I. Background and Objectives

The seminar was organised jointly by the ILO’s Asian Regional Team for Employment Promotion (ARTEP) and the Bangladesh Institute of Development Studies and was held in Dhaka, Bangladesh during 7-9 April 1984. The basic objective of the seminar was to analyse the effectiveness of anti-poverty policies and programmes in selected Asian countries with a view to helping them identify alternative policies which are likely to be more effective. The countries that were represented in the seminar were Bangladesh, India, Indonesia, Nepal, Pakistan, the Philippines and Sri Lanka. Participants were mainly drawn from senior planners and policy makers in these countries. Some academicians who have been working on the problem were also invited. In addition, representatives of workers’ and employers’ organisations and other UN and inter-government agencies like the CIRDAP, ESCAP and FAO also participated in the seminar. ILO’s World Employment Programme was also represented.

The seminar was inaugurated by Prof. M.N. Huda, the Chairman of the Bangladesh Social Science Research Council.

Four types of papers were presented in the seminar. One paper provided an overview of the trends in rural poverty in selected Asian countries. Another paper brought out the data gaps and other problems of monitoring trends in poverty. A third paper analysed the preconditions, possibilities and limitations of effective anti-poverty policies in the Asian context. Finally, a set of country papers were presented, each of which attempted to provide a critical review of the various types of anti-poverty policies adopted in the respective countries and to assess the impact of selected policies.

A summary of the major points that emerged during discussions in the seminar is being presented below.

II. A Summary of the Proceedings

Problems of methodology and inadequacy of the data base

A paper on problems of monitoring trends in poverty in rural Asia¹ formed the basis of discussions on this issue. The paper focused on the measurement of poverty based on nutritional norms and pointed out the methodological problem of using a constant calorie-income relationship over a long period of time. There was general agreement that measures based on nutritional norms provide too narrow a view of poverty and that other aspects having a bearing on the ‘quality of life’ should also be considered.

¹ Rizwanul Islam, ‘Monitoring Trends in Rural Asia: Some Problems’. This paper has not been included in the present volume.

The seminar agreed that as far as the monitoring of trends in the incidence of poverty in the countries of the region is concerned, both the official and unofficial bases for generating reliable, timely and comprehensive data desperately need an imaginative improvement. The inadequacy of the data base precludes definitive conclusions about trends in the levels of poverty. Inadequacy of the official data further hampers a quantitative assessment and analysis of the dynamics of changes in the situation regarding poverty. It was suggested that estimates of the incidence of poverty should form a part of the statistical information that is made available by the governments on a regular basis. Monitoring of a small sample on a continuous basis was also considered desirable. Attempt to capture the nature of assets transacted by households was stressed as an important dimension to the desired restructuring of the statistical system for monitoring poverty. It is possible to put into operation a reasonably inexpensive technology of data collection. It was, of course, realised that in order for such ideas to be implemented, the governments need to be politically committed to the task in particular, and to the welfare of the poor as a class in general. However, it was thought that in the absence of such a political dispensation, other organisations (e.g., research institutes, trade unions, etc.) could take up this responsibility. The interesting Filipino case, where an indigenous research institution was engaged in collecting data on trends in poverty (of course using a subjective method having its limitations), was noted as a promising start.

Poverty record in rural Asia

Evidence presented to the seminar¹ as to the changes in the levels of poverty in rural Asia leaves one at best agnostic that the growth, whether of aggregate GDP or agricultural output per capita, has anything to do with the extent of rural poverty. Only in one out of ten case studies², namely Thailand, was there a particularly strong evidence of a reduction of poverty over the 1960s and 1970s. But even this lone success story deserves to be taken with a more than just a pinch of salt, due to the endemic “dissaving rates” characterising the various income size classes in the relevant body of the data. It was noted, however, that in two other cases, namely in Pakistan and the Indian Punjab, while poverty ratio deteriorated during the 1960s, there was an improvement in the 1970s. For all the other case studies, there was a continued deterioration of the poverty situation over these two decades. When this is matched with the fact that several of these economies have experienced a fair extent of growth in aggregate terms or in their agricultural sector, there is a feeling that the nexus between growth and rural poverty has indeed been an agnostic one.

¹ The paper on this topic has been published in Khan and Lee (eds.), op cit. Ch. 1.

² These studies covered Bangladesh, four states of India, Indonesia, Nepal, Pakistan, Sri Lanka and Thailand.

Analysis of the dynamics of rural poverty

It was argued that changes in the levels of rural poverty could be explained largely by an agrarian structure characterised by a high degree of land concentration, steady population growth and little or no extension of cultivated land - all leading to rising landlessness and dependence on wage labour. Added to this are stagnant or declining real wages and a decline in employment in agriculture. In a situation like this agricultural growth by itself was found insufficient for the alleviation of poverty.

On the question of a decline in the demand for labour, it was observed that large-scale mechanisation that took place in a number of countries could be the main reason. Institutional changes that occurred in the mode of labour hiring in some cases (like Java) was also cited as a factor responsible for a decline in the demand for labour.

It was, however, pointed out that in Pakistan, for example, despite the spread of mechanisation, real wages in agriculture improved. Expansion of non-farm employment and large-scale international migration were cited as possible explanatory factors in this case. It was also pointed out that the first factor could have worked in the case of Java and Indian Punjab. The hypothesis was that the growth in agriculture could have led to increased investment in non-farm activities, particularly by the larger landowners.

On the question of a causal relationship between observed changes in the levels of poverty and socio-economic changes - in particular, those involving the institutional and structural framework of these economies - it was noted that very few studies show the precise nexus between factors like inequality in the power structure, dominance of the elites, etc. and the process of immiserisation. Although data on the factors that are thought to contribute to increased impoverishment are often presented, the use of these data were not found to constitute definitive evidence of causality of relationships. While the need for more empirical research was recognised, emphasis was put also on theoretical work designing models for estimating the relative contributions of various factors to the process of increasing impoverishment. Modelling exercises that have included demographic and occupational structure of households were taken note of. But the importance of incorporating structural and institutional factors into such exercises was also emphasised.

Genesis and perpetuation of rural poverty: an analysis

One paper presented at the seminar argued that rural poverty in Asia must be viewed in a historical perspective. The origin of the types of rural poverty observable today forms part of what Gunder Frank calls ‘the development of underdevelopment’. Colonialism and, to a lesser extent, certain forms of imperialism are regarded as mechanisms which, over long periods of time, drained resources from the ‘periphery’ in the Asian countries to the ‘centre’, mainly in Europe. The historical origin and setting of rural poverty simultaneously connect the present to the past and establish boundaries to what is possible in future.

One of the participants, however, opined that the ‘diluted’ version of the ‘centre-periphery’ models of the ‘development of underdevelopment’ contained in the paper under discussion, as well as its idealised version to be found in the writings of Gunder Frank, is unconvincing and misleading as a general theory of poverty and underdevelopment. This is because it totally discounts the role of the internal class dynamics of underdeveloped countries, and imposes an artificial, empirically false and theoretically dangerous dichotomy (centre-periphery) on a reality which is more a spectrum of levels of development. The participant went on to argue that its central conceptual proposition, the drain of surplus from periphery to centre, remains, in spite of various ingenious efforts, confused and problematic. He thought that, there was something spurious in the proposition that poor countries were being ‘drained’ through generating export surpluses during the colonial era, and that the same countries are being ‘drained’ through massive import surpluses.

Even if conceptually relevant, it was asked, what relevance does a recognition of the (colonial) historicity of rural poverty in Asia have for current problems, possibilities and policy options? It was suggested that in a discourse on concrete policies against poverty, it is not so much the genesis as the ongoing process of reproduction of underdevelopment that are of relevance.

Another participant thought that there was an important distinction between Frank’s historical analysis of underdevelopment and poverty on the one hand, and its application to the present-day problem of rural poverty. This lies in that Frank’s analysis was concerned with the generation of absolute poverty in the colonies in a global sense - that is, involving both rural and urban areas - and cannot therefore completely explain issues of rural poverty in a situation where poverty is both absolute and relative. Also, there is the important problem that many of the poor countries (e.g., Afghanistan, Nepal or Burma) did not have any historically sustained colonial exploitation. The participant thought that it is the ‘stratified class structure’ or ‘internal class dynamics’ which are as much, if not more, important in accounting for rural poverty in countries of Asia as exploitation by the countries in the ‘centre’.

However, the paper referred to earlier also contained a ‘class analysis’ of the problem of poverty. It was argued that in discussions on causes and cures of poverty a preoccupation with the universality of atomistic economic agents, and a pervasive community among their dominant material interests is bound to be futile in a situation where the society is divided into identifiable groups or classes which have different and frequently competing interests. It was generally agreed that a view of the socio-economic polity stressing the stratified nature of the economic and political power provides richer insights into the process of poverty than a technocratic view that obfuscates over social stratification. However, what one means by class analysis and by ‘radical economists’ remained in important question. It was pointed out that the use of imprecise and loose nomenclature is likely to be misleading and therefore counter-productive.

Regarding the role of state, it was argued that since governments represent particular classes and groups, the state is bound to be non-neutral. In a situation like this, it is little wonder that government policies are often designed to promote interests of classes whom it represents.

Whether it is in the design and implementation of development projects or in the patterns of social expenditure on health and education, it is above all the interest of the entrenched elite that is usually protected and promoted. The conclusion suggested itself that, given this dominance of the ruling classes in policy making, the need for poverty alleviation within a reasonable time frame in these Asian countries, must be seen as having very little hope, unless fundamental structural changes could be brought about by the poor.

One participant made an interesting point on the philosophical basis of state policy. According to him, it is one thing to say that the state as an institution is used as a device by the powerful; but to say that this suggests that the powerful will be always or mostly unlikely to adopt policies which benefit the poor involves an implicit assumption. If the problem of alleviation of poverty is a zero-sum game in which the poor can only be better off at the cost of the rich being worse off, the position that the rich will never help the poor follows quite naturally. The question, however, is: are there policies having non zero-sum (NZS) outcomes, in which both the rich and poor benefit from the policies, but where the poor benefit proportionately more than the rich? Whether such NZS policies exist or are perceptibly in evidence is a matter of empirical judgement. For Bangladesh, the Food for Works Programme (FWP) was cited as one such programme. It was admitted that the assets created through this programme are not owned by the workers involved. Even so, the participant argued that the nature of the assets created and the modalities of the programme are such that it makes a clearly positive impact on the short-term welfare of they very poor, as well as on the long-run productivity of the labour force. It is significant that FWP does also benefit the rural rich.

However, as against this specific illustration, it was pointed out that FWP is not strictly a NZS policy in that it is largely funded by foreign donors. This suggests that resources involved are not being taken from the well-off within the country but being supplied to the ruling class as a cost-free transfer. It would be a NZS policy if such public works programme were funded through the mobilisation of domestic resources. However, the FWP has the arguably negative aspect in the particularly insidious form of dependence on the part of the country as well as of the poor in the rural area on the donor countries. Finally, by strengthening the political and economic power of the rural rich, the latter are being given a lever with which to perpetuate zero-sum perceptions in the context of anti-poverty policy.

Definition of anti-poverty programmes

Discussion on anti-poverty policies noted a definitional problem about the package of such programmes. This arose because of the commonly observed tendency to equate an area development plan or even attempts to increase farm output in general with an anti-poverty programme. The core of the problem of rural poverty is lack of income arising mostly from assetlessness. Hence it was suggested that anti-poverty programmes should mean only those activities which consciously transfer or create income generating assets for the rural poor either on individual or on group basis. It may also include specially designed employment programmes guaranteeing at least a minimum period of employment with the creation of such assets which would ensure a continuing flow of benefits to the poverty groups. It was, however, felt that even this definition contained no reference, in quantitative terms, to the need for anti-poverty programmes to seek to influence the actual extent of relative or absolute poverty. It was also argued that a really comprehensive anti-poverty programme should (a) reduce absolute poverty, (b) reduce relative poverty and, (c) equip the poor (by creating income-generating assets and/or by providing support through financial institutions) so as to generate and enjoy growing incomes on a sustained basis on the basis of self-reliance. Such a definition implies that projects like public works of a distress relief variety have to be sharply distinguished from anti-poverty programmes.

Rhetorics vs. reality of poverty alleviation in development plans

It was noted that there is much national and international rhetoric about the alleviation of poverty. The reality of the situation is that there is inadequate specificity of allocation of resources and a conspicuous lack of integration of these rhetorical programmes into the structure of the plan. Often, the financing of these programmes is left largely to foreign aid - a fact that attests to the ceremonial contents of these anti-poverty proclamations. Moreover, explicit time frames are rarely introduced. It was widely felt that greater internal mobilisation of resources and greater use of explicit time frame in planning for poverty alleviation is essential.

Policies needed

It was suggested that a reasonable anti-poverty package must include the following: (i) measures to change the existing balance of power so that the poor can be freed from violence and repression; (ii) an egalitarian redistribution of productive assets; (iii) measures that redistribute flows of incomes; and (iv) public investment that specifically targets itself at the poor. But the deliberations in the seminar brought sharply to the fore the tension between the felt need to bring the poor up to the ability to wield political power and the incontrovertible evidence that the pervasive lack of poor’s political power is not due to covert political repression. There is the tension between the need to accord the most fundamental priority to effective political envigoration of the poor in any conceivable agenda for anti-poverty action and the need to recognise that, on the basis of past record, progress likely to be made on this particular ground is bound to be slow, if feasible at all. De Jure political enfranchisement has been an aspect of the policy in the Asian countries for many decades, but the problem is that de facto political integration requires financial clout. There is, however, a vicious circle. The condition of the poor cannot improve without financial power and organisation, and yet they can not get financial power and organisation without becoming less poor than they are.

It was pointed out that some of the countries in the region have had a disturbing measure of lawlessness on the part of the police. One of the participants, therefore, recommended disarming of the police. While this was described by others as simplistic, especially because violence could also be committed just as often by people other than the police against the poor, one participant reminded the seminar that judiciously-timed but temporary banning of the use of fire arms can facilitate peaceful execution of reforms which raise tension. At the time of ‘Operation Barga’¹ in West Bengal, there was a temporary suspension of gun licenses, and this made a significant contribution to an orderly completion of the potentially explosive reform.

¹ A programme of tenancy reform.

The viewpoint, voiced by one of the participants that real political enfranchisement is more fundamental than distribution of productive assets, was questioned vigorously. It was argued that asset redistribution, if it could be somehow achieved, would potentially have greater impact on the economic and eventually political conditions of the poor.

The idea of redistribution of productive assets across the entire rural population has dimensions that are contentious. For example, in Bangladesh, a thorough-going redistribution of land would transfer no more than 1.25 acres in favour of any one landless family. And yet many rural households have non-farm occupations yielding them an income not less than what they would get from cultivating a land plot of that size. Another aspect of the desired redistribution of productive assets that was stressed was that of the periodicity. An once-and-for-all redistribution was seen as inadequate, as in conditions allowing private ownership rich farmers can buy land back. Unless communal type of arrangements was being envisaged, there would perhaps have to be periodic redistribution of land. While such a periodic redistribution is not considered politically or administratively feasible, the governments could scarcely be advised or expected to adopt a hands-off policy. Following asset redistribution, the government would have to maintain a constant presence as regards extension, cooperative building, credit supply, etc., to ensure maximum output and a minimum flow of agricultural surplus to the cities.

It was recognised widely that attempts to redistribute incomes without affecting asset distribution are likely to be of little help.

As for poverty-focussed public investment, while the poor may conceivably benefit from increases in output due to poverty-focussed public investment, it was stressed that the position that the pattern of public investment could be easily changed to favour the poor without affecting aggregate rate of investment and the pattern of private consumption is not to be taken as proven. Poverty-focussed public investment was seen as operating at a margin of very insignificant proportion of any conceivable country’s GDP. Also, in the absence of appropriate changes involving property rights, such changes in the composition of public investment could only have minimal impact.

Lessons from anti-poverty programmes

Since land is the most important productive asset in the rural economy, it appeared that any income-generating asset endowment programme excluding land was likely to have only minimal effect. Even for activities based on non-land assets, a minimum land base was considered to be essential. It was maintained that even a miniscule land base would enhance the investment absorptive and retentive capacity and would ensure optimal use of other non-land inputs. It was, therefore, argued that redistributive land reform has to be made an integral part of poverty alleviation programmes. While this view was widely held, the question was also asked whether land reform was a sufficient or even a necessary condition for the alleviation of rural poverty. As a supporting evidence for this view, it was pointed out that in the cases where reduction in poverty has been reported (in the paper on trends in poverty), the phenomenon was not associated with any effective land reform.

For programmes oriented towards poor households, an optimal credit-subsidy mix was considered essential. While pointing out the inadequacy of relying on collateral-oriented commercial banks for providing credit to the landless or near-landless, the importance of specially designed financial institutions like the Grameen Bank of Bangladesh was underlined. In this connection, mention was also made of the National Bank of Agriculture and Rural Development in India which provides reference to commercial banks for financing integrated rural development programmes (IRDPs). Promotion of collective or group activity amongst the beneficiaries of IRDP type or other poverty-focussed credit programmes was also thought to be important for the purpose of enhancing the bargaining power of the poor - both as sellers and as buyers. The right selection of beneficiaries for target-group oriented programmes was found to be difficult. It was, however, argued that public scrutiny of the list of poor in a village could act as a check against wrong selection. It was also pointed out that local self-government with representation from the poor would increase the possibility of their participation in anti-poverty programmes.

It was argued that the supplemental employment programme has to be a continuing one with an element of guarantee built into it. Besides discarding the adhocism of distress relief, the aim should be to create durable assets which would not only strengthen rural infrastructure but provide direct continuing economic benefit to the poorest groups. It is possible to design such a programme particularly around social forestry, minor irrigation projects, communally-owned sericulture projects, etc. Wherever possible, the landless who are involved in the creation of some assets should be made the owners of such assets.

Income-employment generating programme for the rural poor should have back-up support of programmes providing basic needs in relation to health, family planning, education and public distribution of essential items of consumption. Experience of Sri Lanka in the early and mid-seventies was considered relevant in this regard. The desirability of linking receipts of benefits under the category of interventions to the adoption of family planning by the recipients was also considered.

It was suggested that the duplication of effort among various agencies be avoided. In terms of the operations of the banks and other financial institutions, responsibility and authority should be judiciously blended.

Regarding employment creation schemes of works programme type, it was suggested that in order to be useful as anti-poverty policies they should (i) lead to the creation of productive and durable assets and (ii) develop mechanisms to pass on the ownership of such assets to those involved in their creation. This kind of programme was considered useful in situations where lack of political manoeuvrability makes changes in the pattern of investment the only feasible anti-poverty programme. However, the implementation of this type of policy cannot be equally easy for all types of potential public works projects. In what way can workers who build, say, a 3 mile long road benefit from owning it? By having tollgates and charging users? If they construct an irrigation channel, how exactly are they to deny access to ‘free riders’? Further, how can workers be expected to form owning and administrative entities, if they come from distant areas? On the other hand, there could be public works projects, not necessarily of the conventional variety but such as social forestry or fishing projects, communally-owned sericulture (both have been tried with some success in India) where the implementation of such policies may be possible. Even so, it was asked in some scepticism, if the rich are determined to neutralise the property rights endowed to the poor through rural works programme, will the state be able and willing to protect the poor? In other words, are measures possible which would change the balance of power in the countryside?

While still on the question of rural works programme, a generalised observation was made as regards the financing of such programmes. In several countries, such schemes, as also major rural development projects, are financed by foreign aid. Is this a cause for concern? Does it have implications for the continuation of such projects in the event of a reduction or stoppage of the flow of aid? Several other questions were raised in this context. The first was whether one could see in this fact a conscious desire on the part of the donors to use the carefully-nursed power of the poor to force the hands of the government on certain unknown policy measures, or it was a novel way of establishing a bridgehead by the donors in the rural power structure. The second question was whether the control enjoyed by the rural rich over the poor increased as a result of the possible economic inequality engendered as a result of such aid-funded activities. Whether the dominant political forces in the various countries were really interested in eradicating poverty was also questioned.

The seminar found it difficult to come to any conclusion on that question. On the one hand, it was emphatically asserted that, far from being non-zero-sum games from which both the poor and the rich derived benefits (with the former doing so more than the latter) reliance on foreign aid for anti-poverty programmes tends to create a particularly insidious dependence on external help and saps the vitals of the indigenous political will for genuinely egalitarian and anti-poverty social engineering. On the other hand, it was asserted equally emphatically that so long as it is possible to ensure that the bulk of the external resources are tied to anti-poverty programmes, and as long as such resources can create durable and productive assets which in the short run create employment opportunities for the poor and in the long run raise farm productivity (which could benefit all groups), there is no inherent reason for aid-dependence to cause any serious concern.

Another question of international relevance was the possibility of learning from the experience of socialist Third World countries, particularly about (a) organising the poor for their own development; and (b) the advantages and disadvantages of various types of communal tenure systems. In this regard, however, some scepticism was expressed about the relevance of such lessons for countries having vastly different political systems.

Finally, it was pointed out that successful anti-poverty programmes, by transferring to the poor the productive assets that are created, would necessarily mean a tilting of the balance of power in favour of the poor. A situation of conflict may also be created. For the poor to be able to face such a situation, the essential element of an anti-poverty programme was thought to be organisations of the poor which could conscientise and mobilise them.

It was also felt that the assessment of the actual effect of anti-poverty programmes on the alleviation of poverty remains inadequate. Whether these programmes lead to a reduction in the incidence of poverty, particularly at the national level, needs to be ascertained more carefully. In this context, it was pointed out that while evaluating the impact of particular programmes it is important to isolate the benefits generated by each individual programme.

III. Priorities for Future Work

Two broad lines were identified for future research in the field of monitoring poverty and evaluating anti-poverty policies: (i) descriptive statistics; (ii) analytical issues.

On the first, i.e., on descriptive statistics, it was thought that although producing more facts about trends in poverty is no longer pioneering, it remains an essential task; and useful work could be done in the area. On the methodology of doing so, question was raised about the applicability of nutritional norms (in terms of calorie requirement) suggested by organisations like the WHO, FAO, etc. to poor countries of the Asian region. This raised some controversy. Some of the participants thought that poverty (particularly genesis of it) should be looked at basically as a social phenomenon rather than as a physiological one and that for the purpose of examining trends over time, the precise norm used does not matter much so long as the selected one is used consistently. The limitations of calorie requirement as an indicator of nutritional requirement was also questioned and the need to look into the symptoms of poverty was emphasised. Others argued that even for monitoring trends the norm used may be important because the use of different norms could provide with different trends.

Regarding tools of analysis it was suggested that disaggregated rather than summary indicators should be used. Emphasis was placed on building profiles of the poor in terms of income or expenditure, measures of the quality of life (e.g., literacy, infant mortality rates, etc.), indicators of working conditions (e.g., wage rates) and the satisfaction of basic needs. Scepticism was expressed about the subjective or attitudinal measures of the incidence of poverty. In this context, the limitations of survey research for generating data on income and expenditure and the usefulness of anthropological methods and longitudinal surveys were also pointed out.

Importance of monitoring the impact on poverty of projects intended to benefit primarily the poor was noted. On this the need for developing proper indicators for monitoring purposes and for generating information on such indicators was emphasised.

Finally, it was also suggested that research on poverty should be ‘class-based’, in other words, it should concentrate on particular identifiable groups, such as, landless labourers, non-agricultural specialized workers, food-deficit farmers, etc.

On the second, i.e., the analytical issue, a number of questions were raised.

On poverty in general, two questions came up.

First, the importance of understanding the reasons for increasing poverty despite high growth of GNP in a number of countries was pointed out. It was thought that despite substantial contributions in the field, there is still a lack of a set of realistic and verifiable hypotheses. It was suggested that elements common to various countries should be brought within a framework of analysis to be built in order to understand the poverty situation. The need to look at wider issues of international nature in understanding the poverty of individual nations was indicated. The impact of multinationals in the agricultural sector on land tenure and the process of accumulation was also thought to be important for understanding the dynamics of poverty.

Second, it was suggested that analysis should be made of cases where the incidence of poverty declined despite the fact that movements in real wages would not lead one to expect this.

On labour market, five important questions were raised.

First was the causes of rising landlessness. In this regard, the impact of factors like population growth (as well as the determinants of population growth), and the importance of looking into the phenomenon of age and life-cycle were thought to be important.

Second was the question of the failure of real wage rates to rise over extended periods of time in many areas.

Third, the effects on the rural labour, market of large-scale international migration and the associated inflow of remittances were also thought to be important.

Fourth, the need to examine the inter-linkage (or isolation) between factor markets was pointed out.

Fifth, an analysis of the link between the market for foodgrains and the labour market was considered important from the point of view of the determination of the grain equivalent of wages.

The third set of analytical questions related to non-farm activities. Here the important question of whether the proliferation of such activities was a symptom of growing poverty or an indicator of dynamic growth was raised.

Whether the research agenda discussed was comprehensive enough to cover the problems of identification and evaluation of anti-poverty policies was, however, questioned. Although it was accepted that an analysis of the causes of poverty will throw useful light on these issues, suggestion for directing research specifically towards anti-poverty policies was also made.