Official non-recognition of contributions to the national as much as to the household economy obviously leads to non-recognition in policy making, planning, allocation of resources, the provision of support services and information, and of course in the distribution of the benefits of development. The failure to recognize much of the work which women do is therefore a failure to take women into account in all these areas. A "gender-sensitive" review of national policies and plans, and of the growing "gender gap" measured by common social and economic indicators will quickly confirm that women are major victims of official negation of their economic role and contributions, as much in the national product and human development statistics as in micro data on household economies and social progress.
Reasons for the failure to properly account for women's contributions range from the narrow definitions of work and economic activity, to sex biases and stereotypes among those who design questionnaires and the enumerators who collect the data at the field level. Rural women are still perceived as farmers' wives, helpers, housewives or agricultural labourers. Their home gardens which produce anything from food to medicines to fibres and dyes, are dismissed as a hobby or at best a nutritional supplement to family meals. In fact women may produce products at least equal in value to the paddy or wheat in the fields. When this is so, a rural woman is a farmer in her own right, and needs to be recognized as such so that her productivity and her product can be measured and fairly reflected in statistics and data bases.
The need to count women's work is not merely a matter of justice. It is a gross distortion of an economy to select only the contributions of half the world's farmers upon which to base agricultural policy and plans, and it is both an injustice and a distortion of reality to direct the flow of inputs only to those farmers. The constraints imposed upon women farmers by such biases has exacerbated an already truncated production base for women simply by reason of their sex and concomitant gender roles which assign most household work and services to them, with very little help from men. This double burden is an argument for more, not less recognition and support to women farmers. As men abandon agriculture therefore in favour of more remunerative work in other sectors, this increasingly holds true, and already in India women farmers outnumber men by a significant margin (Shiva, 1991).
The FAO Plan of Action for Women in Development (1990) includes among three principal areas of activity the gathering of statistical data and the conduct of research studies on all issues related to women in agricultural development. Initial work in this area has focussed on the disaggregation of as much data as possible, both by sex and by rural-urban. This helps tremendously to uncover discrepancies and systematic, institutionalized discrimination against rural women, but it is not enough. Disaggregation does not reveal distortions derived from inadequate definitions, conceptual biases, gender stereotyping, the inadequacy of research instruments or research design, the methods used in data collection and processing, nor constraints imposed by the economic framework used in the United Nations System of National Accounts (UNSNA). Figure 1 shows how a redefinition of economic activity changes the profile of women's work.
Many of the definitions in current use are exclusive i.e. they exclude much of women's work, impose invisibility and act as barriers to the clear explanation of time use and productive activity. The terms work, worker, economic and social activity, head of household, land holder and holding, primary activity, formal or informal, public and private, productive and reproductive are examples of exclusive terms used according to narrow and inadequate definitions. Similarly problems are caused when an enumerator fails to clarify whether a respondent has control over as well as access to a given resource. There is a need therefore to redefine or extend accepted definitions.
The failure to recognize housework, fetching water and fuelwood, and household services is a case in point. Housework is not regarded as "work" by ILO definitions, until someone has to pay for it, at which point it becomes a part of the GDP. Child care is called "Social or Reproductive Activity" and accorded no economic value until a child is placed in a care centre to enable a mother to undertake paid work. At this moment both the child-mincer and the mother become economically active. This is a nonsense.
There is a need not only to review definitions, concepts and instruments, but to engender a commitment on the part of those who hold decision-making power and control over information and data bases, to lend weight and will to the rhetoric of gender - sensitivity in statistics. This should begin with a new data base on women in agriculture. That this is difficult, time-consuming and expensive are frequently-proffered reasons for not addressing the problems, but these must be weighed against the cost of not doing so. As the feminization of farming in Asia goes hand in hand with a widening gender gap and the pauperization of rural women, the risk of ignoring women's work threatens the basis of food security itself. Ways and means simply have to be found to provide, by efficient cost-benefit formulae, ways of improving the official data base as well as providing complementary, accurate information.
Figure 1. Illustration showing time spent working according to traditional definitions of economic activity, and using new definitions:
The former is being tackled mainly through disaggregation, and the addition of a few gender-related social and economic indicators. The separation of data by sex highlights existing situations and biases, and provides a basis for challenging the status quo, but it does not reveal distortions of reality caused by definitional, conceptual and methodological problems. Alternative data collection on women's work has therefore been undertaken in sub-national research and surveys in several Asian countries using new definitions, techniques and instruments. These have generally been in small pilot projects carried out in representative agro-ecological zones over major seasons in the farming calendar. One of the most revealing has been the time use survey in which definitions of work were not limited to the ILO definitions of economic activity. Work for a rural women for example includes all those activities which have economic value for the family not only in terms of their market value but in also in terms of opportunity and/or replacement cost:
- Market cost is the cost of tasks done commercially in the local area
- Opportunity cost is the equivalent loss of earnings outside, if a "housewife" is kept at home
- Replacement cost is the local cost of hiring a person to do similar or the same tasks
Household surveys using new definitions of work which value the product in terms economists understand give a legitimacy to women's work as the enormous contribution it is, both to households and to the gross domestic product. Economists who apparently had no difficulty computing the output and profit in a milch cow are now able, for example, to compute the economic value of a mother breast-feeding her infant by valuing all the net gains against the costs of bottle-feeding:
* the market value of milk produced or the cost of breast milk substitutes (BMS)
* savings in energy when mothers need not sterilise bottles, teats, and the water used for mixing powdered BMS
* savings by not needing to purchase BMS feeding equipment
* lower infant and maternal mortality and morbidity rates, shorter hospital stays, fewer medicines required by breastfed infants and breast-feeding mothers
* fewer trips to markets to purchase BMS and equipment
* stronger psycho-social bonding between mother and infant
* natural birth spacing by lactating mothers, and reduced total fertility
In the same way, a women's "home gardening" can be valued not only for the market value of its products but also for the positive impact it has been shown to have on nutrition, variety in diet, savings in transport, reduction in infectious disease (morbidity) rates and social satisfaction. Far from being merely a women's hobby, food production of this type among poor families can make the difference between household food security and hunger, between poverty and destitution, survival and peril.
A useful tool to begin counting women's work has been the time use survey. Data on the use of time by women and men can be collected in a number of ways, so to find out which methods provide the most accurate information pilot surveys were carried out in 1990 - 91 in four Asian countries - India, Malaysia, Pakistan and Thailand - to test 6-8 different methods:
* rapid appraisal by checklist
* diary method
* interview questionnaire
* participant observation
* non-participant observation
* 24 hour recall
* group discussion (using checklist)
* group feedback analysis
Of these, the least accurate data about women came from traditional interview questionnaires administered by male agricultural extension staff to male farmers. They were particularly inaccurate when men asked male farmers about women's work. Female enumerators interviewing women farmers increased accuracy on their work significantly, but the most effective tool for gathering information on women's work was found to be by non-participant observation, especially when
- women observed women; men observed men;
- the observer was known to and trusted by the "observee";
- men were kept away or busy while women were observed (preferably men were observed simultaneously by a male observer);
- adequate explanations were given to the observed persons beforehand, especially regarding non-participation in activities by the observer (women tend, for example, to pass an infant to a female observer when he/she cries, or wants to play while she is busy, but this will distort time use data).
Time use surveys also present special problems for women's work. Unlike men, rural women seldom engage in only one activity at any one time. Along with work on the farm or in the house a women is often also taking care of children; on her way to and from the fields she may also gather fuelwood or water; with an infant on her back and a child at her heels a mother often organizes help from her older children as she works in the fields herself, at the most boring and menial unpaid tasks. Fully two thirds of unpaid work is carried out by women -more if 24 hour child care is included. Proponents of the new, feminist economics are therefore pressing not only for recognition of "recognized economic activity", but for economic value to be accorded much of the activity that is currently labelled social or "reproductive" activity.
While these surely present a challenge to statisticians and economists, the justifications for excluding women's work - complexity, difficulty, expense - can only be entertained against the cost of basing policy, plans and resource allocations on wrong - often grossly wrong - and misleading information. The justification for including such data is surely a straight-forward one of mobilising human capital around policies and plans made for all farmers, male and female. If the abundance of the past has allowed official exclusion of half the working force, growing competition between people, natural resources and productivity will no longer allow it. It is time to recognize the inadequacies and gaps in official agricultural census data, and to cast about for complementary information which will help complete the picture.
It is not however necessary to undertake costly nation-wide time use surveys. A small, carefully stratified sample in each major agro-ecological zone surveyed over major seasons of the farming year will provide sufficient information to complement traditional, disaggregated statistics, albeit with a few added indicators. These data will alert policy makers and planners to the facts of similarity and differences in men's and women's work, uncovering discrepancies which shackle women to low productivity due to fragmentation of their time, dual and triple work burdens, an ever-lengthening work day with previous little personal time or leisure, and the indignity of non-recognition for all that she accomplishes. The results of one time use survey in Pakistan illustrate the problems in Figure 2.
Practical problems in the field always crop up and must be dealt with. However in seeking to correct gender biases in data collection, the careful selection and training of enumerators is extremely important. As mentioned earlier, accuracy is considerably enhanced when enumerators are of the same sex as the interviewee or observed person, so most data collection teams need to include more women.
Language also presents problems. When enumeration is conducted through an interpreter, the meaning and nuances may be lost in translation. Further, in many cultures concepts of gender, day off, or economic activity simply do not exist in the minds of workers. Enumerators therefore should speak or understand the local dialect and customs in order to elicit meaningful responses. When these are not available, time will be needed to build rapport in order to increase understanding.
In time use surveys the prescence of observers and data collection activities can be quite disruptive to normal household activity. Where distortions are thought to be significant, a return visit after more explanations may be necessary, in order to normalize behaviour. If the enumerator is skilled and sensitive however these distortions will be minimal. The selection and training prerequisites thus are of paramount importance. A case in one country illustrates another gender problem. Some male enumerators were found unable to accept the authority of a female team leader, and deliberately manipulated responses to accord with their own gender stereotypes. Supervisors must be on the lookout for this source of bias.
Fatigue through continuous observation of as many as six 24 hour sessions in a week also presents problems. The job of a non-participant observer is exhausting, and no more than four such sessions should be assigned weekly. The categorization of activities may also present difficulties, and pre-testing is an important part of the training of enumerators so that a consensus can be reached on appropriate categories of activity for a given sample.
Figure 2 Differences in Time Use by Five Rural Women and Five Men
In interviews, problems arise with interviewee reticence, especially apparent among rural women who are more timid and less educated than men. Many interview questionnaires are too long, and boredom, frequent interruptions and the prescence of other family members may influence responses, as do inconsistent standards, insensitivity to cultural norms and sex biases in enumerators.
For all field teams, the climate and seasonal factors must be taken into account, and physical obstacles such as poor transport, floods or snow overcome. Agricultural surveys are comparable only when conducted in the same seasonal segments, and so in quantifying women's work in the home or on the farm, seasonal factors apply.
When women's work is recognized and appreciated for the contributions it truly represents, many things become possible. Some of these benefit women themselves. Many more benefit the entire family and the nation:
1. Household food security is enhanced by the provision of an accurate profile of production at the household and farm levels because bottlenecks and constraints are more readily identified and thus addressed.
2. Women regain dignity and status accorded to them by dint of their vital contributions to the economy, growth and development.
3. National and sub-national policies, plans and resource allocations based on correct information reflect more nearly the needs of all farmers, thus broadening the agricultural productivity and human capital base.
4. Constraints to women's agricultural productivity caused by their dual and triple roles in the home and their community are more readily identified, enabling the adoption of appropriate solutions - technology, training, child care, information, extension services, marketing, credit, access to land and water.
5. Availability of correct information enhances the design of remedies in development planning.
6. Management is enhanced by knowing who is responsible for which activities, and where inputs are best directed. Corrective action is made easier and bottlenecks can be more easily removed when responsibility is clearly assigned i.e. roles are known.
7. Resources are more likely to be mobilized and available on time when all those involved in the action have been identified in the plans.
8. The benefits of growth and development are more equitably shared when all beneficiaries are mentioned in programme plans.
9. Growth is likely to be more sustainable when both men and women are recognized and consulted.
10. Institutions such as agricultural cooperatives are more democratic (and less corrupt) when management and membership is monitored using gender-sensitive indicators, data collection and analysis in order to respond to gender issues.
While existing data collection systems are inadequate to count rural women's work, the time use surveys by non-participant observation demand too many resources to be undertaken in every country nation-wide. Adequate coverage however using a small, carefully stratified sample of rural households randomly selected from each major agroecological zone, and carried out over representative seasons in the farming year will provide an indispensible complementary data base for those seeking to redress gender-based discrimination. Such information will not only challenge but will enrich conventional statistics on women. New data will extend and enhance understanding of women's role in the economic and social fabric of a society, providing together with the tables produced by agricultural statisticians, sets of data which will have statistical validity of a quite different character.
New data on women gathered at sub-national level therefore will provide the kind of information needed to develop gender-sensitive policies, gender-responsive programme plans and gender-just distribution of benefits. This will enhance rural women's status by recognizing their value, but more than that for economists it will have a dual impact of facilitating the clear articulation of gender issues and thereby providing the basis for sound remedial policy and planning initiatives. Such therefore is not only for the sake of women but for the sake of agriculture and food security in the years to come.