State‐sponsored Land Settlement Policies: Theory and Practice

State-sponsored Land Settlement Policies: Theory and Practice

David Hulme

INTRODUCTION

The issue examined in this article is the persistence and popularity of state-sponsored land settlement schemes' as a development strategy in poor countries. Such initiatives have a history stretching back to the last century and have occurred in most African, Asian, Latin American and Pacific nations. Land settlement schemes were a common strategy under colonial administrations and have re- tained their appeal for independent governments. In recent decades the approach has been utilized by regimes of both left- and right- wing political persuasions and empowered by democratic or military means. A cursory review of contemporary government development plans, the reports of international development agencies and the academic literature reveals the continuing popularity of the settlement approach. Schemes are presently being promoted in at least seventeen countries2 and are commonly supported by foreign assistance. These projects have a wide variety of stated objectives including population redistribution, giving land to the landless, relieving environmental problems, improving the welfare of target groups, sedentarizing nomads, assisting disaster victims, increasing food production, raising the value of exports, creating jobs and fostering national security.

Given the persistence of the strategy and its continuing ability to win support from multilateral and bilateral aid agencies, it might seem reasonable to suppose that state-sponsored land settlement schemes have been a relatively successful3 approach to development. Nothing, however, could be further from the truth if the findings of major reviews on land settlement scheme performance are to be b e l i e ~ e d . ~ Nelson (1973,1977) examined settlement

Development and Change (SAGE, London, Newbury Park, Beverly Hills and New Delhi), Vol. 18 (1987), 413-436.

414 David Hulme

scheme performance in Latin America and concluded that ‘few spheres of economic development have a history of, or reputation for, failure to match that of government-sponsored colonization in the humid tropics’ (Nelson, 1977: 98). Chambers’s (1969: 7) detailed analysis of settlement schemes in Africa found that ‘in social and economic terms . . . the record of past schemes has been dis- couraging . . . in almost all countries settlement schemes have been criticized . . . for failure to achieve their social, agricultural and economic objectives, and for their absorption of scarce resources which might have been put to better use’. A World Bank (1978: 16) study of Bank experience revealed that ‘typically, evaluation of settlement projects three to five years after the start of implementation shows economic rates of return at least 50 per cent below those in project appraisal documents’. More recently, a United Nations sponsored review of experience in the Third World concluded that ‘very few programmes have achieved their stated objectives’ (Oberai, 1986: 58) and lists a whole range of common problems. Hirschmann’s (1963) observation that settlement schemes are ‘particularly failure-prone just as some people are accident-prone’ seems as relevant now as it did a quarter of a century ago. The efforts of countless researchers who have produced thousands of recommendations, the refinement of planning models and management procedures and the establishment of monitoring and evaluation units to advise on scheme progress seems to have done Little to help produce more satisfactory outcomes.

Whilst the study of state-sponsored settlement schemes, by both academics and practitioners, has proliferated since the 1960s, little attention has been paid to the key issues of asking why settlement schemes continue to be popular in spite of past experience and why well understood mistakes are repeated. Typically, ex post studies of settlement schemes stand by themselves and focus on narrow technical, economic or social matters but avoid broader i s ~ u e s . ~ In the 1950s W. Arthur Lewis (1954) pointed to the influence that factors such as site selection, settler selection, site preparation, settler capital, group organization and land tenure conditions had on scheme performance. Most subsequent studies have reiterated the importance of such factors but have not taken the discussion a stage further and asked ‘Why aren’t the lessons learned from previous schemes utilized?’ This article explores these broader issues by applying a framework outlined by Clay and Schaffer (1984) to the study of state-sponsored schemes. It is partially based upon re-

State-sponsored Land Settlement Policies 415

search findings from Papua New Guinea (Hulme, 1984) but attempts to frame generalizations applicable to state-sponsored land settlement schemes throughout the Third World.6

RATIONAL MODELS OF POLICY FORMULATION AND PROJECT PLANNING’

A basic premise of most research on land settlement schemes is that these initiatives derive from a rational ends-means process and that ex post studies can improve the effectiveness of this process by identifying ‘lessons from experience’ that can be incorporated into future practice. There are a variety of rational or linear models of public policy of the type outlined by Clay and Schaffer (1984: 4) and of project cycles (Baum, 1978) and even project spirals (Baird and Potts, 1978) which seek to demonstrate the way in which policies and projects are formulated. Pryor (1979: 277) has illustrated the sequence that such models suggest would be followed in the formulation of population redistribution and land settlement policies. Despite the differing geometries employed, the main features of these models and the assumptions on which they are based are similar. All are characterized by a ‘design’ stage and an ‘implementation’ stage (Figure 1). The recognition of a problem or the setting of a goal for public policy is followed by technical and economic studies of the subject area (data collection and analysis) which generates a set of alternative policy options or potential projects. Appraisal of these permits the selection of the ‘best’ option, in the light of stated objectives and available information, and after some final adjustments and refinements the design is completed. Implementation now ensues. The implementing agency monitors the inputs and outcomes, and ultimately the policy or project is evaluated to provide information that will be valuable in future policy formulation or project planning exercises.

Research into land settlement schemes that assumes such models are in operation leads these studies to incorporate the premises underpinning the models themselves. Hence it is assumed that scheme objectives have been clearly defined and are as stated; that plans are believed to be technically and economically feasible by those who prepared and appraised them; that the likely individual and social responses of scheme participants have been considered; that alternative initiatives have been examined; that the institution-

416

--- SECOND STAGE: IMPLEMENTATION

David Hulme

- - - - - - - - -___------_

Figure 1. A Linear Model of the Policy Cycle and Project Cycle

F I R S T STAGE: DESIGN

n a t i o n a l and s e c t o r a l o b j e c t i v e s

- - _ _ - - - - - - - -------_ S t a r t of a n a l y s i s of new p o l i c y l p r o j e c t , rev iew of

n a t i o n a l and s e c t o r a l o b j e c t i v e s , i n c o r p o r a t i o n o f f i n d i n g s of e v a l u a t i o n r e p o r t s

Source: Adapted from Clay and Schaffer (1984).


a1 and managerial capacity to implement the policy or project is believed to exist; and that past experience has been incorporated into the design so that known problems will be avoided or mini- mized. Political interventions and the practice of bureaucratic politics in policy or plan formulation are seen as external to the process and, although they may be recognized as influencing events, they are not treated as an integral part of the process. As Johnson (1985: 351) writes, ‘the dominant image suggests that projects are instruments fashioned with impartial technical expertise for the hands of bureaucrats and politicians’.

POLICY AND PLANNING PRACTICE

The rational models of policy and planning procedure have much to recommend them. Their internal logic is consistent; they provide a means for accurately defining the roles, activities and contributions of various individuals and agencies in an initiative; and they facili- tate the programming of events, funds and materials. However, when the actual experience with regard to land settlement schemes is analysed it is found that these normative models have generally not been followed. This has certainly been and continues to be the case in Papua New Guinea (Hulme, 1984)’ and Sri Lanka (Dunham, 1982). Commonly, an examination of policy and planning practice in relation to settlement activities will reveal that part of the linear model sequence has been omitted, particularly problem identification and the generation of a set of alternative policy or project options; that the sequencing of events is at odds with the model, for example the capacity to implement the initiative is only considered after the determination to proceed; and that feedback about problems and outcomes is treated in a partisan fashion by organizations interested in new projects so that only a portion of what has been learned is used to reshape future activities. However, as Schaffer (1984: 144) points out, those involved in such initiatives only admit that rational models were not operating on rare occasions and ‘in special circumstances’.

Rather than arising out of the application of rational policy and plan formulation procedures the determination to undertake new land settlement schemes, extend existing schemes and rehabilitate old schemes emanates from a complex, amorphous and constantly changing set of inter-organizational, intra-organizational and inter-

418 David Hulme

personal exchanges. To use Biggs’s (1984: 71) expression, such initiatives arise out of the ‘institutional and political stream of events’. At any particular time, debate about the types of activities to be pursued focuses upon a critical zone of what is relatively unobjectionable from the viewpoints of the principal protagonists recognized by the debate. The nature of these viewpoints is deter- mined partly by considerations of personal and group interests and partly by professional paradigms. Gradually, the zone of the unobjectionable is narrowed down until a part of what is unobjectionable becomes unavoidable. A collective process operates with a set of objectives quite different from those that may be appended to policy and project documents. The central quest is for stability and the maintenance of the political and administrative status quo, ‘keeping some in power and the offices in being’’ (Schaffer, 1984:

The interplay of policy and planning practice with the rational models that are supposed to be operating facilitates this quest for stability. Three main mechanisms can be seen to operate. These are the manipulation of the agenda for public action so that items that are potentially objectionable to any of the major protagonists are excluded; the avoidance of substantive challenges, by reference to procedural descriptions, and the creation of new problem-handling institutions; and the avoidance of responsibility for policy or project outcomes by the use of a set of ‘escape hatches’. Some examples of the ways in which these mechanisms operate in relation to settlement schemes are useful at this juncture.

To illustrate the first, reference may be made to the common ‘land problems’ articulated in developing countries, or in certain regions in developing countries, particularly land shortage, population pressure on land and landlessness. Clearly, there are popular expectations that those in power should do something and that these issues will be included in discourses on what public actions are to be pursued. The drawing up of a policy or planning agenda that includes state-sponsored land settlement schemes and that uses data in a selective fashion can draw attention away from alternative courses of action that are objectionable to protagonists and could threaten the stability of the existing political and bureaucratic status quo. In this way the contentious issue of land reform may be avoided and the possibility of fostering spontaneous settlement omitted. The latter is objectionable from a bureaucratic standpoint as it threatens to create a ‘free good’ out of a resource that state-

162).


sponsored settlement converts into an administratively rationed commodity. Its exclusion from the agenda can be explained, any- way, by a ‘lack of information’. Control and selective manipulation of the data available on ‘land problems’ reduces the likelihood of the policy or planning debate noting that, in all but the rarest cases, state-sponsored settlement schemes are only a minor palliative to the problems of land shortage, landlessness and population pressure on resources. The use of settlement schemes to narrow an agenda can be seen in Latin America (Thiesenhusen, 1971), in the actions of British colonial governments in Africa and the Austra- lians in Papua New Guinea.”

There are an array of procedural and institutional means by which the inclusion of land settlement schemes on a policy agenda can be facilitated and by which challenges to their inclusion can be countered. A common example of the former is the references that scheme protagonists make to the long ‘waiting lists’ of eligible applicants for settlement scheme blocks, left over from earlier projects. From this it is inferred that more schemes must be de- veloped to satisfy public demand and ensure fair treatment for all who applied for holdings. The recent introduction of a ‘social planning’ component for settlement scheme appraisal provides a useful illustration of the way in which the creation of sub-institutions can be used to manipulate policy and planning discourse. For decades, anthropologists, sociologists and other socially concerned obser- vers have objected to the initiation of settlement schemes because of their adverse social consequences. These include the increased morbidity and mortality of settlers, nutritional problems, the split- ting up of families, the breakdown of traditional social units, second-generation problems, increased workloads for women and settler-host community violence. l 1 The case made by such critics presents substantive grounds for considering the abandonment of a settlement scheme approach in many circumstances. However, the evolution of the sub-discipline of social planning has provided the opportunity for bureaucracies to inject a social planning input, either from a social planning unit or a consultant social planner, into the rational planning framework. Consequently, challenges to the initiation of new settlements or the expansion of earlier projects on the grounds of the likely adverse social impacts can be side-stepped by reference to these matters being dealt with by the ‘responsible agency’. Skilful administrative practiceI2 can ensure that the social planning unit does not have the influence or power to modify the

420 David Hulme

agenda for action. The social planners themselves are given the quite impossible task of ‘planning’ such matters as settler-host community relations, community organization, the sexual division of new financial and labour budgets within families, inter- generational relationships, nutrition patterns and the reduction of alcohol abuse. These matters divert the social planner from the substantive issue, ‘Should there be a settlement scheme?’ to a secondary issue, ‘What do we do about the likely social consequences of this settlement scheme?’ In this way the institutional- ization and depoliticization of social concern about settlement schemes removes a substantive challenge from the agenda.

The third mechanism for the maintenance of stability is the use of escape hatches to avoid responsibility for poor outcomes and explain why the outcomes of future initiatives will be better. The rational model’s separation of ‘design’ and ‘implementation’ means that agencies on each side of the divide can avoid responsibility by arguing that the results were caused by ‘poor planning’ or ‘poor management’. Hardly ever is it admitted that the strategy that was selected may have been inappropriate. The escape hatches are numerous, but the World Bank (1978: 7), unwittingly, provides a summary of some of the most common for land settlement schemes - ‘management and staff inadequacy, organizational deficiency, overambitious physical targets, underestimates of development costs and difficulties of cost recovery’. The corollary of such diag- noses is that future initiatives will overcome these problems through better management and more skilled manpower, improved organization, less ambitious physical targets, realistic estimates of development costs and effective procedures for cost recovery. This is not all rhetoric, some (and perhaps a great deal of) attention may be paid to these factors, but it serves to change the nature of the policy or planning debate. Rather than focusing on the ‘fun- damental and crucial question . . . whether there should be settlement schemes at all?’ (Chambers, 1969: 250), discourse has moved on to asking ‘How will we get better outcomes on the new schemes?’

THE ATTRACTIONS OF LAND SETTLEMENT SCHEMES

So far the argument in this article has sought to demonstrate the way in which the interplay between the theory and practice of policy and planning formulation contributes to a process that favours the re-

State-sponsored Land Settlement Policies 42 1

tention of state-sponsored settlement schemes on agenda for action, often in spite of previous experience. But why should those groups who can influence the policy or planning debate wish to introduce this item to the agenda or keep it on the agenda? It might be possible to explain this, perhaps, by means of some form of class analysis of the type attempted by Barnett (1977) in a study of the Gezira Scheme. However, such an analysis would require remark- ably powerful explanatory ability to generalize on class processes over the entire history of state-sponsored settlement initiatives and under the auspices of the political spectrum of regimes that have employed this approach. Unless a major analytical breakthrough was achieved, such an undertaking would find itself led into the impasse that neo-Marxist development sociology currently occu- pies (Booth, 1985).

As an alternative the approach adopted here is to examine the appeal of settlement schemes from the standpoint of four main groups that influence policy and planning practice. These are insti- tutionalized political groups, the bureaucracy, international development agencies and private sector interests. All can exert some degree of control over the inclusion or exclusion of specific items on the agenda for public action.

For politicians, political parties and governments, land settlement schemes can be a supreme rhetorical device. At one and the same time it can be claimed that settlement schemes can relieve overpopulation, give land to the landless, solve the social problems subsequent to a natural disaster, improve the welfare of target groups (youths, ex-servicemen, urban squatters, ethnic minorities), reduce unemployment, sedentarize nomads, expand food production and ameliorate the balance-of-payments situation by increasing export revenues. As a panacea for economic and social problems settlement schemes are a strategy almost without rival. If the popular expectation is that those in power should ‘do something’, then schemes are a most effective means of being seen to ‘do something’. Their high visibility (new roads, new offices and houses, large-scale forest clearance, fleets of vehicles) allows them to be used ‘by regimes as ideal public relations exercises, chosen to generate positive publicity’13 (Pratt and Boyden, 1985: 184). It is not surprising that aspiring settlers are commonly found to have unrealistically high expectations (Oberai, 1986; Ploeg, 1972). These expectations should not be interpreted, as is usual, as a misunderstanding on the part of settlers, rather they should be seen as an integral part of the

422 David Hulme

process of maximizing the political dividends accruing from a settlement initiative. In the long term inflated expectations may have adverse consequences on settler morale, but in the short term they represent a bonus for those holding political power.

The previous paragraph infers that the public may be able to exert little influence over the setting of the agenda for action. Evidence from many countries, particularly where repressive regimes are in power, provides support for such a case. Even in nations with more liberal regimes the ability of the public to gain involvement in policy and planning debate, through interaction with politicians and political parties, is often limited. At the local level public opinion about state-sponsored settlement can be manipulated by powerful individuals - traditional leaders, traders, landlords, councillors - who may view settlement as a strategy that serves their interests. Public opinion may be so hardened and cynical, through experience, that initiatives offering some benefits for a small part of the public may seem the best that can be expected. Those with the greatest cause to object to state-sponsored land settlement, tribal groups and ethnic minorities occupying the ‘vacant’ land that is to be colonized are generally peripheral to the local and national political calculi. The lack of concern that national governments have shown about the deleterious impacts of settlement on tribal groups has been widely reported and is documented in Bangladesh (Anti-Slavery Society, 1984), Brazil (Fearnside, 1986), Indonesia (Colchester, 1986a) and Peru (Chase Smith, 1985).

While having rhetorical and publicity value, land settlement schemes can also be utilized to divert attention from other substantive issues that would be more difficult to deal with. Their principal use for this purpose relates to the avoidance or delaying of debate on land reform. The destabilizing effects of aggravating those who would lose out through a major land reform ensure that settlement is ‘politically more desirable, more ,expedient, and easier to carry out than other agrarian reform measures’ (Oberai, 1986: 158). State-sponsored settlement schemes have the added political advantage of creating a good that can be partially used for personal gain or patronage. Finally, from the perspective of those holding formal political power, settlement schemes can be used to achieve both overt and covert political objectives. These include the assertion of sovereign rights over border areas (Budiardjo, 1986), the assimilation of cultural minorities into nations (Colchester, 1986b) and the reassertion of control over


lands held by armed opposition (McBeth, 1981). The factors mentioned in the previous paragraphs partially ex-

plain the continuance of political support for settlement initiatives. It is not argued that settlement schemes have to occur when land is available, nor that there is a general consensus by those in power that schemes should be undertaken. Rather, it is argued that the nature of state-sponsored settlement is such that it has a strong likelihood of gaining access to the agenda for action and, having gained that access, has a propensity to prevail in the chaos of policy issues and influences until it becomes unavoidable. Whether as a central development strategy or a supplementary strategy, land settlement schemes can be used to legitimate those who hold power by demonstrating, in a highly visible fashion, that something is being done to alleviate rural problems.

The second group influencing policy and planning practice are bureaucrats and bureaucratic agencies. Although the bureaucratic standpoint encompasses a range of perceptions of the merits or otherwise of settlement schemes, and is complicated by inter- agency competition, there are grounds to believe that such activities are relatively unobjectionable and commonly desirable. Paramount amongst the attractions of state-sponsored settlement from a bureaucratic stance is the support that settlement initiatives provide for ‘keeping the offices in being’. They provide a rationale for the maintenance and often the expansion of arms of the public service. l4 Ministries of Agriculture and Lands and special-purpose parastatal organizations are the main bureaucratic beneficiaries of new schemes, but in major ventures there may be ‘pickings’ for many other ministries. The foreign assistance often associated with new projects provides a basis for bigger budgets, increases in staff numbers and sometimes the creation of new project management units, and hence new bureaucratic sub-empires.

The activities involved in land settlement schemes pose no challenge or threat to bureaucratic behaviour or practice. Indeed settlement schemes conform with the classic model of bureaucratic public policy described by Schaffer (1984). A public good, state land, is converted into a scarce, divisible good, to which applicants can gain access through a bureaucratically regulated queue. A mechanism is created by which access to the good, settlement blocks, is granted to some whilst it is denied to others. The creation of a scarce public commodity of this nature and the inability to meet demand for this commodity establishes the conditions for the persistence of the

424 David Hulme

administrative institution involved in its production and distribution.

The final basis for proposing that state-sponsored settlement is relatively unobjectionable from the bureaucratic standpoint relates to the technocratic appeal of settlement schemes for those involved in their planning and management. From the planner’s perspective settlement schemes are very tidy, compared to interventions in complex in situ situation^,'^ as it is usual to assume a tabula rasa in terms of land use, settler society and the local economy. Settlers, settler families and settlement blocks are generally assumed to be homogeneous. The resource endowments of settlers are treated as being equal16 (farm size, land quality, access to inputs, capital and labour), settler management ability is assumed to be uniform (fol- lowing a training course) and the structure of settler families is assumed to reside between specified minima and maxima. Once a model farm budget has been established then project-level analysis merely requires the multiplication of the various input and output parameters by the anticipated number of settlement blocks, and the costing of infrastructure. So, for the planner or planning agency under pressure to appraise projects speedily, settlements have a clear attraction as they lend themselves to complex methodological analysis once a small number of assumptions that simplify the planned economic and social system are made.

For agencies responsible for rural administration, settlement schemes have certain attractions, although these are often confined to the early years. The agricultural bureaucracy, in particular, may be able to expand in both budget and staffing terms and be able to operate in the preferred environment of an enclave situation in which both land use and small-farmer husbandry methods can be ‘controlled’. The agricultural regimes that are planned for settlement schemes have their roots in technical appraisals that place formally trained agricultural officers in environments in which their knowledge and experience subordinate that of settlers. Recom- mended practices that require the planting of exotic crops, mono- culture rather than intercropping and reliance upon externally acquired inputs ensure that, at least initially, schemes mirror the professional biases of the agricultural administrator. The literature on settlement schemes provides numerous examples of scheme managements treating settlers as low-level staff who are to be controlled through a series of rules and sanctions - stopping pre-production subsistence allowances, withholding planting


materials, threatening lease forfeiture and stopping loans. Deviance from prescribed plans is to be punished and settler initiatives, such as experimenting with alternative crops, taking on off- scheme jobs, developing small-scale industries and forming interest groups are often viewed as subversive behaviour which is to be discouraged. On schemes that perform relatively well in economic terms and on which a major public service presence is maintained, then there is a constant tension between the administrative per- ception of model settler behaviour and actual settler behaviour. On schemes where economic performance is poor, then the general pattern is for a gradual administrative withdrawal, leaving behind a skeleton staff clamouring for more funds and personnel and exhort- ing higher authorities to sanction harsh disciplinary measures on project beneficiaries who do not conform to management direct- ives. At a later stage underperforming schemes are often ‘rehabili- tated’ by the drawing up of a new plan and the infusion of finance and manpower.

International aid agencies, both bilateral and multilateral, are the third set of protagonists in policy and planning discourse. These have a continuing interest in settlement schemes (Dunham, 1982). Scudder (1985: 151) has reported on the World Bank’s expansion of lending for schemes since the late 1970s. Although settlement initiatives receive only a fraction of total loans the lending profiles of most agencies include allocations for such schemes. What tempts such agencies to persist in financing state-sponsored settlements in the light of the poor results that are commonly produced?” One reason is that international aid is about shifting funds to Third World governments and, on occasion, land settlement schemes are a useful vehicle for effecting such transfers. It is not uncommon for lending agencies and their staff to encounter difficulty in identifying projects in the rural sector that are ‘bankable’, that yield a projected economic rate of return that meets with lending agency criteria. Although such a situation is only to be expected, given the limited state of knowledge, the complexity and the dynamics of the physical, economic and social situation in rural areas and the fluctuat- ing nature of national and international systems, it places a con- siderable strain on certain divisions and staff in lending agencies who are largely appraised in terms of their ability to disburse pro- grammed flows of aid to specific countries according to inflexible timetables.” Hunt (1984) has shown that even for ‘new wave’ agencies, such as the International Fund for Agricultural Develop-

426 David Hulme

ment (IFAD), disbursing funds on schedule is a primary objective that has serious consequences for the types of activities that the Fund sponsors. Such pressures tend to favour the inclusion of settlement initiatives on the agenda for lending despite their relatively high-risk nature in terms of economic results and adverse social impacts.

Three settlement scheme characteristics assist their progression through international agency project appraisal methodologies - speed of planning, scale of funding and staffing. First, as long as public land is already available, settlement schemes can be appraised relatively quickly. They are a classic ‘off the shelf’ project type which can rapidly be packaged together by a large number of competent consultancy groups. By assuming that a simple socio- economic system is to be created, for example by planning for individualized rather than communal land tenure, the planner’s task of forecasting outputs is facilitated. Commonly feasibility studies from earlier proposals that were not adopted can be reworked so that the need for time-consuming field research is min- imized. The cabinets of agricultural and engineering consultancy firms around the world contain feasibility studies of land development projects that did not proceed, but which can be rapidly reworked if a client requires. Secondly, after appraisal, proposals for land settlement schemes usually indicate the need for loans of an order of magnitude preferred by the major multilateral agencies, say at least US$lO million. The proportionally higher administrative costs of small loans mean that they are not favoured by lending agencies. Despite the rhetoric of a preference for low cost per beneficiary projects (Asian Development Bank, 1979; World Bank, 1978) the practice of selecting high-cost approaches converts even small schemes into significant loans. For example, the World Bank sponsored Bura Irrigation Scheme in Kenya is estimated to be costing US$40,000 per settler family (Gunnell, quoted in Oberai, 1986: 153), whilst the rainfed Cape Rodney Agricultural Development Scheme in Papua New Guinea, supported by the Asian Develop- ment Bank, has costs in excess of US$20,000 per family. Thirdly, the ‘bounded site’ nature of schemes means that the potentially complicated task of relating project staffing to existing administrative structures, which are often believed to be ineffective, can be bypassed through proposing an autonomous or semi-autonomous project management unit. For the lending agency this can simplify the administration of the loan. The common ‘staffing difficulties’

State-sponsored Land Settfement Policies 427

that lead to such units being led by expatriates can also ease the administrative burden incurred by lending agencies.

These three factors provide a basis for explaining the continuing interest of aid agencies in settlement schemes and illustrate the way in which lending agency involvement in policy and planning discourses is likely, on some occasions, to favour settlement scheme inclusion on agenda for action. However, they do not lend support to the conspiracy theory types of claims made by Nugent (1985) who proposes that poor project outcomes are the intention of international lending agencies. Nugent’s assertion that ‘aid donors may have intended that colonists fail in order to open up attractive investment opportunities’ (in Hemming, 1985: 2) posits a knowledge of donor motivation which is unsubstantiated.

Formal commercial interests are the fourth group that can be seen to influence policy and planning practice to support the notion of state-sponsored settlement. These are many, varied and sometimes competing, but commonly the use of a settlement scheme approach is to their benefit. As a form of public sector activity settlement schemes have a high propensity to generate demand for formal sector inputs. Settlement schemes are usually dependent on external sources of knowledge and expertise and produce contracts for a wide range of consulting groups, civil engineering companies and construction firms as well as wholesalers and suppliers. The nature of settlement scheme plans and the desire of administrators to see that, at least on the expenditure side, plans are followed, favours formal-sector products in comparison to small-scale or locally produced goods in which the informal sector is more likely to be involved. For plantation concerns, such as Harrisons and Cros- field, SIPEF and the Commonwealth Development Corporation (CDC), outgrower settlement schemes have proved a useful means of enhancing the public acceptability of new plantings, on a nucleus estate-outgrower model, during an era in which foreign agribusi- ness has been viewed with suspicion. Clearly the influence of domestic and international commercial interests on policy and planning practice varies within and between countries, but the probabil- ity that they will see state-sponsored settlement schemes as an unobjectionable policy option is high. Business seeks to make pro- fit, and profits can be made out of settlement schemes.

From the standpoint of these four major groups involved in determining the public agenda for action a land settlement scheme approach is commonly unobjectionable.” There is the problem of

428 David Hulme

past experience but, as Schaffer (1984: 170) has observed, memory can be discouraged and feedback obliterated. The failure of earlier schemes to achieve their stated objectives or produce reasonable outcomes can be explained by the usual escape hatches. Future schemes will be more successful because of more data, better planning, increased funding, realistic implementation schedules and more competent management. If the conditions are right then at one and the same time each of the main actors in the policy and planning process can pursue their objectives - regimes can be legitimated, the offices can be kept in being, loan disbursement targets can be met and contracts can be won.

WHAT CAN BE DONE?

This article has argued that state-sponsored land settlement schemes should not be treated, as is common, as rationally selected ‘best option’ policy or planning decisions which subsequently encounter problems because of poor planning or poor management or both. Rather, they should be seen as evolving out of a complex and often chaotic series of inter-organizational and intra-organizational dealings that focus attention on a limited agenda for action that is relatively unobjectionable from the standpoints of the major protagonists. Depending on circumstances and influences a portion of what is unobjectionable filters through policy and planning processes until it is unavoidable. The nature of state-sponsored land settlement initiatives is such that they have a strong likelihood of gaining access to policy and planning agenda and, ultimately, of becoming an inevitable public action. This is not to argue that economic and technical appraisal is irrelevant, nor that learning from experience does not occur. Nor is it intended to serve as an ‘exposk’, of settlement schemes or to suggest that schemes auto- matically achieve poor outcomes. However it does point to the shortcomings of evaluating schemes and framing lessons from experience solely in terms of the rational models that are supposed to be operating. If this finding is to pass beyond the stage of being mere ‘negative social science’ (Chambers, 1983) then the question ‘What can be done?’ must be tackled. Is it possible to identify the ‘room for manoeuvre’ to which Clay and Schaffer (1984) allude?

One possibility is to accept the inevitability of the initiation of schemes that have not been fully thought out, and to seek to


mitigate the adverse consequences of such schemes by detailed study of the ways in which scheme operations might be modified to better serve the needs of those who are settled and of their host communities. This is the approach adopted by Scudder (n.d., 1985) in his USAID-sponsored review of settlement scheme planning and management. It offers the opportunity of improving the outcomes for ‘project beneficiaries’ but denies the possibility of redirecting resources towards alternative activities. One cannot help but feel that by choosing this route one of the potentially most effective critics of state-sponsored land settlement has been co-opted into tinkering with an orthodoxy that he might otherwise have been challenging. The important conclusions that arise from work such as Scudder’s may all too easily be used by some involved in the policy and planning debate as a procedural mechanism for blocking substantive challenge to settlement policies.

A second possibility is to attempt to ‘improve’ policy formulation and planning methodologies. There are a variety of options ranging from proposals for totally new normative methodologies to more modest attempts to improve the performance of the weakest stage in the existing linear models. Korten (1980) has presented radical proposals for the abandonment of ‘blueprint’ planning approaches and the adoption of a ‘learning process’ approach to development activity. Johnston and Clark (1982) and Rondinelli (1983) have partially adopted Korten’s suggestions and attempt to fuse them with linear policy and planning models to produce a more ex- perimental and adaptive planning methodology. Smith (1987), fol- lowing a teleological systems approach and focusing on the project level, argues that the generation of more relevant initial project concepts, that take account of the multifarious, autonomous decision-takers involved in project activity, would be a major ad- vance for project planning methodologies. These proposals have genuine merit, but all make the assumption that initiatives arise out of rational methodological frameworks. The evidence in relation to land settlement schemes refutes such an assumption. If the proposals that these writers make are to be of utility, then they must be related to policy and planning practice, as well as theory. Political and organizational influences on public action must be incorporated rather than ignored or externalized. The ‘group think’ of which Smith (1987) cautions is not merely ‘a very real factor in suppressing a critical reassessment of the [project identification] methodology currently in use’, it is a major part of policy and planning

430 David Hulme

practice, regardless of normative methodological ideals. A third possibility has been outlined by Chambers (1983: 190)

who argues that ‘all can do something’ and that, whatever the situation, there is scope for personal choice and action. The degree of choice and the nature of possible actions in relation to state- sponsored settlement schemes clearly varies between different individuals in different positions and situations. It is not feasible to attempt to review comprehensively all alternatives in this article, but a number of the choices that exist can be noted. For those involved in the formulation and planning of settlement schemes then the ‘self awareness’21 to which Clay and Schaffer (1984) refer, especially in relation to the actuality of one’s role in the practice of policy and plan selection, could expose limited opportunities for modifying the agenda for action. In particular, by reorganizing the way in which policy and planning practice may seek to ignore past experience, such individuals might identify opportunities for the introduction of alternative data sets based on empirical evidence rather than technocratic projections.22 There is the opportunity, in certain situations, for individuals to release information on the actions of agencies with which they are involved, but personal considerations of career, livelihood and colleagues, allied to no- tions of professional ethics, make this unlikely. The position in which academics find themselves may be less constrained, although exercising the opportunities that arise can have personal consequences in terms of consultancies for which an individual might become ineligible and entry visas that might be denied. Academics studying settlement initiatives could take action by ensuring that they take cognizance of the practices by which settlement policies and plans are selected rather than adopting the common approach of assuming that idealized models are in operation and regarding ‘political influences’ as a non-rational factor meriting no more than a passing comment. At the least, they might seek to widen the land settlement debate to include spontaneous settlement and raise the possibility of spontaneous settlement being viewed as a potential solution to development problems rather than as a law and order issue.23 Scudder (1985: 126) has pointed out the ‘impressive evidence that spontaneous settlers time and time again make better farmers’. Might it not be possible to seek for spontaneous rural settlement to be viewed in a more positive light, in the way that urban squatter settlements and self-help housing now are in many countries? More academic field study of spontaneous settlement,

State-sponsored Land Settlement Policies 43 1

rather than the more easily tackled subject of state-sponsored settlement, could broaden the available information and help strengthen the notion of facilitating spontaneous settlement.

Another alternative, for some, is the adoption of a more activist approach. Within a country, depending on the circumstances, this may take the form of direct political action and lobbying against a policy or project. At the international level this involves the mobilization of lobbies against specific activities undertaken by development agencies. Organizations such as Survival Inter- national, Oxfam and the World Development Movement have spearheaded such initiatives. A dramatic recent example of such action is the denunciation of the Indonesian government’s transmigration programme, and of the World Bank’s financial support for transmigration projects, in the British journal The Ecologist, and a call for widespread individual and group lobbying. Such activities are only likely to influence the behaviour of international agencies if they are based upon accurate information. The supply of such information to activist groups provides another opportunity for some to exercise personal choice.

There are no easy solutions to the search for room for manoeuvre in relation to state-sponsored land settlement policies and plans. An essential initial step for those who are concerned about the effective use of resources and the outcomes of settlement initiatives, including planners, managers and evaluators, is to recognize that such schemes rarely have their origins in a linear policy or planning methodology. One cannot be optimistic, though, about the likelihood of policy and planning practice concerning state-sponsored settlement schemes being greatly modified over a short period of time. The continuance of the strategy over several decades and across many nations points to one firm but paradoxical conclusion. Whilst countless studies by academics and practitioners, based upon stated policy or project objectives, have indicated that settlement schemes are commonly ‘failure-prone’, from the standpoint of the major protagonists determining the policy and planning agenda, state-sponsored land settlement schemes are a tried and proven success.

NOTES

I am grateful to David Marshall, Martin Minogue, Peter Smith and Mark Turner for comments on an earlier draft.

432 David Hulme

1. State-sponsored land settlement schemes are not a natural category and a wide range of activities fall under such a heading. For the purposes of this article a state-sponsored land settlement scheme may be defined as the movement of people to areas that are perceived to have underutilized agricultural potential with government support or guidance. This includes projects on both rainfed and irrigated lands.

2. Brazil, Colombia, Egypt, Ethiopia, Indonesia, Malaysia, Mexico, Nepal, Nigeria, Papua New Guinea, Peru, the Philippines, Somalia, Sri Lanka, Sudan, Thailand and Zimbabwe.

3. Determining the ’success’ or ‘failure’ of an initiative is a complex task and the pronouncement may vary with the standpoint that is taken. For an examination of the problems of evaluation see Apthorpe and Gasper (1982). However, for many land settlement schemes it is found that stated objectives are not achieved, pressing problems are occurring, settler abandonment rates are high and staff are disillu- sioned. For the purposes of this article such results are regarded as constituting a ‘failure’.

4. The one major exception to this statement is the work of the Federal Land Development Authority (FELDA) in Malaysia.

5 . There are two exceptions to this narrow approach. Palmer (1974) has proposed that land settlement schemes can be understood in terms of a model of ecological imperialism. The central tenet of this model is that orthodox prescriptions for agricultural development merely serve to further ‘the elaboration of a world organization that is centred in industrial societies and degrades the ecosystems of the agrarian societies it absorbs’ (Palmer, 1974: 240). The argument is grounded in conspiracy theory and is supported by the use of selective data. It is deterministic and if carried through to its logical conclusion would be critical of any change that increases an agrarian group’s involvement with national or international economic systems. Dunham (1982) has pointed to the crucial importance of political factors in explaining Sri Lanka’s settlement policies, and although he indicates (p. 58) that some form of class analysis might assist in understanding the country’s settlement history, he does not elaborate.

6. State-sponsored land settlement schemes have also been promoted in advanced nations-Spain, Italy, the Soviet Union and Australia-but a discussionof the land settlement experience in such countries is beyond the scope of this article.

7. The dividing line between the ‘decision’ to pursue a settlement policy and the ‘decision’ to undertake a specific project is difficult to distinguish as commonly a settlement policy is comprised of a sequence of decisions to pursue specific projects.

8 . The most recent example concerns the Cape Rodney AgriCUkUrdl Develop- ment Scheme towards which the Asian Development Bank is loaning US$20,000,000. No alternatives were considered by the planning team, project identification was skipped, known problems in relation to managerial capacity, land acquisition and project beneficiary behaviour were glossed over, and previous experience was selectively ignored. Two years into the loan the ‘project beneficiaries’ have still not been identified, attempts are being made to resite the scheme to a new area and the government is considering the possibility of ‘project closure’.

9. The late Bernard Schaffer (1984) elaborates on this issue in some detail in his eloquent but obfuscating style.

10. In the early 1970s the Australian administration hastily inaugurated several land settlement schemes of a type known to be highly problematic, in the East New


Britain province of Papua New Guinea. This was intended to reduce the strenuous demands from the indigenous population for the return of foreign-owned plantation lands. The thorny issue of land reform was thus avoided for a year or two and the powerful planter lobby was not antagonized.

11. For a review of the literature on the social consequences of settlement schemes see Hulme (1984).

12. This includes incorporating social planners as sub-sections of large divisions, job descriptions that impose heavy workloads and appointing women as social planners in societies where women are not treated as equals in decision-making processes.

13. Showcase schemes cannot only generate domestic political popularity, but can also be used to produce international political advantage. The Hoskins Scheme in Papua New Guinea helped the Australian government to improve its standing with visiting United Nations missions observing the decolonization process (Hulme, 1984).

14. In countries where ‘land settlement divisions’ have been established in the public service, then proposing the development of new schemes is often essential for the long-term survival of such units. Such units were established in the majority of African and Asian countries during the colonial period.

15. For an example of this see Maos’s (1984) treatise on the physical planning of settlement schemes. He writes of the advantages of being able to plan without the ‘additional complexities of traditional settlement’.

16. Occasionally, as on the irrigated El Nahda settlements to the west of Alexan- dria in Egypt, a crude ‘two class’settler system is utilized in which ‘advanced’ farmers (former public servants with degrees and personal capital) receive large blocks of land, whilst ‘ordinary’ farmers (the landless, farm labourers, etc.) are allocated small blocks.

17. The extent of the sanctions that scheme managements can impose can be very wide. For example, in Papua New Guinea one scheme manager used to rein in underperforming settlers with outside jobs by telephoning their employers and requesting that the miscreants be sacked. Social and commercial relationships between the scheme manager and local businessmen ensured that such requests were met.

18. It is not suggested here that all aid-assisted settlements produce disappointing results. For example, World Bank loans to Malaysia and Papua New Guinea have helped to establish projects that have achieved their objectives. In contrast Bank loans to Indonesia and Kenya for settlement initiatives have been associated with schemes that have produced unsatisfactory outcomes.

19. Rigid disbursement schedules evolve because of the internal structuring of lending agencies and the pressures that member states exert on these agencies to disburse funds.

20. This is not to suggest that the interests of these groups always coincide in a cosy manner with regard to land settlement policies. Rather, on frequent occasions proposals for settlement will be broadly acceptable to all groups.

Clay and Schaffer (1984) refer to self-awareness in a positive light, but it should be noted that self-awareness could also lead to individuals adopting an attitude of resigned cynicism and a total acceptance of existing organizational processes.

22. For example, the practice of forecasting settler outputs of tree crops as a

21.

434 David Hulme

percentage of good plantation yields could be queried, and it could be argued that projections should be based upon actual experience in comparable schemes.

23. On the day of writing’this I note a report in The Times (2 October 1986) that 600 Brazilian military police have been called into the state of Rio Gande do Sul to stop landless farmworkers from ‘illegally seizing land’.

REFERENCES

Anti-Slavery Society (1984) The Chittagong Hill Tracts. London: Anti-Slavery

Apthorpe, R. and Gasper, D. (1982) ‘Policy Evaluation and Meta-evaluation: The

Asian Development Bank (1979) Sector Paper on Agriculture and Rural Develop-

Baird, A. and Potts, D. (1978) ‘The Project Spiral’, Discussion Paper No. 24, Project

Barnett, T. (1977) The Gezira: An Illusion of Development. London: Frank Cass. Baum, W. (1978) ‘The World Bank Project Cycle’, Finance and Development 4 (15) :

Biggs, S. (1984) ‘Awkward but Common Themes in Agricultural Policy’, in E.J. Clay and B.B. Schaffer (eds) Room for Manoeuvre: An Exploration of Public Policy in Agricultural and Rural Development, pp. 59-74. London: Heinemann.

Booth, D. (1985) ‘Marxism and Development Sociology: Interpreting the Impasse’, World Development 13 (7): 761-87.

Budiardjo, C. (1986) ‘The Politics of Transmigration’, The Ecologist 16 (2 and 3):

Carruthers, I.D. and Clayton, E.S. (1977) ‘Ex Post Evaluation of Agricultural Projects: Its Implications for Planning’, Journal of Agricultural Economics 28 (3):

Chambers, R. (1969) Settlement Schemes in Tropical Africa. London: Routledge &

Chambers, R. (1983) Rural Development: Putting the Last First. London: Longman. Chase Smith, R. (1985) ‘Indigenous Communities and Peru’s Conquest of the

Amazon: The Distribution of Land and Resources in the Pichis-Palcazu’, in J. Hemming (ed.) The Frontier After a Decade of Colonisation, pp. 207-15. Man- Chester: Manchester University Press.

Clay, E.J. and Schaffer, B.B. (1984) ‘Room for Manoeuvre: The Premise of Public Policy’, in E.J. Clay and B.B. Schaffer (eds) Room for Manoeuvre: An Explora- tion of Public Policy in Agricultural and Rural Development, pp. 1-12. London: Heinemann.

Colchester, M. (1986a) ‘Unity and Diversity: Indonesia’s Policy towards Tribal Peoples’, The Ecologist 16 (2 and 3): 89-98.

Colchester, M. (1986b) ‘Banking on Disaster: International Support for Transmigra- tion’, The Ecologist 16 (2 and 3): 61-70.

Dunham, D. (1982) ‘Politics and Land Settlement Schemes in Sri Lanka’, Develop- ment and Change 13 (1): 43-61.

Fearnside, P.M. (1986) ‘Agricultural Plans for Brazil’s GrandeCarjas Program: Lost

Society.

Case of Rural Cooperatives’, World Development 10 (8): 651-68.

ment. Manila: ADB.

Planning Centre, University of Bradford.

10-17.

111-16.

305-18.

Kegan Paul.


Opportunity for Sustainable Local Development?’, World Development 14 (3):

Hemming, J. (ed.) (1985) The Frontier After a Decade of Colonisation. Manchester: Manchester University Press.

Hirschmann, A.O. (1963) Journeys Towards Progress. New York: 20th Century Fund.

Hulme, D. (1984) Land Settlement Schemes and Rural Development in Papua New Guinea, unpublished PhD thesis, James Cook University, Australia.

Hunt, D. (1984) ‘Policy “Implementation”: IFAD and Rural Poverty in Kenya’, in E.J. Clay and B.B. Schaffer (eds) Room for Manoeuvre: An Exploration of Public Policy in Agricultural and Rural Development, pp. 101-26. London: Heinemann.

Johnson, K. (1985) ‘Conflicting Images and Realities in Project Planning’, Third World Planning Review 7 (4): 351-5.

Johnston, B. and Clark, W. (1982) Redesigning Rural Development: A Strategic Perspective. Baltimore, MD: Johns Hopkins Press.

Korten, D.C. (1980) ‘Community Organization and Rural Development: A Learn- ing Process Approach’, Public Administration Review 40 (5): 480-511.

Lewis, W.A. (1954) ‘Thoughts on Land Settlement’, Journal of Agricultural Econo- mics 11: 3-11.

McBeth, J. (1981) ‘The Bulldozer Invasion’, Far Eastern Economic Review 8 May

Maos, J.O. (1984) The Spatial Organization of New Land Settlement in Latin America. Boulder, CO: Westview Press.

Nelson, M. (1973) The Development of Tropical Lands. Baltimore, MD: Johns Hopkins Press.

Nelson, M. (1977) ‘Twenty-four Settlement Projects in Latin America’, Develop- ment Digest 15 (4): 91-103.

Nugent, S. (1985) ‘Amazonia and Foreign Aid: The Regression of Development’, a summary in J. Hemming (ed.) The Frontier Affera Decadeof Colonization, p. 267. Manchester: Manchester University Press.

Oberai, A.S. (1986) ‘Land Settlement Policies and Population Redistribution in Developing Countries: Performance, Problems and Prospects’, International Labour Review 125 (2): 141-61.

Palmer, G.B. (1974) ‘The Ecology of Resettlement Schemes’, Human Organisation

Ploeg, A. (1972) ‘Sociological Aspects of Kapore Settlement’, in J.P. Longaroux et al. (eds) Hoskins Development: The Role of Oil Palm and Timber, New Guinea Research Bulletin No. 49, pp. 21-1 18. Canberra: Australian National University.

Pratt, B. and Boyden, J . (1985) The Field Directors’ Handbook. Oxford: Oxford University Press for Oxfam.

Pryor, R. J. (1979) ‘Development Planning: Population Redistribution Policies and South-east Asian Development Plans’, in R.J. Pryor (ed.) Migration and Develop- ment in South East Asia, pp. 275-85. Kuala Lumpur: Oxford University Press,

Rondinelli, D.A. (1983) Development Projects as Policy Experiments. London: Methuen.

Schaffer, B.B. (1984) ‘Towards Responsibility: Public Policy in Concept and Prac- tice’, in E.J. Clay and B.B. Schaffer (eds) Room for Manoeuvre: An Exploration of Public Policy in Agricultural and Rural Development, pp. 142-90. London: Heinemann.

385-409.

1981: 26-8.

33 (3): 239-50.

436 David Hulme

Scudder, T. (n .d . ) 'The Development Potential of New Lands Settlement in the Tropics and Subtropics: A Global State of the Art Evaluation with Specific Emphasis on Policy Implications', unpublished report funded by USATD, grant no. AID/DSAN - G - 0140.

Scudder, T. (1985) 'A Sociological Framework for the Analysis of New Land Settlements', in M. Cernea (ed.) Putring People First: Sociological Variables in Rural Development. Oxford: Oxford University Press.

Smith, P. (1987) 'Improving the Project Identification Process in Agricultural De- velopment', Public Administration and Development (forthcoming).

Thiesenhusen, W.C. (1971) 'Colonization: Alternative o r Supplement to Agrarian Reform', in P. Dorner (ed.) Land Reform in Latin America: Issues and Cases. pp. 209-25. Wisconsin: Land Tenure Centre, University of Wisconsin.

World Bank (1978) Agricultural Land Settlement. Washington, DC: World Bank.

David Hulme is a lecturer at the Institute for Development Policy and Management at the University of Manchester. He directs the postgraduate development administration programme. He worked in Papua New Guinea for many years and has published articles about rural development policy, regional planning and agricultural extension.

State‐sponsored Land Settlement Policies: Theory and Practice

Documents