Enabling poor rural people
to overcome poverty



In a primitive economy, a household ensures its food security mainly through subsistence production. As the economy grows and markets develop for a variety of products, subsistence production is gradually replaced by production for the market. This tendency is further strengthened when an economy opens up to the outside world. If this happens at an advanced stage, when the population has already crossed the threshold of hunger, as has been the case in the Western world, the shift from subsistence food production to market production does not pose a serious problem to food security. In fact, it may even enrich the diet of the population by enabling it to obtain a wide variety of food from all over the world. But if market orientation occurs at an early stage, when a large section of the population has yet to secure access to sufficient food to guarantee a minimum required diet, questions are bound to arise regarding its impact on food security.

Questions have indeed arisen in recent years in the context of the macroeconomic reforms currently sweeping the Developing World. Markets are opening up both internally and externally, thus providing incentives to farmers to shift towards cash crops. Structural adjustment programmes are strengthening these incentives by making production for export more profitable than before. Partly as a result of these policy reforms and also because of increasing urbanization, agriculture can be expected to become increasingly diversified and commercialized in coming years.

These macro-level tendencies are being reinforced by certain micro-level interventions. For example, in areas where shifting cultivation comprises a sizeable segment of traditional agriculture, attempts are being made to encourage households to embrace settled agriculture. Concerns about environmental degradation underlie this effort as much as do considerations of greater economic returns. A typical example is offered by the Andhra Pradesh Tribal Development Project in India, in which IFAD is deeply involved. Through this project, tribals are being given an incentive to practise horticulture on mountain slopes where shifting cultivation was previously employed.

Even interventions that are not directly related to agriculture can have the effect of reducing subsistence production. A case in point is offered by credit programmes of the kind being supported by IFAD and other agencies. By encouraging households to take up self-employment activities in the off-farm sector, such programmes may well induce these households to spend less time on subsistence production.

The consequences of market orientation: a priori arguments

Is this shift a matter of concern from the point of view of HFS? The question is of considerable practical importance since the agencies responsible for policy-making ought to be aware of how their actions are likely to affect HFS to the extent that they induce a shift towards market-oriented production among the poorest households. On purely a priori grounds, the issue is a fairly contentious one, as plausible arguments can be advanced on both sides of the debate.10 In terms of the conceptual framework discussed in the preceding chapter, the arguments revolve around two important determinants of food acquirement: entitlement mapping and women’s control. What impact does market integration have on the entitlement mapping of the poor, that is, what happens to the productivity, rate of return and purchasing power of the assets of the poor, including the labour power of poor households? What is the impact on women’s control over household resources? These are some of the questions that arise in such circumstances.

On the positive side, the following arguments are usually cited:

(a) Insofar as the shift towards the market occurs in the first place because such a choice entails higher rates of return on land and labour, it will increase household income, and this will, in turn, improve HFS.

(b) Experience shows that the benefits of the improved technology, inputs and infrastructure that are usually pressed into service for the sake of cash crop production can rub off on the production of food crops as well, so that total food production may not decline even if land and labour inputs are shifted away from it. (See Kennedy & Cogill 1987; von Braun, Hotchkiss & Immink 1989; Bouis & Haddad 1990.)

(c) Even if food production declines, increased export earnings from cash crops should enable a country or region to import enough food to maintain the overall level of food availability.

(d) Insofar as the shift is partial, that is, subsistence production is not given up entirely, market orientation will contribute towards a diversification in livelihoods and thereby improve the ability of households to cope with lean times and occasional crises.

But there are some weighty arguments on the other side of the debate as well.

(a) In the first place, whatever the gain may be for the producers of cash crops, the landless and near-landless people who have to depend on the market for their food may suffer from reduced availability and higher prices. The argument that higher export earnings will enable a country to maintain the level of food supply through food imports can well become cold comfort for these people, because they may not have the purchasing power to induce the market to import basic foodstuffs rather than the relative luxuries that cash crop producers and others may desire. Thus, unless the commercialization of agriculture is accompanied by public intervention to modify the market supply of food (something that is not looked upon particularly favourably in the new economic environment), there is a distinct possibility that some land-poor households might lose out at least in the short run.

(b) Secondly, the shift from subsistence to commercial crops sometimes brings with it a reduction in household income. This can happen if women are relatively more involved in subsistence production and men are more involved with cash crops, as is the case in many parts of Africa, or if women lose their title to land in the process of the conversion of land from traditional food crops to modern cash crops. If this happens, then household food security may decline in spite of a rise in income.

(c) Finally, households shifting towards the market in pursuit of higher income might end up exposing themselves even more to sudden shocks to food security because of uncertainties in the market.

Against the final possibility c mentioned above, it might be argued that, if rational agents are consciously shifting away from subsistence production without any external coercion and if they value their food security highly (which they presumably do), then there is no need for concern. In this case, these agents will ensure that their own food security is not compromised. But this argument assumes that households have perfect information and perfect foresight about future changes in the economic environment. This is untrue. In fact, precisely because of the imprecision of information and because of the uncertainty about the future, one rarely finds poor households shifting entirely to market-oriented production. Even the most commercialized households tend to maintain a certain amount of subsistence production as a hedge against uncertainty. If the agents are rational, then the extent of hedging would be optimally based on the information they have and the constraints they face. But optimality in this sense does not preclude the possibility that food security might degenerate; it only means that things might have been worse still. Rationality and free choice are therefore no guarantee against diminished food security.

  Evidence of the impact of market integration on food security

Evidence from the survey areas in India and Nepal demonstrates that such a fear is by no means unfounded. In order to investigate the relationship between market orientation and food security, the analysis involved the classification of households into two groups: those whose market dependence had increased in the five years preceding the survey and those whose dependence on the market had diminished during that time. It is thus possible to note which group performed more effectively in improving their food security during the five years preceding the survey.

Market integration has been measured from the viewpoint of both consumption and production. On the consumption side, the households were asked two questions: (i) in the preceding five years, what had happened to the share of self-provisioning (that is, their own production as opposed to market purchases) of cereal in their total cereal consumption, and (ii) what had happened to the share of self-provisioning in total food consumption, including that of cereals and non-cereals. Those who reported a reduced share of self-provisioning were classified as having become more market dependent.

As Table 7 indicates, some 28% of the households in the combined sample reported increased market dependence in respect to cereals. Almost an equal percentage reported a decline, and over 40% reported no change. With respect to overall food consumption,:and only 20% of households reported a decline.

There appears to be a strong negative relationship between the food security of a household and the dependence of the household on the market for food. First of all, consider the level of food consumption. As Table 8 shows, improvement in the level of food consumption was more common among households that had become less market dependent in the five years preceding the survey than it was among those that had become more market dependent. In the former group, 71% of households reported improvement in food consumption, while only 46% of the latter did so. Correspondingly, a decline in food consumption was experienced more often (25%) in households where market dependence had increased than it was in those where market dependence had diminished (only 18%). Thus, while reduced market dependence was more commonly associated with increased food consumption, increased dependence was more liable to be associated with a decline in consumption.

Greater dependence on the market also seems to be related negatively to the ability to cope with fluctuations in access to food. Among the households where market dependence had gone down in the five years preceding the survey, as many as 82% reported improved ability to cope with lean months, whereas, among the households where market dependence had gone up, only 52% so reported (Table 9). On the other hand, 43% of the households where market dependence had gone up reported deterioration in their ability to cope with lean months; this compared with only 11% of the households where market dependence had gone down. Thus, while a reduced market dependence was more commonly associated with an increased ability to cope with fluctuations in food consumption, an increased dependence was more liable to be associated with a reduced ability to cope.11

It is thus evident that households that had become more market dependent for acquiring food tended to be less well off in terms of food security than were those that had achieved food self-sufficiency. While this is a significant finding, it does not necessarily suggest that market orientation in livelihood is intrinsically damaging for a household’s food security. A negative association between market dependence and food security may well arise from a variety of extraneous causes. For instance, if households become more market dependent as a result of losing endowments such as land, labour and livestock (all of which are essential for maintaining subsistence production), they are very likely to suffer from greater food insecurity. This would be a consequence of endowment loss rather than of market orientation as such. In terms of the conceptual framework developed in the preceding section, the crucial issue here is the distinction between entitlement mapping and endowments as determinants of food security. Market orientation may have a favourable impact on entitlement mapping and, through it, on food security, but turn out to have a negative relationship with food security because of an endowment loss that might have forced increased market orientation in the first place.

Therefore, in an assessment of the impact of market orientation on food security, it may be misleading to focus exclusively on a household’s market dependence in terms of consumption. It is necessary to consider also the household production structure and the circumstances leading to market orientation on the production side.

Market orientation: cause and effect

First, consider the relationship between subsistence production and food security. Table 10 shows that households whose production of subsistence crops had gone up in the five years preceding the survey also managed to improve their level of consumption more than did the rest. They were likewise more successful in coping with fluctuations (Table 11).

This does not mean, however, that the increased production of subsistence crops, as opposed to cash crops, is essential for improving food security. As Tables 12 and 13 show, households that increased their production of cash crops were more successful in achieving improved food security – in terms of the level of consumption and the ability to cope with fluctuations – than were those where the production of cash crops had declined or remained unchanged.

Thus, increased production of either kind of crop – subsistence or cash – was generally associated with a higher degree of food security. This is not really surprising since higher income should improve a household’s ability to acquire food no matter whether the income is derived from subsistence or cash crops. But the question still remains: relatively speaking, which is the more secure means of improving food security: subsistence income or market-based income? In other words, while both sources of income may be helpful for food security, it is still useful to know if one source is superior to the other in terms of improving entitlement mapping.

Table 14 throws some light on this issue. The table considers only those households for which per capita income went up in the five years preceding the survey, and they are classified into three categories depending on which source of income increased the most: subsistence crop, other types of self-employment, or wage labour. Among those for which subsistence income increased the most, 54% indicated improved access to food, compared with 66% among those for which income from other self-employment activities rose more. Thus, higher income from market-oriented sources was, if anything, more strongly associated with improved access to food than was higher subsistence income.

It is also worth noting that among the households for which subsistence income increased the most, a much higher proportion (46%) saw their access to food decline over the five-year period compared with those for which income from other self-employment activities increased the most (18%). Thus, in considering both improvement and deterioration in access to food, one must conclude that access was more secure when the increased income was derived from market-based self-employment rather than from subsistence production.12

It is, however, important to note an asymmetry here between rising and falling income. Table 14 considers only those households in which income increased during the five years preceding the survey. Table 15, by contrast, considers only those households in which income decreased. These households are classified according to which source of income declined the most. The table shows that those households where subsistence income declined the most also suffered more in terms of food insecurity. By contrast, among the households where income from other types of self-employment declined the most, more than half actually managed to improve their food consumption, while less than one third found their consumption on the decline. Loss of subsistence income was thus far more damaging to food security than was loss of income from market-based self-employment.

The same conclusion is reached by considering a household’s ability to cope with fluctuations. As Table 16 indicates, the ability to cope through lean months showed a stronger decline among households in which the income from subsistence crops dropped the most than among households in which income from other types of self-employment dropped. Nearly two thirds of the households in the former category reported a reduced ability to cope through lean months as against only one quarter of the households in the latter category.

So the asymmetry can be summed up as follows: Almost an equal proportion – one third – of these households found that their level of food consumption had increased, diminished or remained unchanged, respectively.

In order to gain further insight into the importance of subsistence income on the ‘down’ side, it is necessary to consider the forces that are responsible for reducing subsistence income. Two kinds of forces need to be distinguished here. They may be referred to as ‘push’ forces and ‘pull’ forces. Pull forces are those that divert household resources from subsistence production to potentially more attractive market-oriented activities. Push forces operate when the loss of resources (such as land, labour and capital) compel households to cut down on subsistence activities. These two forces must be distinguished because the loss of subsistence income is arguably more likely to entail losses in food security when it is caused by push forces rather than by pull forces. Table 17 confirms this hypothesis. It shows that losses in food consumption were more than twice as prevalent among households that had lost subsistence income as a result of push forces than among those that had lost subsistence income as a result of pull forces.

One final issue remains to be considered: why does market income turn out to be more favourable to food security on the ‘up’ side? One possible reason is that market integration offers an opportunity for diversifying the livelihood structure. As argued before, a diversified source of livelihood can improve food security through two favourable influences on entitlement mapping: (i) it can raise the rates of return to family labour (through the specialization of household resources according to comparative advantage) and thereby improve the level of food consumption, and (ii) it can render some stability to entitlement mapping since the risk of entitlement loss is spread over more than one source of income, thereby improving the ability to cope with lean seasons and other shocks. To the extent that the market acts as a vehicle for diversifying a household’s livelihood structure, market integration may thus have a beneficial effect on food security.

This hypothesis has been tested in the present study through the construction of a diversification index to measure the extent to which a household derives its food from diverse sources. The index takes on values from 0 to 1, a higher value indicating greater diversification.13 The households are then classified into three categories of diversification: ‘high’ for those for which the index value is above 0.5, ‘medium’ for those for which the index is between 0.25 and 0.5, and ‘low’ for those for which the index is below 0.25. The food security situation of these three groups is compared in Tables 18 and 19.

As Table 18 shows, households in the medium and high diversification categories were more successful in improving their level of food consumption during the five years preceding the survey than were those in the low diversification category. The difference is particularly striking with regard to the ability to cope with fluctuations (Table 19). Nearly three quarters of the households in the medium and high diversification categories experienced improved ability to cope through lean months, while only slightly over half the households in the low category did so. On the other hand, 38% of households in the low category experienced a reduced ability to cope with lean months, while only 21-24% in the medium and high categories experienced such a reduction. Thus, greater diversification was commonly associated with an increased ability to cope with fluctuations, while lower diversification was more liable to be associated with a reduced ability to do so. It would therefore appear that increased market orientation fosters an improvement in food security on the up side, mainly by helping to diversify the livelihood structure. Presumably, greater diversification helps to improve both the level and the stability of entitlement mapping.

The preceding discussion is based on bivariate relationships between food security on the one hand and some aspect of market integration, such as diversification or loss of subsistence income, on the other. Bivariate relationships can provide useful insights, but they cannot be fully trusted in a causal analysis because of the possible influence of other factors. For instance, from the mere fact that higher food security has been associated with greater diversification, one cannot conclude that diversification has actually caused an improvement in food security. There may not be any causal relationship between the two, though a positive association may be observed because some third factor – say, participation in IFAD-assisted projects – may have contributed positively to both diversification and food security. Therefore, in order to make causal inferences, one must isolate the effects of all possible variables that may have an influence on food security. This has been done in the Technical Appendix with the help of multiple regression analysis.

Through this analysis one finds that, even after isolating the effects of other variables, the diversification of livelihood structure does have a positive effect on both the level of food consumption and the ability to cope with bad times. One also finds that the ability to cope with bad times tends to suffer when income is lost mainly in subsistence production. Furthermore, when the loss of subsistence occurs mainly as a result of push forces, the level of food consumption also tends to fall.

Thus, the evidence confirms the point made earlier that increased market orientation can have two opposing effects on HFS. By allowing increased diversification in the livelihood structure, it improves both the level of food consumption in normal times and the ability to cope during bad times. However, if market orientation is accompanied by a big fall in subsistence production, caused especially by push factors, it can have a deleterious effect on food security. Policies that are likely to increase the market orientation of livelihood structures ought to take these two food security implications into account.


10/ For helpful overviews of this debate, see von Braun, Kennedy & Bouis (1989) and Binswanger & von Braun (1991).

11/ A similar relationship exists between market dependence and the reported ability to cope with major food crises.

12/ The least favourable source of income in this respect was income from wage labour.

13/ Details on how this index has been constructed are given in the Technical Appendix.