LIVING OFF THE NORTH

Alejandro I. Canales
Department of Regional Studies University of Guadalajara.

Introduction

Migration between Mexico and the United States is a complex process, backed up by the formation of social, cultural and family networks underpinning the creation of a system of exchange and circulation of people, money, goods and information between migrant settlements on both sides of the border (Rouse, 1992; Smith, 1994). Hence, the migration process does not entail the dissolution of the home communities but, rather, the shaping of such communities into transnational communities. (Guarnizo and Smith, 1998; Kearney and Nagengast, 1989). Within this general framework, remittances represent a vital element in the reproduction and economic sustenance of these "de-territorialized" migrant communities. This situation also involves families who devise forms of social and economic reproduction based predominantly on the income earned from by migrants in the United States (Canales and Zlolniski, 2001).

Yet, even when all the households comprising a transnational community partake in the migration phenomenon on an equal basis, the amounts, pace and frequency of remittance flows is not the same for all. It would seem that the significant disparities that characterize periods and conditions under which a given household becomes a recipient of remittances are determined by social, demographic and economic factors as well as by family structure, composition of the household, profile of family members, development of family cycle and by family and individual migration patterns and tradition. For the same reason, the presence of migrants in a household is not enough to assume that such a household is more likely to receive remittances. The family and domestic arrangements underpinning each particular migration pattern must be determined, as well as the way in which such arrangements translate into forms of socio-demographic profiling and different socio-economic requirements.

Thus, in some cases, the temporary migration of the head of the household entails the establishment of specific family arrangements. In such arrangements, the home of the parents or the in-laws of the head of the family plays a vital role, especially if there are dependent children. (Woo, 2001). In other cases, migration leads to the establishment of transnational family units that depend for their preservation on the continued flow of people and of material and symbolic goods becomes.(Moctezuma, 1999). In any case, the important thing is that the periods within which the remittances are sent and received depend largely on the nature of the domestic arrangement underpinning the migration process in each family.

In this respect, the objective of this research is to analyze the differentiation factors associated with households that receive remittances, as well as the characteristics of each family and domestic arrangement that play a most important role to determine the timing and frequency of remittances. Our aim is therefore to identify the profiles of households receiving remittances in high-migration communities in Western Mexico, on the basis of specific statistical models

I. Migration and Remittances in Mexico

From the 1980s onwards, the magnitude and modalities of migration between Mexico and the Unities States have revealed significant changes in the dynamics and composition of migration flows. According to data from the United States Bureau of the Census, the total volume of individuals born in Mexico and residing in the United States grew from a little under 2.2 million in 1980 to almost 4.5 million in 1990 and to 7.9 million in the year 2000. These figures indicate an annual net balance of 226 thousand people in the eighties and 343 thousand people by the 1990s.

These changes in the magnitude of migration flows affect directly the flow of remittances sent by emigrants to their home communities in Mexico (Lozano, 1998). In actuality, total international transfers, as well as family remittances, have exhibited continued growth in the last two decades, registering in 1999 an eightfold increase compared to the levels recorded in the early 1980s. By the same token, family remittances averaged almost 80%of total international transfers, later topping 93%in the 1990s (Canales, 2002a).

However, an intense debate has developed in political, academic and even civil society circles regarding the size of remittances. In this regard, we are interested in highlighting two aspects: on the one hand, the role and impact of remittances at the macro-economic level and at the level of households and local economies; and on the other hand, the individual, family, community and structural factors determining the transfer and receipt of remittances¹.

i) As regards the impact of remittances, earlier studies concentrated on gauging the importance of remittances themselves in relation to macroeconomic indicators. This allowed for an initial approximate estimate of the relative importance of remittances. Subsequently, more accurate estimates have been made regarding the impact of remittances on the national economy, based on social accounting macro-economic models. The purpose of such models is to estimate the multiplying effects of each dollar entering the national economy in the form of remittances.

However, the weakness of these models derives from the quality of the information required for the effective design of social accounting matrices. As is often the case, the information needed may not be available or may be highly fragmented. In view of this situation, specific surveys are usually conducted at the local or regional levels to ensure that the different indicators and coefficients required by the social accounting econometric models may be directly gauged. In this sense, pioneering work has been carried out by Adelman and Taylor (1990), who estimated that the multiplying effect of these remittances on local and regional economies was equivalent to 2.9. In other words, each dollar entering the regional economy leads to a 2.9-dollar increase of the GDP.
In addition to these social accounting matrices, probability models have also been used to estimate the impact of remittances on income distribution. These are econometric models that make it possible to estimate the rates at which poverty and inequality indexes vary in income distribution as a result of remittance flows. (Taylor, 1992).

In the case of Mexico, in particular, the results are not conclusive in the sense that, while remittance flows are large, they actually account for less than 3%of household income, thereby substantially minimizing any effect they might have on income distribution. At the regional and local levels, however, the role of remittances does seem to be significant and, generally speaking, remittances may be said to contribute to lessening economic inequalities. Canales (2002b), in particular, found that, in the case of Western Mexico, remittance flows tend to contribute to raising income by between 5% and 15%, depending on the indicator used to measure economic inequality.

At the same time, analyses of the size of remittances and their potential multiplying effects have given rise to a line of debate concerning the role of remittances as a powerful source of local and regional economic development. Viewed from this perspective, remittances may be conceived as a form of migrant saving, earmarked for financing productive investment projects and the incorporation and consolidation of economic businesses and establishments for migrants.

Thus, Durand and Arias (1997), for example, in a study on San Francisco del Rincón, Guanajuato, documented the establishment of shoe factories with backing from migration dollars. According to these and other authors, this example, along with numerous other cases, illustrates that international migration does not represent a drain of resources for the Mexican economy but that, on the contrary, it may even constitute a significant source of productive capital and a dynamic engine for promoting business activity, for forming businesses, and for spurring economic growth, at least at the local and regional levels. (Massey and Parrado, 1994; Durand, 1988).

However, other authors take a different stand, pointing out that this optimism may be ill-founded, in that, conceptually speaking, it is highly confusing and is not backed by precise, adequate and reliable data on the complexities of the migration trend in general and of remittances in particular. In fact, if we consider that the migration of Mexicans to the United States is an eminently work-related phenomenon, then there can be no doubt that the income received from migrants represents a wage fund which, like any other, is often used preferably to support migrants' families and communities on a daily basis. (Canales, 2001). The difference lies in the fact that migrants wages are channeled to families under the heading of "international transfers" which, in economic parlance, suggests a form of "external saving" but which are actually far from being a kind of "saving" per se, as they do not possess the properties traditionally associated with savings and are used as such.

ii) An alternative approach to studying and understanding remittances is the analysis of the factors that determine them. In this regard, we can distinguish two types of approaches: first, an analysis of the factors determining the economic nature of remittances and, on the other, the individual, cultural and family-related factors determining the transfer of remittances.

Macro-economically speaking, the studies conducted have concentrated on assessing the volumes of remittances in terms of the performance of different macroeconomic variables and indicators. In this regard, analyses are performed of time series and longitudinal models which allow for an estimation of the sensitivity (elasticity) of remittances in relation to variations in each macroeconomic variable. These studies, however, seem inconclusive to the extent that the way in which certain macroeconomic variables influence the propensity to send and receive remittances depends largely on the economic conditions in both the country of origin and the country of destination (Russell, 1986, Taylor, 1999).

Nonetheless, a large number of studies agree on some basic macroeconomic relationships. In this regard, Lianos (1997), for example, believes that per capita income levels in the countries of destination, exchange rates, possible devaluations of local currencies in the countries of destination, as well as the length of stay, all are variables that best explain the variations in remittance flows. On the other hand, inflation, interest rates and income levels in the countries of origin do not seem to determine remittance flows to a significant degree.

At the micro-social level, the objective is to determine the individual, family and social conditions underpinning the act of sending remittances, as well as the amounts of money sent to families and/or home communities. With regard to this approach, remittances are linked to the circumstances which, in one way or another, are responsible for shaping the lives of migrants. The studies conducted are based on the assumption that the primary motivation behind sending remittances is determined by the socio-cultural factors influencing the loyalty and commitment of migrants toward their families back in their home countries. In addition, they particularly stress the vital role of family commitments and of the presence or absence of immediate relatives (mainly dependent children) in home countries as well as in countries of destination. (López, 2001).

However, the periods, types and amounts of remittances depend on specific circumstances which include not only family-related factors but also individual and contextual factors. Consequently, Massey and Bassem (1992), for example, point out that the decision to send remittances is directly determined both by family life cycles and by the characteristics of home communities. These authors maintain, in particular, that the propensity to remit money tends to fall in communities possessing more abundant resources and displaying more favorable economic conditions.

The transfer of remittances also seems to be associated with specific forms of migration but, most notably, with the nature of the existing ties between migrants and their home communities. In this regard, López (2001) points to two interesting findings. First, the author stresses that the tendency to send remittances, as well as the amounts sent, is usually greater among temporary and circular migrants but lower among migrants who have chosen to take up stable or permanent residency in the United States.

On the other hand, López holds that the factors determining the sending of remittances are even contradictory in each case. Thus, with regard to temporary migrants, for example, the status of the migrant within the family structure is essential to the transfer of remittances. To be more specific, pressure from families for remittances to be sent seems to be statistically significant only for those who are heads of households. This demonstrates that the transfer of remittances is linked to the immediate responsibilities of migrants toward their families and households back in their home countries. However, with respect to permanent migrants, their status within the family structure is not statistically significant in determining the transfer or not of remittances. This is so because, in this case, it is very likely that the migrant's family lives in the United States, a fact that implies a substantial slackening of the bonds and commitments of the migrant toward his home community.

Similar conclusions have been drawn by Menjivar et al (1998), who point out that whenever migrants plan to stay for more protracted periods -or even definitely- in the countries of destination, the tendency to send remittances usually decreases significantly. In this case, family and socioeconomic resources are used to ensure a more stable and lasting stay that may include the migrant's family members. However, whenever migrants plan to return to their home countries, it is more likely that remittances will be sent to ensure an improvement in status and living conditions in their home communities.

The Lozano analysis (1997) also confirms these findings in concluding that male migrants who arrive in the United States for the first time and who have property and close relatives (primarily parents, wives and/or children) in Mexico, are more inclined to send larger amounts of money. On the contrary, immigrants who were granted amnesty by the IRCA, and who chose to remain permanently in the United States, reduced the average amount of their remittances even though they have may not have experienced any changes in real income levels or other economic conditions.

Now, a complementary approach to these studies on the factors determining the sending of remittances is the one which we are offering in this paper and which focuses on the analysis of the characteristics of the remittance-receiving households in home communities. As we have pointed out, remittances form part of the migration system set up by communities across territories. In a high-migration community, almost all households are linked to the social process of migration by means of family, community, and friendship networks, etc. However, not all households have active migrants at all times nor do all households receive remittances at all times. On the contrary, the periods and spaces both for migration and for receiving remittances are determined by the periods and spaces determined by the different family arrangements made at each stage of the life cycle of the household unit, as well as by the specific migration patterns and traditions prevailing in each household.

Within this general context, any analysis of the conditions of remittance-receiving households must consider one dynamic perspective of the phenomenon. Indeed, the family arrangements established in accordance with the social migration process pose specific demands as far as the remittance flows needed to keep up with these arrangements are concerned. Yet, these arrangements are not static. Rather, they evolve in line with the dynamics of the family life cycle and with the dynamics of individual and family migration patterns and traditions. Remittances are likewise not static. Rather, they contribute to changing the conditions for reproduction and development in the household by influencing the dynamics of the family cycle and migration patterns and traditions.

This, therefore, relates undoubtedly to the effects of a reciprocal and dynamic decision. On the one hand, the conditions and needs governing the receipt of remittances vary at each stage of the family cycle, depending both on the domestic arrangements and on the individual and family migration pattern adopted by each family unit. But, on the other hand, if, at any given time remittances seem to be determined by these family arrangements, at some other time these very remittances will contribute to transforming conditions by modifying the economic base for the reproduction of the households.

Now, in view of the limitations of cross-sectional analysis, intended to highlight dynamic and diachronic relationships, this research concentrates on analyzing the differences between households that receive remittances and household that do not receive remittances, in a high-migration community in Los Altos de Jalisco². Here, the objective is to be able to gage and assess the statistical significance of each component of this differentiation structure between remittance-receiving and non-remittance-receiving households. For this purpose, we first offer a description of the role of remittances in the home community and, later, we analyze the socio-demographic profile of remittance-receiving households, on the basis of the application of a logistical regression model.

II.     Remittances and Households in Teocaltiche, Jalisco³

Jalisco is an area with a long migration tradition. For this reason, it should come as no surprise that in the year 2000 this area accounted for more than 10% of remittance-receiving households, with monthly remittance income averaging more than two thousand pesos. These migration conditions and remittances are even more visible in certain other regions and municipalities of Jalisco than in the Los Altos region, which joined the migration process in the late 19th century. In this regard, it is interesting to study the conditions that characterize both migration and the receipt of remittances in these regions where the centuries-long tradition has led to the formation and consolidation of transnational social networks.

In these communities, migration is a social phenomenon that unfolds horizontally. The social and family network system makes it possible for almost all households to be directly or indirectly linked to the social process of migration. However, the participation of each household cannot be viewed in static, but rather in dynamic, changing terms, in light of the different roles played by migration and by remittances at each stage of the family life cycle. In fact, family arrangement made on the basis of migration tend to be defined by their own dynamics and by development patterns, in accordance with the migration options opened to family members and with the need for remittances at each stage of the family cycle.

Thus, while remittances are one of the primary leit- motiv of the migration process, fully understanding them requires a structural outlook that goes far beyond the short periods of migration of each migration trend. Not all households with active migrants receive remittances. Likewise, not all households not engaged in active migration are necessarily exempt from the sending and receiving of remittances. The likelihood of receiving remittances is undoubtedly greater in the case of households with active migrants; however, it cannot be denied that, in the other households, remittances are also vital to socio-economic organization.

In this sense, the Teocaltiche case is illustrative. In this municipality, only 10% of all households seem not to have any link with the networks that have developed in the social process of migration. These are households without active or "retired" migrants, without permanent migrants and without immediate relatives of the head of the household in the United States⁴. On the contrary, 9 out of every 10 households are directly linked to the migrant social networks, though they are not necessarily characterized by active migration or by more recent migration trends.

Actually, 39% of all households have at least one active migrant, that is to say, an individual whose most recent movement took place over the last five years (1996-2000). Likewise, 20% of all households have permanent emigrants; that is, individuals who were once household members but who currently reside in the United States. At the same time, another 29% of all households have members who are immediate relatives (parents, siblings, grandchildren and/or grandparents) of the household head, whose habitual country of residence is the United States. Finally, only 3% of al households possess only "retired" migrants and no other link (active migrants, emigrants or relatives of the head of the household).⁵

Now, when data on the conditions of migration is compared with data on the conditions governing the receipt of remittances, a very peculiar situation may be observed. Actually, 15% of households without any direct links to the migration process receive remittances, a situation that increases in the case of households which are "not engaged" in migration circuits, but 30% of which receive remittances. Likewise, practically one third of all households with active migrants did not receive remittances during the year 2000 (see table). Thus, while remittances tend to be concentrated among households directly linked to migration circuits, the migrant social networks enable many households that (at least temporarily) are not involved in the dynamics of migration to nonetheless participate in the sending and receipt of remittances.

In other words, participation in migration flows is marked by particular rhythms and frequencies that are neither static nor permanent but dynamic and changing. Both participation in, and exclusion from, the migration process are therefore temporary phenomena which may vary in accordance with family and personal cycles. Nevertheless, participation in the migrant social networks is a more structural factor which is not directly contingent upon the level of participation in migration flows at a given time, but upon a life-long tradition of migration, at both the personal and family levels.

In view of the foregoing, we next wish to present some initial data on the role of remittances in the household structure, as well as in family and residential arrangements established at different stages of the family cycle. An initially vital point has to do with the role of remittances in the household income structure. In aggregate terms, remittances account for 20% of total household income in Teocaltiche for the year 2000. This figure is indeed significant because it highlights the special importance of remittances and migration in the economic dynamics of both the homes of migrants and the local and regional economy.

A comparison of the monetary income of households in accordance with their status as remittance recipients throws light on the essential role played by remittances in recipient households. In fact, in households not receiving remittances the main source of income is earned income, which accounts for 67% of the family income, followed by business income which accounts for another 28%. In the case of households receiving remittances, however, these remittances account for 40% of the family income, equaling the level of earned income but far exceeding the other sources of income (see table).

One relevant piece of data is that there seem to be no significant differences in the income received by households whether they receive remittances or not. In both cases, per capita income is approximately 1,100 pesos per month (see table). In this regard, we can gage the relative importance of the remittances in recipient households. Without the remittances, per capita income in these households will fall dramatically to under 680 pesos per person per month. It cannot be doubted that remittances play a fundamental role in the reduction of economic inequalities and of poverty among the population, especially if we consider that 48% of households were remittance recipients in the year 2000.

Additional information that corroborates this role of remittances relates to the specific way in which these resources are used in recipient households. Far from constituting a source of saving, which could help generate productive income, remittances play the same role as wages by serving as the resources needed for the day-to-day maintenance of households and their members. In fact, upward of 43% of remittances are earmarked for the purchase of direct consumer goods (foodstuffs, beverages and the like), while another 17% is allocated to household expenses (electricity, water, gas, household appliances, etc.) (See table).

The high spending on health care, as well as the low spending on education, is also noteworthy. This may be explained by the presence of children and of elderly people living in recipient households and, to some degree, accounts for the kinds of family arrangements established through remittances and migration. Thus, a somewhat recurrent phenomenon, for example, is what we have come to designate as the "like water for chocolate" syndrome. This has to do with households in which some of the older children have migrated to the United States, while one of the younger daughters has remained at home with the parents to take care of them. In this case, the daughter generally does not have a stable job and the family economy is practically maintained by the flow of remittances sent by the other children to support the other family members living in the home community. Likewise, it is not uncommon that a large proportion of the remittances sent is earmarked for the parents' medical expenses.



At the same time, the low spending on education in remittance recipient households is not due to the fact that such households do not possess young children but, rather, that the area does not offer a broad option where private education facilities are concerned. Consequently, the bulk of school-age children attend public schools which cost less. In the same way, when the head of a household with school-age children migrates, it is not uncommon for the grandparents to partially contribute to the household expenses, including the education of their grandchildren. In fact, when the head of the household migrates to the United States, the wife usually returns to her parents' home, at least temporarily, until the migration and occupational situation of the head of the household stabilizes.

It is therefore important to verify the different degrees of importance that remittances have in accordance with the family arrangements and with individual and family migration patterns. Both the size of remittances and the specific uses to which they are put change over time in keeping with the varying demands and requirements that are posed by the domestic arrangements made at each stage of the family cycle.

In view of the above, we will analyze the different kinds of family arrangements and the socio-economic and socio-demographic conditions of households, in accordance with their status as remittance recipients, with the purpose of establishing a differentiated profile for each type of household. To do this, we have used the logistical regression model which allows us to determine the ways in which remittance-recipient households differ from non-remittance-recipient households and the areas in which there are no relevant differences between the two.

Before analyzing the findings of the model, however, we wish to point out a number of methodological constraints arising from the use of this model for this particular purpose. In general, it is usually assumed that the purpose of logistical regression analyses is to determine the possibility of predicting the development of a given event on the basis of its relationship to other events taking place in the past (Vivanco, 1999). However, the predictive nature of these models is based on a series of conditions that do not always correspond to those derived from the methodological designs underpinning each research. This is why it is usually advisable that the different parameters estimated by the logistical model should be interpreted differently in accordance with whether the research is conducted longitudinally or cross-sectionally.

In the cross-sectional studies, in particular, the different relationships have been interpreted in terms of correlations and associations between the variables analyzed, while in the longitudinal studies, it is more relevant to use the effect-, prediction-, and determination-based concepts (Jovell, 1995, Vivanco, 1999). This limiting factor stems from the fact that, with regard to cross-sectional designs, the condition of dependent or independent variables cannot be determined a priori, since the measurements of both variables tend to coincide over time⁶.

In our particular case, the logistical regression analysis enables us to establish a socio-demographic profile for households receiving remittances, though we are unable to define the causal or deterministic nature of this profile. In other words, the results of the logistical regression model may be interpreted in terms of the statistical significance of the socio-demographic disparities between remittance-recipient and non-remittance-recipient households. This approach enables us to observe the socio-demographic categories in which remittance-recipient households differ from non-remittance-recipient households, but we cannot offer an explanatory or causal reason to the status governing the receipt of remittances in each household.

¹ A third point relates to the measurement and quantification of remittances. We have, however, omitted this point for reasons of space and relevance. For further details, please consult Lozano, 1993.

² Further on, we point out certain methodological constraints which must be taken into account in analyzing the results of the multivariate statistical model used in this particular case.

³ The data used in the section are derived from a survey conducted on households in February 2001 in the Teocaltiche municipality located in Los Altos de Jalisco. They correspond to a probability sample of 367 households, that were randomly selected by urban and rural strata. This survey forms part of the project entitled "Estimation of the Economic Impact of remittances on High Emigration Communities in West-Central Mexico", and financed by the CONACyT and the University of Guadalajara.

⁴ Strictly speaking, these are households that are not completely divorced from the social migration process, given that they form part of a migrant community where networks formed by friends and other community relationships are also important. However, the methodological requirements of the tools applied made it necessary to exclude these kinds of social networks.

⁵ The low percentage of these kinds of households does not mean that there are no "retired" migrants in the community. The problem is more one of methodology because, while at the individual level, can talk about "exclusion" from the migration circuit, at the level of household, however, this withdrawal in fact relates to situation whereby the migrant is substituted or relieved by other household members.

⁶ As Joel (1995:85) points out, "interpretation and measurement of the probability that has been estimated makes sense only in longitudinal-type studies, while, in cross-section-type studies, it is more appropriate to employ the concept of proportion or prevalence. This is due to the fact that cross-section-type designs make it possible to estimate isolated events or developments ... that may be attributable to future -or probable- situations".

Previous-Next 
Top

http://www.sela.org
sela@sela.org
   SELA,  Permanent Secretariat
Torre Europa, Fourth floor, Urb. Campo Alegre, Av Francisco de Miranda,
Caracas 1060- Venezuela
Tlf: (58) (212) 955.71.11 Fax: (58) (212) 951.52.92