of the 1980s. The figures also show that the flow of Mexicans to the United States has remained large, albeit with fluctuations, over that same period. The changes detected when figures on stocks of migrants are examined suggest that there has been a transition, which is unlikely to be reversed, away from the historical pattern of south-south migration within the Mexico and Central America subregion and towards a pattern of south-north migration. The growing importance of the United States and Canada as the destinations of preference for migrants originating in the countries to the south tends to corroborate this finding. It is also found that the scale of population interchange between the United States and Canada has diminished. On the other hand, despite the new preponderance of the south-south [sic] pattern, the figures show that migration between the Central American countries and Mexico grew substantially between 1970 and 1990.

Of course, figures obtained from population censuses do not enable us to evaluate the scale of movements of a temporary or cyclical nature, which have undoubtedly become considerably more frequent with the passage of time; it is probable that these movements have increased greatly due to the effects of economic agreements signed between the countries, such as the North American Free Trade Agreement. To the data on stocks of migrants must be added those relating to people who were displaced in Central America during the years of escalating violence; although the figures for these are not known with any certainty, they were undoubtedly very numerous, and they produced substantial effects both in the countries of Central America themselves - mainly Costa Rica as a recipient and El Salvador, Nicaragua and Guatemala as countries of emigration - and in Mexico and the other countries of North America. Again, mention must be made of the intensive repatriation process initiated at the end of the 1 980s, which speaks volumes about the fruitful efforts towards mutual cooperation made by the countries of the region, and the support given by international bodies.

3. Sociodemographic profile of immigrants from the region into the United States.

Although the United States, as the place where the south-north and north-north patterns of the American continent converge, is the leading destination country for migration within the region, it would be wrong to assume that the different currents of which this migration consists go to make up a homogeneous whole. Among those immigrating into the United States there are similarities, but also differences, and the differences become particularly apparent when we distinguish, for example, between flows originating in Mexico and those coming from the Central American countries. Of course, the characteristics of Canadian immigrants are specific enough for these to qualify as a group apart. Although scanty, the information available on the characteristics of migrants is enough to enable certain sociodemographic attributes which differentiate them to be identified, and these will be briefly described below (tables 6 to 10) 5.

As regards composition by sex, it is found that at the time of the 1970 census the cumulative total of immigrants of Mexican and Central American origin in the United States showed a predominance of women, and the same situation is reflected in the gender breakdowns for all the subregional flows. This information casts doubt upon the supposition that intraregional immigration into the United States has recently become "feminized" 6. The data available show that, as far as migration originating from Central America and Mexico is concerned, this phenomenon is not a new one and nor does it represent a growing trend. Furthermore, the striking thing is that in 1980 this situation is found to have been reversed in the case of Mexicans, with men accounting for a growing proportion over time. Something similar is found when we compare flows originating from different Central American countries (table 6). In both 1990 and 1996, stocks of Mexican migrants show a clear majority of men, an attribute which is shared, although not to the same extent, by those originating from El Salvador (tables 7 and 8). Among the other Central American populations living in the United States during the 1 990s, by contrast, women still predominate on the whole, although not to the same degree as in 1970 (tables 6 and 9). In the case of Canadians, perhaps because this is an ageing stock, there is a growing majority of women (table 10).

Another feature that can be identified concerns the age structure of migrants from within the region recorded as living in the United States. Although it might have been expected that this profile would reflect the demographic changes that have taken place in the areas of origin, i.e. a gradual reduction in the percentage of young people due to a decline in fertility, as well as changes in the type of migratory selectivity, depending on whether the migration is for the purpose of work or of family unification, the data do not show a clearly defined trend. One feature that is observed, both among immigrants originating in Mexico and among those from Central America, is a persistently high percentage of people of active and reproductive ages. By contrast, immigration from Canada is made up to a large extent of older people.

The age structure characteristics of migratory stocks from Mexico and Central America that have accumulated in the United States translate into a high level of participation in the labour market; as was to be expected, this participation is much lower in the case of immigrants from Canada. Given the extent of Mexican and Central American participation in the United States labour market, it is worth inquiring into the qualification profiles of these immigrants. A rough idea of the situation in this respect is provided by data on people's levels of education. It has traditionally been maintained that immigrants from the south have a very low level of schooling, and that this situation arose in the 1970s and 1980s as immigration flows took on a mass character. This perception is supported only in part by the data for Mexicans; although their cross-border migration patterns are closely associated with agricultural activities, there can be no denying that this flow contains large contingents of people with university education. Central American migrants, again, are characterized by increasingly high levels of education, although this is less true of Salvadorans than of the natives of other Central American states; in fact, according to data from the United States ongoing Population Survey for 1996, as of that year more than 10% of Central Americans (excluding Salvadorans) aged 25 and over had university education. Of course, the educational profile of immigrants of Canadian origin is much higher than that of other immigrants.

Two further indicators give a rough picture of the social situation of migrants: the incidence of poverty, and home ownership. According to data from the 1990 census in the United States, in the case of immigrants originating from Central America one in every five households was living in poverty; this proportion was somewhat higher among households whose members were born in El Salvador, and higher still among the households of Mexican immigrants. Data on individuals provided by the 1996 ongoing population census suggest that this situation persists, or even that it is worsening. In relation to the second of the two indicators mentioned, the 1990 census showed that one quarter of Central American immigrant households owned their own home; among those of Mexican origin the percentage was somewhat higher. In more recent years, as the data from the 1996 survey show, home ownership appears to have become more widespread among people from Central America, Salvadorans in particular. Generally speaking,' though, immigrants from the south suffer from a high incidence of poverty and a low level of property ownership, home ownership being an example of this. This situation contrasts with the low indices of poverty and high indices of home ownership found among Canadian immigrants in the United States.

It is likely that the socio-economic profiles of migrants have changed in ways that are considerably more complex than what can be deduced from the data referred to. These changes will have been linked both to shifts in the demand for labour in the United States, associated with "industrial restructuring" and the application of flexible labour criteria, and to increases in the educational level of the population of origin. Furthermore, in terms of the factors that bring about emigration, violence must have had an influence on the socio-economic profiles of Central American immigrants into the United States.

Despite its limitations, the brief description provided makes it clear that there has been an increase in the number of intraregional migrants over recent decades, that the pattern of migration fiows has shown a change in course, and that the sociodemographic composition of these has become rather more complex. Of course, migration is a much more far-reaching phenomenon than would be suggested by examination of migratory stocks alone, since there are other forms of mobility in which the people participating tend to maintain a foothold in their communities of origin, as exemplified by seasonal and circular movements 7. In addition, as long-distance migration has become increasingly commonplace and migrants have entered cultural environments different from those they come from, there has been a tendency for "transnational" spaces, communities and families to come into being. This diversity is creating a situation different from the one that policies have traditionally been designed to address, i.e. permanent residence and naturalization of foreigners.

The category of "illegals" merits a separate mention 8. The very nature of this type of migration makes it difficult to establish its magnitude. Although there are illegal immigrants in all the countries of the region, these are more visible, and there are more of them, in the United States. The Immigration and Naturalization Service estimates that there were around five million illegal immigrants in that country in October 1996 (around 2% of the United States population) and their numbers were increasing at a rate of two hundred and seventy-five thousand people a year (INS, 1997). According to the same source, 54% of illegal immigrants (around 2.7 million people) were from Mexico, and more than 13% were from El Salvador (335 thousand), Guatemala (165 thousand), Honduras (90 thousand) and Nicaragua (70 thousand); most of these people had come into the United States without passing through obligatory inspection procedures at the port of entry. Some of these illegal immigrants, including in particular many Salvadorans, are people whose deportation deadline has been extended or who are applying for the amnesty created by the 1986 law on migration reform and control (IRCA), and who se residence status is still pending (ibid.). It should be added that among the countries of origin for illegal immigrants, the INS puts Canada in fourth place (with 200 thousand people). To sum up, around 70% of all illegal immigrants in the United States appear to come from the other countries in the North and Central America region (ibid.).

III. Factors in the development process that affect migration

1. Demographic processes: labour supply.

If migration is to be interpreted and forecast, it is vital to have an understanding of how demographic factors operate, in combination with others such as political and economic developments both locally and internationally and the struggle between those sociocultural patterns that are becoming global in nature and those that maintain the identities of individual peoples. The changing migration experience of. Western Europe illustrates the importance of these factors. The great streams of emigration that carne out of that region between the second half of the eighteenth century and the first half of the twentieth, of which the main destinations were the Americas and, to a lesser extent, Oceania, were due in some degree to rising population growth which could not be absorbed productively by the societies and economies of these countries. Likewise, the immigration flows that have gone into Western Europe in the second half of the twentieth century can be attributed not just to the socioeconomic attractions of that region, but also to the demographic processes taking place both in the countries of emigration, with rapidly rising populations and an excess of labour, and in the receiving countries, where labour was in short supply during the periods of economic growth in the 1960s and 1970s and there was a dearth of workers to carry out lower-skilled activities, which were often disdained by the indigenous labour force.

The European example demonstrates the advisability of examining what the role of demographic processes might be in the countries of the North and Central America region in terms of their effects in stimulating or inhibiting international migration. One of the first things that comes to light is what a diversity of situations there is: the populations of some countries have reached an advanced stage in the demographic transition process (Canada and the United States), others are experiencing this transition in full force (Costa Rica, Mexico and Panama) and a third group are experiencing it in a more moderate way (Belize, El Salvador, Guatemala, Honduras and Nicaragua) 9 Over and above the technical characteristics of these individual situations, when international migration is being examined it is particularly useful to consider the effects of these changes on the dynamics of the labour force, as people of working age make up the bulk of migrants. In this respect, it should be noted that the demographic transition in the United States and Canada is giving rise to a drop in the growth rate of the population of working age; the offspring of the baby boom, which took place approximately between 1945 and 1960, have already been incorporated into the workforce, and the replacement rate is low, partly due to the baby bust after 1960; this is throwing up challenges in a number of areas, among them social security financing. Although in the other countries of the region the trend has been for the birth rate to drop, the effects of this in terms of growth in the population of working age have only just begun to be felt, and in some cases still cannot be discerned with any clarity. 10

In countries where the decline in fertility dates has been rapid and is of some years standing, the forecast is for a gradual reduction in the rate of growth of the population reaching working age; in Mexico, for example, this rate is expected to fall by half between 1996 and 2010 (Gómez de León and Tuirán, 1997). This trend opens up a demographic scenario in which migration tendencies are likely to be eased; thus, the fall in the rate of growth in the population of working age in Mexico could lead to the excess in the labour supply being cut, making it easier for people to find employment in the local market (CONAPO, 1997). Nonetheless, this conclusion is simplistic, since it fails to take account of the powerful attraction exercised by the labour markets of the United States and Canada, particularly at a time when the working population inherited from the baby boom will be beginning to retire from work. Furthermore, the conclusion is illusory: the actual trend in the labour force depends not only on demographic factors, but also on developments in labour force participation rates 11 . Further still, in countries which have suffered internal warfare, and which are experiencing the demographic transition only to a moderate degree, no short-term reduction is expected in the growth rate of the population of working age. Persistently high levels of fertility, combined with a diminution of the direct effects of violence in terms of forced emigration and mortality, will lead to the ranks of new job-seekers being swelled.

Another subject for conjecture is the possibility of emigration pressures being eased due to a change in the pattern of fertility decline that differs by socio-economic and ethnic groups. This decline has been particularly pronounced among groups with a higher level of education and better living conditions; by contrast, it has barely crept in among poorer groups, rural and indigenous people and those with a low level of schooling (ECLACICELADE, 1998). It might be expected, however, that properly focused programmes and increasing pressures from the social environment may help to speed up the drop in fertility in these groups that lag behind the rest of society. Given that a large proportion of all emigration appears to be accounted for by these groups, such a reduction in fertility would imply a falling-off in the potential for emigration in the medium and long term (CONAPO, 1997).

Notwithstanding the importance of demographic factors in explaining international migration, there is a consensus that the influence of these factors is felt through the medium of other aspects of development. Thus, the impact of demographic variables on migration takes on significance by virtue of their interconnections with trends in employment, institutional factors and inequality of income distribution in countries of emigration. Similarly, demographic variables influence migration in conjunction with the dynamism of demand for labour and the extent to which labour markets are open in countries of immigrations.

2. Economic factors: demand for labour. The 1980s brought with them not just economic crises, stagnation of employment and rising levels of poverty in developing countries, but also profound changes in the economic structures of all the countries of North and Central America. Of course, the direction and effects of this economic (and labour market) reorganization have differed. In Mexico and Central America, economic growth and employment have recovered, albeit still only modestly and with violent fluctuations (ECLAC, 1996). The pattern of recent economic development reveals clear shortcomings: rates of growth in output are below the historical norm (between 1945 and the begriming of the 1 970s); the situation is still precarious, as is demonstrated by high current account deficits, the need to carry out periodic adjustments .and the fragility of financial systems; and savings and investment rates are still low ~CLAC, 1 997a). Economic restructuring has exacerbated long-standing structural discrepancies, as productivity differences have widened between "modern" firms and non-dynamic traditional activities, which account for the bulk of employment (ibid.).

After 1975, when the economies of Western Europe and East Asia were increasing their market share, it became apparent that in the United States there was a need to raise the international competitiveness of the economy. This quest for competitiveness led to more thorough-going application of technical advances, including the introduction of new company organization and management principles, which gave rise to "industrial restructuring" and geographical adjustments in the location of production facilities, all of which are linked to job losses, less equitable income distribution and erosion of real wages (Farley, 1996; Sassen, 1988; Jaifee, 1986). As a result of this transformation, the demand for labour carne to be "polarized": the relative numbers of jobs requiring "medium" levels of qualifications diminished, while there was an increase in the number of jobs available both for the more highly skilled and for those with little training, these latter jobs being characterized by a lack of worker protection (World Bank, 1995; Levy and Murnane, 1992). Within this latter segment of the labour market, which includes high-turnover temporary jobs, niches have opened up for immigrant workers; since these jobs do not provide any additional benefits to compensate for the relatively low wages they pay, they are likely to be unattractive to the native population (Haas and Litan, 1998).

There can be no doubt that the dynamism of job creation in the United States and the substantially higher wages that are paid there by comparison with neighboring countries to the south have acted as a powerful magnet for migration. Although economic restructuring in the United States during the 1980s and at the beginning of the 1990s led to a slowdown in the demand for labour, since 1992 there has been a sustained recovery (Zuckerman, 1998). It is estimated that the average number of non-agricultural jobs generated each year between 1994 and 1996 was around 3 million; since the average annual growth in the population of working age during this three-year period was 1.3 million people, demand for labour apparently outstripped the domestic supply (Escobar, 1998, based on statistics from the Monthly Review of Labor). These conditions have reinforced the standing of the United States as the centre of attraction for the regional labour force, as well as for the rest of the hemisphere. Since it is highly improbable that the low rate of growth in the United States population of working age will pick up in the near term - it is more likely that it will decline because of the baby bust - and since growth in rates of participation in economic activity has slowed down over recent years, forecasts suggest that there will continue to be an historic discrepancy between labour supply and demand. In Canada, the gap between the labour supply and the demand for labour is much smaller, but it is considered that migrants will continue to be required in order for that country's economy and society to function properly.

By contrast with the developments being seen in the United States and Canada, employment trends in Mexico and the Central American countries have been severely affected by the poor performance of their economies during the 1 980s (the "lost decade"); the subsequent recovery has been rather meagre (table 11). It appears to be unquestionable that during the 1980s "the fragile employment equilibrium which had been successfully maintained during the previous period of growth was broken" (ECLAC, 1996); the manifestations of this were: "real wages fell, open unemployment increased and the proportion of jobs that were in sectors with lower average productivity rose" (ibid.). The restrictions, weaknesses and rigidities of labour markets appear to have persisted even once the worst moments of the crisis and the adjustment process of the 1 980s had been coped with; in reality, the growth that began again in the 1990s has not been characterized by significant advances in terms of job creation and improved equity (ibid.). Against this background we need to set the effects of recessive interludes, like the one suffered by Mexico between 1995 and 1996, whose consequences have quickly spilled over into higher unemployment (table 12). Although open unemployment rates do not seem to be exceptionally high in Mexico and Central America, except in Panama and Nicaragua, the low productivity of many jobs, particularly in agriculture and services, and the low earnings provided by wages, indicate that labour is being underutilized.

To sum up, the quantitative dynamics of labour markets enable us to identify both forces of attraction in the more developed countries of the region and forces of expulsion in the other states. To the incentive to migration which these factors provide needs to be added the important role played by the institutional conditions under which these markets operate, especially in the United States. The pressures inherent in the search for greater competitiveness have led to greater labour flexibility, involving a high turnover of employment and efforts to find ways of reducing labour costs. The use of migrant labour is generally one of the means employed to help achieve this objective, and the hiring of "foreign" supervisors tends to facilitate the recruitment of nonnative workers (Martin, cited by Escobar, 1998). Another aspect of the way United States labour markets operate which is also a factor in the employment of migrant workers is the issue of compliance with legal requirements. Although official agencies have become more and more rigorous in enforcing application of controls, it has been found that a far from negligible proportion of employers elude these regulations (Fraser, 1994).

The qualitative characteristics of the demand for migrant labour appear to be crucial when consideration is given to the scope for replacing these workers with domestic ones. A high proportion of migrants to the United States work in agricultural activities which, because of their seasonal nature, are markedly cyclical in terms of the demand for labour; this circumstance, together with considerations of cost, leads employers to hire migrant workers (13ustamante, 1997). Another large and growing proportion of migrants work in services and manufacturing activities that do not require high qualifications (CONAPO, 1997); although jobs of this type tend to pay wages lower than the United States average, which makes them unattractive to the native workforce, the pay is far higher than what could be earned in similar occupations in the countries from which the migrants originate. As may be gathered from the above, the qualitative aspects of migrant labour and of the demand for work requiring low levels of qualifications tend to produce rigidities and inertia which make it difficult to replace migrants smoothly by native workers.

The distinction between migrant agricultural and non-agricultural workers is a factor that might have some influence on the stability of migration and the qualitative perception of its effects by the society that receives the migrants. The historical demand for migrant labour for agricultural activities, especially in the south-east of the United States, does not appear to be directly responsible for the current visibility of migrants in society. Because of the seasonal nature of the work, migrant workers in agriculture are distinguished by the circularity of their movements. Furthermore, as these activities take place in thinly populated rural areas, the presence of the migrants is not particularly noticeable (Bustamante, 1997). The visibility referred to arises rather when migrant workers are employed in manufacturing and services occupations, particularly in metropolitan areas, whether these be in the south and south-east of the United States or in New York and Washington (Pellegrino, 1995). These workers settle in cities and tend to form colonies of permanent residents, where they succeed in maintaining their original cultural identity. Notwithstanding, migrants of this more "permanent" type undergo further-reaching processes of assimilation, which gives them greater prospects for participation in the sociocultural life of the United States, and this participation, which includes higher consumption of services and access to social benefits, is a factor that is instrumental in producing the impression that the costs involved in migration are greater than the benefits it brings.

3. Political factors. There can be no doubt that political factors, manifested in the way power is

shared out in society and in the patterns of inclusion and exclusion associated with this, are also linked to migration. The channels through which this linkage operates are complex; not only &e they mixed with other aspects (social, economic and cultural) of the development process, but they also depend on historical circumstances, the degree of national autonomy, the ways in which different countries relate to one another, and the solidity of institutions. Among the traditional manifestations of the effects that political factors have on migration within the region is the practice of giving asylum to elites or governing groups (the United States, Mexico, Costa Rica and Panama have been particularly prominent in giving sanctuary to these); this practice has helped to keep various types of ideological movements in being, as many of these migrants return to the political arena once they are back in their countries of origin. A different situation was experienced by the region between the second half of the 1970s arid the end of the 1980s, the difference lying in the massive nature of the emigration and the multiplicity of its repercussions. The escalation of political conflict, and in particular the violent confrontations between warring factions, created a powerful expulsive force in El Salvador, Guatemala and Nicaragua; of course, the effects of this violence were felt in the other countries of the region, mainly due to the movements of displaced persons and to the granting of asylum and refugee status, which took place on an unprecedented scale.

With peaceful forms of political competition now reestablished, it may be said that ideological conflict is no longer such a determining factor in emigration. The restoration of peace in Central America and the establishment of more solid forms of political participation in Mexico mean that economic factors once again occupy a special place among the causes of international migratory movements. Even so, armed conflicts and their consequences have provided at least two lessons. The first is the need to take measures to incorporate social actors effectively and efficaciously into the decision-making process by consolidating democracy arid opening up real spaces for popular participation. Besides their intrinsically positive character and their merits in terms of public ethics, since they are directed at securing respect for rights that are considered to be universal, such measures can prevent political action from deteriorating to the point where it gives way to military activity; furthermore, they help to involve the different social groups in a commitment to national development and to fair distribution of the benefits it produces. All of these things would be helpful in reducing incentives to emigration.

The second lesson concerns the way migrants are regarded by the States and civil societies of their countries of origin. Although the political factors which led to massive emigration may be described as essentially transitory, their effects have been more lasting, since a proportion of migrants ended up settling "permanently" in the place of destination (especially when that place was the United States). After consolidating their position, they have sought to reunite their families in the destination country, and this has turned into another factor producing migration. Notwithstanding this, these migrants have tended to form into colonies, in an effort to preserve their cultural identity, and they keep in contact with friends and family in their areas of origin and make cash remittances. Furthermore, it is likely that some will return to their countries of origin, and that others who do not do so will wish to maintain a close relationship with their countries. This situation means that there is a need to redouble efforts to welcome back those who do return and maintain close links with those who decide to stay in the destination country.

Another political factor which can be crucial for international migration is the situation obtaining in the destination countries, as regards both the provisions they apply to immigration and social attitudes to the issue. By and large, the abundant flow of Central Americans entering the United States in the 1970s and 1980s enjoyed a high degree of social and governmental approval (as manifested in part by the 1986 law on migration reform and control, IRCA). Subsequently, migration legislation become stricter and border controls were tightened; these measures can affect migration both directly, through the detention of illegal migrants and refusal of entry applications, and indirectly, by dissuasion. However, there is a degree of consensus about the limitations of these mechanisms as instruments for checking the migration process, particularly when they are not accompanied by measures to change the factors that trigger this off (Espenshade, Baraka and Huber, 1997).

4. Social factors. Being a social process, migration is closely related with all the other components of this aspect of development; in this section, however, brief reference will be made to two factors that it is linked with in a particularly direct way. The first of these is that of the prosperity and living conditions of the population, which are related to income distribution and access to basic services. The second factor is the social structure of migration, in terms of the linkage between people living abroad and their communities in their countries of origin, and the social capital of migrants.

In the North and Central America Region there are stark inequalities in living conditions and levels of prosperity. The first difference, which is self-evident, is the one that exists between the more developed countries (the United States and Canada) and the countries in the Mexico and Central America subregion. This difference is not just in their average levels of prosperity, but also in the way the benefits of material progress are distributed. As has already been mentioned, Mexico and the Central American countries, without ignoring the differences between them, have inherited a situation of social inequity which grew more acute during the crisis of the 1980s; in the first half of the 1 990s this was corrected only very slowly, as is shown by the high incidence of poverty (which is 30% of all households or more in all the countries except Costa Rica and Panama) and of inequality and rigidity in the distribution of income. These circumstances combine to create a context which is favourable to emigration.

Some general remarks need to be made about the effects of living conditions on international migration. The first is that any deterioration in these conditions acts as a stimulant to migration; except in extreme situations, when survival itself is at risk, emigration will only actually occur if there are places that offer better opportunities than the current place of residence, and information is available - often from friends or relatives - about these alternatives. Secondly, any improvement in living conditions can have a bipolar effect on international emigration. In the long term, it is possible that it may lead to a reduction in the tendency to emigrate if it makes it easier for people to satisfy their needs. In the short and medium terms, however, it may cause emigration to increase if it results in people being more exposed, and more receptive, to information about alternative locations, and this effect is reinforced when transport and communications improve. Furthermore, rising living conditions can cause people to raise their expectations; if these are not satisfied locally, the resultant discrepancy becomes an incentive to emigration. The case of Mexico illustrates these propositions: although it is true that economic crises have stimulated emigration to the United States, the progress made in people's living conditions does not appear to have been effective in containing this emigration.

As regards social networks, these may be described as informal organizations representing the "microstructures of migration"; as well as being the means for reinforcing social cohesion and cultural identity, they are institutions whose workings make it plain that decisions to migrate do not come down to individual calculations but are made within a collective context (Arizpe, 1978; Massey and others, 1987). This is one of the reasons why many migrants move in groups and not as isolated individuals; it is also one of the factors which, over and above trends in the demand for labour, explain why these groups have preferences for certain specific locations in the destination country (Portes and Bach, 1985). Networks are an important source of support, as they provide information, a welcome, social relationships and recourses, which are necessary both for crossing borders and for obtaining property and employment. Consequently, they help to minimize the risks and costs of movement, something that is particularly vital for illegal migrants.

Restrictive immigration policies and tighter border controls increase the incentive to belong to and use these networks. The strategies which these have designed have the function of obviating the effects of measures aimed at restricting immigration 12. Furthermore, with the improved communications and transport conditions that now exist, these networks enable migrants to remain in contact with their communities of origin and channel resources towards them, which makes it more likely that they will return to those communities. Thus, if these networks were weakened or destroyed the likelihood of migrants being able to make a successful place for themselves in the destination country could diminish, but at the same time it would be more difficult for them to keep in contact with their communities of origin and to return to these, and thus there would be less pro aspect of migration being beneficial to the development of those communities.

5. Environmental conditions. Although work on the effects of environmental factors on migration does not yet seem to be sufficiently advanced to provide conclusive and incontrovertible evidence, information is available about the limitations that different environmental problems pose for the settlement or permanent establishment of populations. This type of relationship has been discerned with greater clarity in internal population movements. Conditions such as damage to soil (by erosion, salification or desertification), contamination (of the air or water) or natural disasters tend to produce expulsion of the population groups directly affected, in both rural and urban areas.