Evaluation of sites and services projects : the experience from Lusaka, Zambia

Evaluation of Sites and Services Projects The Experience from Lusaka, Zambia Michael Bamberger Bishwapura Sanyal Nelson Valverde WORLD BANK STAFF WORKING PAPERS Number 548  WORLD BANK STAFF WORKING PAPERS Number 548 Evaluation of Sites and Services Projects The Experience from Lusaka, Zambia Michael Bamberger Bishwapura Sanyal Nelson Valverde The World Bank Washington, D.C., U.S.A.  Copyright i 1982 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America This is a working document published informally by The World Bank. To present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and The World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The views and interpretations in this document are those of the author(s) and should not be attributed to The World Bank, to its affiliated organizations, or to any individual acting on their behalf. Any maps used have been prepared solely for the convenience of the readers; the denominations used and the boundaries shown do not imply, on the part of The World Bank and its affiliates, any judgment on the legal status of any territory or any endorsement or acceptance of such boundaries. The full range of The World Bank publications is described in the Catalog of World Bank Publications; the continuing research program of the Bank is outlined in World Bank Research Program: Abstracts of Current Studies. Both booklets are updated annually; the most recent edition of each is available without charge from the Publications Distribution Unit of the Bank in Washington or from the European Office of the Bank, 66, avenue d'Iena, 75116 Paris, France. Library of Congress Cataloging in Publication Data Bamberger, Michael. Evaluation of sites and services projects. (World Bank staff working papers ; no. 548) Bibliography: p. 1. Housing policy--Zambia--Lusaka. 2. Public housing --Zambia--Lusaka. 3. Housing subsidies--Zambia-- Lusaka. 4. Lusaka (Zambia)--Social policy. I. Sanyal, Bishwapura, 1951- . II. Valverde, Nelson, 1954- . III. Title. IV. Series. HD7374.6.L85B35 1982 363.5'8 82-20010 ISBN 0-8213-0115-2 Michael Bamberger is a sociologist with the Urban Development Department of The World Bank; Bishwapura Sanyal is a research associate at the Institute of African Studies at the University of California, Los Angeles and a consultant to the Bank; Nelson Valverde, also a Bank consultant, is with the Management Analysis Center, Wharton School of Finance and Commerce, University of Pennsylvania.  ABSTRACT This report presents the findings of a five-year evaluation of the First Lusaka Upgrading and Sites and Services Project. The evaluation formed part of a four-country cooperative research project supported jointly by the World Bank and the International Development Research Centre (IDRC) of Canada. The evaluation, which was conducted by a specially created Evaluation Team, within the Project Implementing Agency, included both short-term operationally oriented studies and more comprehensive socio-economic impact studies. The report contains four main parts and an Executive Summary. Part 1 discusses the Project and the political and economic context within which it evolved. The remaining parts present the main evaluation findings with respect to design and organization, efficiency of project implementation and project impact.  ACKNOWLEDGEMENTS This report is based on research conducted by the Lusaka Housing Project Unit Evaluation Team, and particularly by Mathias Banda and Richard Singini. Many other people and institutions in Zambia helped with the preparation of the report. Particular thanks are due to Pomulo Muyangwa, Director of the Housing Project Unit and to the mayors and staff of the Lusaka Urban District Council. Jack Bahal and Terry Higham of the Provincial and Local Government Administration of the Office of the Prime Minister, Dirk Jol of the National Housing Authority, the staff of the Institute of African Studies and the members of the Evaluation Steering Committee, were all most helpful. Extensive comments and suggestions were given by Richard Martin, former Deputy Director of the Housing Project Unit and by John English of the World Bank. Callisto Madavo and the staff of the (then) Africa Division of Urban Projects were most helpful throughout the research. Within the Urban and Regional Economics Division constant help and guidance was received from Douglas Keare, the Division Chief; and the help of three succesive research assistants -- Judy Bobb, Susan Pabst and Robert Lynn -- is gratefully acknowledged. We would also like to thank Mary Ann Heraud and Gemma Abarcar for their tireless efforts in the typing and revision of the many drafts. Finally the authors wish to acknowledge the support and guidance received from the International Development Research Centre at all stages of the research. This report was edited by Tonia Aminoff.  TABLE OF CONTENTS Page No. EXECUTIVE SUMMARY ......................................... 1 Part 1: THE PROJECT AND ITS CONTEXT Chapter One: The Project and Its Context .... ......... 37 Map of Lusaka Indicating the Location of Project Sites Part 2: DESIGN AND ORGANIZATION Chapter Two: Physical Design ......................... 46 Chapter Three: Organization ............................ 75 Part 3: EFFICIENCY OF PROJECT OPERATIONS Chapter Four: Achievement of Project Objectives ....... 89 Chapter Five: Land Acquisition ......................... 93 Chapter Six: Selection of Participants ................ 93 Chapter Seven: Community Involvement in Planning and Implementation ................... 102 Chapter Eight: House Construction and Consolidation .... 111 Chapter Nine: Project Finance .......................... 127 Part 4: PROJECT IMPACT Chapter Ten: Impact on Participants .... .............. 147 Chapter Eleven: Impact on the City of Lusaka .... ........ 159 Chapter Twelve: Impact on National Housing Policies ...... 168 References Cited in the Text .179 Annexes 1. Household Heads in the Lusaka Sites and Services Housing Areas: Who They Are, What They Do and How They Live (prepared by Susan Papst). ......................................... 182 2. List of Reports Prepared by the Lusaka Housing Project Evaluation Team. ............... 198  EXECUTIVE SUMMARY This Executive Summary presents the main findings in the evaluation of the First Lusaka Upgrading and Sites and Services Project, and suggests lessons and issues which may be helpful in the planning of future shelter projects in East Africa. We begin with an outline of the political and economic context within which the Lusaka Project evolved, and then present a set of indicators used in the Project's evaluation. The final section discusses the possible contributions of this project and evaluation experience to future development practice. The Project Context Prior to achieving Independence in 1964, Zambia had been a junior partner in British East Africa. Lusaka had been designed as a spacious "garden city" for Europeans and Asians, with Africans being permitted to live there only while they were employed. African workers lived either on the property of their employer or in "compounds" and in most cases could not by accompanied by their families. Once employment was terminated, a worker was required to leave the city and return to his village. The housing policy called for demolition of unauthorized structures; although "illegal" communities did develop outside the city limits, no provision was made for them by local authorities. The years following Independence saw a rapid growth in Lusaka's population. This growth was due to a boom caused by high copper prices, economic diversification, administrative expansion, and a relaxation of controls on internal migration and new settlements.  -2- Migration to the city was fueled by families coming to join spouses already established there, and by a lack of employment opportunities in the rural areas. Inevitibably, most of the new migrants moved to the squatter compounds growing up around the periphery of the city. The newly independent public administration soon found itself responsible for two separate "cities." One was the former European and Asian city with its well-planned layout and high service levels. Administration of this sector, much of which had previously been directed from London, was difficult enough because many expatriate administrators left at the inception of Independence, but at least plans and procedures had been established. The other "city" consisted of the rapidly swelling squatter areas which, after the expansion of Lusaka from 67 km2 to 360 km2 in 1970, represented about 40 percent of the total urban population. No clear guidelines existed for the development of these squatter areas. After initial attempts to continue the colonial policy of demolition, it was accepted that eradication of the settlements was neither feasible nor consistent with the official philosophy of Humanism. The situation was complicated by the fact that the quasi-legal status of the squatter areas meant they produced almost no revenue for the city to help cover the costs of their development. The incorporation of these areas contributed substantially to the decline in per capita revenue of the City of Lusaka from 3.57 Kwacha in 1965 to 1.38 Kwacha in 1972. The policy options were constrained further by the abrupt end of the period of economic growth enjoyed in the 1960s. The combined  Flooding was a constant hazard in many squatter areas prior to the start of the Project.  - 4 - effects of falling copper prices and the trade disruptions and military expenditures caused by conflicts in Zimbabwe, Zaire and Angola were so serious and prolonged that the per capita GDP of Zambia was lower in 1979 than it had been in 1965 or 1972. The deteriorating economic situation made it imperative to seek less costly solutions to the rapidly worsening shelter problems. In Lusaka, as in many other parts of the world, the squatter settlements were strongholds of major political parties that became the ma^in protectors and advocates of the many thousands of people whose right to live in the city was still only partially recognized by the public authorities. Both the African National Congress and the United National Independence Party developed very strong grass roots organiza- tions and became very effective advocates of the rights of the "New Majority." The right to a house and basic services was proclaimed as one of the rewards of Independence, and stamping out housing inequali- ties and geographical segregation were seen as part of the battle against the vestiges of colonialism. After Zambia became a one-party democracy, UNIP established a very effective organizational structure, based on 25-family sections, in almost all the squatter areas. A key theme throughout the development of the Lusaka Project has been interaction between UNIP, the Housing Project Unit, and the Lusaka City Council (now the Lusaka Urban District Council). While the objective of the LCC was to integrate the project areas into the normal administrative and financial systems of the city, UNIP saw its role as ensuring that the underprivileged squatters received an equitable share  of the rewards of Independence. The difficulties of establishing an effective cost recovery system clearly illustrate the tension between these two points of view. The equity issue was complicated further by the policy of providing housing subsidies for employees of government, local authorities and parastatal organizations. The subsidies, which originated as a colonial policy for attracting expatriate staff, continue to be highly regressive and to benefit mainly the middle and upper income groups. It is estimated that in 1978 about 44 percent of all housing subsidies to public employees were received by the 10 percent with the highest incomes, while the poorest 30 percent received only 1.8 percent of total subsidies. In total, the poorest 50 percent of workers received only 10 percent of the subsidies. 1/ It was also found that while housing subsidies represented less than 15 percent of the total employment benefit package for the poorest 30 percent of government employees, they represented over 40 percent of the benefits for the 10 percent with highest incomes. In addition to being a considerable drain on government resources, the long tradition of housing subsidies seemed to render questionable the social justice of the complete cost recovery requirement built into the World Bank-s Project designed to serve the poorest urban groups. It was estimated, for example, that many of the highest-income families were paying only about 10 percent of the economic rent for their housing. 1/ Hinh, Sanyal, Valverde and Bamberger, "Housing Subsidies in the Public Sector in Zambia," Urban and Regional Report No. 81-6, Urban and Regional Economics Division, World Bank, May, 1981.  - 6 - The factors detailed above determined the political climate within which the Lusaka Project was planned and developed. The newly independent Zambia wished to remove the vestiges of colonialism and to give Africans the same benefits previously enjoyed by expatriates. The philosophy of Humanism reinforced this sentiment by making social and econonomic equality an important political objective. These ideals, together with urban planning traditions, created pressures to provide good, high-standard housing for everyone. Further, it was felt that housing is a right, and that standards should be determined by needs as much as by ability to pay. It was assumed that the pursuit of this policy implied subsidies for the economically weaker groups. The substantial housing subsidies provided to most public employees strengthened the case for new categorical shelter relief. The combined effects of the economic depression spanning most of the 1970s and the rapid increase in demand for public services put an increasing pressure on the resources available to the central and local government authorities. These pressures lead to the more pragmatic conclusion that lower standards must be accepted and that innovative approaches, such as squatter upgrading and sites and services, must become the basis of national housing policy. The potential contra- dictions between political objectives and the economic realities have continued to play an important part in the Project-s development. The Administrative Challenge To get an objective view of the Project-s achievements, it is necessary to appreciate the enormity of the administrative tasks it  -7- entailed. On top of the pressures created by Independence, departure of expatriate staff, and assumption of functions previously administered from London, the Lusaka City Council and central government agencies were required to develop completely new administrative systems to handle the 40 percent of the population living in the formerly "illegal" squatter areas. Besides the pressures these events placed on an understaffed bureaucracy, the squatter areas also gave rise to strong political influences,due to the activities of the two major parties (and of UNIP after Zambia became a one-party democracy). The guiding policy of the local party organization was to obtain as many services as possible for the squatter areas from a recalcitrant public administration. Although the party structure rendered considerable assistance in mobilizing popular participation in planning and implementation, it created some opposition to the integration of the areas into the normal administra- tive structure of the City Council. Attempts to integrate the areas more closely into the LCC nexus through Ward Development Committees had to be abandoned due to the much more effective organization already existing in the UNIP local party structure. Subsequently, the local party organization has often opposed use of the stronger sanctions, such as cutting off water or evictions, to improve cost recovery. Most public housing and service programs have very poor cost recovery records. In 1979, it was estimated that the seven largest local authorities in Zambia had a total deficit of 28 million Kwacha on  their housing accounts. 1/ The reasons for the deficit are a combina- tion of poor administration and opposition to the use of sactions to enforce payment. The fact that the Central Government is often one of the worst defaulters complicates the situation further. Poor cost recovery performance has created the widespread belief that high defaults are inevitable for housing programs and consequently make it more difficult to encourage the use of sanctions in the present Project. Despite the poor cost recovery performance of the Project, its record is much better than that of many other LUDC housing programs. The matter of land acquisition illustrates well the adminis- trative difficulties that confronted the Project. One of the major political and economic reforms of the new regime was the elimination of freehold land and the limitation of land value to its unexhausted productive capacity. Although it was assumed that the increased regulatory power of the government would greatly facilitate the process of land acquisition, most of the new legislation was still untested at the time the Project began. In fact, land acquisition proved to be the most difficult obstacle to project implementation, causing delays of more than two years in some phases of the process. In retrospect it is possible to see how the process could have been accelerated: delays were due in large part to the newness of the legislation and the HPU's inexperience in applying it. 2/ 1/ World Bank Data. 2/ These factors should not make us overlook the effect of the bureau- cratic delays which prolonged the process still further (see Chapter Five).  - 9 Criteria for Project Evaluation Tables 1 through 5 summarize the main findings of the report with respect to Project organization and design, physical implementa- tion, efficiency, and impact on participants, the city, and housing policies. Each of these areas is discussed in detail in the main body of the report. Table 1 shows that in general the Project was well designed. Sites were well selected and house designs were accepted by most participants. With respect to infrastructure, water and sewerage were satisfactory, but street lighting and garbage collection were inadequate. There was some dissatisfaction over issues related to the extension of the road networks and the design of community facilities. Table 2 shows that the internal organization of the HPU was good but that problems arose in coordination with the City Council and other agencies. Table 3 shows that almost all of the targeted upgrading and sites and services plots were serviced, but that a considerable proportion of the serviced sites had not been developed. Similar shortfalls can be observed with respect to the provision of community facilities. Table 4 presents indicators of the efficiency of Project implementation. The most efficiently executed components of the Project included selection of participants for sites and services, community participation in Project planning, the process of house construction, and the provisions ensuring Project affordability. At the other  -10- Progressive developmint is a flexible strategy that permits the construction of houses of a wide range of costs and qualities. (Photo: Valverde)  - 11 - extreme, significant problems arose with land acquisition and cost recovery. Experience was mixed with mutual help construction, the use of participants own labor in construction, and the operation of material stores, the materials loan program, and Project maintenance. Finally, Table 5 presents indicators of Project impact. Among participating households, there has been some impact on income through rental use of portions of housing. In the short run there was also some additional wage income during the process of house construction. The Project has generally improved the quality of housing and access to water and gardens in which to grow food. There is no evidence that it has stimulated rural to urban migration or lead to overcrowding. At the city level, the Project has yielded a dramatic increase in housing stock. There has also been an important political impact from the integration of the urban poor into the formal political system. Impacts on national housing policy are more difficult to evaluate, but considerable reductions in standards have been accepted, and both sites and services and upgrading now form a central part of national housing plans. The Project has also created an impulse toward the rationalization of land policies and has raised the equity issue related to current housing subsidies. Lessons and Issues for Future Projects One of the most useful functions of the evaluation has been to draw conclusions which may be helpful to planners in other countries considering urban shelter options. The issues are presented here in the order in which they will normally occur in the design and planning process.  - 12 - Table 1: EVALUATION PROJECT DESIGN (CHAPTER 2) COMPONENT EVALUATION COMMENTS Selection of Lusaka for Good Demonstration effect. Simplifies coordination. Only disadvantage contradicts invest- first project ment decentralization policy. Selection of Sites in Lusaka Good Sites close to employment and services. Unexpected land acquisition delays in George. Problems with higher water table in Site 6. Plot size Good Demonstrates smaller plots widely accepted. Floor plans Good Wide acceptance Building materials Mixed Essential to encourage house upgrading as no other sources of finance. May have encouraged use of more expensive materials such as cement blocks as cheaper options not always sold. Insufficient funds to cover demand. Infrastruture Water supply Good General satisfaction. Some maintenance problems partly due to cutting off water as part of cost recovery strategy. Roads Mixed Road network greatly improved access to the city. Considerable maintenance problems. Lack of interior roads contributed to garbage collection problems. Street lights Poor Agreed level of lighting decreased. Considerable dissatisfaction with lighting level. Sewerage Good General satisfaction with pit latrines. Complaints of insanitary conditions when shared. Garbage collection Poor Dissatisfaction. Garbage often not collected from houses which do not have access to roads where trucks enter. Lack of initial briefing to persuade households to carry bins to main roads. Community Facilities Markets Poor Markets develop spontaneously and there is probably no justification for including them in the project. Materials stores Mixed Material stores an essential comiponent of the project as difficult to obtain materials on the open market. High incidence of robberies suggest stores should be located closer to other facilities rather than isolated. Community centers Mixed May have been better to let community plan, build and control centers. Many under- utililized. Multi-purpose center may not be consistent with cultural separation of the sexes.  - 13 - IL~~~~~~~~~~~~ Meetings with participants took place at all stages of the project. (Photo: Courtesy Housing Project Unit)  - 14 - It had been assumed that most families would build for themselves, but an unexpectedly high proportion hired labor for at least part of the construction. (Photo: Valverde)  - 15 - Table 2: EVALUATION OF PROJECT ORGANIZATION (CHAPTER 3) COMPONENT EVALUATION COMMIENTS Establishing HPU as Good Efficient pioject implementation by dynamic management team avoided special unit many of the bureuacratic constraints facing the city council. Able to recruit higher level staff. Separation may have contributed to cost recovery difficulties and long term maintenance of services. Interaction with Mixed HPU treated as outside agency. Cost recovery example of coordination City Council problems. Interagency Mixed Inter-agency coordinating committee not effective. Land acquisition Coordination serious example of coordination breakdown. HPU-Community Good Field team approach worked well. Potentially disruptive project relations accepted with minimal protest by communities. Satisfactory handling of political organization of the communities. Road planning groups worked very well. Cost recovery major example of coordination problems. Internal HPU Good Management team approach worked well. Many management innovations organization developed.  Most families preferred to purchase ready made building blocks rather than make their own. (Photo: Valverde)  - 17 - Table 3: COMPARISON OF PHYSICAL OBJECTIVES AND ACHIEVEMENTS AS OF JUNE 1981 Appraisal Achievement Achievement as Target (as of 6/81) of Target A. Physical Implementation 1. Residential units serviced (dwellings or plots) 1/ a. Squatter Upgrading 16924 19916 118 b. Overspill areas 7588 7775 102 c. Phased normal (basic) sites 1197 974 81 d. Normal low-cost sites 1204 1188 99 e. Normal medium cost sites 1938 1503 77 TOTAL 28851 31355 109 2. Community Facilicites (Number completed) a. Schools 20 8 40 b. Hiealth centers 3 3 100 c. Markets 17 6 35 d. Community Centers 17 7 41 B. Socio-Economic Improvement 1. Residential IJUnits Allocated (dwellings or plots) a. Squatter upgrading 16924 19916 118 b. Overspill Areas 7588 5550 73 c. Phased normal (basic sites) 1197 888 74 d. Normal low-cost sites 1204 1188 99 e. Normal medium-cost sites 1938 587 30 2. Allocation of Site and Services Plot to Target Group (No. of plots developed vs. % of total developed plots) a. % or plots allocated to households earning less than K70 per month. 50 20% b. No. of plots allocated to households earning less than K300 per month n.a. 2240 c. No. of plots allocated to households earning more than K300 per month n.a. 174 3. Core Units Erected (No. of Units) a. Overspill area 7588 4277 56 b. Sites and services area 4339 1823 42 4. Building Materials Loans X5,609,000 4,474,000 79 Total value of materials issued Source: World Bank Data  - 18 - Table 4; EVALUATING THE EFFICIENCY OF PROJECT IM-PLEMENTATION (Chapters 5-6) COMPONENT EVALUATION COMMENTS Land acquisition Poor Unanticipated delays of 2-3 years slowed project and produced cost escalation. HPUJ slow to follow-up on administrative steps to speed acquisition process. Selection of Good System fair and accessible to low income groups. After initial delays participants for efficiency improved considerably. sites and services Community participation Good Road planning groups worked well. lIPU field teams worked efficiently. in planning Mutual help Mixed Mutual help used much less than planned in construction. Often inefficient construction and difficult to organize. However, worked very well in several cases where groups continued to work on other projects. Use of own labor for Mixed Used less than expected as many families preferred to hire labor for good house construction economic reasons. Own labor was uiseful option for some poorer families. House construction/ Good Most families were able to build good quality house in short period of time. consolidation Material loans and Mixed Loans essential part of project as few alternative sources of housing materials stores finance. Problems and delays in administration of material stores. Some later improvements. Cost recovery Poor Migh default rates due to a combination of administrative problems with lack of political will to collect. Project Affordability Good Upgrading affordable to virtually all families. Sites and services offered wide range of options so affordable to wide range of households. Maintenance Mixed Problems developing due to lack of financing for maintenance. Potential problems with deterioration of roads, garbage collection and blocking of drains.  - 19 - Table 5; INDICATORS OF PROJECT IMPACT TYPE OF IMPACT INDICATORS Impact on Project Participants Rental income At least 14 percent of upgrading households receive rental income with higher proportions for poorer households. Other income 38 percent in Sites and Services believe project had positive impact on their income. House construction generated 8,000 person/months of work and over $1 million in wages paid by participants to hired labor. Housing costs 38 percent in Lilanda (Sites and Services) say housing costs have fallen. Housing quality 75 percent believe housing quality improved. Access to water 69 percent believe access to water improved. Access to gardens 63 percent believe better access to gardens. Population density Not affected by project Rural to urban No evidence of project inducing migration. migration Impact on City Housing stock Increased by 11,500 units with 20,000 upgraded. Value of housing stock increased by at least 15 million Kwacha. City finances Considerable potential increase in tax base. Project currently produces financial burden due to poor cost recovery and demand for higher level of services not covered by project financing. Political impact Integration of urban poor into political system. Social impact Social integration due to wide income spectrum in projects and to location of urban poor in areas close to the city. Transport Substantial improvements in transport systems partly through stimulation of private transport companies. Impact on Housing Policy Housing standards Acceptance of lower standards. Ingration of sites and services and upgrading into housing policies. Land Increased awareness of land as constraint on housing development and helped develop new land legislation. Equity Increased awareness of the inequitable system of housing subisidies.  - 20 - 1. The Role of Shelter in National Development Strategies A number of basic policy issues emerged during the development of the First Lusaka Project, especially when the question of a possible second project arose. The issues that will be common to many country situations will be highlighted here: * Is the provision of low-cost urban shelter inconsistent with the objective of stimulating rural development? That the provision of low-cost housing will encourage economically unproductive groups to stay in the city, or worse, encourage rural to urban migration, is a common development concern. Existing evidence suggests that people migrate in search of better employment opportun- ities, rather than for improved housing. It is unlikely that the provision of housing per se will induce new migration. However, as housing projects are often quite expensive, they may use scarce resources that might otherwise be invested in rural areas. The major justification for low-cost urban shelter programs is two-fold. First, even if migration from rural areas stops completely, the demand for urban housing will continue to grow due to natural population increase. Therefore, there is a strong case for providing some basic level of shelter for the existing urban population. Secondly, a well-planned project, with  - 21 - low standards and effective cost recovery, can almost pay for itself, and consequently need not be a heavy financial burden. * Should the provision of housing be treated as an employ- ment benefit? Until recently, most housing in Zambia has been provided at a subsidized rate as part of an employ- ment package. Consequently, higher-paid workers tend to receive higher housing subsidies. Although it may be justified in certain cases, this system is not to be recommended. The first argument against it is that it amounts to a bias against the self- employed. As housing tends to be provided at a subsidized and controlled price by employers, there has been no incentive for a private housing market to develop, and the self-employed, unless they happen to be wealthy, have no access to a housing market. The second argument against this system is that it has tended to provide high levels of housing support to upper and middle income groups and thus to produce regressive income distribution effects. 2. The Appropriate Organizational Structure for Project Implementation At both the planning and evaluation stages, the effect of setting up a special project execution unit with a considerable degree of autonomy was a key concern. There was a clear trade-off between increased efficiency in staff recruitment and project implementation on  - 22 - the one hand, and long-term institutional development and project replication on the other. There is no simple resolution of the issue. Given the innovative nature of many prospective projects, and the cumbersome, rigid structure of many municipal authorities, project implementation may be seriously prejudiced by operating through traditional channels. It is vital that first projects succeed if they are to demonstrate the feasibility of a new approach. The Lusaka experience suggests that the following factors should be taken into consideration if a special project implementation unit is created. -Q Special units often have more flexibility in terms of salaries. If project unit staff are particularly well paid, consideration must be given to what will happen when the unit is merged with the municipality, so as to avoid the departure of staff due to lower salary scales. * Coordinating mechanisms must be set up early on to ensure a smooth transition. * Lines of responsibility for cost recovery and project maintenance must be clearly defined as these components will be under the long-term supervision of the municipality. 3. Land Acquisition In the Lusaka Project, as in many others, administrative and political complexities bearing on the process of land acquisition have been the major cause of delays, and hence of cost escalation. The  - 23 - Lusaka experience suggests that the following issues must be considered very early in project planning: * Does the acquisition of land require any special legislation? If it does, it is essential to begin the legislative process as early as possible. * Expert advice should be sought on problems likely to arise in getting legislation approved and implemented. One common problem is that nationalization tends to lower the value of land and thus make owners less willing to sell. Another source of problems is the likelihood of political opposition from groups such as squatters, cooperatives, or small farmers who presently occupy the land. e As land acquisition is likely to be the key factor in determining the speed of project implementation, it may be useful to set up a high-level committee authorized to take expediting action. The processes of land surveying and title search can be time- consuming and require a considerable number of professional staff. In the case of Lusaka, these professionals were not available, adding to delays. A check should be made to ensure the required professionals will be available, and provision made for the training of staff or the use of expatriates or consultants, if the need arises.  - 24 - 4. Issues Related to Project Finance Several issues arose with respect to the system of Project finance. Cost recovery proved to have an important impact on Project implementation and long-term maintenance, as well as replicability. The following are some of the issues likely to be faced in other countries as well as Zambia: * Unless the issue of existing housing subsidies is resolved very early, it is likely to give rise to political opposition to cost recovery measures. * Cost recovery problems in the Lusaka Project were due partly to administrative difficulties and partly to a lack of political will in pursuit of defaulters. Both of these liabilities are likely to be encountered in other projects. There are several ways of coping with them. These include: actively involving the community in the cost recovery process, beginning at the planning stage; gaining the support of the local political organizations which can persuade project residents to cooperate in making payments, or taking administrative steps to compensate for the lack of such support; adequate information provided to the community on the amounts and timing of payments and the reasons for them; efficient administration of payment collection, including prompt, accurate information to those who owe, and proper payment facilities, such as convenient office hours.  - 25 - * Materials loans have been an important facet of the Lusaka Project and others, as most households have very limited access to other sources of finance. Appropriate loan estimation is a key factor here. Families must be provided sufficient resources to complete a livable dwelling unit without assuming financial burdens they cannot handle. Another issue is whether loans should be given in the form of materials vouchers redeemable only in project-related stores or in the form of funds they can spend at their discretion. In the case of Zambia, tied loans and material stores were appropriate because the unavailability of materials would have made it very difficult for families to procure needed supplies on the open market. On the other hand, this system apparently forced families to buy relatively more expensive materials than they might have wished, because these were what the project stores made available. In general, where a relatively free building materials market exists, it may be better not to tie purchases to a particular outlet. A further question is whether vouchers are preferable to cash disbursements. While there is a general suspicion that cash may be used for purposes other than house construction, it is also true that cash allows families the option of hiring labor to help in the construction process. In Lusaka, an unexpectedly high proportion of households did hire labor,  - 26 - suggesting that consideration should be given to cash disbursements or to vouchers that could be used to hire labor. 5. Issues Related to Physical Design The first issue of physical design to provoke extensive discussion in Lusaka was that of plot size. It was apparent on the one hand that land was relatively abundant in Lusaka, so planners felt that households should be provided plots of a size appropriate both to housing construction and to the growing of their own food. On the other hand, too large a plot size would increase the cost of providing services and hasten the time when land within the city would actually become scarce. The results of the evaluation suggest that there is a considerable variation in the plot sizes families desire. Many families have not in fact used the area surrounding their house for growing food. It seems that the optimum planning option would be for a project to offer a range of plot sizes with economic pricing so that families desiring larger plot sizes would either have to pay more or be willing to live on the periphery of the city, where land is cheaper. It does not appear to be economical to provide large plots in centrally located areas merely to provide families the option of growing food only to have them decline to do so. Research has also shown that provision of large plots is not a very effective way to ensure lower population densities. Households will begin to sublet or to have additional relatives move in, so that  - 27 - densities will increase wherever low-cost housing is constructed in prime urban locations. There was also considerable debate about the appropriate level of services. Because there is a direct relationship between service level and cost, it is necessary to offer a low service level as one option in order to avoid excluding large groups on grounds of affordability. Probably the ideal design is to begin with a low level of services and to provide households with the option of upgrading at such time as they are able to afford it. It will often be found that as the demand for rental accomodation increases, many families will wish to upgrade services in order to charge higher rents, so the process of upgrading may be faster than expected. There must be an effective system of cost recovery if upgrading of services is to be self-financing, and households must not be permitted to solicit service upgrading unless they are up to date with all their service and loan payments. Another disputed design issue was the question of the Project road network. To reduce costs, the Project was originally designed with many areas not directly accessible by roads suitable for vehicles. Problems arose because garbage trucks were not able to reach all areas of the Project, whereas the original intention was that households should carry their garbage to collection points on the main roads. Similar problems have arisen in several other countries where the requirements of modern garbage collection technology have dictated construction of large and extremely expensive road networks. Unless  - 28 - there are strong reasons for an extensive road system, it would seem better to insist that households transport their refuse the relatively short distances to collection points to achieve a considerable cost saving. 6. Issues Related to Community Participation In Zambia and elsewhere, considerable importance was attached to achieving a high level of community participation in Project planning and implementation. Such an effort was to further the national goal of grassroots participation in political life and to enhance the efficiency of Project implementation. Chapter 7 of the Report describes how popular participation was effectively organized at the planning stage, where it served to promote general acceptance of what could have been a very disruptive upgrading process, but there was much less success in organizing mutual help construction. The use of community pressure was even less effective as part of the cost recovery process. The following are some of the important issues that emerged from the Zambia context but are relevant to many countries: * It is necessary to define clearly whether the community is to have an active role in decision-making as to project design and priorities or whether the objective is simply to ensure community support for decisions which have already been made. In Zambia, as elsewhere, there was a certain ambivalence on this point. If the cooperation of the community is to be sought in securing a high rate of cost  - 29 - recovery, it must be involved in the selection of services and must be given a full understanding of the financial structure of the project. * With respect to mutual help construction, it must be decided whether this is to be a training experience in community organization, a method of reducing project costs, or an opportunity for the community to earn money which can then be spent on additional resources the community wants to develop. In Zambia, there was some confusion in this area, and in some cases residents were pressured to participate in mutual help efforts without a very clear idea of objectives. Although mutual help activities can be an effective way of providing practical training in group organization, experience suggests that it should probably not be made compulsory. Instead, community groups should be given the opportunity to participate if they so desire. The financial incentives used in the Zambia Project seems to be a useful means of encouraging families to participate. * Careful thought needs to be given to the role of the community in the cost recovery process. If there is to be community involvement, it should begin early enough to allow prospective participants some power in the project design process. It will almost never be possible to involve community members simply as collection agents, for they have no incentive to pressure defaulters to pay. As the Zambia experience showed,  - 30 - if local political groups are involved in the collection process, there is always the possibility that they will disagree with the use of sanctions and may thus make it even more difficult to collect payments. The community can either be given a very active role, with its attendant risks, or its role can be kept to a minimum and cost recovery organized as a purely administrative task. * An issue that began to emerge in Zambia at the time the evaluation ended was that of project maintenance. There were some indications that the community perceived maintenance to be entirely the reponsibility of the Project administration or the city government, with very little feeling of participant responsibility for upkeep of public facilities (water taps, drainage ditches, etc.). As with cost recovery, experience suggests that the community must be involved in maintenance considerations at a very early stage, rather than after problems have already developed. It should be possible to instill a sense of pride and proprietariness on the part of the community so that its members will accept joint responsibility with the government for maintenance. It is essential to avoid creating either of two impressions: that the government will solve all problems (a false promise local politicians the world over cannot resist making), or that families in poor areas will be expected to perform for  - 31 - themselves maintenance functions that are apparently provided free to middle-income areas of the city. 7. Issues Related to the Construction Process In general the construction process was very successful, with most households demonstrating their ability to produce an adequate dwelling unit in a relatively short period of time. The following were some of the attendant observations and lessons. * A higher proportion of families than expected decided to hire labor rather than build for themselves. This can be explained partly by a much lower level of unemployment than expected among project participants, and partly by the fact that most families either lacked the skills to build their own house or preferred to invest in skilled labor in order to guarantee a better quality structure. Project design should provide for both options. Families who wish to use their own labor should be helped to do so (through plans and technical assistance), but those who prefer to hire labor should also be encouraged to do so. One provision for such encouragement might be to permit part of the construction loan to be used for hiring labor. * As noted earlier, mutual help should probably not be a required part of the construction process, although it might be offered as an option for those groups who are interested. In several countries, including Kenya and El Salvador, voluntary mutual help groups have proved an effective means of  - 32 - house construction (as opposed to the experience with infrastructure installation in Zambia). 8. Issues Related to Income-Generating Project Components Research has shown that, although the specific income- generating components of the Lusaka Project did not have significant quantitative impacts, the construction process itself generated a significant demand for labor. In addition, the Project appears to have increased rental income and income from food production, although in neither case has it been possible to quantify the effect precisely. A number of important considerations for furture projects emerged from the Zambia experience: * Although subletting and the addition of rental space were discouraged to avoid "landlordism", a high proportion of households may have sublet part or all of their house. This appears to have been particularly important as a source of income for poorer families. Many projects have met with difficulty in defining policy with regard to rentals and sales. On the one hand, rentals are a potentially significant way of increasing the income of poorer households, and thus enhancing project affordability. On the other hand, there are grounds for concern that poorer sectors will be exploited, either by having to pay excessive rent, or by being forced out of the project. In practice, it has proved impossible to regulate subletting, although some reasonably effective  - 33 - controls have been placed on sales in sites and services areas. Although the appropriate policy will depend in part on the social and political philosophy of a project's host country, it is possible to suggest some general guidelines regarding rentals and sales: * Rental income is one of the few ways in which most projects have been able to affect significantly the incomes of their target populations. As long as measures can be taken to avoid the worst abuses, subletting should be permitted. An inter- esting example is the Dandora Project in Nairobi, where poorer families are actively encouraged to sublet part of their house and part of their construction loan may be used to make this possible. Although subletting should be permitted, households should also be required to register the fact that they are renting space, and should perhaps be required to pay a slightly higher service charge to reflect the greater use of services and the greater capital gain they will enjoy. * In sites and services projects, it will generally be approp- riate to regulate sales during the first few years so as to permit families to appreciate the value and benefits of their new dwellings before they consider selling them. Selling should not necessarily be prohibited, but should be subject to special approval.  -34- * In an upgrading project, selling should probably be permitted with only minimal controls. However, some type of capital gains tax should probably also be levied so that the local government can recoup part of the value of any subsidies included in the project. 9. Issues Related to Project Evaluation The Lusaka Project included both a monitoring and an evaluation unit. The two units operated independently with relatively little coordination of activities. Both proved to be useful to project management, although there were a number of problems that arose in the design and administration of various studies (See Chapter 1). Some general conclusions about project evaluation can be drawn from the Lusaka experience: * Evaluation activities have two distinct audiences. The first is project management, which requires mainly short- and medium-range studies to help with the implementation of the project in progress. The second comprises the national planning agencies that are interested in longer-term and more general studies which can help with the definition of future urban shelter strategies. The two types of research should be clearly separated. * Monitoring and evaluation should be considered as one inte- grated system, rather than being conducted separately. When they are separated, a number of important studies may not get done because they do not fall clearly within the terms of reference of either the monitoring or evaluation groups.  - 35 - * The evaluation program should be designed initially to be very simple and economical, so that it can be developed on a small budget, without requiring large numbers of specialist researchers. A Basic Evaluation System of this kind should be closely tied to project development and planning cycles, and should yield brief policy-related reports providing specific answers to the problems managers face at different stages of project development. * Longitudinal impact studies should normally not be included in the intitial evaluation programs unless they can be conducted independently by a national agency, such as the Ministry of Planning.  / t/~~~~ ~~/ \ < J d (>stsbihzct;~~~~~~~~on ponlds / E~~~~~~~~~ RVEab \|4 1G /-1 # Maortrunk seer neorth ojc 1/ /\ l ~~~~~~~~~~~~~~~~~~~~~~~': :O eroetspill g areasere / \ / z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ao trun sewes before the project / / ) / ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Project major trunksewr /~ ~ ~ ~ ~ ~ ~~ ~ ~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~~S.g tramn facliie befe hers-ec \ -t1 J~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Prjc DVEwageL trAtmetEA fACDEEh!77' 7 lWie / t l Vf of id E's .t.e =,