Children's work, earnings, and nutrition in urban Mexican shantytowns

2 downloads 0 Views 113KB Size Report
Jun 16, 2009 - 2Department of Anthropology, University of Georgia ... their own or their younger siblings' nutritional status assessed anthropometrically.
AMERICAN JOURNAL OF HUMAN BIOLOGY 22:60–68 (2010)

Original Research Article

Children’s Work, Earnings, and Nutrition in Urban Mexican Shantytowns. ALEXANDRA BREWIS1 AND SARAH LEE2 1 School of Human Evolution and Social Change, Arizona State University 2 Department of Anthropology, University of Georgia

ABSTRACT For many children living in conditions of urban poverty, earning money can provide additional resources to them and their families, and this raises interesting questions about the potential biological consequences (costs and benefits) of children’s work in ‘modern’ settings. This study uses time allocation, ethnographic, dietary, and anthropometric data collected with 96 urban Mexican shantytown children (aged 8–12 years) and their older and younger siblings (aged 1–18 years) to test hypotheses related to the effects of children’s cash earning and cash contributions to their households for their own and their sibs’ nutritional status. Regression models show that children’s contributions to household income and the time they allocate to working outside the home makes no difference to their own or their younger siblings’ nutritional status assessed anthropometrically. Dietary quality, based on food recalls, is worse in working than non-working children, even taking household income into account. Children’s allocation of time to work and their cash contributions to the household do however significantly improve the weight of their older siblings, especially sisters. This suggests children’s work in urban ecologies might have different constraints and opportunities for their own and siblings’ growth and nutrition than typically observed in subsistence settings. Am. J. Hum. Biol. 22:60–68, 2010. ' 2009 Wiley-Liss, Inc.

As populations urbanize and the market economy expands globally, children’s participation in the cash economy is growing, but comparatively little is known of the possible human biological consequences of children’s work outside of traditional subsistence settings. Here we test how time allocation and relative earning capacity in middle childhood predicts variation in the nutritional and growth status of shantytown children in urban Mexico. The study is guided by two key ideas: (A) Children’s income from work outside the household may have consequences under conditions of urban poverty if it helps maintain adequate nutrition for growth by increasing access to food. (B) Children (at least in subsistence settings) sometimes function as net producers for the household through provision of labor for either agricultural or child care (Bock 2002, Kramer 2004, 2005), creating the possibility that children’s earnings might have consequences for their siblings as well as or even instead of themselves. We use the case of 8–12-year-old children living in shantytowns in urban Xalapa, Mexico. Xalapa, the capital city of the state of Veracruz, is home to approximately 500,000 people according to the last census in 2000, which almost doubled over the preceding decade. This dramatic increase was largely fueled by tumbling coffee prices, forcing campesinos to leave the countryside in search of work. Many of these families settled on the outskirts of Xalapa, constructing temporary-to-permanent housing in mostly illegal and unserviced shantytowns. Acquiring sufficient nutritious food is a concern for this population. All but a few household incomes fall well below the government’s standard of minimum income for a 4.6 person household of $5,110 pesos (US$480), and half the households in our study listed food as the greatest household expense. Most of the families buy their food at local shops and informal market stalls, with some households supplementing their diet through home gardens and fruit trees. A few charitable organizations provide day-old bread for free or at reduced prices. Milk (at 10 to 12 pesos a liter) is considC 2009 V

Wiley-Liss, Inc.

ered a luxury item and meat is usually only used for flavoring family meals. Eggs, however, are fairly cheap and readily available. Most family meals include an abundance of boiled beans, tortillas, a tomato-based salsa, and a small serving of offal, skin, or feet. When households are strapped for cash (usually towards the end of the fifteenday pay cycle), they will switch from machine-prepared tortillas (8 pesos/kg) to making their own (which averaged about 2 pesos/kg). When children earn and reserve money for their own use in these communities, it is almost all spent on food purchases. Families with two working parents or single-parent households also struggle to deal with child care, given many migrant families are nuclear and the cost of daycare is prohibitive. Children in middle childhood thus provide the majority of childcare for their younger siblings if mothers are working, and contribute significantly to housework. Time allocated to sibling care and housework means less possibility of outside sources of income, or other opportunities for supplementary (nonhousehold) food sourcing (like begging, foraging, running errands for food). The children in these shanty colonias, almost all of whom attend school, none-the-less have significant discretionary control of their time to allocate to different activities including work for money in both the formal and informal economy, which sometimes is combined with childcare. Contrary to the experience of some working children world-wide (See Bock 2002), children engage in Contract Grant sponsor: National Science Foundation; Contract grant number: 0350119; Contract Grant sponsor: Wenner Gren Foundation for Anthropological Research. *Correspondence to: Prof. Alexandra Brewis, Arizona State University, School of Human Evolution and Social Change, ASU, Tempe, AZ 852872402. E-mail: [email protected] Received 6 November 2008; Revision received 7 April 2009; Accepted 8 April 2009 DOI 10.1002/ajhb.20954 Published online 16 June 2009 in Wiley InterScience (www.interscience. wiley.com).

61

SHANTYTOWN CHILDREN’S NUTRITION

work without significant educational cost, as the school system is structured around a four-hour school day, 8–12 a.m. or 2–6 p.m., and a relatively short school year. Combined with common teacher labor strikes, all children have significant time that can be allocated to work and earning, even while in school full-time. In this case, as many others (at least in Mexico: Binder and Scrogin (1999)), children’s work occurs around rather than instead of time spent at school. Cost to the children is thus a loss of leisure time, rather than educational access. And only a very few children we observed during the total study were ever seen or reported to give most or all of their entire earnings to an adult, including parents. Under these conditions, we hypothesized that: H1: Children who (a) spend more time working and (b) retain earned money for discretionary purchasing will have better nutritional status compared to other children in the sample. H2: Children who allocate more time to child care and household chores will have worse nutritional status compared to children who spend proportionally less time in childcare and household chores. By better or worse nutritional status, we mean weightand height-for-age, weight-for-height, and skinfolds, as well as dietary quality and diversity. However, we would expect to see greatest advantage in short-run measures of nutritional adequacy, specifically the dietary scores, weight-for-height and skinfolds, rather than longer-run measures, such as height-for-age. In testing these hypotheses, we also considered how the children’s gender and birth order might act to shape the adaptive contexts of their work. Ecology and gender shape if and at what age a child becomes a net-producer in the household (Bock 2002, Hames and Draper 2004, Kramer 2004, Kaplan et al. 2000, Miles 1993, Yamanaka and Ashworth 2002). Drawing on the idea of ‘helpers at the nest,’ (Crogneir et al. 2001, Hames and Draper 2004, Kramer 2005, Niewenhuys 1996, Turke 1988), we can also propose that their work could provide dietary/resource benefits that accrue to siblings. Thus: H3: Children’s increased cash contributions to the household and allocation of time to working and to child care will predict better nutritional status of younger siblings. METHODS Study Samples Data were collected in Xalapa, Mexico over a thirteen month time span (September 2003 to October 2004). The study included 96 children aged 8 to 12 years and their households, drawn from 30 low-income/shantytown neighborhoods. The average household size was 5.9, and average number of children in households was 3.6. One target (focus) child in the specified age range from each participating household was selected. If there were more than one child in this age range, as frequently occurred, then the child with the most recent birthday was selected as the target. The primary caregiver, usually the mother, was interviewed for household data, including income and expenditures, and baseline diet of the household. Target children were observed ethologically, provided dietary recall, and measured anthropometrically. All other children aged 2–18 years living in the household provided the same basic anthropometric and health measurements,

resulting in a total sample of children (targets and their siblings) of 282. Following the research protocol established by IRB 2004-101-88 at the University of Georgia, all participants gave written or verbal consent (parents) or assent (children). Assessments of Children’s Time Allocation to Earning and Non-Earning Work Target children were focally observed on three occasions for five hours each, totaling fifteen hours of structured observation for each of the 96 target children. Other child time allocation researchers have variously used 24-hour recalls (Binder and Scrogin 1999), spot observation (Munroe et al. 1984), and short-duration (e.g., ten minutes) systematic observation (Draper and Cashdan 1988), as appropriate to the questions being addressed. Based on pilot research, we determined a focal follows in five hours blocks to be appropriate to capture the range and frequency of child time allocation to work outside the home, formal and informal work and foraging (or scavenging) for food. Also, given the urban environment and distance between households, scan sampling was not feasible. This amount of observation (15 hours/child total) is also consistent with recently published studies by anthropologists using similar methods of focal observation to study child time allocation (Bliege Bird and Bird 2002 [14 hours], Hawkes et al. 1995 [12 hours], and Yamanaka and Ashworth 2002 [12 hours]). All observations were based on a time allocation ethogram developed using standard ethological principles (Lehner 1996), and conducted outside of school hours and during daylight [Observations were limited to daylight hours for reasons of safety]. Coding for time allocation included child care, household chores, work for a family business, work for others in exchange for food, work for money, free time, foraging, and homework. This included fine-grained distinction of kind and degree of childcare (sole caretaker for one or more younger siblings, child ‘‘minding’’ in a group with other siblings, sibling care in the presence of a parent). The grid was conceived in such a way that child could simultaneously in be several categories at once. For example: ‘‘Juan’’ tends the family market stall, providing childcare for his two year old brother, who he is feeding a sandwich bought with his own money. This would be coded as informal work, sole childcare, and a food sharing event with a younger sibling. We followed the target child for five hours in each block, never being out of sight for more than a few minutes, and taking detailed descriptive notes in addition to checking on the ethogram grid. We paid close attention to duration of behavior, any food or money obtained, shared or spent and took note of the context of the behavior (home, on the street, etc.). Observation of children’s time allocation was confirmed by their own and mother’s reports of work and child care activities during household interviews and time allocation recalls. These interviews provided a valuable means of ensuring that the focal follows did in fact represent what the child did with their time when they were not being observed. However, these data were not used in the time allocation analysis discussed here. For the analyses, children were then categorized based on the percentage the fifteen hours of observed time alloAmerican Journal of Human Biology

62

A. BREWIS AND S. LEE

cated to earning money outside the home, either formally or informally (time was measured in minutes and seconds by the researchers’ wrist watch.). Three levels were distinguished based on whether they were never observed to work outside the home to earn money (46.9% of children), worked some (>0 and less than 30% of time: 30.2% of children), or worked more than 30% of their time (22.9% of children). Children were also categorized based on the percentage of non-school time they were observed to spend engaged in the care of their younger siblings. The three levels of the variables were: low or no child care (less than 5% of observed time), some child care (5–20% of observed time), and more child care (>20% of observed time). In an analysis of time allocation data from studies in 14 countries, Scrimshaw et al. (1996) estimate that boys aged 10– 14 averaged about 4 hours in productive labor and one hour in domestic labor per day, and girls 2.5 hours of each (5 hours, or one third of a 15 hour day). Thus, Xalapa children appear to be somewhere in the middle of variation in children’s time allocation to work, viewed cross culturally. Assessments of Target Children’s Cash Earning and Household Contributions Children’s earnings were assessed through children’s and maternal reports. Children’s monthly estimated incomes ranged from 0-1960 pesos; the mean was 176 pesos (17 USD at the time) (6 369), although the majority of children (72%) had no ‘normal’ monthly cash income. Twenty-two of the children contributed money to the household, and in most cases the contribution is significant compared to total household income from all sources (most contributed greater than 10%, some as much as 55%, of monthly household income). The children who worked for money did so by working in the local bakeries, washing cars, bagging groceries, selling in market stalls or on the street, running errands or doing other chores for money, informal construction, and seasonal work (both agricultural and mercantile). Children’s work, and therefore income, was very fluid, and might vary dramatically month to month. Children’s earnings need to be placed in the context of household income. Monthly household income data need to be treated with some caution, given it tends to be erratic (making seasonal sweets for sale, temporary construction jobs, seasonal agricultural labor). Mothers did not always know the exact amount of their husband’s income, but could report how much money was available to pay bills and buy groceries. Therefore, the amount of money spent for food per household member was used in models as a proxy for household income. Almost half of the families received federal support, in the form of a small stipend for each child between the ages of 7–18 years, paid every other month. We created a two level variable describing whether each child had discretionary control of any earned income, which was coded as 1 if they retained any cash they earned for their own use and 0 otherwise. This categoric variable also acts as a proxy for whether children engage in independent spending on food outside the home because, as almost all earned cash is spent on food items, and in the short term (i.e., money is not saved). We also created a variable related to target child’s contribution of cash to the household. This two-level variable was based on whether the child contributed any earned cash to the American Journal of Human Biology

household or not. For the purposes of analysis, we estimated not only how much money children contributed to the household but also its ratio to total estimated household income. Children could thus provide similar amounts of cash to a household; in one case it would be nearly all the household income but in another only a small supplement. Children were categorized based on whether a child contributed more than 10% of the total household income from independently earned income (22.9% of children) or not. Assessment of Anthropometric Status of Children (Targets and their Siblings) Anthropometric data collection procedures followed the conventions outlined by Frisancho (1990). Each target child was assessed anthropometrically on two different occasions at the time of their entry to the study and a follow-up six to eight weeks later. The second set are those reported here. Measurements taken for all target children included weight, height, mid upper-arm circumference, and skin folds (subscapular, supra-pelvic, triceps and biceps). We measured siblings for weight, height, midupper arm circumference, and skin folds (triceps and biceps). Children were weighed using a Detecto digital scale, height was taken with a Cooper Tools six foot measure and skin folds were assessed using Lange skin fold calipers. Each measure was taken three times and the average was used in analysis after checking for outliers. A caveat is that the scale and stadiometer both require a level floor, and it was sometimes a challenge to find an exactly level area on the dirt floors of many homes. Each child was also given a physical examination to identify edema and characteristic hair discoloration (clinically significant for protein-energy malnutrition), as well as iron deficiency anemia via examination of the inner surface of the lower eye-lid (Jelliffe 1966). For target children and their siblings body mass index (BMI: weight/height2), height-for-age z-scores, and weight-for-age z-scores were calculated by comparison to median values on revised international WHO reference curves. Low height for-age provides a gross measure of long-run under-nutrition, particularly caloric insufficiency, based on the understanding that chronic under nutrition forestalls growth in stature, but is also associated with repeated infections (usually gastrointestinal and systemic infections) (Digby et al. 1987, Frongillo 1999). Children were categorized as ‘stunted’ if they exhibited z-scores of