Jump to... Introduction Food security Project areas Market orientation Women's work burden Conclusions Tables Technical appendix References

Programmes and policies have direct and indirect effects on household food security. The four dimensions of food security and their first-order determinants given in the flow chart (Figure 1) in Chapter 3 provide a convenient framework for thinking through these effects because these are the pathways through which the effects will be felt.

The impact on the level of food acquirement. IFAD-assisted programmes of the kind studied in this report affect the level of food acquirement through all three first-order determinants noted in Figure 1: endowment, entitlement mapping and women’s control over income and spending.

Some components of these projects enable women to acquire new assets, such as livestock, a village shop, or machines and implements. By augmenting the endowment set in this manner, the components enable households to achieve a higher level of food acquirement. Other components of programme interventions such as those that aim at improving the education and health of household members, especially women, can have a similar effect by augmenting a household’s human capital.

Even if no new assets are created, certain kinds of interventions can still have a favourable effect on food security by improving entitlement mapping. One example is the case where credit is provided for use as working capital. Poor women engaged in cottage industries or petty trading will often obtain the necessary working capital from village moneylenders or big merchants at an exorbitant interest rate, resulting in a very low rate of return for their labour. The replacement of such high-cost credit with relatively low-cost facilities or with special credit programmes, such as those for women in Tamil Nadu and Nepal, can raise the rate of return for women’s labour. This can then enable households to acquire more food for the same expenditure of labour power. Another example is the support provided to small producers in the marketing of their inputs and outputs, such as the support the Girijan Cooperative Corporation of the Andhra Pradesh Tribal Development Project is supposed to give. By freeing small producers from the extortionate demands of private middlemen, such support should in principle improve the rate of return on the investments of the producers and thereby improve household food security.

But raising the level of income – whether through the creation of new assets, or through raising the rate of return – is not the only way these programmes affect the level of food acquirement. To the extent that women are enabled to engage in income-earning activities of their own or to contribute to joint family enterprises by serving as a conduit for credit, the status of women within the household is expected to improve. This will help them gain greater control over the disposal of income, which will tend to raise the level of household food consumption, in addition to any benefit that may accrue from higher levels of income.

This implies that there can be a favourable impact on food security even in the extreme case when programme intervention does not generate any new income. For example, if women engage in productive activities, men may decide to withdraw their labour and enjoy more leisure. Chapter 6 shows that this is more than a mere possibility. In such a case, the impact on household income may be nil, and yet food security may improve if, by becoming the major breadwinner, women come to acquire greater control over the disposal of income.

The impact on the ability to cope with shocks to acquirement. Programme interventions can also improve the ability of a household to cope with shocks to food acquirement through the two first-order determinants noted in Figure 1, that is, the diversification of livelihood structure and the scope for consumption-smoothing.

The women’s income-earning activities promoted by these projects can serve as useful vehicles for the diversification of the livelihood structure of households. As noted in Chapter 4, this can have a favourable effect on a household coping ability. If the programmes also contain a component for group savings (either in the form of cash, as in the Tamil Nadu and Nepal projects, or in the form of grain banks, as in the case of the Andhra Pradesh project), that is, if they create funds that the members can draw upon from time to time in hours of need, then the coping ability can improve further as a result of the greater scope for consumption-smoothing.

The impact on the utilization of food. The impact on food utilization depends on the nature of the programme interventions. If the programme focuses exclusively on women’s income-earning activities, requiring them to devote more time to these pursuits, then there might be a problem. Since poor women are already hard pressed for time, any further demand on their time will oblige them either to spend less time on domestic chores, including food preparation and child-care activities, or to subject themselves to extreme drudgery, with potentially deleterious effects on their health.

Whichever option the poor women choose, it will compromise the quality of food utilization because, as shown in Figure 1, the proper utilization of available food depends, inter alia, on both the time constraint and the health constraint faced by women. If the women choose to cut down on domestic chores, then their poor food utilization will be the result of the time constraint. If they choose to push themselves even harder and thereby compromise their health, then their poor utilization will be the result of the health constraint. In either case, the quality of food utilization will suffer.

Of course, the net effect on food security may still be positive, given the potentially positive impact on food acquirement. But the point that needs to be emphasized here is that interventions aimed at increasing incomes may fail to yield the full potential in terms of household food security unless they include components that ease the time and health constraints faced by women.

The preceding discussion shows how the conceptual framework developed in Chapter 3 can help identify the pathways through which programme interventions might affect household food security. However, as mentioned earlier, it has not been possible in this study to investigate all the dimensions of food security or all the pathways that affect these dimensions. In particular, it has not been possible to collect systematic quantitative data on the utilization aspect of food security, whereas the information collected on the acquirement aspect permits at least a partial analysis of the impact of project intervention. For this purpose, the sample households have been classified into three groups: those in which there were no woman participating in the projects, those in which there were women who had been participating for less than a year and those in which there were women who had been participating for more than a year. Table 20 shows how the food security of these three groups of households changed during the five years preceding the survey.

As the table indicates, participating households did much better than non-participating households in terms of the improvement in food security. Thus, the level of food consumption improved in 54% to 64% of the participating households, compared with only 30% among the non-participating households. Similarly, the ability to cope during lean months and major crises improved in over 60% of the participating households, compared with less than 30% among the non-participating households.

It is interesting to note that, among the participants, those who had been involved in IFAD-assisted projects for less than a year seem to have done even better than those who had been involved for more than a year. This is somewhat counter-intuitive since, if involvement in these projects does, in fact, bring any benefit in terms of food security, it should do more for the relatively long-standing participants. There are a number of possible explanations for this apparent anomaly.

One possibility is that project effectiveness has improved over time, and consequently new participants are able to gain more. Alternatively, it is also possible that the projects have a tendency to lose vitality over time so that any gain made by longer-term participants during an initial stage gradually dissipates. Or it may simply be the case that, after serving the poorest groups during the initial stage, the projects are now bringing in relatively more well off clients. In other words, if new participants are seen to have done better in terms of improvements in HFS, it is simply because they happened to be more well off to begin with compared with the longer-term participants. Without further investigation, it is not possible to judge which, if any, of these alternative explanations is closer to the truth.

A multivariate analysis is required to capture the effect of project participation after isolating the effects of all possible intervening variables. This has been done in the Technical Appendix, where the length of project participation is used as one of several variables to explain changes in food security over time. After controlling for the effects of other variables, one finds through this analysis that (a) participants have improved their food security more than have non-participants and (b) among the participants those who have been involved longer have done better.

Project intervention has thus clearly had a beneficial effect on HFS. The remainder of this section attempts to identify the elements of intervention that have had a singular effect. Specifically, it investigates the importance of focusing on women. In Tamil Nadu and Nepal, the programmes have been targeted exclusively to women, and, although the general focus of the Andhra Pradesh project has been on the household as a whole, it possessed some women-oriented components, too. This report argues that the focus on women has made a significant contribution towards HFS.

The argument proceeds as follows. First, it is shown that there are two main channels through which project intervention has contributed to household food security: by increasing women’s access to resources and by increasing their control over decision-making within the household. The physical and financial resources provided by the projects have played a role in improving women’s access and control and, through them, in improving household food security. Finally, it is argued that the most crucial aspect of the interventions has been the effort that has been put into organizing women through the formation of women’s groups, because group mobilization has played a central role in improving women’s access and control.

Table 21 shows the effect of project participation on women’s access to resources and control over decision-making. Among the project participants, over 70% of women were able to increase their access to resources in the five years preceding the survey, while only 28% of the non-participants were able to do so. Likewise, among the participants, the control over decision-making increased for 64% to 84% of the women, depending on their length of participation. This compares to a mere 22% among the non-participants. Thus, participation in IFAD-assisted projects has evidently enabled women to improve both their access to resources and their control within the household. The consequence of this for food security is discussed below.

Women’s access to resources and household food security

Improvement in women’s access to resources can contribute to household food security in a variety of ways. Using the conceptual framework of Chapter 3, one can identify a number of pathways. First, greater access to productive resources such as credit, agricultural inputs and markets can improve food security by improving entitlement mapping. Second, greater access to resources such as fuel wood and water can help economize on women’s time, which can then be devoted to either income-earning activities (resulting in improvement in entitlement mapping and hence a higher level of food acquirement) or to more domestic care (resulting in better food utilization). Third, greater access to health-care facilities and greater access to fuel wood and water can raise the health status of women, and this, too, can contribute to household food security by enabling women to become more engaged in both income-earning and domestic care activities. Finally, to the extent that greater access to resources removes some of the socioeconomic constraints on women and thereby enables them to gain greater control over decision-making within the household, it should contribute to HFS, too.

The investigation of the relationship between women’s access to resources and HFS has involved the classification of the sample households into two groups: those in which women’s overall access to resources increased during the five years preceding the survey and those in which women’s access diminished or remained unchanged. The food security situations of these two groups are compared in Tables 22 and 23.

Table 22 shows that the group in which women’s access to resources increased did better in improving their access to cereals, as measured by the number of months during which cereal consumption was adequate. The number went up by 1.1 month for this group, from 6.9 months five years earlier to 7.9 months at the time of the survey. For the other group, in contrast, the number of months during which cereal consumption was adequate declined by about two weeks.

The former group also did better in terms of overall food consumption (both cereals and non-cereals). As can be seen in Table 23, the improvement in the level of food consumption was twice as high in the households where women’s access to resources had increased compared with the other group of households. The ability to cope during lean months and major food crises also improved more among the former group of households than among the latter. These results show that women’s greater access to resources does have a strong positive relationship with HFS in all its dimensions.

The same conclusion holds when women’s access to resources is defined somewhat narrowly to include only the assets owned by them. The sample households have been classified into two groups depending on whether or not women’s ownership of assets increased during the five years preceding the survey, and Table 24 compares how the food security of these two groups of households changed during this period. As the table shows, the level of food consumption improved much more among the households in which women’s assets increased than it did among those in which women’s assets declined or remained unchanged. The ability to cope during lean months also improved much more significantly among the former group of households.

Thus, whether resources are defined narrowly to include only the assets owned by women or more broadly to include all kinds of resources, both private and public, it is evident that greater access to resources on the part of women is strongly associated with more household food security. But how exactly does this improvement come about? Some light is thrown on this question by the respondents’ own perceptions about the most important factors contributing to their improved access to resources (Table 25). By far the most important factor turns out to be the strength of organization promoted by the projects and by NGOs. This factor alone accounted for more than half of all responses, whereas a general improvement in the economic environment accounted for only about one fifth. This indicates that an improvement in access, especially women’s access to resources, is not simply a matter of providing services by government or non-governmental agencies. It is only when women are sufficiently organized that they have the necessary confidence to come forward and claim their right to these resources.

This point is seen clearly from the pattern of interregional variation. Greater organizational strength was the most common reason cited among the women in Tamil Nadu and Nepal for improved access to resources, but not so in Andhra Pradesh (Table 25). It is no coincidence that women’s access to resources showed less improvement in Andhra Pradesh than it did in the other two regions: only 50% of the women in Andhra Pradesh reported improved access, as against 65% in Nepal and 93% in Tamil Nadu. Thus, poorer organization clearly led to poorer access to resources among women in Andhra Pradesh. These differences must account at least in part for the pattern noted in Chapter 3, that food security improved the most in Tamil Nadu and the least in Andhra Pradesh.

Women’s control over decision-making and household food security

The significance for household food security of women’s control over decision-making is discussed in Chapter 2. It is noted that, at any given income level, a household is likely to enjoy more food security if it is the women who exercise greater control over decision-making, especially over the decisions involving the disposal of household income. This is because women are likely to spend a greater proportion of household income on food compared with men. The evidence gathered by this study from the surveys in India and Nepal testifies to the importance of women’s control over decision-making.

The respondents were asked how their overall control over decision-making within the household had changed during the preceding five years. Depending on the responses, the households were classified into two groups: those in which women’s control had increased and those in which women’s control had decreased or remained unchanged. The food security situations of these two groups are compared in Tables 26 and 27.

Table 26 shows that, in terms of the adequacy of cereal consumption, food security improved more among the group in which women’s control had increased. Five years previous to the survey, the two groups differed little in terms of the number of months during which household members ate sufficient cereal. But over the five-year period, the number increased (by 1.1 months) for the group in which women’s control had increased and declined (by 0.6 months) for the group in which women’s control had declined.

The former group also did better in terms of overall food consumption (both cereals and non-cereals). As Table 27 shows, the level of food consumption improved in two thirds of the households in which women’s control increased, as compared with only one third of the households in the other group.

A similar relationship is observed in respect of the ability of households to cope with shocks. As can be seen in Table 26, the ability to cope with both lean months and major food crises improved much more among the households in which women’s control had increased, and deteriorated much more among the households in which women’s control had decreased or remained unchanged.

Thus, both the level of annual food consumption and the ability to cope with fluctuations improved more among the households in which women had gained greater control over the decision-making process. From the point of view of policy, it is important to ask about the factors that enabled these women to gain greater control. The answer, in terms of the women’s own perceptions, is given in Table 28.

As in the case of women’s access to resources, the power of organization again comes to the fore as the primary explanatory factor, accounting for nearly half of all responses in the three survey areas combined. The next most important factor revolves around the fact that in many households, women had become the de jure heads of household in the absence of husbands. Excluding these special cases where control had been gained by default and considering only those households in which both wife and husband were present, organization turns out to be by far the most important factor behind the improvement in women’s control within the household.

It is often argued that involvement in income-earning activities is an important determinant of women’s status within the family. In terms of the so-called bargaining models of household behaviour, the argument goes as follows:¹⁴ Men and women have different preferences (e.g., women want to spend more on food), and they each try to mould household decisions according to their own preferences. The conflict of preference is resolved through an implicit bargaining process. Whether the final solution will be closer to men’s or women’s preferences depends on their respective controls over decision-making, which in turn depends on their respective bargaining strengths. One of the determinants of bargaining strength is each spouse’s fall-back position, i.e., the position each can achieve on his/her own in the event of a breakdown in the relationship. Women’s involvement in income-earning activities is supposed to improve their fall-back position, which in turn should improve their bargaining strength and control over decision-making within the family.

However, it can be seen from Table 28 that involvement in income-earning activities accounted for only 10% of the responses given as an explanation of the improvement in women’s control, despite the fact that the vast majority of the women in the sample were involved in income-earning activities, thanks largely to project interventions. Obviously, mere involvement in income-earning activities is not enough to raise women’s status within the family or to improve their control over decision-making. This is not to suggest that income-earning activities have no role to play here, but clearly something more is needed to overcome the cultural norms of a patriarchal society if women are to gain greater control over decision-making. It is in recognition of this point that Sen (1990) has argued, in an extension of the bargaining model of the household, that the perception of women is what matters most. It is only when women are able to perceive their self-interest and become conscious of their real contribution to the family that they will be able to exercise greater control.

Mobilizing the collective strength of women and raising their consciousness through the formation of women’s organizations – be it in the guise of a credit society, or a savings group, or some other functional grouping – may be just the catalyst that is needed for this purpose. This is the reason that it is the benefit of organization, rather than involvement in income-earning activities as such, that women cite as the most common explanation of greater control over decision-making.

Once again, the pattern of regional variation confirms the importance of organization (Table 28). Andhra Pradesh lags far behind the other two regions in terms of the importance of organization in determining women’s control; correspondingly, in Andhra Pradesh women have gained the least in terms of control over decision-making.

Reasons for change in coping ability as perceived by households

It is argued above that project interventions have improved food security by improving women’s access to resources and their control over decision-making. This has been inferred by comparing the food security status of women in terms of improvements in their access to and control over resources. Furthermore, since organization has been found to be a main factor behind improvements in access and control, it is inferred here that organization is also a key to food security. These indirect inferences are confirmed through an analysis of what women themselves perceive to have been the most important factors behind improvements in their food security.

The respondents who claimed to have acquired greater ability to cope during lean seasons were asked to identify the major contributory factors to this outcome. Among the answers they gave, three sets of factors stand out: (i) greater command over savings and assets, which accounts for 32% of all responses, (ii) easier access to credit, which accounts for 28% of the responses and (iii) higher demand for wage labour, which accounted for another 24% (Table 29). Of these, the first two represent to a large extent the contribution of project interventions. By providing poor rural women with access to credit, by encouraging group savings among women members and by enabling them to accumulate assets by opening up income-earning opportunities in the off-farm sector, these interventions have led to improvements in women’s access to resources and helped raise the status of women within the family. All this has resulted in a greater coping ability among poor rural households.

There are, however, some instructive interregional variations in this regard. The role of savings/assets and access to credit, the two principal mechanisms through which project interventions have contributed to coping ability, is much more pronounced in Tamil Nadu and Nepal than in Andhra Pradesh (Table 28). These two factors accounted for two thirds of all responses in Tamil Nadu and Nepal, but less than one third in Andhra Pradesh, where higher demand for wage labour was by far the most important contributory factor.

This is a reflection of the nature of the interventions. As mentioned earlier, in both Tamil Nadu and Nepal, the interventions are focused specifically on women, delivering credit to them for productive activities and encouraging them to form savings groups. By contrast, the major thrust in the Andhra Pradesh Tribal Development Project is on area development and on strengthening the agricultural base of households. Relatively less emphasis is placed on organizing women and on delivering services directly to them.

The fact that organization represents a powerful contribution to food security is apparently appreciated by the women themselves. Those women who claimed to have experienced improved coping ability during the five years preceding the survey were asked to give their opinion on whether the better organization of women had helped them to achieve this improvement. Over 80% of the respondents acknowledged the contribution of organization in Tamil Nadu and Nepal, but less than one third did so in Andhra Pradesh. It can hardly be a coincidence that food security also improved the least in Andhra Pradesh.¹⁵

These findings strongly underline the need to organize women as an integral part of any programme for improving household food security. This is not to demean the role of area development and agriculture-focused interventions of the kind adopted in Andhra Pradesh. Given the precarious nature of the shifting cultivation on which the tribals depend for their livelihood, these efforts have much to recommend them. This is especially so in view of the importance, as noted in Chapter 4, of maintaining the ability of poor households to acquire a certain minimum amount of food through subsistence production. But these efforts ought to be supplemented by strong organizational activities that are focused especially on women.

---

14/ Discussions of these models can be found in Manser & Brown (1980) and McElroy & Horney (1981).

15/ The inferior performance of Andhra Pradesh cannot be attributed to any area-specific geographical, economic or environmental disadvantage, or to any other non-project-related disadvantage, since the regional dummies that have been introduced in the regression on food security are statistically insignificant. (See the Technical Appendix.)