Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Household food distribution
The analysis of time allocation and activity
patterns in nutrition and rural development planning
Women's role in domestic food acquisition and food use in India: A case study of lowincome urban households
The analysis of time allocation and activity patterns in nutrition and rural development planning
Food and Agriculture Organization of the United Nations, Rome, Italy
People's time allocation is a relevant indicator for those concerned with nutrition, food consumption, and food supply. First, when poor nutritional status results from insufficient food supply, policies and programmes meant to increase production and availability of deficient food items need to address the right people to be effective: A picture of the division of food-related tasks within the family-a common output of time-allocation surveys-may serve the purpose and avoid, for example, the design of intensive rural extension programmes explicitly meant to increase food production and specifically addressed to adult men in areas where food production is a woman's responsibility.
Second, nutritionists have been increasingly concerned with the negative effects of development: In particular, undesirable side effects of development on household food supply and consumption often occur because projects do change the time-allocation patterns of the people they affect. Projects especially modify time use of women whose activities are insufficiently documented during project preparation, despite the major contribution they make to household food supplies. In this respect, a better knowledge of women's role in foodrelated activity, and of their contribution to household food supplies can help prevent undesired consequences of development projects and the threat they may present for the nutritional status of the affected families.
Third, when they assess the adequacy of the food consumption of a population, nutritionists need to compare food and energy intakes to energy requirement estimates. These are calculated on the basis of assumptions about the individual's physical activity levels, which are largely undocumented, except for adult males engaged in military or industrial activities. Data on time allocation and activity patterns may provide a valid basis to confirm or disprove the current hypothesis.
This paper presents three time-allocation surveys; illustrates their main findings and their relevance to anthropologists, economists, and nutritionists; and emphasizes the major rote women play in economic and especially food-related activities.
WOMEN'S WORK-LOAD: AN OBSTACLE TO DEVELOPMENT IN CENTRAL AFRICA
Women's role in food-related activities and in the securing of household food supply has already been emphasized by the results of a number of field surveys. Leaving aside the many micro-surveys conducted over the last 30 years at the village level on a few households and those conducted in developed countries, larger surveys conducted in the Central African Republic, Nepal, and the Ivory Coast, for example, confirm the major role women play in food-related activities.
A survey conducted in the Central African Republic more than 20 years ago, with the stated objective of identifying "savable" time in the farmers' lives as "an indispensable preliminary to any serious operation of rural modernization," arrived at some unexpected conclusions: first, the significance of constant moving and travelling in the Bandas' lives, from homestead to farm or garden plot and water points or from village to market or farther, and second, the relevance of women in agriculture and especially food production (1).
These pioneering conclusions were the result of a survey conducted in two Banda villages of the Ouaka cotton area. The villages had been chosen to represent two diametrically opposed situations in relation to modernization: Pouyamba as an almost untouched, traditional village, far from channels of communication, and Madomale, well-situated on a main road and proud of signs of modernization in both daily life patterns and agricultural techniques.
The survey on the activities of the Banda villages was only part of studies conducted to define a project for the improvement of agricultural, and especially cotton, production in the area. However, in order to identify savable time, the project designers thought it wise to record at/ the activities of the persons covering 24 hours every day, and not-as was generally the case in such surveys-the duration of agricultural production activities only. "An understanding of the relations between productive activities and non-productive activities" was even one of the objectives of the survey. All household members 15 years old and over, considered adults in the report, were included in order to gain an idea of time allocation at the household level. The survey was conducted with a balanced combination of observational and interview techniques: 25 persons in Madomale and 30 in Pouyamba were visited every evening for ten consecutive months to interview them on their activities of the day; surveyors also watched their activities during the day, in the field or in the village. Altogether, after two months of psychological preparation and ten months of survey, data on the activities of a total of more than 15,000 person-days were recorded.
Time Spent in Visits: "A Threat to Productivity"
Table 1 summarizes the global time-allocation patterns in the villages and reveals first the relevance of time spent away from village life: 16.2 per cent of time on average spent in visit, travelling, or illness lasting more than a day in the wellconnected village of Madomale, 7.4 per cent in the more isolated Pouyamba. The report incidentally suggests that visits "correspond to the pleasure inhabitants find in long walks, but threaten productivity" (1).
Women cannot marry within the clan, and the village is structured around the clan. Because the details of activities performed outside village limits-or in isolation at home during long periods of illness-were difficult to obtain and would not be representative of an individual's productivity, they were not recorded. This is a problem that must always be faced in time-allocation surveys.
TABLE 1. Time Allocation by Sex and Village Expressed as Percentage of Total Time over a Ten-Month Period
|Rest and sleep||55.7||60.9||46.0||52.7||50.0||56.6|
|Total recorded activities||88.0||93.8||80.8||91.5||83.8||92.6|
|Visits and illness||12.0||6.2||19.2||8.5||16.2||7.4|
Work represents about one-quarter of total time, but the average daily 6 hours vary markedly with the seasons. In contrast, time for sleeping and resting is very constant and oscillates around 12 hours per day: seasonal variations of work activities are compensated by non-work activities (leisure, games, dances, visits, etc.).
Work Hours for Men and Women
Besides the importance of time spent in travelling and visiting, the main conclusions of that survey concerned the difference between men's and women's work-load. When they are present, women work on average 8 hours a day, men 51/2 stable 2). Domestic tasks occupy more than 31/2 hours of women's days but are negligible for men. In addition, women's time devoted to agricultural activities is equal to men's in the traditional Pouyamba village and even greater in the "advanced" village of Madomale
More precisely, they spend a comparable amount of time cultivating cotton-the same 2 hours in Madomale, and only 18 minutes less than men in Pouyamba-but time devoted by women to food production is one and a half times as much in Pouyamba and four times as much in Madomale as the time men spend in such non-cash cultivation-an indication, perhaps, that money earned on cotton does not go to food purchases stable 3) and surely that women provide most of the household food supply.
Men work more instead in the "other activities" groups, such as house construction, crafts, and hunting, amounting to a total of over 2 hours every day. All this leaves men with more time for non-work activities such as short visits, dances, games, and other social activities, a constant that we will see in other time allocation surveys. Men, furthermore, have rest and leisure for 11/2 hours longer than women do every day.
TABLE 2. Average Time Allocation Per Day (Hours and Minutes) in the Village, Based on a Ten-Month Period of Observation Extrapolated to One Year
|Agriculture||2 h 28 min||3 h 57 min||3 h 17 min||3 h 17 min|
|Domestic activities||17 min||3 h 37 min||10 min||3 h 40 min|
|Other work||2 h 44 min||36 min||2 h 5 min||44 min|
|Total work||5 h 29 min||8 h 10 min||5 h 32 min||7 h 41 min|
|Non-work activities||3 h 18 min||2 h 9 min||2 h 52 min||2 h 28 min|
|Rest and sleep||15 h 13 min||13 h 41 min||15 h 36 min||13 h 51 min|
TABLE 3. Time Devoted to Cotton, Food Crops, and Other Crops-Daily Average over the Year (Hours and Minutes)
|Cotton||2 h||2 h||2 h 26 min||2 h 8 min|
|Food crops||23 min||1 h 54 min||39 min||1 h 5 min|
|Other crops||5 min||3 min||12 min||4 min|
|Total||2 h 28 min||3 h 57 min||3 h 17 min||3 h 17 min|
The Conclusion That Development Adds to Women's Work-load
During the last decade the question of women's role has become a cliche, but their heavy work-load was recognized much earlier. Twenty years ago, the final BDPA report stated that "women participate almost equally as men to cotton production, but in all other productions (cassava, peanuts) play a more important role." Moreover, "when a rural development programme requests from the farmer an increased workload, women are those who perform it. If women play such a role in agricultural production, it is paradoxical that rural extension addresses husbands or brothers but never women in the first place" (1).
Intensification of agricultural practices was, in conclusion, impossible if extension agents were not able to convince men to work more in agriculture and also to relieve women from part of their work-load. Otherwise, in view of the major role women played in food-related activities, household food supply could have been threatened by a reallocation of women's time. Or, more probably, in view of the wise priority people generally give to their household food supply, time reallocation was not to take place, and there was no way to increase cotton production.
THE NEPAL SURVEY
A survey conducted more recently in Nepal with a completely different data collection methodology reached similar conclusions on the relevance of women's food related work, besides revealing, at least on a quantitative basis, the heavy work-load of children. It found that women spend 60 per cent of their household time on food-related activities. The survey-specifically intended "to highlight the role of women in Nepalese society"-was conducted in eight villages representing all ecological zones of Nepal (2).
Time-use data were collected for 192 households comprising about 1,200 individuals. Half of these people were observed for six months and the other half for one year. Data collection methodology was based on visits by a researcher at hours that had been selected in advance at random between 4 a.m. and 8 p.m.
Time for sleeping was assumed to be an eight-hour constant-from 8 p.m. to 4 a.m.-and data were therefore collected only during the daytime. Each household was visited on alternate days for 26 or 52 weeks. Thus, each household was visited 78 times in the six-month studies and 156 times in the one-year studies. Data collection by this method represented frequency of activity observations. This was taken as, and assumed to reflect, frequency of time distribution. The resulting time allocation data for the person were thereby derived. A total of over 90,000 activity observations were recorded.
The survey's "attempt to capture the full subsistence production of the household" (2)) by estimating the economic value of food produced for household consumption was innovative. The researchers "tried to capture the physical production within the household to the maximum extent and to document which family members are responsible for that production. To do this they developed detailed schedules on household production and food processing," Trade food-i.e., food items available in the market-were valued at the prevailing market prices. Non-trade food items were valued with a conservative replacement cost approach: for example, dried green vegetables were valued at the price of the cheapest vegetable in the off-season. A special procedure was adopted to put a value on food processing done at home, which generally involves home-produced raw materials.
TABLE 4. Comparative Average Daily Time Allocation by Sex and Age Group, Expressed in Hours and Decimals
15 Years and Older
10 to 14 Years
5 to 9 Years
|Labour force participation|
|Hunting and gathering||0.17||0.05||0.11||0.08||0.03||0.03||0.02||0.03||2.01|
|Cleanig dishes and pots||0.03||0.39||0.23||0.03||0.23||0.15||0.01||0.07||0.04|
|Child care and rearing||0.16||0.06||0.45||0.22||0.35||0.29||0.15||0.34||0.24|
|Total work burden||7.51||10.81||9.32||4.83||7.31||6.23||2.33||3.39||2.87|
Source: Acharya and Bennett, ref. 3.
Sleep has been assumed as accupying eight hours a day for everybody-almost certainly an underestimate as judged by the other studies cited in this paper. (Sub-totals and totals calculated before rounding to two decimal places.)
The Work-load of Children
The Nepal time-use survey, conducted to illustrate the status of Nepalese women, also revealed the heavy work-load of children (3). Children between 5 and 9 years of age work nearly three hours a day stable 4), and more than six hours in the 10- to 14-year age group. Education is marginal- about one hour a day on average.
From the age of 5 onward, girls are at a disadvantage. Females were found to work more than males in all age groups: almost 31/2 hours for a girl 5 to 9 compared to less than 2 hours a day for a boy of the same age. Female children 10 to 14 years old spend the same time in animal care, but more time on agriculture, and globally more time in all activity groups than do male children in the same age group. In particular, girls spend more time replenishing the water supply, collecting fuel, and processing food than boys do.
In this respect, time allocation data do confirm interviews: "Neither the cost of education nor the conservatism of the parents is the main cause for the significantly lower percentage of female enrollment encountered in the sample villages. Rather, it is the family's dependence on girls' labour at home and in the fields that is the primary reason for keeping girls out of school" (3). Children's participation in production is a recurrent obstacle to their school enrolment in rural societies, and indeed, affects children of both sexes.
The Work-load of Women
To return to women's time allocation in Nepal, as in the Central African Republic, women work longer hours than men do: 10 hours and 48 minutes compared to 7 hours and 30 minutes stable 4). Again, domestic activities and meal preparation make up part of the difference: 4 hours a day for women and just over 1 hour for men. Women, however, also work 1 hour longer in food-related activities, and they spend less time on personal maintenance, education, social activities, market activities, and leisure or rest: less than 4 hours compared to more than 6 hours of Nepalese men.
Women's major share in food-related activities is illustrated in table 5. They spend more than 1 hour per day carrying water and wood up and down the hilly slopes of Nepal, almost 3 hours (the same as men) on agriculture, and an additional 1 hour on food processing. Men prevail in hunting and gathering, and in animal husbandry they spend almost 1 1/2 hours, only 30 minutes more than women do.
The relative importance of women's food-related activities is confirmed in table 6, which shows the share women contribute in performing various kinds of work: some 90 per cent of time spent on collecting or processing food, while still spending 55 and 45 per cent of time on agriculture and animal husbandry respectively, totalling 60 per cent of time devoted to food-related activities, not taking into account the preparation of meals.
Women and Decision- making
The Nepal survey also inquired into the sensitive question of distribution of decision-making responsibility between the sexes for certain important steps in the agricultural production process. Men were found to have a slight lead in decisions regarding labour allocation, except in the area of arranging for parma tan exchange labour that is shared equally between the sexes). But in the area of agricultural decisions, women have a substantial lead over men. Seed selection, in particular, is clearly a female responsibility (2).
TABLE 5. Time Devoted to Food-Related Activities by Nepalese Males and Females 15 Years Old and Older, Expressed in Minutes and as Percentage
TABLE 6. Contribution of Nepalese Male and Female Adults in Activities or Activity Groups Listed as "Productive Activity" (Share Expressed as Percentage)
|Total productive work burden||36||64|
Women's Contribution to Cash Income
Finally, it was found that "rural Nepalese women not only contribute more time, but also generate more income than men for the total household economy" (2). Men have the lead (72 per cent) over women as far as outside cash income is concerned, but outside cash earnings represent only 18.6 per cent of total income. In the major area of home production, women contribute 54 per cent of the total. When both home production and outside cash earnings are considered together, women contribute 50 per cent of the total household income compared to 44 per cent for men, and children contribute a surprising 6 per cent.
However, the final report concluded that "women's role in subsistence agriculture and in the market economy, including their considerable decision-making responsibilities, are not reflected in any development agency strategies for extension, training, credit, employment, etc. Instead, these strategies are targeted exclusive/y to ward men" (italics added). This not only results in a failure to mobilize the full productive human potential but also entails the risk of failure of these very policies and, in particular, of any rural or food production development projects, since these do not address those who actually are primarily responsible for food-related activities: women.
The final report, therefore, recommends programmes to mobilize women fully in the development process, to assign targets concerning women in each development sector, and to reduce the current work burden of women by introducing appropriate technologies for food processing, water collection, food storage, compost making, seed selection, and crop processing.
Besides noteworthy revelations of the work-load borne by children and the role of women in rural Nepalese society, a review of the Nepal survey suggests three conclusions: First, women's participation in household activities is underestimated by conventional definitions of economic activity. Second, the real productive activities of children have gone largely unsurveyed. Third, home production-a very important element of rural household total income-is difficult to identify and measure, especially through interviews.
MULTIPLE USE OF TIME-ALLOCATION DATA
Time-use surveys have been for a long time the privilege of social anthropologists interested in the division of labour by age and sex. They have collected data through intensive investigation of small communities using participant observation as the main technique. Unfortunately, until recently, very few social anthropologists were oriented toward collecting and analysing quantitative data, and they presented mostly descriptive information (4). In Nepal, the research team combined for field studies "quantitative data gathering techniques with an in-depth anthropological approach requiring extended residence in the study villages" (3). Also recently anthropologists have become more familiar with quantitative data handling and more concerned with participating in an operational use of their findings. They do participate increasingly, therefore, in development planning. Agronomists and development planners are also interested in time-use surveys, focused and often limited to agricultural production tasks, to quantify productivity and identify savable time in specific areas during the preparation of development projects.
The final report of the time-use survey conducted in the Central African Republic 20 years ago concluded with a statement going far beyond the scope of the study: Rural modernization for improving productivity increases women's work-load, and reduces men's working hours. "In these conditions, any programme of rural modernization will soon reach its limits, unless planners can force men to work more on agriculture and also to release women from part of their work-load" (1). In other words, time-women's time in particular-can be the scarce production factor in a development process. Yet at the same time, conventional statistics do not reflect women's time as "saturated" at all.
If we consider only the Nepalese census definition of economic activities (3), women in Nepal are less active than men. They work one hour less and represent only 49 per cent of economic activities in Nepal (table 7). However, they represent 74 per cent of the activities that, in addition to food production, are necessary for the subsistence of the households and that would be purchased with cash in the form of final products (fuel, water, processed food, etc.) if family labour were not available. In total, women perform 56 per cent of the "enlarged economic activities" of the household, all of which correspond to goods production, and these are valuable in monetary terms. If we would add to that (as would the new household economists) activities that constitute services for the household (and again that could be substituted in some societies by employing a person for remuneration), women would bear as much as 64 per cent of the household's total work burden.
The problem is that, in most conventional employment surveys, the category of "economically active" includes all those who perform activities called work in the market sector or who are self-employed in a family enterprise (4). In general, only activities that can be readily measured in money terms through wage rates or prices of the resulting products are considered economic. As a result, work is most often conventionally defined as a receipt of cash or in-kind income in the form of salary, wages, or profit for activities in agriculture, trade, industry, or external services. Unpaid domestic work is even specifically excluded from the work definition in the UN census manuals. The 1968 UN Handbook on Methods of Analysing Census Data on Economic Activity of the Population (5) specifies, "Housewives occupied only with domestic duties are to be excluded from the measure of the labour force because the goods and services they produce are not considered as economic, just as the value of their products is excluded from the measure of income in national accounts. "
In largely monetized economies, unpaid domestic work corresponds essentially to the running of the households. Applied to less monetized economies, it becomes a residual category, including-besides cooking, child care, and domestic services-food processing, kitchen gardening, animal tending, fuel and water fetching, fishing, hunting, gathering, and manufacturing for home use.
This restrictive definition of the active population has resulted in an under-estimation of the amount of work performed in the subsistence sector and in a subsequent overestimate of non-usefully-employed time, especially in rural societies. This error has resulted in the failure of many projects built on the assumption that such "wasted" time, in reality time devoted to activities not recognized as economic, could instead be recovered for gainful activities.
Activities compete with each other for the time of household members. Total time available for the household from its members is the ultimate major source of income and at times a very scarce factor. Households and individuals economize on their time by selecting activities (4). In the rural areas of the Central African Republic, for example, water collection is at least as essential and economically as important as selling cotton. The response of households and individuals to development proposals-all proposals of change for, or threats to, their time-allocation pattern-is an expression of the criteria with which they allocate their ultimate production factor: time.
In this respect, not only is the contribution of women to total household activities largely under-estimated, but also that of children. This has serious implications for the success of educational programmes. In most countries, there is an age up to which school is compulsory, and an age below which remunerated work is illegal. Employment surveys prudently do not inquire about economic activities below that age. As to the unpaid household work children perform, determination of the amount of work required for the inclusion of family workers in the category of the economically active population is difficult. The 1962 UN Report on Sex and Age Patterns of Populations in Economic Activities (6) indicates, without reference, that "international bodies" have recommended that family helpers who do not receive specific pay for their work should be counted as economically active if their contribution equals a third or more of a normal Pay period. Nepalese children would qualify. A specific time-allocation survey revealed that fuel and water collection or food processing were found to be major activities beginning at age five in the Nepalese communities surveyed (table 4).
Within the household time-allocation system, children's work contribution, therefore, represents another major element of decision-making for households faced with development proposals-including education or family planning choices-and another element, too, that planners do not or cannot appreciate unless specific time-allocation studies are conducted in the project area.
TABLE 7. Contribution of Women and Men to Different Levels of Economic Activities (in Hours and Minutes and in Share Expressed as Percentage)
|Average Hours per Day||Weighted Work Share (%) per Sex|
|Conventional economic activities|
|(labour-force participation)||5 h 49 min||4 h 37 min||51||49|
|Subsistence economic activities||54 min||2 h 10 min||26||74|
|Expanded economic activities||6 h 43 min||6 h 47 min||44||56|
|Domestic||47 min||4 h 20 min||14||86|
|Total productive work burden||7 h 30 min||10 h 49 min||36||64|
Women's Role in Food-Related Activities and Contribution to Household Food Supply: Evaluation of Energy Expenditure
Women and children contribute a large share of food-related activities and of the household food supply as well Yet these tasks are not precisely identified or quantified, and they are not taken sufficiently into account by planners. Therefore, since time is a finite resource, and change, when attempted, results in a reallocation of available household time, development may and does threaten household food-related activities and overall food supply, especially when a large share of it comes from home production.
In addition, for nutritionists, the ultimate objective of collecting data on food consumption is to compare them with estimates of requirements. To take the example of energy, energy requirements correspond to energy expenditure, 28 to 52 per cent of which is the level of physical activities of the individual (7). This is another reason for nutritionists to be interested in time-allocation studies, because they provide data on the physical activities of people, a largely undocumented subject, especially for women and children.
Contents - Previous - Next