Social Security in South Asia
Social Security in South Asia
Abstract and Keywords
This chapter evaluates the anti-poverty measures in different parts of South Asia. It classifies these measures under the following four headings: security through control over land, security through self-employment, security through wage employment, and security through public provision of basic needs. The chapter explains that the objective is to analyse the experience in these areas so as to learn something about the possibilities and limitations of offering social security through these channels. Furthermore, instead of covering the South Asian region as a whole, it looks at only three countries, namely India, Bangladesh, and Sri Lanka.
Social security in the Western sense occupies a marginal place in the social and economic policies of the South Asian countries. The reasons for this are not far to seek. In the first place, there is the question of scale. If income support of the kind given through the Western social-security system were to be given to over half the population rather than to 10 to 15 per cent of it as in the West, it would require an incomparably higher level of fiscal commitment in relation to resources. Secondly, given the structure of these economies, some of the conventional social-security measures may not even be operationally feasible. For example, where the labour-force is still predominantly self-employed and where unemployment is manifest mainly in the form of underemployment, it is not easy to devise a scheme of unemployment benefit; and where the proneness to falling sick is endemic due to poor food and hygiene, the idea of a sickness insurance scheme may turn out to be an actuarial disaster.
There is no doubt, however, that the millions of poor in South Asia need social security of some sort. If is often pointed out that one particularly venomous instance of insecurity, namely famine, has been largely conquered in South Asia in recent times—the sole exception being the Bangladesh famine of 1974. There is none the less an enormous problem of endemic insecurity that still remains. The problem is not merely that a vast number of people continue to live below some norm of poverty line; more disturbingly, many of those who live on the edge of subsistence are in constant danger of falling further into a state of destitution. While the eradication of poverty may have to be a long-term goal, depending as much on overall growth as on direct action targeted to the poor, there is certainly an immediate need for a safety net for those who do not even have the security of hanging on to a subsistence level of living.
What is not so clear, however, is which kinds of security measures are best suited for this purpose.1 It is perhaps wise to consider an inclusive set of policies to begin with, gradually discarding those which are found wanting on (p.306) closer scrutiny. Keeping in mind the nature of South Asian economies and the experience of various anti-poverty programmes pursued in this region, we present below a tentative classification as a starting point. One may think of three broad categories: (1) policies relating to land: these include land reforms (both ownership and tenancy reforms), assistance for small and marginal farmers, crop insurance, and so on; (2) policies relating to employment: these may include creating opportunities for either self-employment or employment through co-operatives in non-agricultural activities, providing wage employment through public works, and reserving employment for members of the disadvantaged groups in salaried employment; and (3) direct public provision: this may include both income support of various kinds as well as public provision of basic needs such as food, housing, health, and so on.
All these measures have been tried to varying extents in different parts of South Asia. We cannot, however, attempt a comprehensive evaluation of all of them within the space of a single chapter. Instead we pick out a few of them for close scrutiny and classify them under the following four headings: security through control over land, security through self-employment, security through wage employment, and security through public provision of basic needs.2 The objective is to analyse the experience in these areas so as to learn something about the possibilities and limitations of offering social security through these channels. Furthermore, instead of covering the South Asian region as a whole, we shall look at only three countries, namely India, Bangladesh, and Sri Lanka.
2. Security Through Control Over Land
In the predominantly agrarian economies of South Asia, security of livelihood for a majority of people clearly depends on control over land and its produce. This naturally accords a position of pre-eminence to the policies relating to land control in any discussion of social security in this region. Three types of policies are going to be discussed in this context, namely, ceiling-cum-redistribution policy, tenancy reform, and the alienation of State-owned land.
The ceiling-cum-redistribution policy is by far the most radical in nature, and also the one that has been least successful in practice. Attempts have been made in all three countries to acquire the excess of land beyond a certain ceiling from large landowners and distribute it to the land-poor people.
By the mid-1980s about 1.5 per cent of the cultivated land in India had been acquired under ceiling laws passed in the 1950s, and less than 80 per cent of it (p.307) was actually distributed.3 In the process, nearly 3.4 million persons received on an average a little over 1.3 acres of land. This has no doubt enhanced the security of livelihood for the beneficiary families, especially since more than half of the beneficiaries belong to the scheduled castes and scheduled tribes who are among the most disadvantaged social groups in India (Bandyopadhyay 1986). However, since the amount of land available for distribution was itself so small, the total number of beneficiaries amounts to a minuscule proportion of land-poor households. The record of Bangladesh is even more dismal. The excess of land over the stipulated ceiling would have amounted to no more than 1 per cent of the cultivated land even if the ceiling were strictly enforced. In the event, only 15 per cent of the potential surplus has been acquired and out of the acquired land no more than a quarter has been distributed (Siddiqui 1979).
Sri Lanka’s achievement appears far more remarkable at first sight—as much as 20 per cent of cultivated land was promptly acquired following the legislations of 1972 and 1975. But the landless and small peasants gained very little; only 12 per cent of the land acquired, which amounts to 2.4 per cent of cultivated land, accrued to the peasantry (Wickramasekara 1985 a). The reason for this dissonance between acquisition and redistribution lies in the special nature of land reform enacted in Sri Lanka. The reform was aimed mainly at the plantation sector: 60 per cent of the area planted with tea, 30 per cent of the area planted with rubber and 10 per cent of the area planted with coconut in the plantation sector was taken over by the State, but only just over 1 per cent of paddy land was acquired in the process. A conscious decision was also taken at the time not to distribute plantation land to the peasant sector on the ground of economies of scale. Whatever the merit of this decision, there was clearly an element of irony in this. For since the very early days of the nationalist movement in Sri Lanka, political agitation against the plantation sector has been conducted in the name of the peasantry, especially the peasants of the Kandyan highlands where the plantation estates were mostly situated.4 Perhaps in deference to history, a small amount of estate land was distributed to the Kandyan peasants; but the bulk of it came under State-run corporations which did little to enhance the control of the land-poor over the fruits of the land. Initially an attempt was made to combine peasant control with economies of scale by transferring a part of the acquired land to farmers’ collectives, but the collectivist institutions were abolished by the government which came to power in 1977.
The general picture all over the region is, thus, one of negligible impact as a (p.308) result of redistributive land reform. The proximate reason for this failure lies in the stipulation of an exceedingly high ceiling—so high in Bangladesh (33 acres) and Sri Lanka (25 acres for paddy land) that the proportion of land above these limits is too miniscule even to appear as a separate category in any official statistics on the distribution of land. Even the meagre amount of land that should have been legally available could not be fully acquired as the landowners made use of various legal loopholes to keep possession of their land.
Political imperatives have no doubt played a paramount role in leaving the land of large landowners (outside the plantation sector) virtually untouched in all three countries.5 It is therefore of some interest to note what has happened in two isolated pockets in the region, namely the states of West Bengal and Kerala in India, where the polity has been dominated for a considerable stretch of time by parties with avowedly radical intents.6 Although constrained by the Constitution of India, the State governments were not entirely powerless to implement radical land reforms if they intended to do so, because considerable autonomy was given to the States in matters relating to land reform. As we shall see, both States did in fact take advantage of this autonomy to push through tenancy reforms which were much more radical than anything attempted elsewhere in South Asia; but when it came to redistribution of land through ceiling legislations, both failed like the rest. The ceiling was fixed at 20 acres in Kerala and 25 acres in West Bengal, which yielded a surplus of less than 2.5 per cent of the cultivated land in Kerala (Raj and Tharakan 1981:17) and just over 1 per cent in West Bengal (computed from the figures given in Bandyopadhyay et al 1985: 14).
Ceiling legislations have thus achieved very little in any part of South Asia. It has been claimed, however, that considerably more progress has been made, at least in some parts of the region, in enhancing the security of a certain class of peasants, namely, the tenant-cultivators who lease in land from others. The terms and conditions of lease contracts are known to be highly unfavourable to the tenants; under share-cropping—the most predominant form of tenancy all over South Asia—the tenants usually supply most of the input and retain only half the produce. Moreover, there is no security of even the meagre livelihood eked out in the process as the landlords can evict tenants at any time at their will. The tenants also tend to belong mostly in the smaller landholding groups. Consequently, along with landless labourers, share-croppers are found to be the most deprived section in rural South Asia.
There is much uncertainty about the precise number of the share-croppers in this region, but even the rough estimates are adequate to reveal the (p.309) enormous extent of this practice. Official estimates based on agricultural censuses show that tenants (including owner-cum-tenants) constituted some 35 to 40 per cent of all farmers in Bangladesh in the late 1970s and 53 per cent in Sri Lanka in the early 1960s. In India the extent of tenancy varies widely across the States; we shall consider here only the States of West Bengal and Kerala where tenancy reforms are claimed to have had the most pronounced effect. Tenants (including owner-cum-tenants) constituted 40 per cent of all farmers in West Bengal according to the National Sample Survey of 1971/2, and 52 per cent of all farmers in Kerala according to the Land Reform Survey of 1966/7.7Noting that micro-level surveys reveal that official statistics almost invariably underestimate the extent of tenancy, we may describe the overall picture as one where something like half of the farmers in South Asia cultivate some amount of rented land. Any improvement that can be brought about in the terms of their contracts can thus have a potentially significant effect on the security of livelihood for a large segment of the rural population, provided the same changes do not undermine their access to land.
The issue of tenancy, however, never received any prominence in the political movements of pre-independence Sri Lanka. It began to be discussed more seriously after independence in 1948, but even then the prevailing view did not seem to consider share-cropping to be a particularly exploitative institution. In fact, the majority of social scientists tended to regard it either as a mutually beneficial contract between parties who are socially equal in the village hierarchy but who differ in their relative endowments of land and labour, or a benevolent gesture on the part of the well-off families to help out their less fortunate friends and relatives.8 In so far as the institution was criticized at all, it was mainly because of a presumed disincentive effect on agricultural production. This effect was never quite documented, but the theoretical arguments closely resembled the familiar Marshallian argument about the inefficiency of share-cropping .
It was, however, an entirely different kind of reasoning that eventually led to the first serious attempt at tenancy reform in Sri Lanka. The Marxist Minister for Agriculture, Philip Gunawardena, who was instrumental in attempting this reform through the Paddy Lands Act in 1958, was of the view that peasant production in general, including both tenant and owner cultivation, suffered from the diseconomies of small scale and was hence a drag on production. His long-term objective was therefore to replace peasant production by collective farming; but as a first step he felt it necessary to liberate the tenants from the clutches of ‘feudal’ landlords so that a community of equal peasants could (p.310) eventually make the transition to a collective form of agriculture. With this end in view, sweeping provisions were made for enhancing tenants’ control over land. Evictions were prohibited except under special circumstances, hereditary rights of cultivation were conferred on tenants, and the traditional 50 per cent share of the landlord was reduced to 25 per cent or a fixed amount per acre (allowing for regional variation) whichever was less. More importantly, Cultivation Committees were to be formed in each village, excluding the landlords, initially to ensure that the tenants’ rights were established and eventually to pave the way for collective agriculture. These Committees were also supposed to provide credit and other input facilities for the tenants so that they would not have to turn to the rural elite for help.9
In the event, not even the immediate goal was achieved, not to speak of the distant one. Under intense political pressure, the minister was forced to resign, landlords had to be allowed to join the Cultivation Committees, and conditions were relaxed under which the landlords could resume cultivation. As a result, a spate of evictions was noted throughout the 1960s and the rental share remained virtually unchanged in most parts of the country. Significantly, the only region where the rental share showed some improvement was the Hambantota district, which has long been a centre of radical politics and where, unlike in other parts of Sri Lanka, absentee landlordism was the order of the day. It has been rightly argued that improvement was possible in this region only because of the special circumstance in which the landlords, being absent, were unable to control the Cultivation Committees (Peiris 1976).
In terms of political preconditions for tenancy reform, the situation in pre-independence Bengal (comprising both Bangladesh and West Bengal) was vastly different from that of Sri Lanka. Share-cropping was widely seen here as an exploitative institution. In fact the grievances of share-croppers were made a central pillar of political agitation by the peasant leaders of Marxist persuasion, which led to the famous Tebhaga movement of 1946/7. The movement sought to improve tenurial security and to ensure a two-thirds shared instead of the traditional half in favour of the share-cropper. A Bill was also introduced by the government of Bengal in early 1947 recognizing these demands. But then came independence, and the issue of share-croppers has been shelved ever since in the territory now comprising Bangladesh. A Tenancy Act was indeed passed in the early 1950s, but it was aimed solely at giving relief to the owners of land from the exploitation of rent-receiving intermediaries. As every student of Indian history knows, these intermediaries were the zamindars who were empowered by the colonial government to collect land revenue from the peasants in return for the obligation of handing over a fixed sum to the government. As proteges of the colonial rulers, this class of intermediaries was an immediate target of nationalist attack. With the advent of independence, it (p.311) was politically both expedient and feasible to isolate them and dispossess them of their rent-receiving rights. This obviously eliminated a long-standing source of exploitation of the peasantry, but only to the benefit of those who had legal title to the land, regardless of whether they cultivated or not. The sharecroppers found no relief, as they could not claim ownership rights on leased-in land. Not until 1983 did a Lands Reform Committee deal with this issue for the first time and recommend more security and a higher share of the sharecroppers, but the recommendations still remain only on paper .
While inheriting the same history, West Bengal, however, took a different course. The rent-receiving rights of zamindars were promptly taken away, as in Bangladesh, but the rights of share-croppers were also recognized at the same time. They were not given ownership rights, but eviction was made legally difficult and their share was raised to 60 per cent in the 1950s and then to 75 per cent in the early 1970s.10 Similar laws were enacted also in many other States of India, but everywhere in the 1950s and 1960s they not only failed to improve the tenants’ control over land, but in fact made their condition worse by prompting large-scale eviction as in Sri Lanka.11
Lack of ‘political will’ and ‘bureaucratic lethargy’ have often been blamed for this reverse. But it is being increasingly recognized that the root of failure lay in a basic flaw in the very conception of these reforms. It was assumed by the reformers that the mere existence of a law would encourage tenants to confront landlords with their legitimate rights. This assumption was, however, based on a profound misunderstanding of the socio-economic relationship between the share-croppers and the landlords. When we speak of this relationship as an unfavourable one for the share-croppers, this is mainly in relation to an alternative social scenario where the land would be owned by the sharecroppers themselves. But given the existing ownership rights and the acute land-hunger on the part of the landless and marginal peasants, those who are able to lease in land have reasons to consider themselves privileged compared to their peers. The purely economic contract is enmeshed in a complex web of socio-cultural relationship in which the landlords often provide ‘subsistence insurance’ and other material support to their tenants in return for their acquiescence in the prevailing social hierarchy. In a situation where the majority live a precarious existence owing to adverse man-land ratios, few opportunities of alternative employment, and the vagaries of weather, and where the provision of State-sponsored social security is all but non-existent, a high value is naturally attached to the ‘subsistence insurance’ mediated through traditional owner-tenant relationships.
The reforms called upon the tenants to strike at the roots of these relationships; (p.312) any attempt to alter the prevailing hiearchy of rights in their favour was bound to do so. But since any such attempt would also put their ‘subsistence insurance’ in jeopardy, the tenants can only be expected to rise against the landlords if they can be assured of àn alternative source of security. Two conditions must be fulfilled for this to be possible. First, the tenants must feel assured that their action against the landlords will not boomerang by prompting eviction. The mere existence of law is not enough; the tenants must have countervailing political power at the local level to neutralize the pre-existing superior power of the landlords. Secondly, mere security against eviction will also not be enough; this will only give them access to land They will also need access to new sources of consumption loan, working-capital loan, distress relief, and other elements of subsistence insurance which will no longer be forthcoming from the traditional channel.
It was precisely the non-fulfilment of these two conditions that led to the universal failure of tenancy reform in India. Interestingly, it was also precisely these two conditions that Philip Gunawardena had sought to fulfil through the institution of Cultivation Committees in Sri Lanka. But he failed because the ruling coalition, in which his breakaway Marxist party was only a minor partner, did not share his commitment. As a result, the Cultivation Committees could not find either the political power or the financial resources to carry out their task.12 This perspective also helps to explain how the present Left Front government of West Bengal has achieved a measure of success, albeit a limited one.
The strategy of the Left Front was to fulfil the first condition (namely, the creation of countervailing political power) by a process of what Kohli (1987) has described as the separation of political from social power at the local level. Since no attempt was made to alter the existing pattern of ownership rights, it was expected that the landlords would continue to maintain their superior social power. But it was hoped that if political power was effectively wrested from them at the local level, the tenants would be assured of protection against eviction (that is, the first condition would be met) and would hence begin to assert their rights. However, since the tenants, left to themselves, could hardly be expected to bring about the required separation of power, it was also felt necessary that an external power must act as the catalyst. With this end in view, the dormant gram panchayats (village councils) were first revitalized—not by openly excluding the landlords as Gunawardena had intended to do, but by popular election along party lines. Since the Left Front parties, especially the Communist Party of India (Marxist), enjoyed majority popular support in the countryside, it was hoped that their members and sympathizers would be able to dominate the panchayats. And so they did, which paved the way for the next step.
(p.313) It was realized that since share-cropping was seldom carried out under written contract, it would not be possible to prevent pre-emptive or punitive eviction unless there was documentary evidence of who was share-cropping whose land. This led to the campaign of what has come to be known as the Operation Barga,13 during the course of which the share-croppers were asked to register themselves in public meetings. There was nothing novel about the idea of registration as such; it had been in the statute books for a long time without achieving its intended objectives in any part of India. What was novel, however, was the existence of countervailing power at the local level through the presence of politicized panchayats and party activists who exercised political power for the benefit of tenants. Consequently, registration was a huge success. The exact measure of this success is in some dispute, as the exact number of share-croppers is itself unknown. However, even by conservative estimates, anywhere between 65 and 85 per cent of the share-croppers have been registered14—an achievement unparalleled in other parts of South Asia.
It is now time to ask to what extent all this has helped to improve the security and level of livelihood of share-croppers in West Bengal. Registration and the continued presence of countervailing power at the local level have presumably gone a long way towards improving the security of their tenure. But security in itself is not much of a gain. It is evidently a gain in comparison with the immediate past when evictions had become common. But these evictions were themselves a consequence of attempted reforms that backfired. One must therefore compare the present situation with the status quo ante, when traditional landlord-tenant relationships existed in full force. Eviction as such was not a problem at that time, at any rate not for the whole class of potential tenants as opposed to an individual.15 The problem was in the use that was made of the threat of eviction—to keep the rent high, for example. Presumably, as the threat of eviction was removed, the tenants would then be able to press for a higher share, especially since the law provided for it and countervailing power existed at the local level to enforce the law. Such an outcome would indeed constitute a genuine gain from tenancy reform. But the available evidence does not suggest that this has happened to any significant extent. In the largest survey that has so far been conducted on the consequences of tenurial reform in West Bengal, Bandyopadhyay et al (1985) found that only 20 per cent of the registered share-croppers were able to pay the legally stipulated share; in fact, as much as 71 per cent of the share-croppers (p.314) continued to pay the traditional share of half the produce. In a survey of four villages, Westergaard (1986) observed that the tenants) share had improved in only one village, but even there the share was less than the legal minimum. A contrasting result is presented by Kohli (1987) who found, again in a small survey, that the proportion of tenants paying the legal rent had increased sharply from 13 per cent to 66 per cenŧ It is, however, significant to note two very interesting features of his findings: first, 80 per cent of the landlords in his sample were absentee landlords and secondly, as Kohli himself presents the puzzle, although the Communist Party activists were particularly strong in one of his survey areas (in the Burdwan district), the tenants' share failed to improve in that region.
The implication of all these findings can be summed up as follows. In spite of the fulfilment of the condition of countervailing power, tenants are in general reluctant to break up the traditional landlord-tenant relationship by demanding a higher share, except in the case of absentee landlordism where this relationship did not in any case exist to begin with.16 The clue to this apparent paradox lies partly in the fact that there is no guarantee of continued rule by the Communist Party for an indefinite future. The tenants are naturally apprehensive of the consequences that might befall them once the parties more favourably disposed to the landlords return to power. But partly the explanation also lies in the non-fulfilment of the second condition of a successful tenancy reform, namely, the assurance of an alternative source of material support. The reformers were not altogether unaware of this potential pitfall. In fact, the provision of credit and input facilities for the registered share-croppers was made, at least on paper, an integral part of the reform programme. But all the existing surveys show that the amount and timing of credit was not enough to make the tenants' traditional dependence on landlords economically redundant. Whether adequate resources can be found for this purpose and whether institutional methods can be devised to channel these resources to the right person at the right time in the right amount are difficult questions that cannot be fully answered at present. What is clear, however, is the lesson that tenancy reform is by its very nature caught in a trap. It seeks to improve the security of tenants by freeing them from the clutches of the landlords, but the tenants will refuse to remove themselves from those clutches unless there is prior provision for adequate social security.
It can be argued, and it has been argued forcefully in the context of the West Bengal debate,17 that the solution of this problem lies in the abolition of tenancy altogether by according full ownership rights to tenants. The presumption (p.315) is that once the tenants are able to claim the entire produce of the land, their economic condition will improve sufficiently to eliminate the material basis of their dependence on the landed élite. Spurred by this incentive they would then be willing to take advantage of the existing countervailing power to break the traditional relationship in a way they are unwilling to do while the only alternative is to remain as tenants deprived of traditional patronage. As for the political feasibility of this approach, it is pointed out that the government of Kerala has been able to abolish tenancy through the same political party that rules in West Bengal and under the same constraints imposed by the Constitution of India. If Kerala can, why cannot West Bengal?
This apparently persuasive argument has one rather serious flaw. It ignores the fact that the agrarian structure of Kerala which had emerged over the colonial period gave its tenancy reform an entirely different political and economic content from that which would be entailed by a comparable programme in a State like West Bengal.18 Due to complex historical reasons which we cannot go into here, tenancy had become, by the end of British rule, a widespread phenomenon only in the north-central region of Kerala comprising Malabar and Cochin.19 Nearly two-thirds of all tenants were to be found in this area, whereas 80 per cent of the owner-cultivators were in the south, in Travancore. Tenancy reform therefore essentially affected only the northern part of Kerala.
Several factors contributed to the political feasibility of this reform. First, the landlords in the north were relatively few in number, possessing a vast amount of land; some 80 per cent of all leased-out land came from holdings of 25 acres of more. Secondly, while these landlords did not enjoy full ownership rights on the land in the pre-British era (they only had a superior claim in a hierarchy of customary rights), they were accorded full ownership by the colonial regime and thus became identified as its protégés. Thirdly, unlike the rich landowners in the south, they were themselves never involved in the management of cultivation, and were thus regarded as parasitic landlords. In all these respects—namely, being relatively few in number, being identified as colonial protégés, and being seen as parasitic landlords—the landed aristocracy of Malabar was no different from the zamindars of Bengal. Consequently, the abolition of tenancy in Kerala was of the same order of political feasibility as the abolition of zamindari in Bengal.20 Furthermore, just as the abolition of (p.316) zamindari left the underlying production structure basically untouched (in so far as it failed to tilt the balance in favour of the small peasantry), so in Kerala the abolition of tenancy did very little to alter the vast inequalities in the distribution of landholdings (see Raj and Tharakan 1981). It would thus appear that both the political and the economic implications of the abolition of tenancy in Kerala were essentially similar to those of the successful abolition of zamindari in Bengal and elsewhere.21
The abolition of tenancy in Bengal, however, would have come up against a political problem of an entirely different order of magnitude. It would have had to confront not just a handful of absentee overlords (as in Kerala), but a large number of owners many of whom were small and actively engaged in cultivation. In Kerala the small landlords (possessing no more than 5 acres of land) were exempted from dispossession, and yet most of the land under tenancy went to the tiller. But a similar exemption in West Bengal would leave out nearly 80 per cent of all landlords and as much as 50 per cent of all leased-in land.22 A meaningful ‘land to the tiller’ policy cannot leave out this group, nor can the Communist Party alienate such a huge repository of local power as long as it sticks to the electoral path. It would of course be a different matter if the ‘revolutionary’ path were to be chosen, but that can only be a goal for a distant future.
One cannot, therefore, avoid the conclusion that the immediate political limit of tenancy reform in a State like West Bengal is not much above what has been achieved so far. But this also means that tenancy reforms by themselves cannot expect to achieve a great deal by way of improving the material conditions of living of the share-croppers. As we have concluded in the light of West Bengal’s experience, prior provision of social security is a prerequisite for this purpose.
In addition to ceiling legislation and tenancy reform, a third method of enhancing the land-poor’s control over land consists in the alienation of State-owned land. To some extent this has happened in all three countries of South Asia, but its scope has been very small in both India and Bangladesh. Historically the process of clearing and possessing wasteland was left to private initiative in these two countries. Consequently the State came to control very little of potentially alienable land But in Sri Lanka land policy has evolved differently, under the impetus of the colonial policy of promoting plantation agriculture. Under the Waste Lands Ordinance of 1840, the State took control of all land except that for which pre-existing private ownership could be clearly demonstrated. In the process, nearly 40 per cent of the total land area of the (p.317) country came under direct State control. Initially, the colonial government used this land to promote plantation agriculture; but there was a decisive shift in policy in the first quarter of the present century. By the 1920s population growth had created a serious problem of landlessness and land-hunger among the peasantry; at the same time food production was stagnating at a level that was increasingly inadequate to feed the growing population. As the nationalist movement made the cause of peasantry a rallying cry at this stage, and as the Donoughmore Constitution of 1931 devolved considerable power to national politicians, the land alienation policy shifted decisively in favour of the peasantry. The Land Development Ordinance of 1935 set the stage for the pattern that has been evolving up to the present (see Samaraweera 1981 and Moore 1985).
Most of the alienation has taken place under two major schemes, namely the Dry Land Colonization Scheme and the Village Expansion Scheme. The first involved clearing land and developing irrigation in vast tracts of the sparsely populated dry zone, and the second promoted the creation of new habitats around the existing villages, especially in the densely populated wet zone. Several features of these schemes are worth mentioning. First, the land was given exclusively, and virtually free of cost, to landless or extremely land-poor households. Secondly, the cost of clearing and developing land was borne entirely by the State. In the case of colonization schemes, the State also had to provide health and education facilities in order to attract residents of the wet zone into the malaria-infested wastelands of the dry zone. As a result, the beneficiaries not only got access to fully developed land complete with irrigation facilities, many of them also received the benefit of welfare services specially designed for them. Although the problem of landlessness was by no means eliminated in the process, a significant proportion of the land-poor population clearly gained a security they did not enjoy before. It has been estimated that nearly a quarter of all households in Sri Lanka now possess land granted under the Village Expansion Scheme with an average allotment of 1.3 acres per household (Moore 1985: 42). The Dry Zone Scheme has catered to rather fewer households—roughly 4 to 5 per cent of all households in the country up to 1980 (computed from figures quoted in Wickramasekera 1985è). Since then, the amibitious Mahaweli Accelerated Programme has aimed at more than doubling the number of beneficiaries within less than a decade.
Over the years, many shortcomings of these settlement schemes have been pointed out. For instance, it has been noted that the land alienated under the Village Expansion Scheme was often of inferior quality (Moore 1985); social anthropologists have often pointed out the lack of social cohesion among the uprooted households and the resulting absence of traditional redistributive and mutual insurance mechanisms that provided some degree of social security in the old villages (Moore and Perera 1978); others have noted renewed landlessness and concealed tenancy under onerous terms in dry-zone colonies (p.318) (Wickramasekara 1985b); yet others have pointed out that the value of agricultural output yielded by the newly developed land has not been commensurate with the cost of development (Wimaladharma 1982)
All these problems are genuine enough. But on balance, it must still be concluded that the land alienation policy pursued since 1935 has entitled over 30 per cent of households in Sri Lanka to a level of land-based security that would be a matter of envy to the land-poor population of India and Bangladesh. It is revealing to contrast the Sri Lankan experience with the most recent historical phase of reclamation of wasteland in Bengal, which occurred during the British rule. This reclamation took place in the north-western and southeastern parts of Bengal through the private initiative of wealthy individuals; and it is precisely in these regions that the agrarian structure is the most inequitable, the tenurial condition most oppressive, and the problem of landlessness most acute.23 Unfortunately, history has foreclosed the possibility of replicating the Sri Lankan experience in most parts of India and Bangladesh where very little reclamable wasteland is currently available.
But the Sri Lankan potential is not yet exhausted. Cultivated land still accounts for just over 30 per cent of all land area in the country. Although much of the rest will have to remain under perennial forest, particularly in the central highlands, enormous possibilities still remain in the potentially cultivable dry zone. This zone accounts for 70 per cent of total land but only 30 per cent of cultivated land. However, two serious impediments stand in the way. First, given the huge cost of development, trying to achieve too much too soon may place an unbearable burden on budgetary resources, as the ongoing Mahaweli Accelerated Programme has amply demonstrated. A second, and perhaps more serious, problem is the current ethnic conflict between the Tamils and the Sinhalese, since the Tamils tend to regard dry zone colonization as Sinhalese encroachment on traditional Tamil territory. These problems cast grave doubt on the possibility of enhancing land-based security in the future for the still significant land-poor population of Sri Lanka.
3. Security Through Self-Employment
One possible alternative to land-based security is to enable the land-poor to engage in those self-employed activities which are not crucially dependent on land. These so-called non-farm activities occupy a significant position in the occupational structure of South Asia, accounting for 34 per cent of the rural labour-force in Bangladesh, 46 per cent in Sri Lanka, and 19 per cent in (p.319) India.24 There are also indications that the share of the work-force engaged in this sector is rising over time. The question is to what extent the livelihoood of the land-poor population can be secured through participation in this sector. As it is, a considerable number of them are already engaged in these activities. For instance, sample surveys in Bangladesh have shown that of all the gainful activities (both primary and secondary) pursued by the workers owning no more than 0.5 acres of land, more than 50 per cent belong to the non-farm sector; in contrast, agricultural labour accounts for only a third of the total (Hossain 1986: 126). This does not, however, indicate the extent of their self-employment in these activities since many of them are actually employed as wage labour. Moreover, even among the self-employed category, the members of the land-poor group are usually engaged in less productive lines of activity compared to those who are better endowed with land and other resources (Hossain 1984a; Muqtada and Alam 1983). Presumably, the imperfections in the rural credit market prevent many of the land-poor households from obtaining the resources required for engaging in remunerative activities. The provision of resources through credit and/or subsidy is therefore an essential prerequisite for ensuring security through self-employment.
Over the years a variety of programmes have been tried in order to address this problem throughout the region, the most outstanding among them being the Grameen (Rural) Bank in Bangladesh and the Integrated Rural Development Programme (IRDP) in India. They both took off around 1979–80 with essentially the same objective, but were conceived in a somewhat different manner. There is also a difference in scale. While Grameen Bank covers only about 3 to 4 per cent of all villages in the country, with the hope of extending countrywide at some unspecified time in the future, the IRDP was launched with a countrywide coverage almost from the beginning. Thus the Grameen Bank is still only a model while IRDP is already a full-blown programme. There is, finally, a notable difference in the evaluation of their performance. The Grameen Bank has been widely acclaimed as a huge success, while most of the evaluations of IRDP range from mixed to derisory. It will be instructive to see to what extent their relative performance is related to differences in approach or to the difference in scale, and what lessons they together provide on the possibility of ensuring security through self-employment.
Let us first take a brief look at their relative performances. The Grameen Bank defines all those households who own no more than 0.5 acres of land as its target group. Independent evaluations show that singular success has been achieved in confining the benefits to the genuine target group: only 5 per cent of the beneficiaries owned land above the specified limit, while 52 per cent (p.320) owned no land at all.25 The average pre-intervention income of the recipients of loans (in 1980) was half the national average, which indicates that most of them would belong to the bottom 40 per cent of the income scale . Considering that anywhere between 60 and 80 per cent of the rural population were below the poverty line around that time, the beneficiaries must be considered to constitute the poorest segment of the rural population. There is also clear evidence that the operations of the Grameen Bank have provided new employment opportunities to previously unemployed rural women; the rate of female participation in the labour-force rose from 5 per cent to 25 per cent within the households receiving loans in a period of two years. Women also figured prominently in the overall operations of the Bank—by 1984 nearly 54 per cent of the beneficiaries were females, receiving 42 per cent of all loans.
According to rough estimates based on a sample survey, the per capita real income of aided households increased (net of interest payments) by 32 per cent over a period of two and a half years (starting from 1980), while at the national level per capita income rose by only 2.6 per cent. To what extent this picture is distorted by the well-known difficulties of collecting accurate income data is not clear, but the rates of return estimated for various investment activities give an independent indication that the income gain must have been substantial. In most of the trading activities and a few of the manufacturing activities for which loans were utilized, the return on family labour (after meeting all costs and repaying the loans with interest) was found to be marginally higher than the prevailing agricultural wage rate. On the other hand, in the majority of manufacturing activities as well as in cattle and poultry raising, the return on labour was lower than the agricultural wage rate. However, it is important to note that almost all of these low-return activities are dominated by females, whose scope for employment in agriculture or elsewhere in rural Bangladesh is negligible. Consequently, even though the return is rather low, nearly all of it represents a net addition to family income. As for male-dominated activities, not only is the return on labour (per man-day) higher than the agricultural wage in most cases, the duration of employment is also much longer . All this is not to suggest, however, that the recipients of loans have been able to secure a livelihood above any acceptable norm of poverty line. With 1.5 earners and 3.5 dependants per loan on the average, they could not possibly do so, even with year-round employment, at a level of earning that is below or only slightly above the prevailing agricultural wage rate.
One other aspect of the Grameen Bank’s performance that has been widely commented upon is its success in terms of a purely banking criterion. The recipients are charged an effective interest rate of 18 per cent which is above the (p.321) general rate of interest (between 12 and 15 per cent), and are required to repay the loan in weekly instalments over a year. Despite these stringent conditions, the proportion of overdue loans after two years of loan disbursement was found to be less than 1 per cent, and only 4 per cent of the recipients had not fully repaid the loan by the time it was due. It is thus clear that the Grameen Bank has succeeded in giving income support to the poor without degenerating into the kind of dole-giving institution that so many other poverty-alleviation programmes have tended to become.
In comparison with these achievements of the Grameen Bank, the performance of IRDP appears decidedly poor. One point should, however, be clarified before highlighting the contrasts. The goal of IRDP was pitched at a much more ambitious level than that of the Grameen Bank, and consequently the former is likely to suffer in any comparison between aims and achievement. While the Grameen Bank had the modest goal of providing a more secure source of income for a few assetless people without necessarily eliminating their poverty in the immediate future, IRDP was aimed at lifting the beneficiary households above the poverty line with a one-shot intervention. The Sixth Five Year Plan (1980–5) set the target that over a period of five years as much as 30 per cent of the officially designated poor population would be assisted and lifted out of poverty, starting from the poorest group and moving gradually upwards.
Some of the evaluation studies have indeed claimed a rather spectacular rate of progression above the poverty line; for example, nearly half of the beneficiaries are supposed to have made the transition, according to NABARD (1984) and PEO (1985). But these estimates suffer from some elementary methodological problems, such as comparing current income with a base-line poverty level and estimating net income without adjusting for loan repayment. After some rough adjustments on these counts, Rath (1985) has estimated that not more than 10 per cent of the beneficiaries (or 3 per cent of the base-line poor population) could have made the transition to the status of non-poor over the Sixth Five Year Plan period. Based on a much more systematic countrywide survey for the year 1985–6, Kurian (1987) finds that less than 5 per cent of the beneficiaries may have crossed the poverty line. And even that was possible only because in many cases the assistance went to people just below the poverty line rather than to the poorest, in violation of the declared bottom-up procedure. It was, in fact, demonstrated by Sundaram and Tendulkar (1985) that given the level of assistance and even accepting the highly optimistic capital-output ratio assumed by the Plan, none would be able to cross the poverty line if all the assistance went to the bottom half of the poor population. There was thus a manifest contradiction between the level of assistance and the proclaimed bottom-up procedure on the one hand and the target of transition beyond the poverty line on the other. Consequently, IRDP made itself a sitting duck to be merrily shot down by its numerous critics.
(p.322) Deservedly so, one might say, in view of the contradictions built into the programme. But it is still necessary to evaluate to what extent IRDP did in fact ensure a more secure livelihood for the poor, even if it meant that they continued to live below the official poverty line. The first issue is the degree of success in reaching the target group. In order to capture the poorest group, the eligibility criterion was fixed at a level of income that was some 23 per cent below the official poverty line. After reviewing four major evaluations done during the 1980–5 period, Bagchee (1987) has estimated that about 15 to 20 per cent of the beneficiaries were above the cut-off point to begin with.26 The survey of 1985–6 has also revealed a leakage of some 20 per cent (Kurian 1987). This is a fairly commendable achievement compared with the experience of earlier target-group-oriented programmes, especially if one notes that many of the ‘ineligible’ households would actually have had an income below the official poverty line although above the eligibility criterion.
There is, however, some doubt as to whether the bottom-up procedure has been strictly followed. For example, the results of the countryside survey conducted in 1985–6 show that nearly 60 per cent of the beneficiaries belonged to the top half of the poor population (Kurian 1987.· table 1). For the state of Karnataka during the same period, Rao and Erappa (1987) have shown that landless households constituted less than half of all beneficiaries. In none of these cases is there reason to believe that the potential beneficiaries in the poorest segment were exhausted before the less poor households were catered to. In this respect, there is some similarity with the experience of the Grameen Bank. Although, as we have noted, as much as 95 per cent of the Grameen Bank beneficiaries belong to the target group, the programme has not succeeded in catering to the need of agricultural labourers who (along with sharecroppers) are recognized to be the most deprived group in rural Bangladesh. As Ahmad and Hossain (1985) have pointed out, this group constituted only one-tenth of those receiving loans although they formed one-third of the target group. We shall comment later on the significance of this phenomenon common to both the Grameen Bank and IRDP.
To continue with IRDP’s performance, we next look at its impact on the beneficiary households’ ability to sustain a higher flow of income. The findings in this respect are much less clear-cut, mainly because of the well-known difficulties in estimating income and, to an even greater extent, in finding a suitable reference point for comparison. By analysing the data from a countrywide evaluation for 1985–6 Kurian (1987) has estimated that, over a period of two years, 84 per cent of the assisted households enjoyed some positive increment in family income and as many as 38 per cent did so to the tune of more than 50 per cenŧ. This is taken by Kurian as an indication that IRDP has (p.323) succeeded in giving a significantly higher level of living to a large group of households.
But his conclusion cannot be taken seriously because of certain definitional problems, quite apart from the problems of eliciting information on past income through the recall method. In the first place, the amount of loan repayment is not netted out of gross income when estimating the post-assistance income of the beneficiary families.27 Secondly, it is not clear whether the income data for the two periods have been converted into constant prices. Without access to basic data, it is not possible to correct Kurian’s estimates for the factors mentioned above. However, one can see how a very different picture emerges if some rough corrections are made on his data concerning income generated by the IRDP assets (Kurian 1987: table 11). Some 24 per cent of the assets are found to have generated no income at all. Another 26 per cent of the assets generated an income of up to Rs 1,000, but this income is in current prices and does not exclude loan repayment. Taking the standard loan specifications of the IRDP, namely, an average loan of Rs 2,000 per family, a repayment period of 3 years, and an interest rate of 10 per cent per annum, and assuming 10 per cent inflàtion over the two-year period, it is easy to see that, after loan repayment and price adjustment, an asset yielding an income of up to Rs 1,000 would add nothing or little to a family’s real disposable income. Thus a total of 50 per cent of the assets can be said to have made no or negligible contribution to net disposable income.
As for the remaining 50 per cent of the assets, the evidence of positive contribution must be tempered by the fact that once a family engages on an occupation based on the newly acquired assets, its income from other sources tends to fall because of diversion of labour . Working on the same set of data as used by Kurian, but focusing on the state of Uttar Pradesh alone, Rao and Rangaswamy (1988) have found that the fall of income from other sources was about 20 per cent of the pre-assistance income of an average family. The opportunity cost from the point of view of the family is thus seen to be quite substantial; consequently, many of the assets which did make a positive contribution to disposable income, even after adjustment for loan repayment, may not turn out to do so after this opportunity cost is also accounted for. In the absence of more detailed information, we cannot be sure exactly what picture emerges at the end, but the least one can infer from the preceding (p.324) arguments is that more than half the assets failed to add to the beneficiaries’ net disposable incomes, properly defined.
Rao and Rangaswamy (1988), in their study of Uttar Pradesh, havecorrected their estimates of income for the opportunity cost mentioned above, but not for loan repayment, and it is not clear whether they have made any corrections for price increase. According to their estimates, an average beneficiary enjoyed a net increment of 32 per cent of pre-assistance income over a period of two years. In absolute terms, the increment comes to Rs 1078 (computed from ibid, table 12). On the other hand, the annual burden of loan repayment can be seen to be somewhat higher than this figure, if one applies the three-year repayment period and 10 per cent annual interest on the average loan size (reported to be Rs 2,797, ibid, table 10). Thus the net contribution to disposable income once again turns out to be negligible at best for the average beneficiary.
One major problem with all these studies is that they take a static view of income gain, whereas the crucial issue in the evaluation of IRDP must be whether the beneficiaries can sustain a higher flow of income in the long term after paying all dues and replacing the old assets when they become obsolete. Such long-term effects on income mobility have been rarely explored. One partial effort was made by Subbarao (1985) who calculated an income mobility matrix from the data collected by IFMR (1984) on five districts in southern India. He compared the assisted households with a carefully chosen control group and showed that the upward mobility of the poorest among the assisted households was decidedly superior to that of the poorest non-beneficiary households. His estimates of the proportions of upwardly mobile households is not of much use, since, as rightly pointed out by Swaminathan (1988), he did not adjust the nominal incomes for price increase. But his comparison between beneficiaries and non-beneficiaries is not vitiated by this flaw.
There is, however, a more serious problem with Subbarao’s findings. As once again pointed out by Swaminathan (1988), the mobility he looks at relates to a time gap of only fifteen months, which is too short a period to judge whether the beneficiaries would be able to maintain a higher level of consumption over the long term. They may not be able to keep the assets after some time, or they may find their extra income insufficient to enable them to set aside the money required for replacing the old assets. Swaminathan’s (1988) own micro-level survey of two villages in Tamil Nadu was addressed precisely to this question of the long-term asset position of the assisted households. She found, after six years since the initiation of IRDP, that the upward mobility of the assisted households in terms of assets was no better than that of the sample of unassisted households.28
(p.325) Clearly, more studies of this type are needed to form a clear picture of how far the IRDP has succeeded in improving the long-term economic viability of beneficiary households. But the evidence and arguments presented above do not inspire much optimism. The same large-scale survey from which Kurian (1987) draws a mildly optimistic picture, also shows that 29 per cent of the assets were ‘not intact’ only two years after the receipt of assistance; and as much as 24 per cent of the assets were found to yield no income at all either because the assets were disposed of, or had perished, or had never been procured in the first place. Piecing together a number of earlier surveys both large and small, Rath (1985) found that the proportion of beneficiaries whose assets ceased to exist after two to three years varied between 20 and 50 per cent.
Even in the cases where the assets contined to exist, the correctly measured contribution to disposable income, as we have noted, was often rather low. The most talked-about example of low return is that of milch cattle, which have come to bear the rather sarcastic appellation of the ‘IRDP cow’. In the early years of IRDP, as much as 60 to 80 per cent of the assets consisted of cattle; even in 1985–6 the share of cattle in all assets was 37 per cent (Kurian 1987). There are innumerable accounts, albeit of an informal nature, of how the milch cattle have yielded a pitifully low rate of return either because most of them were of inferior breed, or because fodder and veterinary care were utterly inadequate, or simply because they died of disease or drought. While the problem with milch cattle was especially acute, the other assets were not particularly productive either. Those who framed the Sixth Plan had expected that the IRDP investment would yield an average incremental capital-ouptut ratio of 1.5 :1. In the event this ratio was achieved only in the case of 26 per cent of the assets, and the overall ratio was close to 3 :1 (Kurian 1987). On top of this, one must also take note of the numerous accounts of how the prospective beneficiaries had to incur a substantial initial cost by way of paying the intermediaries (and sometimes the bank officials too) and searching for a suitable assets. According to the National Committee on the Development of Backward Areas, these expenditures often amounted to 50 to 70 per cent of the subsidy component (GOI 1981: 24). Taking all these factors into account— namely, 20 to 50 per cent cases of vanishing assets, low yield of existing assets, and high initial cost—one cannot but be sceptical about IRDP’s success in ensuring a sustained flow of a perceptibly higher level of income.
What about the future? Can the programme be reconstituted so as to offer a genuinely secure source of higher income for resource-poor households? In order to answer this question, one must first identify the sources of IRDP’s weakness. At one level, it can be argued that much of the problem lies in disregarding the basic economics of a massive self-employment programme. If assets are to become a viable basis of self-employment, at least three kinds of (p.326) ‘matching’ must be ensured. First, the structure of assets must be such that the resulting structure of output corresponds to the existing or prospective pattern of demand fqr that output. Secondly, the structure and quantum of assets to be acquired by the households must be matched with the existing or potential supply of assets. Finally, there has to be a matching between the type of asset (as well as the size of investment) and the pre-existing resources (land, labour, and entrepreneurial capacity) of a household. If a mismatch occurs in any of these respects, the yield of investment is bound to suffer. There are enough indications to believe that IRDP suffered from mismatches of all three kinds.
Excessive concentration on milch cattle is a clear example of the neglect of the demand side. Working with the IFMR data, Subbarao (1985) noticed that the only district (out of five) where milch cattle had yielded a satisfactory rate of return was that blessed with a number of milk-chilling plants. The demand generated by these plants, as well as other infrastructural facilities associated with them, obviously helped in maintaining a high level of return. In other districts, there were simply too many heads of cattle in relation to demand.29 Similarly, Kurian (1987) narrates the case of dozens of sewing machines being supplied to different women in a single village; and stories like this can be multiplied. The neglect of the supply side is equally glaring. As Rath (1985) notes, something like 5 million cattle were purchased by the IRDP households during the Sixth Plan period, without there being any arrangement for producing that many high-breed cattle. As a result, not only did the beneficiaries have to purchase inferior cattle, but sometimes even the required number of cattle were not available. As a result, the price of cattle went up, which often induced the purchasing households to sell off the cattle to other households. In this way, the same animal moved from household to household satisfying aggregate demand in the absence of matching supply (Kurian 1987). With more than a tinge of sarcasm, V. M. Dandekar has for this reason described the ‘IRDP cow’ as ‘circulating capital’ (quoted in Kurian 1987: A-175).
Similar failures have been noted in respect of matching IRDP assets to household resources. Kurian (1987), for example, mentions that cases were reported where hundreds of animals were distributed in drought-affected areas although the purchasers did not have the grazing facilities to feed the animals. At a more general level, Rath (1985) has argued that the poorest of the households who have traditionally sold nothing but raw labour may simply not have the entrepreneurial ability to engage in viable self-employed activities. A degree of support for this contention is provided by our earlier observation that agricultural labourers were mostly left out by both IRDP and the Grameen Bank.30 It is, however, difficult to agree with Rath when he goes on to argue (p.327) that the whole idea of promoting self-employment should be de-emphasized and that the expansion of wage employment through public works should constitute the principle direct intervention programme. The notion that the poorest households do not as a rule possess entrepreneurial ability cannot simply be true; for a large number of them are traditionally engaged in various self-employed activities such as cottage industries, petty business, and services; moreover, it has been noted from the IRDP experience that the poorest households who participate in the programme do no necessarily fare worse compared to the less poor ones (Rao and Rangaswamy 1988). However, the essential point that the chosen activity must correspond to the entrepreneurial ability and other resources of the household still remains valid.
The observed mismatch in IRDP has generally been attributed to two factors. The first relates to the absence of a comprehensive local development plan, without which matchings of the various kinds described above could scarcely be ensured. Plans are required for the purpose of both assessing and augmenting the supply of assets on the one hand and demand for their products on the other. This will make it possible to endow the prospective beneficiaries with assets that are both available in required specification and viable enough in terms of the demand for their products. Initially, this is precisely what the programme was supposed to have ensured; hence the name Integrated Rural Development Programme. In actual practice, hotvever, planning for the creation of supply and demand was never integrated in the programme. There was some rough attempt to carry out an assessment of existing supply and demand, but that too was thoroughly inadequate. The second factor relates to the imperative of fulfilling a numerical target within a specified time limit. A severely understaffed local administrative machinery was burdened with the tasks of identifying the eligible beneficiaries, motivating them to participate in the programme, preparing suitable investment projects for individual households, and often identifying and procuring the assets for them. Under pressure of meeting the target, they frequently took the easy option of pushing through whatever assets were readily available, disregarding the need to match described above.
One might pause to ponder at this stage why is it necessary to ensure the required matchings through planned intervention—why cannot one simply transfer some financial resources to the poor households and leave the task of matching to the market process? In this context, it is instructive to contrast the experience of IRDP with that of Grameen Bank. As opposed to the bureaucratic-interventionist approach of IRDP, the Grameen Bank of Bangladesh has adopted what may, with some liberty, be described as the laissez-faire approach. It is true that Bank officials do apply a strict eligibility criterion, motivate the prospective beneficiaries to participate in the programme, and subject them to (p.328) strict discipline; but there is no attempt to determine the structure of activities either in the aggregate or for individual households. The whole underlying philosophy of the Grameen Bank is that most local people have a pretty shrewd idea of what opportunities are potentially available and they also know which of these they are capable of handling; all they need is the access to credit in order to take those opportunities.
It will be wrong, however, to describe it as an entirely hands-off approach, because the Bank does employ a highly imaginative method of indirectly pressurizing the beneficiaries to be particularly careful in the choice and execution of their projects. This is done by enforcing a rigorous discipline of loan repayment. There are a number of components of this strategy. First, the beneficiaries are required to form a group of five people. Although the loan is given to an individual and he or she alone is responsible for repayment, a kind of group pressure is brought to bear upon the recipients by the stipulation that default by one member of the group will make the rest of the members ineligible for loan. Secondly, the repayments are to be made in weekly instalments in local meetings which all beneficiaries in a particular locality are required to attend. This again acts as group pressure. All this is motivated by the basic premiss that if the repayment discipline can be rigorously enforced, the recipients will be careful in seeking out truly viable projects. Thus the combination of people’s personal knowledge of local opportunities and the Bank’s insistence on the discipline of repayment is considered to be an adequate guarantee of securing the productive use of credit. Apparently, this strategy has worked remarkably well.
Does the laissez-faire approach operated by the Grameen Bank, then, offer a viable strategy for self-employment programmes? We believe there are reasons to be sceptical about it. A very important consideration here is the question of scale. As mentioned before, the Grameen Bank has so far covered only 3 to 4 per cent of the villages in the country; even in the limited area of its operation only half of the target group has been covered (Hossain 1984b). While operating at such a low scale, it may not be so essential to worry about supply and demand. The modest purpose at this level is to enable a few of the poor to take advantage of the opportunities that are somehow being created within the economy. In this sense the Grameen Bank does little more than marginally augment the ‘trickle-down’ process. But if a massive self-employment programme of the IRDP type is to be launched, it will not do to disregard the need for integrating credit-giving within a broader framework for planned expansion. There are already indications of this even in the limited scale of the Grameen Bank’s operation. A couple of Hossain’s (1984b) findings are worthy of note in this context. First, the rate of return from repeat loans is generally lower than that of the original loan and secondly, the relatively large-size loans are seldom fully invested. This indicates the presence of limits to the expansion of self-employment within the existing configuration of supply and demand.
(p.329) The need for genuinely integrated rural development cannot, therefore, be denied. One might argue, however, that while macro-matching is ensured through an integrated plan, micro-matching between household and assets may still be left to the individual, as in the Grameen Bank approach. Some degree of individual autonomy is certainly desirable, but its limits should also be recognized. The Grameen Bank may be right in assuming that even the poorest households may have a shrewd idea of what opportunities are available, provided that the local economy is only slowly evolving. But that assumption may no longer hold if massive interventions are to be made at the local level to alter the macro-configurations of supply and demand. The resulting uncertainty about the new economic environment may be difficult to handle even for the well-off entrepreneurs, but at least they can be expected to take their losses along with their profits; in any case their subsistence will not be threatened in the event of a loss. But even a single loss-making venture by a poor household may ruin the slender basis of subsistence it may have had by saddling it with a net liability of loan repayment. That would be the cruelest outcome of a programme whose ostensible purpose is to enhance the security of poor households.
It is therefore clear that a successful programme of security through self-employment will require detailed planning at the local level with regard to all three kinds of matching discussed earlier. It should also be clear that just any kind of integrated rural development programme will not do, for one may have integrated development which caters to the needs of the better-off section. One therefore comes back to the question of countervailing power at the local level for safeguarding the interest of the poor. This question has not been systematically explored in the various evaluations of IRDP that have been carried out in India. But some indication of its importance may be seen in the findings of Rao and Erappa (1987) and Thimmaiah (1988) for Karnataka, and Sharma and Gianchandani (1988) for Rajasthan. They all noted that the opportunities opened up by IRDP were seized more effectively by middle-level households than by the bottom-placed agricultural labourers. It is also interesting to note that by several accounts the performance of IRDP has been somewhat better in West Bengal where the rule of the Communist Party of India (Marxist) has created a countervailing power, however partially.31 A survey conducted by the Programme Evaluation Organisation of the Planning Commission of India showed that the proportion of beneficiaries drawn from the poorest group is the highest in West Bengal (along with Tamil Nadu) among all Indian states (Bandyopadhyay 1985:112). The subsequent countrywide survey of 1985–6 showed that West Bengal was also among the best performers in terms of utilization of resources: while the assets ceased to yield income after two years of loan disbursement in 24 per cent of the cases at the (p.330) national level, this happened in West Bengal in only 7 per cent of the cases (Kurian 1987).32 Finally, the same survey also showed that while the incremental capital-output ratio was 3:1 at the national level, West Bengal achieved the ratio of 1.5 :1 in 70 per cent of the cases—the best performance among all States.
These illustrations are meant to be indicative rather than conclusive. But they do lend credence to the a priori proposition that the presence of countervailing power at the local level is a precondition for an effective programme of social security through self-employment of the poor.33
4. Security Through Wage Employment
A large proportion of the rural labour-force in South Asia depends on wage labour as their principal means of livelihood. The proportion of principally wage-dependent persons (including landless agricultural labourers, marginal farmers, and households engaged in non-cultivating labour) is at least one-third of the rural labour-force in India and well over half in Bangladesh. They are also among the poorest of all households. In Bangladesh as much as 90 per cent of such households are counted as poor, while in India a reasonable estimate would be around 60 per cent.34 This massive poverty is due partly to low wages and partly to lack of secure employment throughout the year. Recognizing that the amount of employment generated through the normal process of development is thoroughly inadequate to meet the challenge, both countries have launched huge programmes of employment generation, largely through public works. It is worthwhile to see to what extent these programmes have been or can be useful in providing security to the poor people.
The major programme in Bangladesh goes under the name of Food For Work (FFW). It came into existence in 1975 and by now it is the most important of all employment-generating programmes. Its distinctive feature, compared with earlier programmes, is the payment of wages in kind, mainly wheat that is obtained as food aid from the Western world.35
Like most other programmes of its kind elsewhere in the developing countries, FFW is not free from the usual allegations of leakage and the misuse of funds; one study has estimated that the underpayment of labourers may have amounted to about 30 per cent of total allocation (BIDS-IFPRI 1983). Yet the benefits derived by the employed workers were not inconsequential. Some (p.331) of the positive features noted by a comprehensive survey (Osmani and Chowdhury 1983) may be summarized as follows: first, there is no doubt that the benefit of employment went mostly to the poorest segment of rural population. Nearly 70 per cent of the employees belonged to what is called the functionally landless category, that is, those who own less than half an acre of cultivated land. Since they constitute about half of the rural population, it is evident that this group has found more than proportionate representation in total employment. Secondly, the net income gain of participating households amounted to roughly 10 per cent of their annual wage income or 7–8 per cent of total household income. Thirdly, the consumption of food (especially foodgrains) improved significantly for those who were more or less regularly involved in the FFW projects.
While noting these positive aspects, it must also be admitted that in terms of its aggregate impact the achievement does not amount to much. An illustrative calculation can give the picture . According to the Labour Force Survey of 1983/4, the total rural labour-force of that year was 24.3 million. It has been observed from various scattered sources of evidence that the extent of underemployment in rural Bangladesh varies from 31 to 40 per cent (Islam 1986). Taking the conservative estimate of 30 per cent, the extent of underemployment in 1983/4 was the full-time equivalent of 7.3 million people unemployed. Against this background, FFW created 101 million man-days in that year, which comes to about 0.34 million man-years (taking a standard man-year to consist of 300 days). This means that FFW was able to eliminate just under 5 per cent of underemployment among the rural labour force in 1983–4.
This is a rather modest achievement, clearly inadequate in relation to need. The question therefore arises: can the scale of operations be expanded to provide complete security of employment? One constraint can be noted immediately. Our illustrative calculation shows that the scale of operation will have to be expanded at least twentyfold in order to provide complete security. Considering that the amount of wheat distributed under FFW constituted roughly 20 per cent of total food imports in recent years, such an expansion would involve a sixfold increase in the quantum of food import from its present level—clearly an unfeasible proposition. There may be other constraints also. For instance, will it be possible to find an adequate number of public-works schemes which are the same time both labour-intensive and sufficiently productive? This question takes us to a much wider range of issues which have relevance not just for Bangladesh but for the rest of the region as well. We shall return to it after taking a brief stock of India’s experience with similar programmes.
After experimenting with various forms of employment generation schemes in the 1960s and the 1970s, India too decided to adopt Food For Work as the major programme in 1977. There was, however, a major difference from Bangladesh. Instead of depending on food aid, India supplied the required (p.332) food-grain from its own rapidly accumulating stock. Indeed, one of the motivations behind shifting from a system of cash payment to a system of partial payment in food was to dispose of this huge stock, which was something of an embarrassment for a country where anywhere between one-third to one-half of the population were reported to be suffering from endemic hunger. In any case the existence of this stock implies that India is in a much better position to expand the programme compared with Bangladesh, provided that other constraints do not become binding.
As it happens, a severe constraint seems already to have emerged as far as the food-grain component of the programme is concerned. This is indicated by the sharp fall in this component after the FFW was merged into a new expanded programme called the National Rural Employment Programme (NREP) at the start of the Sixth Plan (1980–5). In 1980/1 an average of 3.2 kg of food-grain was utilized per person per day, but the average for the next four years was only 0.5 kg, although the official stipulation was that at least 1 kg of food-grain should be given as part of the wage. Difficulties in public distribution and workers’ preference for coarse cereal over the finer cereals (wheat and rice) which are distributed by the programme have been suggested as the main reasons for this phenomenon.36
This did not, however, hinder the progress of the programme as a whole, since the shortfall in grain payment was made up by cash What the planners were worried about was the possibility that the benefit might not be going mostly to the neediest group, namely, the landless population. Consequently, in 1983, a new scheme called the Rural Landless Employement Guarantee Programme (RLEGP) was initiated alongside NREP, with the specific objective of guaranteeing up to 100 days of employment to at least one member of every landless household in the country. These two programmes were merged in 1988 and are together expected to create 490 million man-days in a year during the Seventh Plan period (1985–90).
How significant are these efforts in relation to need? Once again an illustrative calculation will help. Assuming that the Seventh Plan target is met, and taking a standard man-year to consist of 300 days, a total of 1.6 million full-time jobs will have been created in 1986. Exactly what this means in terms of removing unemployment is not certain, as no estimate of unemployment is available for the recent period. However, we can get a rough picture by projecting past estimates. It was noted in the Sixth Plan document that in 1980 some 15.4 million job-seekers in rural India remained unemployed on a typical day, that is to say, total unemployment was the equivalent of 15.4 million full-time jobs. Even if we make the optimistic assumption that total employment has grown since then at the same rate as the labour-force, this will mean that the absolute size of the unemployed labour-force has grown at that rate too. On that basis, (p.333) just as about 9 per cent of unemployment would have been removed by the two programmes in 1986.
This is indeed some achievement, and decidedly better than that of Bangladesh, but still clearly inadequate in relation to need. In order to eliminate unemployment completely through NREP and RLEGP, the scale of operation will have to be expanded at least tenfold. Considering that the Seventh Plan allocates 2.4 per cent of total outlay on these two programmes, such an expansion would imply an allocation of as much as a quarter of total Plan expenditure.
This shows the financial implications of ensuring social security through guaranteed employment. It is worth noting in this context that the concept of guaranteed employment is enshrined in the Indian Constitution where the right to work is given the status of a (non-justiciable) directive principle. However, it was only with the introduction of RLEGP in 1983 that the concept of guarantee was explicitly recognized for the first time in a national-level programme. Even then the recognition was incomplete. First, only one rather than all the working members of each landless household was to have guaranteed employment. Secondly, the guarantee was given only in the form of a ceiling rather than as a floor of minimum employment, namely ‘up to 100 days per year’. This means that the authorities can claim to have honoured the guarantee even by offering just a single day’s employment. In one of the few evaluations of RLEGP done so far, the authors did in fact find that only 5 per cent of the participants received more than 100 days of employment in Punjab (Singh and De, n.d.).
Much more serious commitment to the concept of the guarantee is evident in the State of Maharashtra. The State government had been implementing an Employment Guarantee Scheme (EGS) from its own finances long before RLEGP came into being at the national level. Several other States have also been trying to emulate Maharashtra with schemes of their own. But it is the Maharashtra scheme that has come to be acclaimed as a resounding success. For that reason, it is worthwhile to look into it in some detail.
Although EGS was formally sanctioned only in 1978/9, it had in fact been in operation since the late 1960s. Starting at a low level of activity it began to pick up from 1975, and it has now become the most important social-welfare programme in the State (consuming nearly 10 per cent of the State budget in 1979/80). In its present form, the scheme guarantees that every adult who wants a job in rural areas will be given one, provided he/she is willing to do unskilled manual work on a piece-rate basis. In order to ensure that only truly needy persons will come forward to take up the guarantee, a mechanism of self-selection has been built into the programme in a number of ways. First, no choice of work is given—only unskilled manual work is made available. Secondly, the wage rate is kept below the ongoing agricultural wage; the piece rate is fixed in such a way that with average effort the daily wage would equal (p.334) the minimum of all statutory minimum wage rates obtaining in different parts of the State. In practice, the process of self-selection is further reinforced by the widespread prevalence of underpayment.37 Thirdly, although the official instruction is that work is to be provided preferably within a radius of 5 km from a worker’s residence, in practice the guarantee operates at the block level, which may sometimes require a person to travel a long distance for a few days of temporary work (Bagchee 1984).
On the whole the self-selection mechanism has worked reasonably well, but not with complete success. An evaluation done in 1977/8 found that 45 per cent of the workers came from landless households and 42 per cent belonged to the deprived social groups of the scheduled castes and the scheduled tribes (Dandekar and Sathe 1980). However, according to the same survey, 42 per cent of the workers came from cultivating households owning up to 5 acres of land. While some of them, especially the so-called marginal farmers owning less than 2.5 acres (of mostly unirrigated land), belonged to the neediest group, not all of them did. It has been observed that small, and sometimes not-so-small, farmers participate to a considerable extent in the EGS during the lean season when many of them would otherwise remain idle. This explains the finding of an official evaluation that, as a proportion of those needing work, the small and medium farmers have secured more employment than the landless group (PEO 1980 b).
This is perhaps an unavoidable consequence of offering an open-ended guarantee to everyone regardless of economic status. Perhaps there would not be much wrong with this if it did not happen at the expense of the neediest group. It is, therefore, important to see how far the EGS has succeeded in meeting the employment gap of landless workers and marginal farmers. In sheer absolute size, the employment created under the EGS certainly exceeds by a wide margin anything that has so far been achieved elsewhere in India (or for that matter other parts of South Asia). In 1977/8, for example, an average participant was found to have received 160 days of employment under the EGS (Dandekar and Sathe 1980). By comparison, FFW provided just 44 days of employment per participant all over India (PEO 1980a) and NREP provided 51 and 55 days in Gujarat and Karnataka respectively (ILO 1984). Also, the EGS contributed two-thirds of the annual income of participants (Dandekar and Sathe 1980), while the corresponding contribution of NREP in Gujarat and Karnataka was found to be one-fifth (ILO 1984).
Despite such impressive figures, there remains some doubt as to the achievement of the EGS in relation to need. Dandekar and Sathe (1980) have claimed that in 1977/8 the EGS eliminated as much as 75 per cent of unemployment among the ‘weaker sections’, defined as the set comprising landless labourers and the bottom 10 per cent among cultivators. This estimate (p.335) appears to be highly inflated in view of the doubtful procedures they have employed.38 They assume, for instance, that all employment went to the weaker sections whereas their own survey shows that only 45 per cent of the workers came from the landless group. Of course we do not know how many came from the bottom 10 per cent of the cultivators, only that 42 per cent came from households owning no more than 5 acres. Given this information, it is in fact impossible to deduce how much of the EGS employment went to the weaker sections, and by how much their unemployment was reduced.
As an alternative, we offer an estimate for a subset of the ‘weaker sections’, consisting only of landless labourers. We know from Dandekar and Sathe (1980) that 45 per cent of the EGS workers were landless labourers, but not how much of total employment went to them. It is, however, possible to estimate an upper limit from the finding that an average worker found employment for 160 days in the EGS but hardly anyone exceeded 180 days. If we assume (liberally) that all the landless workers had reached the maximum limit of 180 days, then their share in total employment was 50 per cent (45 × 180/160). Thus the upper limit of the landless labourers’ EGS-employment, out of a total of 117.3 million man-days in 1977/8, turns out to be 58.6 million man-days.
Next, we require an estimate of the level of unemployment that would have prevailed among the landless in the absence of the EGS. For this purpose we use the data from the 1974/5 Rural Labour Enquiry, as the EGS was still operating on a modest scale in that year. Unfortunately no data is available separately for the landless labourers; so we must take the rate of unemployment prevailing among the primarily labour-selling households instead. Assuming a standard man-year to consist of 300 days, the extent of unemployment turns out to be 86 days per worker (average of male and female). From the data presented by Dandekar and Sathe (1980), it appears that in 1977/8 there were about 3.1 million landless workers in the whole of rural Maharashtra.39 This gives a total unemployment figure among the landless labourers of about 266 million man-days. Recalling that a maximum of 58.6 million days of employment came their way by courtesy of the EGS, it would thus appear that the scheme was able to eliminate no more than 22 per cent of unemployment among this group in 1977/8, and that under a host of liberal assumptions. The size of the EGS has of course expanded since then; for example, the volume of employment generated in 1979/81 was 60 per cent higher than that in 1977/8 (Herring and Edwards 1983). But since the landless population has also expanded at the same time (albeit at a slower rate), the total impact could hardly amount to more than a third of unemployment among landless labourers.
(p.336) Illustrative as they are, the preceding calculations show that perhaps too much has been made of the guarantee aspect of the Maharashtra scheme. What is not in question, though, is its relatively superior performance compared with similar schemes elsewhere in South Asia. This, combined with the fact that the scheme is still expanding, invites the question: does not the Maharashtra experience indicate that it is possible to achieve much more by wa V of providing security through employment than has been the experience so far in the rest of the region? In order to form a perspective on this matter, it is first necessary to take cognizance of a special feature of the Maharashtra scheme.
As mentioned before, the EGS was initiated as a fully State-financed project without any financial support from the central government. Half of the finance was raised by means of special taxes legislated for this purpose and the other half came from the general revenues of the State government. It is highly instructive to look at the range of these special taxes. The bulk of the revenue was extracted from the urban sector, particularly from the city of Bombay. Taxes on agriculture were originally supposed to contribute roughly one-fifth of the total EGS revenue. In the event their share turned out to be a maximum of 3 per cent during the period between 1975/6 and 1978/9, while as much as 60 to 70 per cent was collected from the city of Bombay alone (Herring and Edwards 1983: 585). As opposed to this urban orientation of finance and almost total exemption of the rural sector from the sharing of cost, the distribution of benefit was disproportionately in favour of the richer farmers. It is true that the workers also benefited during the course of specific projects but the lasting benefits went to farmers in the form of assets that were created under the EGS. It is of course a commonplace that asset creation through rural works generally benefits the landed people more; but this commonplace assumes a particular poignancy in rural Maharashtra because of the nature of assets created by the EGS. While earth-roads and other non-durable assets usually dominate the portfolio of assets in a typical rural works programme, a predominant share in Maharashtra was taken up by irrigation works of various kinds. As an official evaluation of the EGS showed, these irrigation works and other agriculture-related land development works yielded remarkable benefits to the farmers, specially the middle and large farmers (PEO 1980a)
This contrast in the distribution of costs and benefits of the EGS is not unrelated to the fact that the State politics of Maharashtra are known to be strongly dominated by the richer landowners (MHJ 1980; Herring and Edwards 1983). Of course, they were assisted in this matter by the presence of a very rich urban sector. Maharashtra is one of the wealthiest States in India in terms of per capita income but one of the poorest in terms of rural income, which means that almost all of its enormous wealth originates from the prosperous urban sector. In a spectacular reversal of the ‘urban bias’ hypothesis, the politically powerful rural elite of Maharashtra has succeeded in extracting surplus wealth from the urban sector, combined it with the raw labour of a (p.337) marginalized rural population, and created assets for the benefit of their own land.40
The secret of Maharashtra thus lies in forming a triangle between a dominant rural elite, a materially prosperous urban sector which for some reason was willing to foot the bill, and a poor but abundantly available rural labour-force. The third corner of this triangle can be safely assumed to exist all over South Asia, but one cannot be too certain about the first two. This is what sets the limit to the possibility of emulating Maharashtra in the rest of the region.
There are, of course, other limits too. If an employment programme is to be something more than a pure relief operation, it must create assets that are not only productive but are also capable of creating long-term employment, ideally leading to a stage where special programmes for employment generation will no longer be needed. Experience shows that well-executed projects for irrigation and flood control can have a significant effect on both crop production and long-term employment (BIDS-IFPRI 1985). The FFW in Bangladesh is indeed slanted heavily in favour of such agriculture-related activities, but there is some doubt as to how well they are executed (Asaduzzaman and Huddleston 1983). The situation is somewhat different in India, where the preponderance of ‘roads that are washed away’ have been a source of constant criticism (Basu 1981; Bandyopadhyay 1985).
The transitory nature of the benefit accruing from employment programmes and the dependency they allegedly tend to generate among workers on continuous support from public works, have led some scholars to argue that the creation of self-employment through IRDP-type programmes is a superior mode of supporting the poor (see, for example, Dantwala 1985). In contrast, others, such as Rath (1985) and Dandekar (1986), have argued that wage employment is the only feasible means of supporting the poor since the majority of them do not possess the entrepreneurial skill needed for doing well in IRDP-type programmes. The controversy that has ensued over ‘wage employment versus self-employment’ does seem to be rather sterile, however We have noted in the preceding section that a blanket denial of the entrepreneurial acumen of the poor is hard to sustain; mány of them can and do make a regular living out of self-employed activities. On the other hand, it is also true that many may not want to take the risks and troubles of independent occupations, and may much prefer to seek their livelihood in the labour market. There is no reason why both types of employment cannot be generated in a complementary fashion for furthering the security of livelihood of the poor. This will, however, require truly integrated local-level planning of the kind discussed in the preceding section. Without such planning, neither mode of supporting the poor will have a sustained effect—the programmes for self-employment will fail to be productive enough to raise permanently the asset (p.338) base of the poor, and the programmes for wage-employment will fail to create the assets required for raising the long-term employment-generating capacity of the economy. Whether the countries of South Asia have the ability to undertake such poverty-focused local-level planning on an extensive scale is a moot question. What is clear, however, is the inference that in the absence of such ability the programmes of employment promotion will fail to provide lasting security to the poor. Some temporary relief will indeed be provided and, if operated on a modest scale, even lasting security can be ensured for a small segment of the poor, but not generalized long-term security for the bulk of the poor. In that event, other methods of providing social security will have to be seriously contemplated.
5. Security Through Public Provision of Basic Needs
If the goal of social security is to ensure guaranteed access to a certain minimal standard of living, and the standard of living is itself defined in terms of functionings and capabilities (Sen 1985), then the policies and programmes discussed in the preceding sections are best described as indirect means of achieving that goal. They are supposed to provide the resources with which the beneficiaries will hopefully acquire the inputs—such as food, clothing, housing, health care, education, and so on—which will in turn create the desired ways of functioning, such as living a healthy active life. An alternative, and potentially complementary, approach is to make those inputs more easily accessible to the people so that more of them can be acquired with given resources. When public policy takes this latter, more direct approach we may describe it as social security through public provision of basic needs. The attribute of directness, however, does not necessarily confer upon it the attribute of superiority. But the empirical evidence in South Asia does make a strong case for emphasizing the direct approach. The experience of Sri Lanka and the Indian state of Kerala has already attracted widespread attention in this regard. Before looking at their record of public provision, we may take a glimpse of their achievement from the comparative picture laid out in Table 7.1.
Table 7.1. Comparative quality of life in Bangladesh, all-India, Kerala, and Sri Lanka
Infant mortality rate (per 1,000)
Crude death rate (per 100)
Per capita income (US $)
(a) Reference period varies as follows: 1981 for IMR, 1984 for CDR, 1980 for life expectancy, 1981 for literacy, and 1985 for per capita income.
(b) Except for per capita income for which the reference period is 1985, and see (d), below.
(c) Refers to rural Kerala for the period 1976–80.
(d) References period is 1981 in all cases. The reference population varies as follows: adults of 15 years of age and over in Bangladesh, all people over 10 years of age in Sri Lanka, and all people of 5 years and above in India and Kerala.
Sources: For Bangladesh data on IMR, death rate, life expectancy, and literacy are from World Bank (1988: statistical appendix, tables 1.3, 1.7). Indian data on IMR, death rate, and life expectancy are from GOI (1985: vol. 2, table, 2.11). Literacy rate is from Vaidyanathan (1987: table 1). IMR and literacy data for Kerala are from Vaidyanathan (1987: table 1) death rate and life expectancy are from Kumar (1987: tables 2, 4). Sri Lankan data on IMR, death rate, life expectancy, and literacy are from Rasaputra (1986: app. tables 8,9). Data on per capita income of India, Bangladesh, and Sri Lanka are from Bhatia (1988: table 2). The estimate for Kerala was derived from the figure for India by applying the ratio between the per capita State domestic product of Kerala and the per capita NNP of India for 1984/5 as given in Agarwal et al (1987: 39).
By all the indicators of quality of life presented in Table 7.1, Sri Lanka and Kerala are clearly ahead of the rest of South Asia. Their lead can be appreciated even better by noting how long ago they had achieved the levels that Bangladesh and India (as a whole) are enjoying only now. The levels of infant mortality, the death rate, and life expectancy obtaining in India today were reached by Sri Lanka in the late 1940s and the corresponding level of literacy was achieved as early as the second decade of this century (Rasaputra 1986: app. tables 8, 9). Similarly the current all-India levels of mortality and life expectancy were already achieved in Kerala in the early 1960s, while literacy was achieved even (p.339) earlier41 Differences in the levels of per capita income cannot explain all this. It is true that Sri Lanka enjoys a higher per capita income than the rest of the region, but not high enough to explain the difference in quality of life.42 The dissonance between per capita income and quality of life is manifest even more starkly in the case of Kerala.
As for other possible explanations, our preceding discussion shows that neither land reforms nor the employment-generating programmes could possibly hold the clue. In Sri Lanka, as we have seen, land reforms have had a negligible impact on the poor, and programmes for the generation of employment were not even tried there on the same scale as has been attempted elsewhere in the (p.340) region. Kerala has of course experienced a more effective land reform, but the resulting distribution of land is not any more equitable than that of India as a whole. In fact, in absolute terms the rural poor of Kerala control a much smaller size of land per capita because of the extreme density of population.43Nor is there any evidence that Kerala did spectacularly better in respect of the employment-generating programmes; in any case these are of fairly recent origin while Kerala’s lead dates back long into the past.
It is in respect of public provision of basic needs that Sri Lanka and Kerala stand out with striking distinction. Both tried to ensure for every citizen a guaranteed access to three basic amenities—food, health care, and education. It will be instructive to see how their approach differed in this area from the rest of the region.
The history of public distribution of food-grain in South Asia dates back to the crisis years of the Second World War. The elaborate network of rationing set up during the war was retained afterwards, but since then it has evolved differently in the three countries. In Sri Lanka it was further consolidated and extended so as to cover the entire population; in Bangladesh too it was steadily consolidated but its principal focus remained on the urban population; in India it alternated between periods of consolidation and abandonment (or weakening), but focusing always on the urban sector whenever it was in force (except in Kerala, as we shall see).
The rural segment of the rationing system in Bangladesh has all along been a residual claimant on available food-grain. The priority claim was vested in the urban dwellers, especially those engaged in the formal sector. Even so, rural rationing accounted for about 55 per cent of total distribution in the 1960s. Since then its share has progressively come down, standing at less than 20 per cent in the 1980s (MOF 1986). As a result, the system is now capable of catering to only a small fraction of eligible rural households; and even those receive only a meagre ration To illustrate, a sample survey conducted in the mid-1980s found that the contribution of subsidized ration amounted to only 2 per cent of total income for an average recipient household (MOF 1986).
The official justification for the scaling down of rural rationing has rested partly on the ground that one other component of the distribution system, namely Food for Work, has been making an increasing contribution to food-grain availability in the rural areas . There are at least two problems with this justification. First, even after allowing for the food-grain distributed under Food for Work, only about 6 to 8 per cent of the rural population was estimated to have been covered by public distribution in the second half of the 1970s (MOF 1986: table 7.3). Secondly, an argument based on total availability (p.341) cannot be enough to justify the withdrawal of guaranteed food security for the vulnerable population. While it is true, as we have seen, that FFW not only increases availability but also distributes the benefit mostly to the landless and land-poor population, it cannot ensure security for such households as are handicapped by the absence of required labour power. Moreover, the nature of FFW activities is such that they can be undertaken primarily in the dry season, leaving out a protracted slack period in agriculture during the rainy season. The substitution of rural rationing by FFW cannot adequately take care of the problem of food security in the latter period.
Even in the favoured urban sector, the scope of security given to the neediest people has shrunk in recent years. Since the early 1970s, new ration cards have not been issued in the urban areas except under special circumstances. This has left out of the security net the vast number of rural immigrants who flocked into the urban areas, particularly the capital city, in the immediate post-1971 period. As a result, the proportion of urban residents covered by subsidized rationing has come down drastically from near 100 per cent to just around 50 per cent in the 1980s (MOF 1986: table 7.3). That the left-out segment consists primarily of the poorer group is indicated by a recent finding that the average household income of urban ration-receivers is higher than the average income of all urban households (MOF 1986).
All this has happened as a result of a basic shift that has occurred in the whole underlying philosophy of the public distribution system. Instead of offering targeted security to individual households, the aim now is to ensure a generalized security in times of crisis by stabilizing prices through open market interventions. In this aim some notable success has indeed been achieved. First in 1979 and again in 1984, the capability of the public distribution system to intervene vigorously in the open market helped contain speculative price hikes of the kind that wreaked havoc in the famine of 1974, despite the fact that the potential scarcity was greater in the later two years.44 This story of success however needs to be qualified in two important respects. First, the country has not yet acquired the ability permanently to sustain a level of buffer stock that would guarantee similar success in the future. In both 1979 and 1984 the crisis was averted by taking recourse to huge imports—financed in one instance by unusual generosity on the part of the international aid community, and by the bounty of an unexpected trade surplus in another (Osmani, 1990). Secondly, even if the task of crisis management is adequately performed, there remains the question of that huge number of vulnerable households who do not have the security of a minimum consumption of food in normal times. The public distribution system of Bangladesh was never adequate to face up to this task; today it is utterly irrelevant for this purpose.
(p.342) A shift in policy similar to that of Bangladesh is also evident in India, especially since the dawn of the so-called Green Revolution era (around 1967/ 8). In the earlier phase, the aim of public food distribution, as in Bangladesh, was to offer targeted security to the urban population. Its potential role as a mechanism for price stabilization, and thus as a purveyor of generalized security, first received official recognition during the drought years of 1965/6 and 1966/7. Since then ‘. . . its role has undergone a basic change: from an agency for distributing rations to the target groups under the scheme of statutory and non-statutory rationing, it has now become an instrument of agricultural price support and price stabilization policies of the Government’ (Bhatia 1983: 69). In contrast to Bangladesh, however, India has built up a huge buffer stock from its own production, which gives her a more secure basis for managing crisis. This ability was tested most severely in 1979/80 when the production of food-grains fell by 18 per cent from the previous year’s level, and in the ensuing three years when production stagnated at the level of the mid- 1970s. That this protracted crisis was averted without a serious price hike is largely attributable to the existence of the buffer stock.45
While recognizing this success in assuring some degree of generalized security in times of crisis, one must also point out that, as in Bangladesh, the public distribution system in India is incapable of offering targeted security to those who need such security even in normal times. In the case of India, moreover, this failure assumes an added irony. The buffer stock has arisen in the first place because the development strategy pursued in India has failed to engender enough effective demand among the majority of its people while production was being pushed up by the Green Revolution, and the government was obliged to take up the slack in fulfilling its commitment to agricultural price support.46 It would thus appear that India has acquired the ability to offer generalized security in times of crisis by neglecting the lack of security that remains endemic for a large section of the people.
A striking contrast is presented by the Indian state of Kerala. Its distribution system covers almost the entire population (97 per cent) at a level of per capita off-take that is by far the highest among all the Indian states. In 1977, for example, the per capita off-take in Kerala (61 kg per annum) was more than double that of all the states except two (Jammu and Kashmir with 47 kg and West Bengal with 40 kg). Furthermore, the rural-to-urban ratio of per capita off-take was almost 100 per cent in Kerala, whereas in other states the ratio varied between 2 and 22 per cent. As a result, the rural per capita off-take (p.343) in Kerala stood at 61 kg as against a range of 0.8 to 15.5 kg in other states.47
The sheer scale and spread of public distribution in Kerala means that the rationed food-grains figure prominently in household consumption. A survey conducted in two villages in 1977 showed that the rice purchased from ration shops accounted for two-thirds of all the rice consumed by the poorest 40 per cent of households (George 1979: 32).
An even more ambitious programme of subsidized food distribution was in force in Sri Lanka until recently. The coverage was, as in Kerala, near universal. Food-grain was distributed through the rationing system—in varying combinations of totally free or heavily subsidized amounts at different times. The principle of universal coverage was abandoned in 1978 when the richer 50 per cent of the population became ineligible for subsidized rationing. In 1979, the system of rationing was itself abolished and replaced by a food stamp scheme, targeted once again at the bottom half of the population. These stamps were redeemable in exchange for food-grain (and a few other essential commodities) sold in designated outlets at unsubsidized prices. Initially, the change-over from rationing to food stamp did not make much difference to the bottom quintile of the population in terms of the subsidy received.48 But since the value of stamps was kept fixed in nominal terms, their real value became quickly eroded in the face of rapid inflation; by 1981/2, the real value had nearly halved (Edirisinghe 1987: 18).
During the period of subsidized rationing the amount of rationed food-grains figured prominently, as in Kerala, in the total consumption of Sri Lankan households. In 1969/70, for example, it was estimated from a nationwide survey of household expenditure that rationed rice amounted to 18 per cent of the calories consumed by all households, and to 21 per cent in the case of the bottom quintile (Gavan and Chandrasekera 1979). A similar survey in 1978/9 showed that the value of the ration subsidy was the equivalent of 18 per cent of total expenditure of all households, and 25 per cent of total expenditure of the bottom quintile (Edirisinghe 1987).
An example of how well the rationing system was able to protect the poor is provided by the experience of 1973/4 when rice distribution through rationing was reduced by 20 per cent. Production was above average and total calorie availability was only slightly below the trend level, and yet in 1974 the mortality rate increased from 7.7 to 8.9 (Isenman 1980). It is also significant that the incidence of increased mortality was most pronounced among the estate workers who depended on the rationing system much more heavily than (p.344) the rest of the population. It will perhaps be wrong to attribute the entire increase in mortality to reduced rationing, because during the same period there was also a sharp increase in the price of wheat, which the estate workers tend to consume in large amounts (Gavan and Chandrasekera 1979). Nevertheless the fact that increased mortality was also found among the rest of the population, who consume very little wheat, does point to the presence of a rationing effect as well. Further evidence can be gleaned from the changes in nutritional status that have reportedly occurred after the subsidy was halved within three years of the introduction of food stamps in 1979. The incidence of wasting among children in the age-group of 6 to 60 months has gone up from 8.4 per cent in 1975/6 to 13.8 per cent in 1980–2 (Sahn 1987: 814).
Despite such evidence, there must be some doubt about how far the spectacular indices of mortality and life expectancy achieved by Kerala and Sri Lanka are due to the security provided by subsidized food-grain. This is best appreciated by noting the extent of the contribution made by the subsidy in the total consumption of calories. We have seen earlier that the value of subsidy as a proportion of total income is remarkably high in both Kerala and Sri Lanka. But the value of the subsidy as such, or the calorie-equivalent of that subsidy, does not indicate the amount of calories contributed by the subsidy. The contribution would depend on the marginal propensity to consume calories out of the additional income coming by way of subsidy, plus any possible substitution effect. It would appear from the (admittedly crude) estimates made by George (1979: 37) that the net contribution of the subsidy was about 18 per cent of rice consumption among the bottom 40 per cent of the households. Since rice accounts for about 40 per cent of calories in an average Keralite diet, the net contribution of subsidy amounts to a maximum of 7 to 8 per cent of total calories.49 A similar figure—6 per cent for the bottom quintile of households—seems to emerge from the estimates for Sri Lanka (1969/70) made by Gavan and Chandrasekera (1979: 41).50
Even considering the margin of error that is likely to be associated with these estimates, it will perhaps be fair to conclude that the net contribution of the subsidy to the total intake of calories was nothing spectacular. It is difficult to see how a contribution of this magnitude could explain the phenomenal progress made with respect to mortality and life expectancy. This impression is further reinforced by looking at the actual levels of calorie intake enjoyed by the poorer groups. The overall per capita intake in Sri Lanka, estimated at 2,010 Kcal per day in 1978/9, is somewhat lower than in India; but even more (p.345) importantly, the average intake for the bottom half of the population is only 1,607 Kcal (Anand and Harris 1987: 6a), a figure that is way below any reasonable standard of requirement. In Kerala, the situation is even more precarious according to the official estimates of the National Sample Survey; even the average intake of the total population appears to be some 20 per cent below the national average. There are disputations about these official figures which, according to some studies, do not take sufficient note of the varieties of non-cereal sources from which the Keralites get their calories.51 But even the findings of smaller surveys which have tried to capture the variety in diet do not alter the basic conclusion that the poor in Kerala do not consume more calories than the poor in the rest of India. One such study, looking at some of the South Indian States, found that a poor Keralite adult consumes less calories (1,751) than his or her counterpart in the neighbouring states of Andhra Pradesh (1,847), Karnataka (1,902), and Tamil Nadu (2,019).52
The preceding discussion is not meant to deny the importance of public food-grain distribution in Kerala and Sri Lanka but to put its role in perspective. As the Sri Lankan experience in 1973/4 and again in 1980–2 shows, it has clearly made a difference to the nutritional achievement of the population. However, the extraordinary improvement made in mortality and life expectancy must be attributed largely to other factors.
The most obvious candidate in this respect is the singular effort made by both Kerala and Sri Lanka in the field of public provision of health care and education. Public action in these fields has a long history in both these regions, dating back far into the colonial era.53 In interpreting this history, however, there has sometimes been a tendency to emphasize certain fortuitous circumstances and a concurrent tendency to de-emphasize the role of public provisioning as such. For example, in explaining the reduction in mortality in Sri Lanka, Bhalla (1988b) has made much of the malaria eradication programme that was undertaken in 1946 with the help of the newly available DDT. It is true that this programme had a most dramatic result—the crude death rate fell from 20 to 14 and infant mortality fell from 141 to 101 in a single year—a phenomenon rightly described by the World Health Organization as an unparalleled achievement in world demography (Gunatilleke 1984). Nevertheless, it would be wrong to ascribe Sri Lankan health achievement primarily to this programme. The most careful assessment of the programme made to date shows that it can explain only about 20 to 25 per cent of the improvement in the death rate (p.346) between 1936–45 and 1956–60 (Gray 1974). In fact throughout the period from 1920 up to the present, one can notice a secular improvement in all the health indicators, fortified of course by the structural break of 1946. Furthermore, the tempo of improvement has been more or less maintained in the post-independence (which is also the post-malaria) period. For instance, infant mortality dropped at the annual compound rate of 2.7 per cent between 1920 and 1950, and at the rate of 2.8 per cent between 1950 and 1982. Similarly, the rates of improvement in life expectancy (measured as the reduction in shortfall from a ceiling of 80 years) for the corresponding periods were 2.6 per cent and 2.3 per cent respectively (see Rasaputra 1986: app. table 9).
The case of Kerala is even more clear-cut. It did undertake a malaria eradication programme in the 1950s, but that was part of an all-India campaign and cannot as such explain the difference in the health performance between Kerala and the rest of India. In fact both Sri Lanka and Kerala have achieved what they have, not by a one-off campaign to conquer malaria, but by sustained efforts to provide either free or heavily subsidized health services to their entire populations. Some basic statistics would be illuminating. In the mid-1980s, Sri Lanka had one hospital bed for every 350 persons as against a ratio of one to 1,200 persons in India. More importantly, Sri Lanka has always paid special attention to maternal health care, which has contributed significantly to the reduction of infant mortality. In the mid-1970s, for instance, Sri Lanka had one maternity centre for every 400 expectant mothers as against one for every 3,200 in India; as a result, nearly 98 per cent of Sri Lanka mothers received pre-natal care as against 10 to 15 per cent in India (Gwatkin 1979). Similarly, a comparison with the all-India performance shows Kerala in a distinctly impressive light. For example, in 1977, there were 1,975 hospital beds per million of population in Kerala as against 791 in India as a whole (Jose 1984). More importantly, the health facilities were spread widely across the State instead of being concentrated in the urban areas, as is usually the case in the rest of India as well as other developing countries. As a result, some 80 per cent of Kerala’s population are treated in hospitals every year compared with 40 to 50 per cent in the other States of India (CDS 1975).
It is important to recognize, however, that it is not merely the physical volume of facilities that has made the difference. In fact there are States in India which out-perform Kerala in terms of indices such as per capita expenditure on health, the number of persons per hospital bed, and even rural medical facilities per person, and yet have inferior indices of health compared to Kerala.54 What has made the difference is the efficiency with which medical facilities have been used. There are two aspects to this efficiency. First, in both Kerala and Sri Lanka, at least as much attention has been paid to preventive health care as to curative service, and this proved to be a highly cost-effective (p.347) method of reducing mortality. Secondly, a highly health-conscious population made sure that the existing facilities were effectively utilized.55
The degree of awareness about health was itself the result of a long history of public provision of education. That the spread of education, apart from being a good thing in itself, was also instrumental in bringing about an efficient health care system in both Sri Lanka and Kerala has been noted by many commentators. The Sri Lankan scholars, for example, never tire of pointing out how politicians had to move towards strengthening the welfarist State in response to the felt need of an educated electorate after the Donoughmore Constitution had granted universal adult franchise in 1931.56 In Kerala, an added force came from the high degree of politicization created among the masses by the active presence of communist parties at the grass-roots level. Mencher (1980), for example, has noted from her field experience how the educated and politicized people of rural Kerala make sure that the existing health centres are fully geared to their service, while similar health centres in the neighbouring State of Tamil Nadu remain inactive or underutilized.57
The problem of poverty in South Asia is both massive and acute. The task of providing the security of a minimal standard of living cannot, therefore, be an easy one. This sobering thought recurs time and again in our review of the programmes and policies that have been pursued in this part of the world. The only encouraging note comes from the experience of some parts of the region where a good deal more has been achieved compared with the rest. Correspondingly, some of the policies seem to hold a much greater promise than others. It was indeed the major aim of this chapter to evaluate the relative possibilities and limitations of alternative courses of action so as to identify the more promising ones.
This task cannot by any means be said to have been completed. For each of the policies we have considered, there is need for a much more thorough probing of its political, financial, and organizational implications. Besides, there are a good number of alternatives which we have not considered at all. Nevertheless, it may be useful to lay out some of the tentative conclusions that emerge from this preliminary analysis.
As expected, the most difficult policies to implement are those that try to provide security through control over land. This could hardly be otherwise in a (p.348) political system which is based on the primacy of private ownership rights. But even where a more favourable political condition prevails, as in West Bengal, nothing substantial has been achieved. The share-croppers may have gained a more assured security of tenure, but in the majority of cases their rental share has not improved. Consequently, the security of a better standard of living still remains an unattained goal. We have in fact argued that tenancy reform is not likely to achieve this goal until there is prior provision of social security through other public actions.
Among these other actions, the most popular in current official thinking is the set of programmes whose aim is to create employment—either self-employment in non-crop activities or wage employment in public works . There is no doubt that these programmes have an immediate appeal . Those who receive credit and/or subsidy for self-employed activities, or get wage employment in the slack season certainly receive a temporary benefit. What is crucially missing from these programmes, however, is the ability to provide a basis for a sustained increase in the standard of living. This deficiency is in principle remediable, but this will require a fundamental change in the orientation of these programmes. Instead of being seen purely as a support programme for the poor, the employment-generating schemes must be conceived as an integral part of the development process itself. In order that employment can generate a sustained flow of a higher level of income, it is first necessary to create viable employment opportunities. This in turn requires integrated planning of available resources—both human and material. Obviously, the distinction between social-security programmes and development planning tends to get blurred in the process. But this is inevitable in a situation where security is sought to be provided through income generation, encompassing the majority of the population rather than a few.
The alternative option of providing security through public provision of basic needs has enjoyed much less popularity at the official level. The exceptions are Sri Lanka and the Indian State of Kerala; and both have been outstandingly successful in improving the quality of lives of their people. Their poverty has not been conquered. In fact, in both Sri Lanka and Kerala people are desperately poor. But a long history of public provision of such basic needs as food, health, and education has given them the basic security of avoiding premature mortality due to disease and malnutrition. We have argued that much more significantly than the provision of food, it is the provision of basic health care and education that has played the crucial role. A similar course of action is also likely to prove the most promising way of providing social security to the rest of the region.
But can such a programme be financed by the poor countries of this region without sacrificing long-term growth? This question assumes a particular significance in view of an ongoing debate on whether Sri Lanka has jeopardized her long-run prospects of growth by choosing to become excessively ‘welfarist’, (p.349) that is, by ignoring her development needs in the pursuit of ‘basic needs’. This is not the occasion to enter this lively debate;58 we shall only make a couple of simple observations. First, as the experience of Kerala shows, the success of a health care programme may depend more on how efficiently the facilities are utilized than on the amount of money spent. Secondly, even granting that additional resources will have to be allocated to the welfare programmes, it is not clear why this should be seen as a dent in the resources for growth. ‘Welfare’ and ‘development’ (meaning growth) are not the only heads of expenditure in a typical government budget. It is significant that while spending heavily on welfare programmes, Sri Lanka kept its defence spending at a surprisingly low level of around 2 per cent of total government expenditure (in the 1970s), compared with India’s 10 to 15 per cenŧ Obviously, ‘growth’ need not always be the opportunity cost of social-welfare programmes.
Agarwal, A. N., Verma, H. O., and Gupta, R. C. (1987), India. Economic Information Yearbook 1987–1988, 2nd edn. New Delhi: National Publishing House.
Ahmad, Q. K., and Hossain, M. (1985), ‘An Evaluation of Selected Policies and Programmes for the Alleviation of Rural Poverty in Bangladesh’, in R. Islam (ed.), Strategies for Alleviating Poverty in Rural Asia. Dhaka: BIDS; and Bangkok: ILO-ARTEP.
Alailima, P. (1985), ‘Evaluation of Government Policies and Expenditure on Social Welfare in Sri Lanka During the 20th Century’, mimeo, Ministry of Finance and Planning, Colombo.
Anand, S., and Harris, C. (1987), ‘Changes in Nutrition in Sri Lanka, 1978/79-1981/ 82’, mimeo, World Institute for Development Economics Research, Helsinki.
——— and Kanbur, R. (1990), ‘Public Policy and Basic Needs Provision: Intervention and Achievement in Sri Lanka’, mimeo, to be published in J. P. Drèze, and A. K. Sen (eds.), The Political Economy of Hunger. Oxford University Press.
Asaduzzaman, M., and Huddleston, B. (1983), ‘An Evaluation of Management of Food for Work Programme’, Bangladesh Development Studies, 11/1 and 2.
Bagchee, S. (1984), ‘Employment Guarantee Scheme in Maharashtra’, Economic and Political Weekly, 19/37.
——— (1987), ‘Poverty Alleviation Programmes in the Seventh Plan: An Evaluation’, Economic and Political Weekly, 22/4.
Bandyopadhyay, D. (1985), ‘An Evaluation of Policies and Programmes for the Alleviation of Rural Poverty in India’, in R. Islam (ed.), Strategies for Alleviating Poverty in Rural Asia. Dhaka: BIDS; Bangkok: ILO-ARTEP.
——— (1986), ‘Land Reforms in India: An Analysis’, Economic and Political Weekly, Review of Agriculture, 21/25–6.
——— (1988), ‘Direct Intervention Programmes for Poverty Alleviation: An Appraisal’, Economic and Political Weekly, Review of Agriculture, 23/26.
Bandyopadhyaya, N. etal. (1985), Evaluation of Land Reform Measures in West Bengal, A Report. Bangkok: ILO-ARTEP.
Basu, K. (1981), ‘Food for Work Programmes: Beyond Roads that Get Washed Away’, Economic and Political Weekly, 16/1.
BBS (1985), Statistical Yearbook of Bangladesh 1984/85. Dhaka: Bangladesh Bureau of Statistics.
Bhalla, S. S. (1988a), ‘Is Sri Lanka an Exception? A Comparative Study of Living Standards’, in T. N. Srinivasan and P. K. Bardhan (eds.), Rural Poverty in South Asia. New York: Columbia University Press.
——— (1988b), ‘Sri Lanka’s Achievements: Fact and Fancy’, in T. N. Srinivasan and P. K. Bardhan (eds.), Rural Poverty in South Asia. New York: Columbia University Press.
——— and Glewwe, P. (1987), ‘Growth and Equity in Developing Countries: A Re-interpretation of Sri Lankan Experience’, World Bank Economic Review, 1.
Bhatia, B. M. (1983), A Study In India’s Food Policy. Kuala Lumpur: Asian and Pacific Development Centre.
BIDS-IFPRI (1983), Characteristics and Short-run Effects of the WFP-aided Food for Work Programme in Bangladesh. Dhaka: Bangladesh Institute of Development Studies; Washington DC: International Food Policy Research Institute.
——— (1985), Development Impact of the Food-for-Work Program in Bangladesh, Final Report, Summary. Dhaka: Bangladesh Institute of Development Studies; Washington DC: International Food Policy Research Institute.
Bose, B. (1981), ‘Agrarian Programme of Left Front Government in West Bengal’, Economic and Political Weekly, 16/50.
Bose, S. (1986), Agrarian Bengal: Economy, Social Structure and Politics 1919–1947. Cambridge: Cambridge University Press.
Caldwell, J. C. (1986), ‘Routes to Low Mortality in Poor Countries’, Population and Development Review, 12.
CDS (Centre for Development Studies) (1975), Poverty, Unemployment and Development Policy: A Case Study of Selected Issues with Special Reference to Kerala. New York: United Nations.
Dandekar, K. and Sathe, M. (1980), ‘Employment Guarantee Scheme and Food for Work Programme’, Economic and Political ’Weekly, 15/15.
Dandekar, V. M. (1986), ‘Agriculture, Employment and Poverty’, Economic and Political Weekly, 21/38–9.
Dantawala, M. L. (1985), ‘“Garibi Hatao”: Strategy Options’, Economic and Political Weekly, 20/11.
Dasgupta, B. (1984), ‘Sharecropping in West Bengal: From Independence to Operation Barga’, Economic and Political Weekly, Review of Agriculture, 19/26.
de Janvry, A., and Subbarao, K. (1986), Agricultural Price Policy and Income Distribution in India. New Delhi: Oxford University Press.
Drèze, J. P. (1988), ‘Social Insecurity in India’, paper presented at the STICERD/ WIDER workshop on Social Security in Developing Countries held at the London School of Economics, July 1988.
——— and Sen, A. K. (1989), Hunger and Public Action. Oxford: Clarendon Press.
Edirisinghe, N. (1987), ‘The Food Stamp Scheme in Sri Lanka: Costs, Benefits and Options for Modifications’, Research Report 58, International Food Policy Research Institute, Washington, DC.
Gavan, J. D., and Chandrasekera, I. S. (1979), ‘The Impact of Public Foodgrain Distribution on Food Consumption and Welfare in Sri Lanka’, Research Report 13, International Food Policy Research Institute, Washington, DC.
George, P. S. (1979), ‘Public Distribution of Foodgrain in Kerala: Income Distribution Implications and Effectiveness’, Research Report 7, International Food Policy Research Institute, Washington, DC.
GOI (Government of India) (1959), ‘Report of the Committee on Tenancy Reform’, in Reports of the Committees of the Panel on Land Reforms. New Delhi: Planning Commission, Government of India.
——— (1981), National Committee on the Development of Backward Areas: Report on General Issues. New Delhi: Planning Commission, Government of India.
——— (1985), The Seventh Five Year Plan 1985–1990. New Delhi: Planning Commission, Government of India.
Gray, R. H. (1974), ‘The Decline of Mortality in Ceylon and the Demographic Effects of Malaria Control’, Population Studies, 28.
Guhan, S. (1988). ‘Social Security in India: Looking One Step Ahead’, Bulletin of the Madras Development Seminar Series, 18.
Gunatilleke, G. (ed.), (1984), ‘Intersectoral Linkages and Health Development: Case Studies in India (Kerala State), Jamaica, Norway, Sri Lanka and Thailand’, WHO Offset Publication No. 83, World Health Organization, Geneva.
Gwatkin, D. R. (1979), ‘Food Policy, Nutrition Planning and Survival: The Cases of Kerala and Sri Lanka’, Food Policy, 4.
Hameed, N. D. A. .,etal. (1977), Rice Revolution in Sri Lanka. Geneva: United Nations Research Institute for Social Development.
Herring, R. J. (1981), ‘Embedded Production Relations and Rationality of Tenant Quiescence in Tenure Reform’, Journal of Peasant Studies, 8.
——— and Edwards, R. M. (1983), ‘Guaranteeing Employment to the Rural Poor: Social Functions and Class Interests in the Employment Guarantee Scheme in Western India’, World Development, 11.
Hossain, M. (1984a), ‘Productivity and Profitability in Bangladesh Rural Industries’, Bangladesh Development Studies, 12.
——— (1984b), Credit for the Rural Poor: The Grameen Bank of Bangladesh, Research Monograph No. 4, Bangladesh Institute of Development Studies, Dhaka.
——— (1986), ‘Employment Generation Through Cottage Industries: Potentials and Constraints’, in R. Islam and M. Muqtada (eds.), Bangladesh: Selected Issues in Employment and Development. New Delhi: ILO-ARTEP.
——— (1988), ‘Credit for Alleviation of Rural Poverty: The Grameen Bank in Bangladesh’, Research Report 65, International Food Policy Research Institute, Washington, DC.
IFMR (1984), ‘An Economic Assessment of Poverty Eradication and Rural Employment Programmes and Their Prospects’, mimeo, Institute for Financial Management and Research, Madras.
ILO (1984), ‘Summary Results of the Socio-economic Survey on Sample Projects of the National Employment Programme’, mimeo, Employment and Development Department, International Labour Organization, Geneva.
Isenman, P. (1980), ‘Basic Needs: The Case of Sri Lanka’, World Development, 8.
Islam, R. (1986), ‘Rural Unemployment and Underemployment: A Review’, in R. Islam and M. Muqtada (eds.), Bangladesh: Selected Issues in Employment and Development. New Delhi: ILO-ARTEP.
——— (1987), ‘Rural Industrialisation and Employment in Asia: Issues and Evidence’, in R. Islam (ed.), Rural Industrialisation and Employment in Asia. New Delhi: ILO-ARTEP.
Jose, A. V. (1984), ‘Poverty and Inequality: The Case of Kerala’, in A. R. Khan and E. Lee (eds.), Poverty in Rural Asia. Bangkok: ILO-ARTEP.
Joshi, P. C. (1975), Land Reforms in India: Trends and Perspectives. Bombay: Allied Publishers.
Kohli, A. (1987), The State and Poverty in India. Cambridge: Cambridge University Press.
Kumar, G. (1987), ‘Well-being, Deprivations and Basic Needs: A Perspective on the Development Experience in Kerala’, mimeo, Balliol College, Oxford.
Kurian, N.J. (1987), ‘IRDP: How Relevant Is It?’, Economic and Political Weekly, 22/ 52.
Mencher, J. P. (1980), ‘Lessons and Non-Lessons of Kerala: Agricultural Labourers and Poverty’, Economic and Political Weekly, Special Number, 15/41–3.
MHJ (1980), ‘Who Pays For and Who Gains From EGS?’, Economic and Political Weekly, 17/31.
MOF (1986), ‘The Existing System of Public Foodgrain Distribution in Bangladesh and Proposal for Restructuring’, Draft Report, preparerd by Beacon Consultants for the Ministry of Food, Government of Bangladesh, Dhaka.
Moore, M. (1985), The State and Peasant Politics in Sri Lanka. Cambridge: Cambridge University Press.
——— and Perera, U. L. J. (1978), ‘Land Policy and Village Expansion in Sri Lanka’, Marga, 5.
Muqtada, M., and Alam, M. M. (1983), Hired Labour and Rural Labour Market in Bangladesh. Bangkok: ILO-ARTEP.
NABARD (1984), ‘Study of Implementation of Integrated Rural Development Programme’, mimeo, National Bank for Agricultural and Rural Development, Bombay.
Nag, M. (1985), ‘The Impact of Social and Economic Development on Mortality: Comparative Study of Kerala and West Bengal’, in S. B. Halstead, J. A. Walsh, and K. S. Warren (eds.), Good Health at Low Cost. New York: Rockefeller Foundation.
Osmani, S. R. (1990), ‘The Food Problems of Bangladesh’, to be published in J. P. Drèze and A. K. Sen (eds.), The Political Economy of Hunger. Oxford: Oxford University Press.
——— and Chowdhury, O. H. (1983), ‘Short Run Impacts of Food for Work Programme in Bangladesh’, Bangladesh Development Studies, 11.
Panikar, P. G. K. (1979), ‘Resources Not the Constraint on Health Improvement: A Case Study of Kerala’, Economic and Political Weekly, 14/44.
Peiris, G. H. (1976), ‘Share Tenancy and Tenurial Reform in Sri Lanka’, Ceylon Journal of Historical and Social Studies, NS 6/1.
——— (1978), ‘Land Reform and Agrarian Change in Sri Lanka’, Modern Asian Studies, 12.
PEO (1980a), ‘A Quick Evaluation Study of the Food for Work Programme: An Interim Report’, Programme Evaluation Organization, Planning Commission, Government of India, New Delhi.
——— (1980b), ‘Joint Evaluation Report on Employment Guarantee Scheme of Maharashtra’, Programme Evaluation Organization, Planning Commission, Government of India, New Delhi.
——— (1985), ‘Evaluation Report on Integrated Development Programme’, Programme Evaluation Organization, Planning Commission, Government of India, New Delhi.
Rahman, A., and Haque, T. (1988), ‘Poverty and Inequality in Bangladesh in the Eighties: An Analysis of Some Recent Evidence’, mimeo, Bangladesh Institute of Development Studies, Dhaka.
(p.354) Raj, K. N., and Tharakan, M. (1981), ‘Agrarian Reform in Kerala and Its Impact on the Rural Economy: A Preliminary Assessment’, Working Paper WEP 10–6/WP49, Rural Employment Policy Research Branch, World Employment Programme, International Labour Organization, Geneva.
Rao, C. H. H., and Rangaswamy, P. (1988), ‘Efficiency of Investments in IRDP: A Study of Uttar Pradesh’, Economic and Political Weekly, Review of Agriculture, 23/ 26.
Rao, V. M., and Erappa, S. (1987), ‘IRDP and Rural Diversification: A Study in Karnataka’, Economic and Political Weekly, 22/52.
Rasaputra, W. (1986), ‘Public Policy: An Assessment of the Sri Lankan Experience’, mimeo, World Institute for Development Economics Research, Helsinki.
Rath, N. (1985),’ "Garibi Hatao": Can IRDP Do It?’, Economic and Political Weekly, 20/6.
Rudra, Ä. (1981), One Step Forward, Two Steps Backward’, Economic and Political Weekly, Review of Agriculture, 16/25–6.
Sahn, D. (1987), ‘Changes in Living Standards of the Poor in Sri Lanka During a Period of Macroeconomic Restructuring’, World Development, 15.
Samaraweera, V. (1981), ‘Land, Labour, Capital and Sectional Interests in the National Politics of Sri Lanka’, Modern Asian Studies, 15.
Sanderatne, N. (1972), ‘Tenancy in Ceylon’s Paddy Lands: The 1958 Reforms’, South Asian Review, 15.
Sen, A. K. (1981), ‘Public Action and the Quality of Life in Developing Countries’, Oxford Bulletin of Economics and Statistics, 13.
——— (1985), The Standard of Living. Cambridge: Cambridge University Press.
——— (1988), ‘Sri Lanka’s Achievements: How and When’, in T. N. Srinivasan and P. K. Bardhan (eds.), Rural Poverty in South Asia. New York: Columbia University Press.
Sharma, R., and Gianchandani, D. (1988), ‘IRDP and Rural Diversification’, Economic and Political Weekly, 23/19.
Siddiqui, K. (1979), ‘The Political Economy of Land Reform in Bangladesh’, Journal of Social Studies, 6.
——— (1985), ‘An Evaluation of Grameen Bank Operations’, in S. Mukhopadhyay (ed.),Case Studies on Poverty Programmes in Asia. Kuala Lumpur: Asian and Pacific Development Centre.
Singh, S. and De, N. (n.d.), ‘A Study of Rural Landless Employment Programme in Punjab’, mimeo, Punjab State Institute of Public Administration, Chandigarh.
Subbarao, K. (1985), ‘Regional Variation in Impact of Anti-Poverty Programmes: A Review of Evidence’, Economic and Political Weekly, 20/43.
Sundaram, K. and Tendulkar, S. D. (1985), ‘Anti-Poverty Programmes in India: An Appraisal’, in S. Mukhopadhyay, (ed.), The Poor in Asia: Productivity-Raising Programmes and Strategies. Kuala Lumpur: Asian and Pacific Development Centre.
Swaminathan, M. (1988), ‘Inequality and Economic Mobility: An Analysis of Panel Data from a South Indian Village’, Oxford University D.Phil thesis.
Thimmaiah, G. (1988), ‘Concurrent Evaluation of IRDP’, Economic and Political Weekly, 23/7.
Vaidyanathan, A. (1987), ‘Poverty and Economy: The Regional Dimension’, paper presented at a conference on Poverty in India held at the University of Oxford.
Varghese, T. (1970), Agrarian Change and Economic Consequences: Land Tenures in Kerala 1850–1960. Bombay: Allied Publishers.
Weerawardena, I. K. (1975), ‘Lessons of an Experiment: The Paddy Lands Act of 1958’, Evaluation Studies No. 3, Department of Rural Institutions and Productivity Laws, Colombo.
Westergaard, K. (1986), ‘People’s Participation, Local Government and Rural Development: The Case of West Bengal, India’, Research Report No. 8, Centre for Development Research, Copenhagen.
Wickremeratne, L. A. (1977), ‘The Economy in 1948’, in K. H. de Silva (ed.), Sri Lanka: A Survey. Honolulu: University Press of Hawaii.
Wickramasekara, P. (1985a), ‘Strategies and Programmes for Raising the Productivity of the Rural Poor in Sri Lanka’, in S. Mukhopadhyay (ed.), The Poor in Asia: Productivity-Raising Programmes and Strategies. Kuala Lumpur: Asian and Pacific Development Centre.
——— (1985b), ‘The Role of Land Settlement Programmes in Raising the Productivity of the Poor: A Sri Lankan Case Study’, in S. Mukhopadhyay (ed.), Case Studies on Poverty Programmes in Asia. Kuala Lumpur: Asian and Pacific Development Centre.
Wimaladharma, K. P. (1982), ‘Land Settlement Experiences in Sri Lanka’, mimeo. Colombo.
World Bank (1988), ‘Bangladesh: Adjustment in the Eighties and Short-Term Prospects’, Report No. 7105-BD, Washington, DC.
(*) The author is deeply grateful to V. K. Ramachandran, J. P. Drèze, P. K. Bardhan, and C. Fonseka for incisive comments on an earlier draft. The responsibility for the views and any errors is of course the author’s alone.
(2) Drèze (1988) contains an in-depth case-study of the actual operation of some of these policies in an Indian village. On the feasibility and desirability of other forms of social security, such as old-age pension, survivor benefits, insurance against occupational hazards, etc., again in the Indian context, see the important contributions of GOI (1984) and Guhan (1988).
(3) Out of a total operated area of 163 million hectares (according to the Agricultural Census of 1980/1), only 2.97 million hectares were declared surplus, 2.36 million were taken possession of, and 1.82 million distributed by the mid-1980s (Bandyopadhyay 1986: A-50).
(4) The Kandyan peasants have had a long-standing grievance that plantation agriculture has endangered their livelihood by encroaching on their land. There is, however, a good deal of controversy on the degree to which such encroachment did in fact occur. An interesting account of this controversy is given by Moore (1985).
(6) The Communist Party of India (Marxist) has been controlling the State government of West Bengal since 1977. Two major Communist parties have dominated the polity of Kerala for most of the period since 1956, although often in coalition with more conservative elements.
(7) The estimates of tenancy quoted in this paragraph are taken from BBS (1985) for Bangladesh, Hameed et al. (1977) for Sri Lanka, Bandyopadhyay et al. (1985) for West Bengal, and Raj and Tharakan (1981) for Kerala.
(13) Barga is the local name for the system of share-cropping.
(14) According to official records, nearly 1.3 million share-croppers were registered by the mid- 1980s (Bandyopadhyay 1986: A-50). As against this, estimates of the total number of sharecroppers in West Bengal vary between 1.5 million (according to peasant leaders) and 2.0 million (an early official estimate). See Bandyopadhyaya et al. (1985) for a discussion of these estimates.
(15) If eviction took place at that time, it would usually mean replacing one tenant by another rather than taking back the land for cultivation by hired or family labour. Eviction of the latter kind is much more of a modern phenomenon.
(16) The importance of absentee landlordism in shaping the success of tenancy reform is borne out both by Kohli’s sample and the experience of the Hambantota district in Sri Lanka. It is only fair to mention, though, that Kohli himself does not draw this link; instead he interprets his findings as pointing to the success of Operation Barga as such.
(19) See Varghese (1970) for a detailed account and Raj and Tharakan (1981) for a brief but lucid exposition of this history. All the statistics quoted in this paragraph and the next are taken from the latter source.
(20) The Keralite problem would have been of an altogether different order of feasibility if the politically powerful landowners of Travancore (southern Kerala) had had to be antagonized. Initially, such an antagonism did in fact arise. But it was kept under control by legally allowing the smaller owners to retain their land. This largely took care of the problem, since the prevalence of tenancy was in any case rather small in Travancore.
(21) Raj and Tharakan (1981) have themselves drawn a similar parallel between the post-colonial abolition of tenancy in Kerala and the pre-colonial tenurial reforms in Travancore under the aegis of the princely states of the region.
(24) These figures are quoted from Islam (1987: table 1.1). They refer to the situation prevailing in 1984/5 in Bangladesh (Labour Force Survey) and in 1981 in India and Sri Lanka (Population Censuses).
(25) All the statistics quoted in this paragraph and the following two are taken from Hossain (1984b, 1988)—both these reports are based on the most comprehensive evaluation of the Grameen Bank done so far. For further details, see Siddiqui (1985) and the references cited therein.
(26) There was a good deal of interstate variation in this respect. In Gujarat, for example, the figure was as high as 47%.
(27) Kurian does not mention this, but Rao and Rangaswamy (1988) do while analysing the same set of data for Uttar Pradesh. The latter authors also defend this procedure on the ground that ‘since the concept of net income followed is the same in respect of IRDP as well as other sources (both in the pre-and post-assistance periods), per cent changes in such incomes may approximate to changes in disposable income’ (p. A-71). This argument is tenable only if the pre-assistance income did in fact include a loan repayment component of comparable magnitude. Since this is highly unlikely to have been the case for most beneficiaries, the methodology followed by these authors is bound to give an inflated picture of the change in disposable income.
(28) Although the mobility studied by Swaminathan refers to a time gap of eight years (from 1977 to 1985), there was no serious problem of recall lapse as the recall method was not used to collect past data—a set of households surveyed in 1977 in connection with a different study were resurveyed in 1985, and the data from the two surveys compared.
(30) It is also interesting to note that most of the beneficiaries of the Grammeen Bank scheme engage in those activities where their households have already had some entrepreneurial experience. It is no coincidence that some 40 per cent of the loan is taken for trading activities and some 30 per cent of the beneficiaries had trading as their prior occupation (Hossain 1984b: 125, 29).
(31) See discussion in the preceding section.
(32) Only Gujarat, an infrastructurally developed state, andTripura, another CPI (M)-dominated state, did any better.
(43) This is partly offset by the fact that per acre productivity (value) of land in Kerala is also one of the highest in India. None the less, rural poverty is one of the most acute here. See Vaidyanathan (1987) for more on these regional comparisons.
(45) The key role was played not by the distribution of food as such but by the very existence of the stock which dissuaded speculative hoarding. For more on this, see Bhatia (1983). See also Drèze and Sen (1989: chap. 8), on the primacy of price stabilization over income generation through the Indian Public Foodgrain Distribution System.
(47) De Janvry and Subbarao (1986: table 1.5). It is worth noting in this context that there is some difficulty in comparing rural Kerala with the rural sector of the other states as the rural-urban distinction is not as clear-cut in Kerala as in the rest of India.
(48) The value of food stamps at the time of inception was only 3 per cent less than the ration subsidy received by the bottom quintile, though for all the recipients taken together it was 17 per cent less; Edirisinghe (1987: table 5).
(49) This is probably an overestimate because the share of rice is lower and the share of tapioca higher in the diet of the poorer groups compared to an average Keralite diet.
(50) It should be noted though that 1969/70 was a year of exceptionally good harvest which enabled even the poorer groups to consumer more calories compared to a normal year. Consequendy, the percentage contribution of subsidy to total calorie intake in a normal year would be higher than is suggested by this figure.
(51) For example, the National Sample Surveys suggest that per capita calorie intake in Kerala was as low as 1,620 Kcal in 1961/2 arid 1,618 in 1971/2 (as against the all-India average of 2,445 and 2,263 respectively). But an alternative estimate from the food balance sheets give an average figure of 2,339 Kcal in Kerala for the period between 1961/2 and 1970/1 (CDS 1975).
(57) See also Panikar (1979) on the role of education in Kerala. For a wide-ranging analysis of the role of education, health care, and food distribution in enhancing the quality of life in Kerala, Sri Lanka, and several other regions of the world, see Caldwell (1986).