Journal of Social Development in AfricaJ1999), 14,1,101-119 Participatory Planning: Counterbalancing Centralisation ARNON A BAR-ON* & GERARD PRINSEN + ABSTRACT Many African countries have clearly established central development planning structures, unlike most countries in the North. This article takes acloser look at one such planning machinery, in Botswana, that having remained virtually unchanged for three decades is currently under review to reinvigorate its original commitment to popular participation, which was undermined in the process of national development. The article begins with a brief introduction to Botswana. This is partly because having one of the few stable regimes in Africa, Botswana is little reported on and so remains largely unknown, but mainly because policies and the structures that support them develop against abackdrop of specific historical experiences that shape these policies and structures. Subsequently, following a description of the country's current development planning mechanism and the factors that have led to its re-examination, we analyse an exploratory experiment in grassroots participatory planning, focusing on the attitudes of middle-management civil servants to its effects. The conclusions for participatory planning are generally positive, and may be relevant to other countries in Southern Africa, like Zambia and Zimbabwe, whose planning structures bear a close resemblance to those in Botswana, even though their governments havepursued different ideological paths since independence (see, for example, SNV, 1995). A Cattle Herding Society Transformed Botswana is a landlocked country, larger than France, situated north of South Africa. The Kalahari Desert covers 85% of the land, and although the average annual precipitation is a fairly healthy 475 millimetres, this figure is deceptive as evapo-transpiration far exceeds the rainfall, and droughts, often severe, are more the norm than the exception. It is mainly on the country's eastern rim, therefore, where the environment is more hospitable, that most of the inhabitants reside. In 1997, these people numbered around 1.4 million, which makes the Batswana one of the world's most sparsely populated nations. * Senior Lecturer, DepartmentofSocialWork,UniversityofBotswana>PrivateBag0022, Gaborone, Botswana, Tel/Fax (267)-31321 38; e-mail bar-on@noka.ub.bw + Extension Workers' Training Co-ordinator, Ministry of Finance and Development Planning, Gaborone, Botswana. 102 Arnon A Bar-On & Gerard Prinsen The harsh climate and poor soils have fully determined Botswana's historical legacy. In 1885, with no meaningful opportunities for investment, Britain reluctantly declared the territory a protectorate, solely to offset Germany's expansion from South-West Africa and to prevent Cecil Rhodes from turning its tracts into a private domain in his quest to link Africa by rail. From this time, and for the next eighty years, life went on virtually undisturbed, the routine being regulated by a handful of colonial officials who were made lethargic by the country's inability to cover even their salaries. Indeed, such a backwater was Botswana regarded that it was administered by an ironically-named Resident Governor from outside its borders, and its only note to international fame arose from the marriage of a local royal to a white English woman, which offended the racial assumptions of colonial rule. In 1966 Botswana was given independence, still unable to sustain itself economically or, for that matter, on almost any other front. Officially, the average per capita annual income was US$30, placing Batswana among the poorest of peoples, but even this amount derived mainly from the remittances of relatives working abroad. In fact, so impoverished and underdeveloped was the country that it was served by only three kilometres of tarred road, and the establishment of the Cabinet denuded the embryonic school system of its only university graduates (Harvey & Lewis, 1990; Parsons, et al, 1995). By the early 1970s, however, all this changed. In 1968 extensive diamond pipes were found, and within five years, helped by a re-negotiation of the Southern African Customs Union Agreement, Botswana faced the extraordinary situation for a developing country of achieving a balanced recurrent budget from internal sources. After that, and throughout the 1980s, the economy grew in real terms at an average of 12% a year, exceeding that of Asia's Four Little Dragons, which transformed Botswana into one of Africa's richest countries. With its new found wealth, the state could expand budgets to formal social provision, with which most its people practically had no experience. From 1976 to 1993 this expenditure increased on average by 13.5% a year in real terms - outstripping the growth of the population fourfold- of which basic social spending accounted for over half (Duncan, et al, 1994). This reflected an effort to ensure that virtually everyone has access to basic education, health, and clean water, and teaching the people "to welcome and benefit from [these institutions]" (Government of Botswana, 1968:63). Consequently, facilitated by a stable democratic regime, uncharacteristic of the rest of the African continent, rapid social gains accompanied Botswana's economic improvement. For example, between 1970 and 1990 primary school attendance rates more than doubled (with female enrolment exceeding that of males), and life expectancy rose from 46 to 63 years of age with the halving of infant and under-five mortality rates. Participatory Planning: Counterbalancing Centralisation 103 Planning Processes and Structure In carrying out this work, which by the mid-1990s placed Botswana at the top of the human development improvement league, second only to Malaysia (UNDP, 1995a), government attached from the beginning importance to actively involving its constituents in its planning, long before popular participation became the prevalent approach to development This active involvement, or "consultation" in the local jargon, is described as, "a process whereby the opinions and involvement of the people at village level are taken as the basis, by officials and elected representatives.. for making decisions about.. programmesfor a district as a whole and for each village individually" (Noppen, 1982:1). The intention is that locally produced Village Development Plans should be integrated in District Development Plans, and through the latter into National Development Plans, each of which last for six years with a mid-term review. The origins of this grassroots orientation lie deep in Botswana's past. Traditionally, critical elements of communal life were conducted democratically, bearing in mind that the underlying structure of Tswana society, as in most of Africa, is founded on filial piety and so is essentially authoritarian. In this system, a chief, who had strict obligations, led each community: he held the group's lands in trust and used them and his power of taxation on behalf of his people, especially in time of need. On some of these matters he decided with the help of his senior relatives and the village headmen, but most issues of public policy were decided in open council, or Kgotla, where opposing the chief was quite acceptable. The colonial regime, being largely indifferent to internal affairs, besides keeping the peace, left this system intact to form the basis of Botswana's post-colonial political structure. To use this system for development planning and, at the same time, extend its authority over the country's hitherto self-reliant communities, the government established in every locality a Village Development Committee (VDC) or its urban equivalents, Ward Development Committees. Composed of ninepopularlyelected residents, the VDC is meant to serve first as the executive arm of the Kgotla and the co-ordinating body of other local organisations, such as the Farmers' and Parent-Teachers' Associations, and, secondly, as the link between the community and its formal, mainly governmental, service providers. Accordingly, where planning is concerned, residents are meant to identify their needs and priorities in the Kgotla which the VDC then formulates in greater" detail with the relevant village-level service agencies. 104 Arnon A Bar-On & Gerard Prinsen As a second step, the chairpersons and secretaries of a!l the VDCs in a district convene with the middle-level administrators of the local authority, who link the field to central government through their line ministries. At this meeting, called a District Development Conference, the villages' needs are aggregated sector by sector into a tentative District Development Plan. Also attending are councillors, representing the political system, who - after the broad objectives of the plan have been translated into specific projects and expenditure proposals - have the power to recommend it to government. Once approved by the Districts, this process is repeated at national level. First, the relevant ministries review each plan according to their own priorities and budget ceilings, and combine them to form comprehensive sectorial national plans. Then the Ministry of Finance and Development Planning takes over. It coordinates the different sectors, and adjusts, where necessary, specific projects to meta-national policies, such as that the rate of development expenditure must correspond to the nation's expected implementation capacity. Lastly, after approval by Cabinet, the finalised plan is brought to Parliament whose role, de facto, is reserved to influencing the timing of particular projects, as, for example, bringing forward the scheduled construction of a specific school - but only if another project, in the same locality, can be delayed. In summary, building on its indigenous political legacy, Botswana tried to institute a genuinely "bottom-up" development planning system in which popularly identified local priorities are co-ordinated horizontally and then gradually combined vertically to form a comprehensive national development plan, modified by financial and implementation constraints alone. This ideal, however, has run into considerable difficulties due to several factors that build on each other. For the sake of clarity, and better to understand why the system is currently being reviewed, these factors are explored separately at the village, the district, and the national levels. Consultation Gone Awry Village level Although the Kgotla is thoroughly established in rural society (encompassing 57% of the population), its influence as the ultimate local decision-making body has declined sharply. Before independence, villages and smaller settlements were virtually autonomous. They initiated their own development and implemented it through self-mobilisation, which required unanimity. Later, many of these functions, especially in infrastructure, became the responsibility of government, whose clear standards for service delivery (for example, by the number of people served) reduced the need for consultation. This was both because of these standards and Participatory Planning: Counterbalancing Centralisation 105 because it is obvious that the communities, which no longer have to construct their own services, will always approve them. Consequently, the nature of the Kgotla has changed. First, it has become primarily a forum, "in which politicians and civil servants...try to persuade people to accept plans devised elsewhere" (Rural Sociology Unit, 1981:15). Secondly, and largely because of this top-heavy and largely rhetorical approach, meetings are poorly attended, mainly by elderly persons and people who live near by. Also, given the patriarchal structure of society, among those who do attend, women, youth, and ethnic minorities are seldom encouraged to participate, so that what decision-making as takes place does not represent the sentiments of the community as a whole. A similar fate has befallen the VDCs. For one, mirroring the Kgotla, most VDCs are dominated by the elite, who, believing they effectively articulate the needs and aspirations of all, rarely hold themselves accountable to the community. In turn, marginalised groups have refrained from trying to change this situation, partly because they believe the institution is for the elite, but mainly because of their own debilities caused by a culture in which questioning people in authority outside the Kgotla is unacceptable. Secondly, but unlike the Kgotlu, which with all its weaknesses at least has the advantage of being community-based, VDCs are community-located, which is to say government-established and overseen. One consequence of this is that residents often see them as working to foreign agendas, especially, as with time, government has given them several tasks to perform on its behalf, such as overseeing its public works operations. Finally, while Ihe VDCs potentially could have used their position to build an independent power base, government ensured this would not happen, albeit usually by default In part it accomplished this by paying the members a sitting allowance that, in effect, makes them government employees, but these payments being restricted to only four meetings a year invariably decreases their motivation to meet more frequently. Another strategy has been to confine the VDCs' direct taskperformance to support roles, such as building and maintaining houses for government staff, or devolving to them responsibility for identifying residents in need of social assistance, which is frequently divisive. District level Some difficulties mentioned above are repeated at district level. While it is often a question of which came first - the chicken or the egg?, the major constraints at this level arise from the local authorities' almost total dependence on central government for trained labour and for finance. Additional constraints attach to communication and co-ordination, both with the field and within the administration, and to the attitude of many civil servants towards the people they serve. Together these factors contribute to vicious cycles that reduce the capacity to carrying out bottom-up development planning successfully. 106 Arnon A Bar-On & Gerard Prinsen As previously alluded to, Botswana had little trained labour on the eve of independence. Indeed, as late as 1980 only 80 Batswana were enrolled in the then joint University of Swaziland, Lesotho and Botswana, and only a handful of others were studying elsewhere. The establishment of a central administration virtually denuded, therefore, the country of its trained personnel, leaving the local authorities - where a priori work is less appealing - particularly wanting. To reduce this gap in labour and equalise its distribution, government appropriated the local authorities' responsibility for staffing for all but their support positions, which helped to alleviate the problem slightly (the national supply of trained labour still being low). However, where the desired planning process is concerned, this proved an impediment One problem, derived directly, is that with no local authority overpersonnel, employees inevitably give their loyalty to central government and to their line ministries over all. In fact, this orientation is semiofficial as shown by the title of most district level and direct service providers - "extension workers" - that denotes their affiliation to some outside agent, not the community. Another problem is the high turnover of staff due to an ongoing "circulation'' of personnel between departments and districts, a carry-over from the colonial period when this practice was used to train public servants for a variety of positions, and to prevent them from gaining a strong local footing. Few of the officers have sufficient knowledge, therefore, of the communities they are meant to serve, to the obvious detriment of the planning process. The second structural obstacle is financial. Besides her mineral wealth, which as a capital-intensive industry provides little employment, Botswana remains fundamentally poor. Around half the people live below the official poverty line, already set at the barest minimum for physical survival alone, and although reputed to have one of the highest skewed income distributions in the world (Nteta, et al, 1997), the potential tax base of the private sector is extremely low. Consequently, government is forced to finance wholly all capital projects and between 60% and 95% of the recurrent expenditure of all councils, which gives it total leverage to direct their development Lastly, district-level planning is hampered by poor communications, horizontal as well as vertical. The former difficulty arises primarily from the officers' direct subordination to their line ministries, which critically reduces their incentive to coordinate. For example, a recent study found that few district department heads even knew how often they were meant to convene (Prinsen, et al, 1996). District Development Plans, therefore, are more an assemblage of separate departmental plans, in which integration and analyses of the districts' particular problems and opportunities are virtually absent In contrast, the obstacles in vertical communication are chiefly logistical. This is mainly because many villages are remote, transport scarce, and telephones almost non-existent outside headquarters, so Participatory Planning: Counterbalancing Centralisation 107 district workers and villagers find it difficult to keep in touch. Another reason, however, is many officers' negative attitudes towards villagers because of their lower educational attainments and standard of living (Hutton, 1994), and the villagers' concomitant culturally induced deference to the officers' position. National Level In multi-party democracies, government's tendency to thwart decentralisation is high, especially, as with Botswana, where one party has been in power since independence which blurs the distinction between party and state (Mutahaba, 1989). Add this to the weak local taxation base and the generally low level of local management, and there is an ensured recipe for government officials pronouncing that, "the quality of district development plans continues to decline so much that they no longer can be taken as major inputs into the national plans" (Molale, 1995:19). Positions like these led in the past to suggestions to abolish the local authorities (Mfundisi, 1995). Obviously unfeasible, however, a compipmise was reached by which government's top local official, the District Commissioner, chairs the local development planning process, thus extending the principle-agent model in central-local government relations in which local authorities are extensions of central government, responsible for executing the policies of the latter. Moreover, and because of this centralisation, adherence to the National Development Plans has become an article of faith - extraordinary in a capitalist economy (Parsons, et al, 1995). To summarise briefly, while Botswana's official developmentstrategy incorporates fully a policy of community-led planning, her reliance on an almost single, state-controlled source of income and lack of sufficient social capital have strongly undermined this idea to the extent that her planning process has become chiefly a device for engineering consent to centralised decision-making and giving it credence. Given, however, that this transformation has served the country so well for so long, the crux of the issue is why change it now? The principal answer to this question is that the present mechanisms and their approach have outlived their usefulness. Pressures for Change Developmental and Bureaucratic Imperatives As mentioned above, development planning in Botswana has been infrastructureled to compensate for the country's initial paucity in formal service provision. Today, much of this structural coverage has been achieved, or is in the pipeline. 108 Arnon A Bar-On & Gerard Prinsen Thus 90% of households are within fifteen kilometres of a health facility, with most less than eightkilometres away (UNDP, 1995b), and the proportion with access to non-piped water has declined to 17% (Siwawa-Ndai, 1997). (1) Attention is shifting, therefore, to recurrent service provision and its enrichment, which inevitably lagged behind. For example, while primary school enrolment now averages 85%, untrained staff from the one-year post-secondary school National Service programme still does much of the teaching, especially in rural areas. Also, with no major additional sources of state income in prospect, planning must adjust to tighter public spending, which implies strengthening existing facilities rather than expansion, and relying on more careful targeting and prioritisation in relation to need. All this poses new challenges for the bureaucracy, first in diversifying its outlook beyond its predominantly "hardware" approach and, secondly, in modifying its paternalistic mode of operations, having realised that carrying the entire burden of change is beyond its capacity. Political Imperatives A second need for change is political. Countries experiencing rapid economic growth are likely to face a direct trade-off between the pace of development and community involvement. The faster the development, the lower people's involvement^ Yet while this might be instrumental in the early stages of development, lessened involvement has its costs. In particular, it fosters a culture of passivity among the beneficiaries that restricts their role to producing Christmas-like shopping lists of requests which, increasingly difficult to meet, generates, in turn, disillusionment in the system as a whole. Politicians, like the bureaucracy, have concluded, therefore, that sustainable development requires increased participation in decision-making, if only to lower the level of popular discontent/5,) Revitalising the Process: Participatory Rural Appraisal In 1994 the bureaucratic and political forces converged in the appointment of an inter-ministerial working group to re-examine the country's development planning, ostensibly reflecting concern that its infrastructure and economic trickledown approach had failed to strengthen the rural economy. The major conclusion from this examination was that the existing arrangements and their outputs "had [unwittingly] undermined community leadership structures" and so contributed to "an [unsustainable] syndrome of dependency on government" (MFDP, 1997:1). The working group called, therefore, for "a significant shift of responsibility and control over rural development activities from central and district level to community level" (ibid), labelled a Community Based Strategy. Participatory Planning: Counterbalancing Centralisation 109 Parallel to this development, but quite separately, the government also decided to pilot a participatory planning methodology called Participatory Rural Appraisal (PRA) to assess its feasibility as a standard instrument in the planning process. Developed originally in India and in Kenya, PRA is a process in which service providers co-operate with members of a community to (a) assess their situation in their own idiom, (b) identify problems and opportunities within the community and its environment, and (c) establish a prioritised list of actions to which all the stakeholders commit themselves, called a Community Action Plan (CAP). The entire process usually lasts ten to fourteen days, and employs specially designed group work and visualisation techniques to incorporate all segments of the community into the proceedings (for further details see Chambers, 1994a, 1994b, 1994c and PRA Notes published by the International Institute of Environment and Development). The Botswana project was carried out in four villages of varying size in four of the country's ten districts. It was accompanied by participatory observation research that focused on district level extension workers because of their pivotal role in both the current planning practice and in any new structure that may emerge due to their commanding, middle-management, position in government's interface with the public. In all, 74 district-located, and 62 village-located, civil servants were interviewed individually and in focus group discussions at four intervals: before the PRA application, half-way through the application, immediately at its conclusion, and six months later (Prinsen et al, 1996). Also interviewed were 151 villagers (in a balanced gender proportion), 17 village leaders, and the chief planning officers in 6 government ministries. Finally, to assess the long-term impact of the methodology, the research was extended to 5 additional villages in 3 further districts where PRAs were conducted two years earlier, but where a donor agency recruited a village-based co-ordinator to follow through the CAPs that emerged from the PRAs. In the four "base" villages, no one person was given such a task. The experiences from this projectrevealed two major points. First, unconfirmed many of the obstacles in the current planning process alluded to above, and especially the limited role of villagers and their organisations in this process and the general sense of frustration this engendered. Thus in eight of the nine villages, the residents repeatedly told the investigators that officers "tell us things like children" and that "we can never ask or discuss things." Similarly, but from the service-delivery side, 65% of the officers supported the statement that "government programmes disable villagers by taking responsibility for their lives." The second point the project showed was that participatory planning can provide some of the leverage required to change three critical elements in the district officers' practice of development planning: their communication with, and attitudes to- 110 Arnon A Bar-On & Gerard Prinsen wards, their constituents, their communication among themselves, and their communication with central government. Each of these relationships must be modified if the civil service is to be reoriented, and reorient itself, from being a "benefactor of development" to becoming a "facilitator of development." The following section first reiterates some problems in each of these communication channels and then provides findings from the pilot project that show the effects of participatory methods like PRA on development planning and their implications. Working with the Community 1. Appreciation of Villagers' Knowledge For participatory processes to be meaningful, facilitators must appreciate: (a) the participants' knowledge about the problems that confront them; and (b) their capacity to solve these problems. The project revealed that a clear majority (80%) of the district officers met the first of these preconditions (the constituents' knowledge of their situation), and that this proportion even increased the longer the officers stayed in the field. However, when assessing the villagers' problem-solving skills, a more divergent picture emerged, with half the respondents questioning this capacity even after the PRA application. It can be concluded, therefore, that working in the field directly and intimately does not affect most civil servants' perception of their constituents' problem-solving skills. Indeed, it might even lower their appreciation of this factor as 53% of the village-located workers - who are not only in daily contact with villagers but usually live among them - believed that they do not have such skills, while only 41% believed they did. 2. Representation Before experiencing PRA in action, 61% of the district officers believed that the forums they customarily use for consultation (the Kgotla and VDCs) adequately represent all segments of the population, and only 23% disagreed. However, immediately after their experience, with an alternate model in hand, these two proportions converged (44% and 48%, respectively). Moreover, in the later follow-up, after working on the CAPs, 85% of the officers were already of the opinion that PR A offers better opportunities for the expression of community-wide sentiments than the Kgotla and the VDCs. Personal opinions aside, how representative and active were the village participants in practice? First, the number of persons attending the PRA sessions at any one time varied between 8 and 79, averaging between 26 and 37 people depending on the location. In the light of the fact that there was considerable fluctuation among the participants from day-to-day and between morning and afternoon sessions due to other calls on'their time, it can be estimated that these numbers Participatory Planning: Counterbalancing Centralisation 111 covered between 5 and 18% of each village's population (the largest village, with 2 500 residents, representing the former figure and the smallest village, with 400 residents, representing the latter). Generally, men and women attended all sessions equally, but, for reasons that remain unexplained, persons less than 30 years of age were significantly under-represented. As to the level of the participants' involvement beyond mere physical presence (gauged by the frequency of their remarks and their impact on the proceedings, rated "low," "medium" or "high"), the findings were mixed. During the plenary sessions (excluding the ranking exercises in which all present took part by casting their vote[s]), the number of persons having an impact was limited to around 5. These were usually also men more than 50 years of age, which replicated customary public decision-making practices in Botswana. But in the smaller and more numerous groapwork meetings that generated the bulk of the information for the plenary sessions, women, members of ethnic minorities, and the poor contributed on an equal footing with the older men. Combining these findings suggests, therefore, that PRA enhances grassroots participation significantly by giving a voice to populations who are customarily marginalised. 3. Capacity of Village Organisations As viewed by middle management, implementing policy is more important than formulating it as carrying out policy constitutes most of their work and it is this side of policy for which they are accountable. The officers' perception of the villages' capacity to carry its share in the CAP is of critical significance, therefore, if they are to support participatory planning. At district level, the proportion of officers who believed that village organisations have this capacity and those who believed they do not were evenly split However, at village level, among the direct service providers, 65% judged these organisations are too weak. One explanation for this general pessimism is that with increased exposure to the villages, an increasing number of officers concluded that the inhabitants were too divided among themselves to work together (raising from 43% before the PRA to 61% at its conclusion). Asecondpossibility is that based on their previous personal and workrelated experience, most officers believed before the PRA that women were more responsible than men in dealing with community affairs, implying that they expected women to play a crucial role in the CAP. By the end of the PRA, however, the proportion of officers who held this view fell by 44%, which might have led them to reassess the villages' capacity to carry out the CAP. Yet these findings relate only to the pilot project, which was followed-up before much work could be done on the CAPs. In the remaining five villages, where work on the C APs was well under way, the exact opposite was found. There 90% of the officers noted that they could rely on the village organisations. This suggests that where appreciation of 112 Arnon A Bar-On & Gerard Prinsen local capacity is concerned, long-term implementation that involves joint villageis- officers work is more important than the shorter-term joint identification and prioritisation of unsatisfactory situations. 4. Interaction and Workload Another issue in introducing participatory planning methods is whether they change the relations between their participants for the better, which for the pilot project meant the relations between the officers and the villagers on the one hand, and among the villagers on the other. The findings in this regard were highly mixed, and depended on the position of the respondent Concemingthe first setofrelations (between officers and villagers), 30% of the village-located officers in the base villages reported that more individuals came to their offices following the PRA, and 25% saw an increase in their relations with village organisation. In the 5 follow-up villages the corresponding increases were 55% and 80%. The villagers, on the other hand, usually noted that little had changed as the officers, they claimed, remained aloof. In contrast, the picture on the relations among the villagers was more optimistic. An equal majority of both district and village level officers (76%) believed the PRA had brought the villagers and their customary leaders closer together (going up to 100% in the follow-up villages), and 40% of the village leaders concurred. If we accept the officers' responses to their relations with their clients, even if somewhat exaggerated, then participation increases their workload which, in turn, might decrease its attraction for them. Yet when examining this potential obstacle immediately after the PR As, two-thirds of the officers expected that their investment in PRA would offset their investment in other areas, and consequently reduce their workload overall. In the follow-up it was found that this expectation was optimistic - as only a third admitted such reductions in practice - but it can be generally concluded that at least PRA does not increase civil servants' potential reluctance to participatory planning due to increased workloads. 5. Overall Appreciation of PRA for Planning Both immediately after their experience with PRA and in the later follow-up, officers at both the districtand village level believed that it was a better instrument for planning than the tools they currently use. With clear majorities, varying from 70% to 88%, they forwarded the following three reasons as their major explanations for this preference: (a) PRA leads to a better understanding of the villages and their inhabitants by generating more diverse and encompassing information. (b) PRA offers better opportunities for women, ethnic minorities, and the poor to express their opinions. (c) PRA engenders more and better co-operation from the villagers and their organisations in setting up projects and carrying them through. Participatory Planning: Counterbalancing Centralisation 113 Horizontal Communication As indicated previously, under current extension work each line ministry carries out its own policies through its local offices, and this structure determines wholly the civil servant's career. Unsurprisingly, therefore, cross-departmental planning hardly occurs despite a host of institutionalised mechanisms to promote coordination that involve officers from different departments in an untold number of joint meetings. Consequently, an important side-benefit of PRA would be if it could help alter this situation. To test this hypothesis, the research posed two questions. First, it asked the district officers if they expected their supervisors at central office to endorse more interdepartmental cooperation, expecting that since the current structure does not support such practice de facto, the answer would vary between departments. Surprisingly, however, 85% of the respondents expected to be supported, which begs the question why such co-operation, if not frowned upon, is currently so lax? In contrast, the second question explored the extent of actual co-operation after the PRAs. At district level, three-quarters of the officers reported an increase on this variable and in the villages 385 (going up to 605 where a CAP co-ordinator had been put in place). This confirmed PR A's contribution to co-ordination, but it was the reasons that the officers provided for this that proved of greatest interest One reason was the personal ties they forged during their stay in the field, which supports the importance of informal relations in the workplace. Long-recognised elsewhere, this orientation is quite foreign, however, to Batswana management that is highly formalistic. The second reason, and structurally more significant, is that the C APs usually incorporated some items that demanded interdepartmental co-operation. This suggests that the problem of compartmentahsation may stem more from the lack of practical agendas for working together man from the current state of central-local government relations and its subsequent exclusively intraministerial worker evaluation system. Vertical Communication Lastly, the project assessed PR A' s contribution to the relations between village and district officers and between local and central government The former were expected to change because of ties formed between the district and village officers during the PRA, while local-central relations were expected to change because of upward pressure by middle-management to implement the CAPs, in whose formulation they were personally involved. The results on both counts, however, were disappointing. Regarding the relations between district and village officers, the study found that 38% of the former saw improvements six months after the PRAs, but only 18% of the latter. This was far less than the almost 90% of officers at both levels who 114 Anton A Bar-On & Gerard Prinsen had predicted increased co-operation immediately after the PR As. It would appear, however, that the lower of these perceptions was closer to the truth as less than a fifth of the village officers reported that they were told by their supervisors what their part in implementing the CAPs was to be. Instead, they had to rely for this information on the village organisations, which, in a vertically organised bureaucracy, as is well known, carry little authority. As to the districts' relations with central government, there was a general consensus among the district officers after the PR As that certain priorities in the current five year District Development Plans (DDPs) must be changed and others either added or detracted. In part, this was because some of the priorities identified by the villagers differed significantly from those in the DDPs, and so failed to live up to their sense of justice (that is, that felt needs, popularly agreed upon, must be met). More pragmatically, however, the officers believed that the projects and activities identified in the PRAs would be easier to cany out because of the villagers' active involvement. Yet they also agreed that their and the villagers' hands in instigating the necessary changes were currently tied because of the prevailing fixed budgetary system under which they operate, and so significant Progress could only be achieved in the next DDP, five years hence. Development Planning at a Crossroads Throughout the world, it now seems self-evident that the role of the state as provider of a wide range of public services is ending. Be the reason economic, due to burdening welfare expenditures, or the New Right's reassertion of fundamental (Western) bourgeois values of individual responsibility, both Left and Right arc potting their faith in their search for progress in local, small-scale user-controlled projects and agencies instead of in the "big state." These organisations, close to, and. if possible, emergent fr°ffl> •** People they serve, it is argued, are able better to relate to the diverse needs and identities of specific populations with their particular configurations ofc\ass> culture, gender, ethnicity, and other characteristics (Leonard, 1997). To achieve this transfbrmati011' participatory community planning, management, and monitoring is currency in vogue. The classic model of participation depicts a continuum with "control" at one end and "information" at the other (Arnstein, 1969). "Control" denotes a situation where all stakeholders have the right to be involved in decis*0"8 t h a t e f f e c t t h e i r l i v e s ' a n d ""P^es command of economic, political, and adi"in i s t r a t i v e resources. At the other extreme, "information" denotes telling people of policies after their content and scope have been Signed, usually, as in Bots*3113'8 P r c v a i l m£ planning structures, to the satisfac- Participatory Planning: Counterbalancing Centralisation 115 tion of technocrats (Somolekae, 1993). The community management approach argues in the name of effectiveness, efficiency, and social justice that people should be empowered to get closer to "control," while professionals and technocrats should relinquish their predisposition to engage in "information" in favour of facilitating this empowerment of the community. But this micro-political solution poses its own problems. First, it confronts a deeply-rooted sense of dependency on government by many people that resists change. This resistance is predictable, for as long as (poor) people are dependent on government for many of their basic needs, they are likely better to appreciate the civil servant who can "deliver the goods" than one who facilitates a process in which the direct material benefits are unclear. A second problem is most civil servants' perception of their role as implementors of preconceived government programmes, not facilitators, which the project of modernisation reinforces by legitimising the power of professionals and administrators formally trained in scientific reason while denying such power to people who have other kinds of knowledge. The findings reported above from nine PRA applications in Botswana - albeit only one of many participatory methodologies - show that PRA can tackle some of these obstacles by resocialising communities and the persons who serve them. Regarding the former, it appears that when given the opportunity and helped to reflect on their situation in own idioms, and decide with service providers how they can improve this situation realistically, people of all backgrounds are willing, and perhaps even eager, to do so. The significant number of participants who gave substantial amounts of time to an exercise that took 10 days (especially impressive for women who have more calls on their time man men), and their overwhelming support for its process when compared with the prevalent planning mechanisms, are ample demonstrations of this. Moreover, the only clearly measurable outcome of the PRAs, namely the items contained in the CAPs that related not to attitudes but to concrete plans, also differed sharply from the often criticised "shopping lists of requests" that put the onus of responsibility for change on external agents alone. Thus more than half the CAP items contained a direct resource investment by villagers (in cash or in-kind) to which they committed themselves, either as the only resources required to implement particular items or in combination with external resources. Similar support for PRA was found among most of the civil servants who are charged with development planning. In particular, they reported that it offered them an opportunity to be better acquainted with the communities and connect to their resources, which justified their investment in PRA. Translating this manifest support into practice, however, proved more problematic. On the positive side, there appeared, following the PRAs, increased interaction with individual mem- 116 Arnon A Bar-On & Gerard Prinsen bers of the community and their organisations, and some increase in interdepartmental cooperation. Far less change occurred, however, in the relations between the district officers and their subordinates in the villages who are responsible for the "hands-on" achievement of the CAPs. This suggests that intra-departmental relations are more resistant to change than their officers' relations with the public, and especially in Botswana where public consultation is integral to the mind-set of extension officers (Byram, et al, 1995:24) but does not extend to the realm of supervisors and supervisees. Another explanation might be that because of the generally low level of expertise of their subordinates, district officers are reluctant to delegate them responsibilities beyond their basic duties. The research found, however, that establishing a formal position to follow through the CAP can overcome most of these difficulties. With substantial differences, the villages that had a CAP co-ordinator not only did better in carrying out their plans, but the officers involved had better work relations with both the villagers and among themselves, both horizontally and vertically. Now a donor agency pays for this apparently crucial position in the villages that have it, and it is doubtful if government can institute it nation-wide due to financial constraints. A feasible alternative might be to empower the VDCs to carry out this task, or give it to the community development section of the local Social Welfare Departments that cover the whole country. Conclusions In 1997, the Government of Botswana formally adopted the community-based approach as its chief strategy for rural development, and PR A as the primary tool for its development planning. By doing so, if mainly for pragmatic rather than ideological reasons, the government not only reaffirmed its commitment to bottom-up planning, but provided it with a specific tool expected to generate additional advantages beyond pure planning, such as better intra-governniental coordination. The question that now has to be asked is whether this policy will succeed, or like its predecessor is doomed to fail, especially as the development issues facing the country today are more complex than at independence? The findings reported in this article suggest that at least the middle-management civil servants who are charged with development planning should support it. The process of adopting the policy, that true to its underlining philosophy involved every local authority in the country, also confirmed this. Further, a second constraint - the generally accepted dependency of the rural population on government (Ngwenya, 1992), while not without foundation, appears exaggerated as witnessed by the communities' will- Participatory Planning: Counterbalancing Centralisation 117 ingness to contribute actively to their welfare. (4) Yet these indispensable preconditions for bottom-up planning notwithstanding, several issues remain open. Chief among these issues is that participatory work takes considerable time and energy, both to undertake in itself and to empower the potential participants to follow its output through. It needs to be seen, therefore, if government, with its increased emphasis on productivity, can withstand the political pressure to do "for" people rather than "with" them, which is usually faster. Another challenge is to modify PRA for use in large villages and towns (where 53% of the population resides) while preserving its strengths in mobilising different sub-communities and helping them to articulate their concerns assertively. Overshadowing all else, however, is Botswana's fixed budget policy. As practised today, any change to the six years National Development Plan, which embodies this policy, requires Cabinet and parliamentary approval, no matter its size or significance. Under these circumstances, any meaningful bottom-up planningwill be smothered as such planning.by its very nature, demands a measure of flexibility. A possible solution to this dilemma is to create within the National Development Plan a new vote, or interpret creatively an existing one, to allow funding of small projects that emerge from a PRA. Without such measures, introducing community-based strategies at best will create marginal change, using the currently meagre resources of the communities. At worst, the absence of flexibility will generate further frustration, only now not only at the grassroots but also among the middle-management civil servants who are charged with empowering communities. Notes (1) As distinguished from most developing countries, the rural-urban divide in developments such as these is not that large. One reason for this is that until recently, bureaucrats and politicians invested their salaries in cattle, thus creating a strong pastoral lobby. Botswana gave rise to, therefore, a more rural-linked ruling class, and perhaps more of a national rather than urban bourgeoisie, than elsewhere in Africa (Parsons et al, 1995). (2) A classical example of this phenomenon is currently experienced by South AfricawrKwecorrmiitment,afteracWevingmajorityrule,tocoraultingaUstakeholciers fully is critically slowing its development efforts. To overcome this stalemate, politicians and government officials are increasingly deciding matters for themselves, in a process Good (1997) aptly labels "Accountable to Themselves." (3) A second factor, though more pronounced in Zambia and Zimbabwe, has been the pressure of international donors to democratise the political process and its' institutions. Botswana, having Africa's longest standing democratic regime, and being financially independent, is given far less to such pressure. 118 Arnon A Bar-On & Gerard Prinsen (4) It should be noted that in another PRA application, not covered in this article, the residents of a village - who were resettled by government with the promise that they would be looked after - were unable to understand why they should contribute to their development, arguing that it was akin to having children and telling them to fend for themselves (BOC, 1995). References Arnstein, S R (1969) "A Ladder of Citizen Participation," in Journal of the American Institute of Planners, 35(4), pp224-232. Botswana Orientation Centre (BOC) (1995) Khekhenye: Changing Livelihoods, BOC, Gaborone. Byram, M; Molokomme, I & Kidd, R (1995) "Local Consultation Process," in B A Chilume (ed) District/Urban Development Planning -The Way Forward, DraftReportof theProceedirigsandEvaluationofthe 1995Planners' Seminar, Ministry of Local Government, Lands and Housing, Gaborone. Chambers, R (1994a) "The Origins and Practice of Participatory Rural Appraisal," in World Development, 22(7):953-969. Chambers, R. (1994b) "Participatory Rural Appraisal (PRA): Analysis of Experience," in World Development, 22(9): 1253-1268. Chambers, R (1994c) "Participatory Rural Appraisal: Challenges, Potentials and Paradigm," in World Development, 22(10): 1437-1454. Duncan, T; Jefferis, K & Molutsi, P (1994) Social Development in Botswana: A Retrospective Analysis, Government of Botswana and UNICEF, Gaborone. Good, K A (1997) "Accountable to Themselves: Predominance in Southern Africa," Paper presented at the Department of Political and Administrative Sciences Seminar Series, University of Botswana, October. Government of Botswana (1968) National Development Plan (NDP1): 1968- 1973, Government Printer, Gaborone. Harvey, C & Lewis, S R (1990) Policy Choice and Development Performance in Botswana, Macmillan, London. Hutton, M (1994) "Reshaping the Social Work Curriculum: An African experience," Paper read at the International Association of Schools of Social Work, Amsterdam (July). Leonard, P (1997) Postmodern Welfare: Reconstructing an Emancipatory Project, Sage, London. Mfundisi, A (1995) "Central-Local Government Relations in Botswana: An appraisal of an important controversy?" Paper presented at the Department of Political and Administrative Studies Seminar Series, University of Botswana, 22 February. Participatory Planning: Counterbalancing Centralisation 119 Ministry of Finance and Development Planning (MFDP) (1997) Community Based Strategy for Rural Development, Ministry of Finance and Development Planning, Gaborone. Molale, E (1995) "Central-Local Government Coordination in Development Planning"in B A Chilume (ed) District/Urban Development Planning - The Way Forward, Draft Report of the Proceedings and Evaluation of the 1995 Planners' Seminar, Ministry of Local Government, Lands and Housing, Gaborone. Mutahaba, G (1989) Reforming Public Administration for Development: Experiences from Eastern Africa, Kumarian Press, Hartford, Connecticut Ngwenya, B (1992) "Botswana," in N Hall (ed) The Structural Implications of Structural Adjustment Programmes in Africa, School of Social Work, Harare and International Association of Schools of Social Work, Vienna, Harare. Noppen, D (1982) Consultation and Non-commitment: Planning with the People of Botswana, African Studies Centre, Leiden. Nteta, D; Hermans, J & Jeskova, P (eds) (1997) Poverty and Plenty: The Botswana Experience, The Botswana Society, Gaborone. Parsons, N; Henderson, W & Tlou, T (1995) Seretse Khama: 1921-1980, Macmillan, Braamfontein. Prinsen, G, et al (1996) PR A: Contract and Commitment for Village Development, Republic of Botswana and University of Botswana, Department of Adult Education and Department of Social Work, Gaborone. Rural Sociology Unit (1981) Local Institutions and Development in Botswana, Ministry of Agriculture. Siwawa-Ndai, P (1997) "Some Facts and Figures about Quality of Life in Botswana," in D Nteta; J Hermans & P Jeskova (eds) Poverty and Plenty: The Botswana Experience, The Botswana Society, Gaborone. SNV (Netherlands Development Organisation) (1995) Regional Workshop on Participatory Planning in Government Structures in Botswana, Zambia and Zimbabwe, SNV/ Botswana Publications, Gaborone. Somolekae, G (1993) "Bureaucracy and Democracy in Botswana: What type of relationship?" in S J Stedman (ed) Botswana: The Political Economy of Democratic Development. Lynne Rienner, Boulder. United Nations Development Programme (UNDP) (1995a) Human Development Report 1995, Oxford University Press, New York. United Nations, Development Programme (UNDP) (1995b) The Experience of Botswana, UNDP Series on Sustainable Human Development, UNDP, Gaborone.