The Sociodemographic Effects of the Crisis in Mexico

Chapter Six: Differences in Patterns of Domestic Organization: The Middle and Popular Sectors in Urban Mexico Mercedes González de la Rocha

I. Introduction[1]

It is a commonplace to state that the crisis of the 1980s had a severe impact on the welfare of both the working people and the middle-income sectors in Mexico. But systematic studies that analyze changes in family economics and welfare and external factors, such as the economic crisis of the 1990, which affects household structure are rare. Even scarcer are longitudinal analyses of a panel of households over the entire period of the eighties.[2]

Guadalajara has been the site of a number of studies of labor markets, as well as others which have explored the social dynamics of domestic groups and their transformations throughout the decade. With the support of the Ford Foundation, we have studied the labor market and the "social organization of poverty", by which we mean the social mechanisms that permit people to live, survive and reproduce themselves on very low wages, with few fringe benefits and no assistance from the state Our studies focused on the social impact of the crisis and the ways in which people made use of their resources, (labor, time, household organization, and consumption patterns), stretching their pesos to cope with rising prices and fewer job opportunities.

At the beginning of the 1980s, we began a study of the service sector and the concomitant transformation of social structures that accompanies the process of labor market tertiarization. During the economic boom, the growth of the middle income sectors, comprising a large and heterogeneous group of non-manual workers, technicians and professionals, was clearly linked to urban tertiarization and the increase of employment in producer, social and personal services. Some writers (Samaniego, 1990, Escobar and Roberts, 1991) have suggested that the middle income sectors were the principal beneficiaries of economic growth in the industrial and service sectors during the sixties and seventies. In Mexico City, for example, growth in services from 1950 brought about an increase in nonmanual jobs, primarily for technicians and professionals, and secondly for supervisors and administrative personnel (Muñoz 1975, cited in García et al. 1982. See also Garza (this volume) for an analysis of the trends of growth in the service sector. In Guadalajara, nonmanual jobs increased along with service and tertiary employment in general . For example, professional and technical services employed 10.9% of the economically active population in 1976, but by 1987 the percentage had increased to 13.4% (Alba and Roberts, 1990). Throughout the sixties and seventies, the lifestyles of the middle sectors improved due to this real increase in their incomes (Escobar and Roberts 1991, López Cámara, 1988).

The economic crisis may well have set off explosive social and economic changes. In earlier writings I proposed a model of household dynamics and used it to discuss changes among the popular classes of Guadalajara during the crisis (González de la Rocha 1986a, 1986b). In this chapter, I further analyze households belonging to different occupational categories to compare patterns of social organization and domestic economies between the middle and popular urban sectors, with the aim of deepening our understanding of the ways in which different social sectors and occupational groups fit into the urban economy[3].

Mexico's economic restructuring during the eighties was accompanied by a reorganization of the households of the popular classes which buffered the impact of wage reductions and stagnation in formal sector employment (González de la Rocha, 1988; Chant, 1988, 1990 ; Benería, 1992; Selby et al. 1990a, 1990b). Although studies of the middle sectors are much rarer, there are some indications that their domestic economies also changed (cf. the studies compiled by Loaeza and Stern, 1990, particularly the work of Samaniego and Tarrès). Middle-sector incomes suffered considerably and patterns of consumption adjusted accordingly. In middle sector households, the number of individuals in the work force did not rise, as it as did in the households of the popular sectors, but workers took on second and even third jobs. More women in the popular sectors took on the extra burden of the "double day", and more middle-sector workers moonlighted. Examples abound:, a doctor, in addition to her office appointments, works extra hours in a public clinic or hospital; a school teacher has two or three different teaching jobs in different schools with different time tables; in addition to his legal work, an attorney takes over his wife's business in the afternoon, after she has worked all morning. The intensification of labor on the part of the adults in middle-sector households made it possible for the children to stay in school and perpetuated fundamental differences between the middle and popular sectors in the city.

In this article, I propose that the life styles and living conditions of the lower strata of the middle sector, called here the "lower-middle strata" or "lower-middle households" and the popular (or "manual worker") sectors are, economically, very similar. I hypothesize that this similarity is the product of the impoverishment which afflicted lower-middle sector households throughout the eighties[4] and of these households' relative inflexibility in regard to income generation[5] , especially when compared to the relative successful strategies of survival adopted by households of the popular sectors. Popular households were able to put more workers into the work force to offset the drop in individual incomes, but by no means should the term "relatively successful" suggest that the popular classes did not suffer during the crisis . Nor is it meant to gloss over the grave implications that intensification of work holds for their lives.

This analysis takes as its point of departure the existence of structural, cultural and social factors, which are widely documented in the literature on social stratification and social inequality (Muñoz, Oliveira and Stern, 1977; García, Muñoz and Oliveira, 1982; Wright Mills, 1956; Loaeza and Stern, 1990). These factors correlate with differences in job skills and wages. But differences among groups are also found in lifestyles and social and cultural forms. Ever since Wright Mills' study of the middle class in the United States, it has been taken for granted that it is not access to and control over private property that determines income, material welfare and the exercise of control over one's life, but the position that a person has in the occupational structure. Having a certain kind of job implies having both knowledge and training, which are the determinants of the social class position of any given individual (Wright Mills, 1956). In the case of Mexico, and specifically for the middle sectors, Stern (1990:20-21) has pointed out that occupation is the criteria that determines social class position:

"In order to fill a certain job, both knowledge and abilities are required (general education as well as educational attainment); doing different jobs carries with it different levels of compensation (different incomes, wealth). The three elements mentioned (job, education, income) are not only intercorrelated, but they translate into different life style, as the nature of the job and the level of education of income permit (as measured in objective indicators like residential location, house type, level of consumption etc.). Values and attitudes are associated with these different levels, as well. Both kinds of elements are combined at once in order to yield symbolic complexes which give meaning to groups that share them and afford the basis for collective action..."

Specifically, I am suggesting that, while there are usually qualitative or cultural differences between the groups which income statistics gloss over--the important differences in life style between the school teacher ("maestro de educación"), and the mechanic (the "maistro"), for example--some of these differences have been erased by ten years of economic recession, even though the social distinction implied by the terms "maestro" and "maistro" abides. One of my aims in this study is to ascertain what differences remain, as well as what similarities have been caused by the crisis. To establish the hypothesis of difference, I shall try to answer the following question: do the lower level nonmanual workers and the manual workers belong to the same social group, with a similar origin and with similar characteristics? Can we look at the characteristics of the household and at household dynamics (size and structure, income levels and consumption patterns, social and occupational histories and occupational heterogeneity within the household) to see the more subtle ways in which the "maestro" differs from the "maistro", the school teacher from the mechanic?

II. The Restructuring of Popular Sector Households as a Response to the Crisis

The effects of the crisis on domestic units in Guadalajara, as documented in previous research, can be summarized as follows:[6]

1) Working class households increased the number of members in the work force,[7] , thereby intensifying their use of labor power. In the formal sector, employment did not increase but credentialism did. For this reason, new workers went into the informal sector and into own-account work despite the low wages (Escobar, 1988; García and Oliveira, 1990).

2) Adult women, especially female heads of households, increased their work force participation rate the most. Right behind them were young men under the age of 15. Daughters of the house redoubled their efforts in domestic tasks, thereby making it possible for their mothers to leave home in the search for income. Adult males did not figure in the intensification process, since they were already employed in the years before the crisis.

3) The size of households and the absolute number of extended and composite households increased as more units incorporated adult relatives in order to generate more income and to achieve economies of scale in consumption (see Gilbert 1991 for an analysis of the availability of housing during the crisis).

4) Households produced goods and services which had been purchased in the market previously. This meant more work for all members of the domestic group, especially the women who had more clothes to mend, and more lunches to prepare, among other tasks.[8]

These adjustments in the management of household resources enabled people of the popular sectors in Guadalajara to ameliorate the effects of the crisis to a significant degree. The relative success of these strategies can be seen by comparing the drop in household income to the general decrease in individual wages: household income fell by just 11% compared to a 35% decline in wages (González de la Rocha y Escobar, 1988). But even this relative success has to be considered in the light of changes in consumption patterns.

Without the intensification of labor, other changes in domestic organization and in patterns of consumption would have been much more drastic than they were in fact. During the crisis, all households reduced their consumption of expensive foodstuffs, especially beef and poultry, and increased their consumption of cheaper foods. However larger households with more people in the work force, some of them with older children who went out to work, were much more able to keep up previous levels of consumption. Their consumption patterns changed less than those of households with young children and few members in the work force. Female-headed households, with lower incomes (in constant pesos as well as lower per capita incomes), did not necessarily consume less either before or during the crisis. These women heads were able to channel more resources towards high priority areas of household consumption like meals, even though they lost income (González de la Rocha, 1991).

III. The Middle Sectors: A Different Life Style?

Escobar and Roberts (1991) suggested that the middle class of the 1960s and 1970s developed upscale patterns of consumption, including spacious, well-furnished homes, private education, and foreign vacations. Not only were their salaries many times higher than the wages of the manual workers, but they benefited from many indirect subsidies: low wages paid for domestic help; subsidized urban infrastructure; price subsidies for basic consumer goods enjoyed by the whole population; and easy availability of bank credit . The increase in the incomes the middle classes actually decreased the overall inequity of the national income distribution(Cortés and Rubalcava, 1991). But, foreshadowing what would happen during the crisis, during the 1970s, there were some middle-sector households with incomes as low as some of the better-off manual-worker households. For example, a household living off the wages of a secretary had the same income as one supported by a mechanic (Muñoz, Oliveira and Stern, 1977; García, Muñoz and Oliveira, 1982, Escobar and Roberts, 1991).

Everything indicates that the crisis of the 1980s intensified this process. According to the study of income dynamics carried out by the National Institute for the Consumer (INCO), the middle class-formal sector, consisting of shop assistants, white collar workers and secretaries, lost the greatest proportion of income during the 1980s. Within this sector, household incomes decreased by 26.4% from June 1985 to February 1988 and patterns of consumption changed accordingly. In contrast, manual -worker households, thanks to the increase in the number of labor-force participants per household, managed to offset the reduction in individual wages (De Lara Rangel, 1990). Unfortunately, the study carried out by the INCO started when the crisis was well under way. If data from earlier years had been available, more drastic changes in the situation of the middle sector would have been apparent.

One reason that the formal-middle stratum did not put more members into the work force during the crisis was that employment was already high[9], In addition, there were very few new middle-class jobs during the eighties: the modern sectors of the economy did not grow and rising credentialism meant that the job qualifications for middle-class employment were tighter.[10] Placing more household members in the work force would have entailed taking distant relatives in the household and putting them to work, or removing children from school. This happened occasionally, but never became a general strategy. Instead, moonlighting or evading unemployment by taking a lower-status job were the tactics more often favored in the middle sectors.

Another response to the crisis by the middle-class households in the study was to use public health and education facilities instead of private ones. It had been unusual for a middle class household to send its children to a state school (escuela "de gobierno"), since state schools were believed to have low achievement levels and poor teachers as well as students from the lower classes. But in 1992, high numbers of middle class households reported using public health-care clinics and sending their children to state schools, both state and federal. According to the Guadalajara data, 15.2% of the women in professional/managerial households and 41.3% of those in lower-middle households had a gynecological examination in a public facility, and 45% of professional/managerial households and 77% of lower-middle households used a public hospital. (See also Samaniego, 1990:57.)

IV. The Households of Guadalajara: The Current Study

The data discussed below were collected during the first six months of 1991 via a questionnaire administered to 375 households in three different socioeconomic groups in different residential zones of the city[11]. The sample includes 145 manual worker households, 98 lower-middle stratum households, and 125 professional/managerial households. Households were classified according to the occupation of the principal labor-force participant [12].

In the "manual worker " category, we included workers in the manufacturing, industrial or service sectors; domestic workers; roving vendors; police and other security personnel; agricultural workers; and self-employed workers in small firms and other sectors. The "lower -middle " category includes industrial floor supervisors, mid-level technical personnel, teachers, clerical personnel, white collar workers in manufacturing and industry, sales personnel in large establishments, and "artists". Finally, the "professional/managerial "category is comprised of higher-level administrators in both public and the private sectors, professionals, businessmen, and proprietors of medium and large establishments. In the next section we will compare systematically these social groups' characteristics in terms of incomes, patterns of expenditure and changes over the course of the domestic cycle.

Manual worker Households (145 households)

As shown in Table 1, the 145 manual worker households had the largest households of the three groups, averaging 5.1 members. Within these households, there was an average of 2.1 members employed. (This average is markedly lower than the average number employed found for similar households in studies undertaken by the author in 1985 and 1982[13].) Among these households, 37% are in the expansion phase of the domestic cycle; 53%are in the equilibrium or consolidation phase of the domestic cycle; and 10% in the phase of dispersion[14]. As expected, the average number of labor force participants is much higher in households in the "equilibrium" or "consolidation" phase. This distribution of households among the phases of the domestic cycle is similar to that found in earlier studies (González de la Rocha 1986a, 1986b).

The average number of members and number of labor force participants increases as the household matures, until the last stage when the numbers turn down. Household and per capita incomes show the same inverted "U" pattern, increasing until the final stage, despite the fact that the age dependency ratios are highest for households in the middle stage. As shown in Table 1, the average household income in these manual worker households is 33% higher in the consolidation phase than in the expansion phase stage and drops by 51% in the dispersion phase. The oldest households confront difficult economic conditions; if their members' health and strength permit them to work, they are paid less on account of their age. Their children are less able to help them, since they have left the house to found their own families and are themselves in the stage of expansion, when their expenses are high and their incomes lower than they will be later. The aging of a household often implies impoverishment and vulnerability, particularly among the lower -income families

The average income of manual worker-worker households, including pensions, profits from small businesses and rental properties, is 6.7 minimum salaries a month, when the minimum salary is $375,000 pesos (approximately US$150.) a month. The average worker's income is 3.7 minimum salaries and 63% of the household members above 15 years of age are in the work force. In general and per capita terms, these are the poorest households of the sample, and they show a worker dependency ratio that is the highest of the three occupational types. The majority of these households are nuclear (66%), but there are some "incomplete" nuclear formations where a spouse is absent (13%). In this group, 12% of the households are extended and only 2% are composite households. In all, 36% of the female-headed families in the sample are found in this group. The economic vulnerability of these households stems from two causes: low paid jobs for women and the absence of a man who could contribute to the household budget.[15]

In manual worker workers' households, men contribute the greater proportion of the money for maintenance of house and household[16]. This holds true whether one looks at all households or at only those households where the male head contributes, as shown in Table 2. After the male heads, contributors to the household expenses are spouses and then male children. Contributions from household members living outside the household or in the United States are small. Although kinship relations with absent members are important, in normal circumstances the domestic group is notably independent, sustaining itself through the contributions of its coresident members.[17] Other income sources exist, such as the contributions of daughters and earnings from rental properties and pensions, but they are also small. As is usual, daughters in the homes of the popular sectors in Mexico perform domestic chores and care for younger brothers and sisters. The daughter's role was especially important role during the economic crisis, since it allowed her mother to take paid employment outside the household.

Lower-middle Status Households (98 households)

With 4.8 members per household on average, and a dependency ratio of 2.7 consumers per labor-force participant, the households are smaller than the manual workers' but larger than those in the professional-business group. Household incomes are about the same as the manual worker households (an average of 6.9 minimum salaries per month compared to 6.7 for the latter group), as is income per capita, at 1.64 minimum salaries for lower-middle households compared to 1.55 minimum salaries for between the manual worker households. Although this lower-middle group has a slightly larger number of workers per household than the other groups, there are fewer households in the consolidation phase of the domestic cycle than might be expected. In this group, 49% of the households are in the phase of expansion and only 8% in the dispersion phase. Compared to manual worker households, there are more workers in the households in stages of equilibrium and dispersion, but in both groups the number of workers rises substantially as the households pass from the stage of expansion to that of consolidation or equilibrium, and falls slightly as it passes into the stage of dispersion.

Both manual worker and lower-middle households have the same age-earnings profile during the first two phases of the domestic cycle, but the fall-off in earnings during the last phase is not as severe in lower-middle status households as it is in the manual worker households. Perhaps that older lower-middle status households were able to acquire income-generating resources, such as rental properties, during the period of the oil boom when the middle classes increased in number in Mexico. In addition, some members of the lower-middle group have worked long enough to have fairly good pensions. Both these tendencies would explain why lower-middle household incomes do not fall off as dramatically as those of manual worker households.

In this lower-middle group, there is a high percentage of both extended households and female-headed households, which may also contain other adults. Compared to manual0woriker households, lower-middle households have essentially the same number of workers and the same average wage per worker, and just slightly higher household incomes. They also have slightly higher per-capita incomes (because the average number of persons per household is smaller).

Family structure in the lower-middle status group is very similar to the manual worker group: the majority of the households are nuclear with children present (57%), while the percentage of "incomplete" households and extended household is quite high at 17% each. This group shows a small fraction (2%) of composite households but female-headed households are very frequent, accounting for 41% of the total[18]

The proportion the male head of household contributes to the household budget is smaller in the lower-middle status group than it is in the manual worker group, 52% versus 70%. As a corollary, the contributions of wives and daughters are more important here than in the manual worker workers' households, as are the contributions of relatives living outside the household or in the US, although these are still relatively small. What we term "other contribution," mainly from pensions and from property rentals, is notably more important in the lower-middle status group than in the manual worker group. It seems clear that the smaller contribution of the male heads of household is subsidized by that of wives and daughters, as well as by income from "other sources". [19] In contrast to the manual worker households, daughters in lower-status nonmanuals participate at higher rates in the work force instead of devoting themselves solely to domestic duties.

Professional/Managerial Households (125 households)

The households in this group have an average size of 4.7 members, smallest of the three groups. Since, they have an average of 2.08 members in the work force, they have the lowest dependency ratio of the sample in terms of both consumers per worker (2.6) workers per household member. As noted previously, in professional/managerial households children tend to stay longer in school and to postpone marriage, and are therefore retained longer in the household. In addition, the relatively older childbearing age of the parents helps to explain why the number of members in the work force remains consistently high throughout the phases of consolidation and equilibrium, even into the phase of dispersion[20] .

The ability of the higher-status households to pay for domestic help or private child care also explains why the ratio of workers to adults in the house is so high in this group during the expansion phase, when child care duties press hard on the average Mexican urban household and compel the mother of the children to stay home.[21] Individual incomes in professional/managerial households average 8.5 minimum salaries (7.5 minimum salaries during the expansion stage), while the average household incomes is 12.3 minimum salaries in the expansion phase and rises and climbs to 17.2 minimum salaries during the consolidation phase. As mentioned earlier, the inverted "U" age-earnings profile is found in all groups, with the highest incomes during consolidation phase and the lowest during the phase of dispersion. However, it is striking that even the lowest incomes of the professional/managerial group are much higher than the highest incomes of the manual worker and lower-middle groups There is a relatively high percentage of extended households in the professional/managerial group, the majority headed by adults. Household composition helps to explain the very high ratio of workers to adults but not the great discrepancy in income between this group and the two others. Instead, the vast difference in average household income reflects the difference in per-capita income, which is more than 100% higher for professional/managerial workers than for workers in the manual and lower-middle groups.

Turning now to the social structure of these professional/managerial households, again we find that the majority (64.8%) are nuclear "complete", with children present. While the percentage of extended households is very high (at 18.4%), there are few incomplete households (5.6%), and no composite households at all. In contrast to the other occupational groups, there are relatively few female-headed households (8.5%), perhaps because the relatively better economic conditions of this group inhibit marital conflict . The problems often associated with female headedness are not characteristic of this group: higher-status women have the kind of educational credentials and training which affords them access to better jobs.

As in the other groups, the male heads of household are the primary contributors to the household budget, contributing on the average a higher proportion of the budget than males heads of household in the manual worker group and a much higher proportion than the lower-middle male heads of household. As shown in Table 2, when all households are considered, professional/managerial wives' contributions to the household budget are proportionately higher than those of the manual worker group but slightly lower than those of wives in the lower-middle status group. But in households where women actually make a contribution, their earnings account for 50% of the household budget in the professional/managerial group, 43% in the lower-middle status group, and 31% in the manual worker group.

V. Household Expenditures by Household Groups.

We have noted the similarity of the age-earnings profiles for manual worker and lower-middle households. The analysis of these groups' budgets and consumption patterns also shows striking similarities between these two groups, both of which are outspent doubly, and sometimes triply, by professional/managerial households. [22] (See Table 3.)

The manual worker and lower-middle groups spend about the same amount in all categories, with the exception of education. It is possible that they could each be spending the same amount on food, but purchasing very different things. However, studies by the National Consumer Institute (INCO) show that, during the crisis, the diets of the middle-sector families did indeed change and became similar to those of the lower-income strata. According to the INCO study, the food budget of the "middle -sector formal employment" group fell by some 30%, thus severely inhibiting purchases of high protein and high quality foods (Jusidman, 1987)[23].

As expected, the higher the status of the household, the lower the proportion of income spent on basics such as food. If we look at household expense categories by stage of the domestic cycle, we generally find that the greatest expenses are incurred during the phase of consolidation, when children require the greatest expenditures and the household has its maximum number of members in the work force and enjoys its highest income. An exception to this trend is the level of savings in lower-middle households ; in the last phase of the domestic cycle, they save more than in the earlier stages. This may be because they invested enough during the years of relative plenty to enable themselves to continue saving in the 1980s.

VI. Occupational Heterogeneity within Groups, and Educational Attainment of Household Members.

So far, the data show that the three occupational groups in the study are really only two: the manual worker group and the lower-middle status group are almost identical when they are described in terms of age-earnings profiles and the composition of the household budget: I argue that this homogeneity is a product of the effect of the economic crisis on the lower-middle status group in particular. If the two groups do retain social distinctions attributed to them on the basis of differing occupations (manual versus nonmanual work) their distinctness may be revealed through an analysis of noneconomic variables such as education and the degree of occupational heterogeneity within households.

We find that the educational attainment of fathers of the current household heads in manual worker and lower-middle households is remarkably similar. In manual worker households, 74% of the fathers had manual worker occupations; 33% of the individuals have no schooling; 18% did not complete primary school; 33%completed primary school; and just 18% had education beyond the primary level.

Similarly, the majority of the fathers of the household heads of the lower-middle status group (55%) had manual worker occupations, including that of farmer. However, some fathers of the current heads of household in the lower-middle group did have higher overall levels of educational attainment: only 13.4% had never attended school; 22.3% went to primary school; and 42% completed their primary education (compared to 32% of the manual worker fathers). In all, 22.3% of these individuals went beyond primary education.

Interestingly, 54.5% of the fathers of current heads of household in the professional/managerial group had also been manual workers, suggesting that father's occupation many not predict differences between the lower-middle and the manual worker groups. However, when we look at the educational attainment of the current heads of household, we do find significant differences between the professional/managerial group and the other two groups. Only 10.4% of professional/managerial households are headed by people who never went to school and just 15% reported that they only finished primary school , while 30.2% finished secondary school; 15% have some technical school equivalent; and 30% went on to preparatory school, university or postgraduate training. Clearly, the educational attainment of the current household head, not of his or her parents, predicts occupational category and current status.

The percentage of households heads with no formal education declines from 5.3% in the manual worker group to 1.5% in the lower-middle status group to zero in the professional/managerial group. Similarly, while 14.2% of the male heads of household in the manual worker group have some (but not complete) primary education, this proportions falls to 6.2% in the lower-middle status group and to 0.9% in the professional/managerial group. On the other hand, the percentage of male heads whose educational attainment is minimally secondary school (including technical school, preparatory school, and university) is a little more than 50% in the manual worker group, almost 70% in the lower-middle group and almost 95% in the professional/managerial group, which contains a large number of professionals with university degrees and postgraduate training. Although heads of household in the lower-middle status group have a higher level of education than those in the manual worker group in general, the number of heads in the manual worker group with higher education is surprising: 12% have university education (14 individuals in total, 9 who completed university degrees. These may be those households that experienced downward mobility from nonmanual occupations to manual worker ones.[24]

Differences in educational attainment among women among the three occupational groups were even more marked than those among the male heads of household: the percentage of female heads failing to complete primary schools was 16.8% in the manual worker group, 6.4% in the lower-middle group, and 2.7% in the professional/managerial group. The percentage of women with primary schooling only was 30.5% in the manual worker group, 22% in the lower-middle group, and 11% in the professional/managerial group. In the manual worker group, only 45% of the female heads of households have secondary school or higher education, compared to 70% in the lower-middle group and 86% in the professional/managerial group.

The educational differences between groups found among adults also hold true for children. In professional/managerial group,92.5% of the children completed secondary school or have higher credentials, compared to 77.5% in the lower-middle status group and 74.5% in the manual worker group.

Occupational Heterogeneity within the Households of the Different Occupational Groups

The analysis of the mixture of occupational types within the household was done by comparing the job of the second, third, fourth and fifth worker in the household with that of the first worker, whose occupation was used to define the household's status. Beginning with the second worker, the data indicate that 10 out of 13 second workers in the professional/managerial households have jobs in that sector. Second workers in other groups hold a wider range jobs. Workers with jobs classified as lower-middle status, including teachers, technical personnel and officer workers, are likely to live in professional/managerial households as they are in lower-middle or manual worker households. This is not unexpected, since women in all types of households have access to these jobs, and the second workers tend to be wives of the head of the household. However, in manual-worker households, second workers tend to have manual worker jobs, although there are some manually employed second workers to be found in the households of the other occupational groups.[25] Households where all the workers are of the same type tend to be either manual worker or professional/managerial, that is, at the extremes of the occupational typology, even though manual worker households show the highest degree of heterogeneity.

For third workers, the picture is much the same: occupations of a lower-middle status type are uniformly distributed among the occupational groups, while manual worker third workers are to be found only in manual -worker households. In lower-middle households, the majority of third workers are employed as shop assistants or office personnel. With respect to the fourth and fifth workers, we find increased occupational heterogeneity in the professional/managerial group, where young workers are engaged as shop assistants or manual worker workers to earn a little money to help with their school expenses.

Since women in all three groups tend to hold lower-middle types of jobs, women's occupations do not serve to differentiate among households. Thus, it is the sex of the household head, not only his or her occupation, that determines a household's place in the social structure.

VIII. Perceptions of the Economic Crisis.

An examination of the group members' attitudes regarding the crisis is also helpful in revealing the differential impact the crisis had on these three occupational groups. Household heads' perceptions of the effects of the crisis on their household economies and social structures are, by definition, subjective, but their statements do reflect the objective reality of concrete changes in their households. This analysis of subjective perceptions reveals differences between the lower-middle and manual-worker groups that were not shown by the comparison of their economic circumstances.

The first question was, "What was the effect of the crisis on your household?" Among the households interviewed, 316 out of 375 reported that the decrease in real wages was the most important effect. This response was given by 77% of the households in the professional/managerial group, 86% of the lower-middle status group and 90% of the manual worker group.[26] In addition to the drop in wages, 20% of the professional/managerial households, 11% of the lower-status nonmanuals, and 9% of the manual worker group reported that some household members had lost jobs.[27] Escobar (1992) discovered a similar trend in the Guadalajaran [28]labor market, where he found that professional workers had changed jobs much more frequently in recent years than other kinds of job holders.

The second question asked what the householders did to alleviate the impact of the crisis. In the lower-middle group, the response given by 55% of the households was, "We hunkered down, "[29] followed by the entry of one of more children into the work force (13.2%), the entry of the wife into the work force (5%) , and the male head increasing his contribution to the budget (4%). In the manual-worker group, 54% reported they had "hunkered down", 8.9% of the wives had entered the work force, and 13% of the households put children to work. In 8.9% of the cases the husband had increased his contribution to the household budget. Among professional/managerial households, 40% reported "hunkering down"; 9% said that the woman of the house entered the work force; 8% put children into the work force; and in 4%, the male head took a second job.

These data clearly show that lower-middle households responded differently to the crisis than the other two groups, perhaps because they had fewer options: the husband could not or would not increase his contribution to the household; there were no children of the age to enter the work force; and there were no new lower-middle types of jobs for the women, as we stated earlier. In the other two groups, the frequency with which wives entered the work force was considerably higher.

Data on the kinds of budget cutbacks occurring during the crisis also help to define "hunkering down" for the three occupational groups. ( See Table 4) Among the professional/managerial households, 53% reported that they did not have to make cutbacks in their standard of living. However, 80% of lower-middle households and approximately 70% of the manual labor households reported the necessity of making economies. Lower-middle households, from which few wives entered the work force, were compelled to make cuts with the greatest frequency, and in this sense were the worst off of the three groups. Consistently we find fewer budget cuts in households which put more women into the work force. Among households reporting a budget cutback, 16% of the professional/managerial households cut back their food purchases, while 30% of the lower-middle and 27% of the manual worker households did. The percentages of households making cutbacks in various areas can be seen in the Table 1.

The fact that the proportion of female heads of households entering the work force recently is highest in the professional/managerial group and the manual worker group is consistent with the frequency of the response, "hunkering down." In fact, many of these women had entered the work force during the eighties, as my earlier research in Guadalajara (González de la Rocha, 1988, 1989) and research carried out on the national level by Oliveira and García (1990) have indicated.

X. Conclusion.

This analysis offers evidence that reveals a high degree of economic homogeneity between the lower-middle status group and the manual worker group, as well as striking differences between these two group sand the professional/managerial group. The similarities between the lower-middle and manual worker groups are rooted in the amount of their incomes, the structure of their expenditures, and the percentage of their incomes devoted to different categories of expenditure. This is not to suggest that there are no differences at all between them; the analysis of what they actually consume, as compared to what they expend in different categories will be reported in future works. However, the current analysis suggests a high correlation between the amount spent and what category it is spent on.

Another conclusion that can be drawn from the study is the remarkable similarity of responses from households of all occupational groups. All the households follow the same curve of the domestic cycle in almost all the aspects that we analyzed. It is striking, for example, the degree to which households income declines during the last phase of the domestic cycle in all three groups. Clearly, the domestic cycle effects changes in the way that social rules and practices are instituted, the number of workers, and the rate of incomes and expenditures.

There are some interesting differences among the different occupational groups. For example, during the final phase of the domestic cycle, lower-middle households are much less badly off than manual worker households. It is likely that lower-middle workers are paid higher wages as they get older, at least in comparison with the manual labor workers. It is also possible that lower-middle households benefit from past investments such as rents and interest on bank accounts. Another interesting aspect of the domestic cycle is that the households of the professional/managerial group tend to stay in any given phase longer than the other occupational groups, in large part because the women of these households bear their children at a later age, and keep the children longer in school.

The analysis of the occupations of current heads of households' fathers showed no effect on the occupational group membership of the current group, since the overwhelming majority of the parents of current household heads held manual jobs. However the educational attainment level of the parents has a very strong effect on the overall status of the household: educational attainment level was "low", "intermediate", or "high" in respect to the current status of the households.

The analysis indicates that the economic crisis of the eighties erased differences between the lower-middle and manual-labor group in the two groups in three ways. First, it brought the manual labor group and the lower-middle status group closer in household incomes. Secondly, both groups changed their consumption patterns. Thirdly, as part of new "adjustment," the middle sectors increased their use of state schools and health services, despite unwelcome effects on their self-esteem. The fact that the lower-middle status groups were the ones who most had to "hunker down " indicates that their new condition of poverty obligated them to make budget cuts, especially since they had a little slack in the budget in the first place. They opted to make budget cuts so their children would not have to enter the work force, but made cuts in different areas than the other groups. For example, they reduced clothing expenditures only by 5%, preserving a middle-class appearance as much for themselves as others, and keeping children in school for the same reason.

While the crisis brought the lower-middle status and the manual-worker group closer together in economic terms, it took them father away from the professional/managerial types. The relatively prosperous past of the middle sectors, during which they benefited from the development model based on import substitution and on the creation of a broad internal market (in which the middle sectors played an important role), is the opposite of the present. Right now, the economic condition of the lower strata of the middle sectors classes them together with the masses of low-salaried workers. This abundance of low paid labor will be key to the nature of Mexico's participation in an open international market.

BIBLIOGRAPHY

Alba, Carlos and Bryan Roberts

1990 "Crisis, Adjustment and Employment in the Manufacturing Industry of Jalisco." Working paper no. 24, Commission for the Study of International Migration and Comparative Economic Development. March.

Benería, Lourdes

1992 "The Mexican Debt Crisis: Restructuring the Economy and the Household," in Benería, Lourdes and Shelley Feldman (eds.) Unequal Burden: Economic Crises, Persistent Poverty and Women's Work. Boulder: Westview.

Chant, Sylvia

1991 Women and Survival in Mexican Cities. Perspectives on Gender, Labor Markets and Low Income Households. Manchester: Manchester University Press.

1988 "Women, Households and the Urban Labour Markets in Mexico. Paper presented in London Third World Economic History Group Workshop on Household Income and Resource Management. School of Oriental and African Studies. January.

Cortés, Fernando and Rosa Maria Rubalcava
1991 Autoexplotacion forzada y equidad por empobrecimiento. Mexico: El Colegio de México, Jornadas 120.

de Lara, Rangel, S.

1990 "El impacto socioeconómico de la crisis sobre la clase media" in Loaeza, Soldead and Claudio Stern Las clases medias en la coyuntura actual. Mexico: Cuadernos del CES. La Jolla. Center for US-Mexican Studies.

Escobar, Agustin and Bryan Roberts

1991 "Urban Stratification, the Middle Classes and Economic Change in Mexico" in González de la Rocha, Mercedes and Agustín Escobar (eds.) Social Responses to Mexico's Economic Crisis of the 1980s. La Jolla. Center for US-Mexico Studies.

Escobar, Agustín

1992 "Occupational Mobility in Guadalajara: 1975-1982-1990, or the Myth of the Golden Age of Self-Employment". Paper presented at the seminar "The Sociodemographic Effects of the Crisis in Mexico". Austin, Texas, University of Texas at Austin.

1988 "The Rise and Fall of an Urban Labor Market" Bulletin of Latin American Research. Vol. 7, no. 2.

García, Brígida, Humberto Muñoz and Orlandina de Oliveira

1982 Hogares y Trabajadores. Instituto de Estudios Sociales. UNAM/Colegio de México.

García, Brígida and Orlandina de Oliveira

1990 "Cambios en los determinantes del trabajo femenino". Mexico: El Colegio de México. Mimeo.

Garza, Gustavo

1992 "Crisis del sector servicios de la ciudad de México." Paper presented at the seminar "The Sociodemographic Effects of the Crisis in Mexico". Austin, Texas, University of Texas at Austin.

Gilbert, Alan

1991 "Self-help Housing during Recession: the Mexican Experience". González de la Rocha, Mercedes and Agustín Escobar (eds.) Social Responses to Mexico's Economic Crisis of the 1980s. La Jolla. Center for US-Mexico Studies.

González de la Rocha, Mercedes

1993 "Women-headed households and occupational position: Notes towards a Better Understanding of the Gender and Class Differences in Households in an Urban Context". Paper prepared for the seminar on Well-being and Gender. Institute of the Americas. La Jolla.

1991 "Family Well-Being, Food Consumption and Survival Strategies during Mexico's Economic Crisis" in González de la Rocha, Mercedes and Agustín Escobar (eds.) Social Responses to Mexico's Economic Crisis of the 1980s. La Jolla. Center for US-Mexico Studies.

1989 "Crisis, económica doméstica y trabajo femenino" in Oliveira, Orlandina (ed.) Trabajo, poder y sexualidad. México. Programa Interdisciplinaria de Estudios de la Mujer. Colegio de México.

1988 "Economic Crisis, Domestic Reorganization and Women's Work in Guadalajara, Mexico". Bulletin of Latin American Research 7:2:207-223.

1986a Los recursos de la pobreza: Familias de bajos ingresos in Guadalajara. Guadalajara: El Colegio de Jalisco/CIESAS/SEP.
1986b "Lo público y lo privado: el grupo doméstico frente al mercado de trabajo urbano" in de la Peña, Guillermo and Agustín Escobár (eds.) Cambio regional, mercado de trabajo y vida obrera en Jalisco. Guadalajara: Colegio de Jalisco.

González de la Rocha, Mercedes y Agustín Escobar

1991 Social Responses to Mexico's Economic Crisis of the 1980s. La Jolla: Center for US-Mexican Studies.

1988 "Crisis and Adaptation: Households of Guadalajara". Texas papers on Mexico 88-04 Austin: The Mexican Center of the University of Texas.

González de la Rocha, Mercedes et al.

1990 "Estrategias vs. conflicto: reflexiones para el estudio del grupo doméstico en época de crisis" in de la Peña, Guillermo and Agustín Escobar (eds.) Cambio regional, mercado de trabajo y vida obrera en Jalisco. Guadalajara: Colegio de Jalisco.

Jusidman, Clara

1987 "Evolución del consumo en el medio urbano". Paper presented in the Seminario Científico de la IIa Asamblea general de Asociados de la Fundación Mexicana para la Salud. Mexico, D.F.

Loaeza, Soldedad and Claudio Stern

1990 Las clases medias en la coyuntura actual. México. Cuadernos del CES. El Colegio de México.

López Cámara, Francisco

1988 Las clases medias en la era del populismo. Mexico: CRIM, Coordinación de Humanidades. UNAM y Miguel Angel Porrúa.

Muñoz, Humberto

1975 "Occupational and Earning Inequalities in Mexico City; A Sectoral Analysis of the Labor Force" Unpublished Ph.D.

dissertation University of Texas (Austin).

Muñoz, Humberto, Orlandina de Oliveira and Claudio Stern

1975 Migración y desigualdad en México.

Murphy, Arthur (ed.)

1991 City in Crisis; The Case of Oaxaca, Mexico. Urban Anthropology 20:1.

Murphy, Arthur and Alex Stepick

1991 Social Inequality in Oaxaca: a History of Resistance and Change. Philadelphia. Temple University Press.

Oliveira, Orlandina and Brígida García

1990 "Recesión económica y cambio en los determinantes de trabajo femenino". Colegio de México. Mimeo.

Samaniego, Norma

1990 "Algunas reflexiones sobre el impacto económico de la crisis en las clases medias" in Loaeza, Soldedad and Claudio Stern Las clases medias en la coyuntura actual. México. Cuadernos del CES. El Colegio de México.

Selby, Henry et al.

1990a "La familia urbana mexicana frente a la crisis in de la Peña, Guillermo and Agustín Escobar (eds.) Cambio regional, mercado de trabajo y vida obrera en Jalisco. Guadalajara: Colegio de Jalisco.

1990b The Mexican Urban Household: Organizing for Self-Defense. Austin: University of Texas Press.

Stern, Claudio

1990 "Notas para la delimitación de las clases medias en México" in Loaeza, Soldedad and Claudio Stern Las clases medias en la coyuntura actual. México. Cuadernos del CES. El Colegio de México.

Tarrès, María Luisa

1990 "La oposición política y la idea de democracia entre las clases medias en la coyuntura actual" in Loaeza, Soldedad and Claudio Stern Las clases medias en la coyuntura actual. México. Cuadernos del CES. El Colegio de México.

Wright Mills, C.

1956 White Collar: The American Middle Classes. New York. Galaxy.

Wright, Erick Olin

1985 Classes. Great Britain: Verso.

1989 "The comparative project on Class Structure and Class Consciousness: An Overview". Acta Sociológica. Vol. 32.

Table 1

Characteristics of Guadalajaran Households by Occupation

and by Stage of Domestic Cycle

Occupational Group

Characteristic Manual worker Lower-middle Prof./Mgr

N=145 N=98 N=125

Household Size 5.11 4.79 4.66

Expansion Phase 5.19 4.42 4.19

Equilibrium Phase 5.4 5.31 5.3

Dispersion Phase 3.8 3.62 4.7

Members in Work Force 2.08 2.12 2.08

Expansion Phase 1.64 1.61 1.78.

Equilibrium Phase 2.42 2.56 2.28

Dispersion Phase 2.0 .12 2.4

Dependency Ratio[30] 3.06 2.69 2.63

Expansion Phase 3.7 3.23 2.64

Equilibrium Phase 2.89 2.35 2.85

Dispersion Phase 2.02 1.85 2.19

Adult Dependency Ratio[31] 63.3% 63% 63.7%

Expansion Phase 68.2% 67.3% 76.5%

Equilibrium Phase 58.9% 59.5% 49.7%

Dispersion Phase 67.9% 61.9% 55.5%

MonthlyHousehold Income[32] 2540.93 2606.9 5336.57

Expansion Phase 2069.43 2266.88 4362.64

Equilibrium Phase 3099.66 2996.41 6463.60

Dispersion Phase 1516.00 2055.00 3792.0

Monthly Per Capita Income[33] 582. 615 1278

Expansion Phase 509 577 1295

Equilibrium Phase 646 651 1338

Dispersion Phase 519 708 772

Table 3
Monthly Expenditures By Category and by Phase of Domestic Cycle

(shown in thousands of pesos)

Occupational Group

Characteristic Manual worker Lower-middle Prof./Mgr

N=145 N=98 N=125

Housing 297 371 562

Expansion Phase 239 400 523

Equilibrium Phase 339 353 523

Dispersion Phase 303 320 249

Food 721 756 1021

Expansion Phase 611 694 895

Equilibrium Phase 828 827 1221

Dispersion Phase 598 657 840

Food per Worker 20 27 42

Expansion Phase 19 20 37

Equilibrium Phase 25 37 54

Dispersion Phase 0 2 8

Restaurants 39 38 98

Expansion phase 43 16 119

Equilibrium phase 44 61 75

Dispersion phase 7 9 35

Transportation 128 119 212

Expansion phase 106 99 184

Equilibrium phase 154 144 261

Dispersion phase 80 75 159

Clothing 119 123 193

Expansion phase 66 101 215

Equilibrium phase 173 152 204

Dispersion phase 31 66 64

Footwear 83 74 118

Expansion phase 56 69 133

Equilibrium phase 113 82 120

Dispersion phase 27 47 58

Entertainment 93 92 202

Expansion phase 58 80 162

Equilibrium phase 128 98 261

Dispersion phase 46 120 162

Medical 51 45 113

Expansion phase 46 37 85

Equilibrium phase 59 47 149

Dispersion phase 39 81 92

Education

Expansion phase 46 81 210

Equilibrium phase 138 116 540

Dispersion phase 27 19 244

Savings

Expansion phase 161 151 616

Equilibrium phase 242 217 924

Dispersion phase 70 306 328

Expenditures per capita 469 537 1029

Expansion phase 449 567 1175

Equilibrium phase 495 503 973

Dispersion phase 402 582 500

Table 2

Contributions of Household Members to Budget

(shown in thousands of pesos)

Occupational Group

Potential Contributor to Manual worker Lower-middle Prof/Mgr

Household Budget N=145 N=98 N=125

Husband[34]

All households 1146.20 653.02 1322.82

% of budget 68% 51.6% 68.1%

Husband contributing 1396.63 901.35 1589.93

% of budget 79% 69.6% 78.2%

Wife

All households 261.64 230.38 397.04

% of total budget 15.5% 18.25 16.8%

Wife contributing 758.76 664.05 1341.35

% of total budget 30.6% 42,5% 51.8%

Daughters

All households 77.71 156.02 87.60

% of total budget 4.6% 12.4% 4.5%

Daughters contributing 388.59 477.81 730.00

% of total budget 20.5% 32.1% 41.1%

Sons

All households 121.47 78.67 103.84

% of total budget 7.2% 6.2% 5.3%

Sons contributing 476.03 350.451 590.04

% total budget 30.7% 28.5% 29.8%

Relative living outside the household

All households 15.37 19.48 0.80

% of total budget 0.9% 1.5% 0.04%

Relative contributing 287.75 382.00 100.00

% of total budget 21.3% 22.7% 8.3%

Relatives in the United States

All households 1.79 6063 0.00

% of total budget 0.1% 0.5% 0.0

Relative contributing 130.00 325.00 0.0

% of total budget 14.9% 23.1% 0.0%

Other Sources

All Households 61.06 118.93 29.56

% of total budget 3.6% 9.4% 1.5%

Others contributing 553.43 728.50 461.87

% of total budget 10.8% 50.6% 32.6%

Table 4

Household Budget Cuts by Category of Expense

Occupational Group

Type of Cut Manual worker Lower-middle Prof./Mgr

N=145 N=98 N=125

Groceries 27% 30% 16%

Clothing/Footwear 6% 5% 13%

Health 1% 0% 7%

Entertainment 37% 29% 30%

Domestic Help 25% 29% 30%

Other 4% 7% 4%

Total 100% 100% 100%

[1] This article is based on information which was collected during the research project "El sector terciario en Guadalajara: fuerza de trabajo y bienestar urbano" which I am doing in collaboration with Agustín Escobar and Bryan Roberts. It has been funded by the Ford Foundation (Mexico) and the Center for Advanced Studies in Social Anthropology (CIESAS). The analysis and data interpretation was carried out at the University of Texas at Austin, where I was a Visiting Scholar during the academic year 1991-1992. The financial support of the Fulbright Commission and the Center for Advanced Studies in Social Anthropology made this sabbatical leave possible. I had the good fortune to have the useful commentary and the support of Bryan Roberts, Henry Selby, Harley Browning, Agustín Escobar and Orlandina de Oliveira. Alana Gómez was absolutely key in the collection and coding of the data. Selena Solís, Guadalupe Serna and Wendy Eisenstein did a beautiful job of helping to mount the data in Austin. My thanks to all of them.

[2] The work of Selby, Murphy and Stepick in Oaxaca, that of Sylvia Chant en Querétaro , as well as the studies of the Instituto Nacional del Consumidor in Mexico City, and the work of the author in Guadalajara are clear exceptions.

[3] The debate about the class membership of workers in nonmanual occupations is still being carried on in the sociological literature, and there is no agreement as to whether one should refer to them as "middle class" or "middle sectors". Part of this debate revolves around the question as to whether nonmanual workers are members of the traditional class structure of capitalist societies or whether they are actually in a position that contradicts the traditional description of the class structure (Olin Wright, 1985, 1989). I do not wish to enter fully into this theoretical debate, but rather add to our sparse store of knowledge regarding the organization and lifestyle of nonmanual workers' domestic groups, in comparison with the households of manual workers.

[4] Cortés and Rubalcava have called this process "equity through poverty" ("equidad por empobrecimiento").

[5] In contrast to popular sector households, where the children quickly take jobs requiring little formal education, the middle sectors have not given up on education. Middle sector parents will not lightly give up on their children's prospects in order to increase household income. A definitional point: the degree of "flexibility" in the household refers to its relative elasticity of labor supply.

[6] See González de la Rocha 1986, 1988 and 1990, González de la Rocha and Escobar 1988, González de la Rocha et al. 1990, and González de la Rocha 1991 for greater detail. The studies of Chant (1988), Murphy and Stepick (1991), Murphy (19921), Selby et al. (1990a and 1990b) and Benería (1992) display results that coincide with those of the author in Guadalajara.

[7] The average number of labor force participants per household increased from 2.3 in 1982 to 2.69 in 1985, and then declined slightly to 2.59 in 1987. This last can be interpreted as a response to the slowing in the rate of wage reductions in those years.

[8] Men also intensified their work in the household, by carrying out repairs in plumbing, electrical work and carpentry. But, because it has always been difficult for a poor family to buy a home, men's work in house construction was common practice among the popular sectors in the city even before the crisis. Therefore, despite some intensification of labor by males, I contend that it was the women who had to triple their work.

[9] According to the information gathered by the Survey of Household Incomes and Expenditures in 1977, the upper-middle and upper-income strata had monthly household incomes, respectively, of from 4.3 and 8.1 minimum salaries. These groups had an average of two work-force participants per household, the highest number of employed persons per household (cf. Samaniego, 1990). As noted earlier, the rate of female participation in the work force increased significantly during the seventies(García and Oliveira, 1990).

[10] García and Oliveira (1992:10) indicates that there was a significant increase in the number of women in employment as professionals, technical personnel and office workers from 1976 to 1986. However, the same authors note that from 1982 to 1987 the relative participation of middle-level workers in the work force dropped for all the job categories previously mentioned, even when the level of education of the workers was taken into account.

[11] Originally we had planned to make use of the sample frame of the ENEU (The National Survey of Urban Employment), which is done quarterly by the National Statistical Institute (INEGI). However the interviewers encountered repeated refusals: respondents were tired of being interviewed by INEGI as well as by Sabritas, Coca Cola and Teléfonos de México. Therefore, we adopted a network sampling strategy which sacrifices representativeness for the sake of reliability in the data collected. We took particular care to ensure that the sample included the full range of occupations, residential areas (where quotas were used), and households in every stage of the domestic cycle.

[12] In the case that there was an adult male in the household who was earning money, that man was considered the principal labor force participant even if he was self-employed . In female-headed households where the head of household was in the work force, her occupation was used. When female heads were not in the work force, the occupation of the eldest child was used or that of the person designated by the head as the principal participant. When a male head was retired, currently unemployed, or did not state his occupation, his wife's occupation was used, or, if she was not a labor-force participant, that of their child. There were seven cases in which the male head was retired and there was no other labor-force participants in the household: these households were dropped from the analysis.

[13] However, it should be noted that the 1982, 1985 and 1987 samples were the same, and 1991 was different.

[14] The concept of domestic cycle is defined here the same way it was in previous studies of the popular sectors in Guadalajara by the author. The first phase , "expansion", is defined as continuing until the senior female in the household reaches the age of 39. The second phase, "equilibrium" or "consolidation", ends when the senior female reaches the age of 60. From the senior female's age of 61 on, the household is in the "dispersion" stage.

[15] The analysis of female headed households will be undertaken in González de la Rocha (1993).

[16] Money for the maintenance of house and family is "gasto" in Spanish. The "gasto" is the sum of all the contributions of the different members given over to the housekeeper ("ama de casa") for her to administer. Normally the "gasto" is for the purchase of food, cleaning products for the home and articles of personal hygiene.

[17] In future studies the kinship relations of households in the different occupational categories will be analyzed. But we can anticipate one result which is endorsed by the case studies and by the survey information, which supports the idea that the household is not isolated and without access to social networks of support. This is true for the households of the manual workers, and for the households of nonmanual workers, both low and high-status. The household's expenditures are the coresident household's business, but in times of extreme emergency such as family crisis, unemployment, and children's sickness, family members can count on the support of their extended networks.

[18] I should make clear that female headed households fall into two distinct forms: incomplete nuclear households made up of a mother and her children, and extended female-headed households where there are other members, relatives or nonrelatives in addition to the mother and her children. The percentage of female headed households in the different occupational groups is considered apart from the "incomplete" households, or the extended households.

[19] Ownership of rental property by middle-class the families is one of the more important ways of reducing the impact of the crisis ,according to Tarrés (op. cit.). The property yields crucial income to the household budget and also represents an important investment.

[20] The analysis of mothers' ages at the birth of their first child for all the occupational groups shows that the age of women at the fist of their first child averaged 26 years for managerial and professional mothers ,compared to 24 years for the lower-middle group, and 22 years for the manual labor group.

[21] Many employed female heads of households in this occupational group have domestic servants (nanas) who are hired to take care of the children. Women in this group also have more frequent access to child-care centers than those in other groups.

[22] In this chapter I am considering only the composition of the budget; the analysis of consumption will be carried out in subsequent reports.

[23] This study compared two income levels (low and high) and two sectors of employment of household head (informal and formal). The low income formal sector group had reduced its expenditure for food 23% over the period of the study, while the low income informals reduced their expenditures for food between 12% and 16%.

[24] The analyses of Escobar (1992) show that during the eighties there were important changes in the patterns of occupational mobility. One such change was the downward movement from nonmanual to manual occupations.

[25] In general, there is a high frequency of female workers in jobs categorized as "lower-middle," so it is not surprising that the majority of the female-headed households fall into this group.

[26] Note that there were fewer households in the professional/managerial group that experienced a reduction in incomes, compared to the other two groups.

[27] The higher income levels of the managerial/professional group should not obscure the fact that they did experience job loss land lower wages, as shown in case studies. Moonlighting served to maintain the level of household income in professional/managerial households: this practice should be viewed as a type of the intensification of work. See also Tarrès (1990) in Ciudad Satélite.

[28

29] In Spanish "nos hemos apretado". (Translator's note).

[30] Ratio of number of members in the work force divided by the number of members not in the work force. This is sometimes referred to as the "worker dependency ratio".

[31] The percentage represented by the number of members who are in the work force divided by the number of adults in the household.

[32] In thousands of pesos.

[33] In thousands of pesos.

[34] In each subtable, the category "all households" includes households in which the household member or relative made no contribution The third row of each subtable includes only households in which a contribution was made by the household member or relative identified in the subtable title.