Document of The World Bank FILE COpy FOR OFFICIAL USE ONLY Report No. 1034 MANAGEMENT POLICY REVIEW: DELAYS IN PROJECT PREPARATION February 18, 1976 Operations Evaluation Department This document has a restricted distribution and nay be used by recipients only in the performance of their oflicial duties. Its contents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USE ONLY PREFACE Early in 1974 the Operations Evaluation Department added a new dimension to its assignment, consisting of focusing on a believed problem area in Bank operations, examining selected recent cases where the problem seemed to manifest itself and trying to identify constructive approaches to these problems. 'Delays in project preparation' was one of two problems initially selected for treatment by this technique of so-called Management Policy Review, the other being Delays in Loan and Credit Effectiveness, as covered in the Operations Evaluation Department report by that name issued on July 22, 1975. Together the reports cover quite fully the initial period in the project cycle through commencement of disbursements, even though neither focuses specifically on the relatively much-analyzed stage from field appraisal to Board approval. This study also followed again the relatively economical method of focusing heavily on detailed studies of a small number of cases, smaller in this case than in the Loan Effectiveness study because of the complexity of the analysis needed, but complemented with reviews of much wider samples of projects from old and from recent lending. Nonetheless, the subject of project preparation is so complicated, and the amount of general work previously done on it in the Bank so small, that the conclusions are inevitably more tentative and general than in the case of the Loan Effectiveness study, and this paper must be seen in a sense as an interim report. The Operations Evaluation Department plans to continue study on this aspect of the Bank's operations, particularly in the framework of the Project Completion Report and Project Performance Audit system and focusing more on particular sectors, but it believes that this first general incursion into the subject yields findings which warrant consideration by the Bank in a wider context. The substantial contribution of many Bank and IDA borrowers to this effort, particularly in discussing the underlying case studies dealing with loans to them, answering questionnaires about their view of the subject, making numerous helpful suggestions for the Bank and commenting on the draft of this report, as well as of many Bank Executive Directors and staff members, is very warmly appreciated; many of the ideas come from these various com- mentators and it is hoped that they will in turn find the report useful. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization TABLE OF CONTENTS Page Summary i I. The Problem 1 II. The Facts 5 III. The Reasons 10 Bank Procedural Delays 10 Government Decision-making Delays 12 Delays in Getting Appropriate Studies Done 13 Assessment of Aggregate Potential Time Savings 15 The Evidence of the Questionnaire Survey 16 The Evidence of the Project Performance Audits 17 IV. The Issues 17 Defining Requirements 18 Local Involvement 21 Bank Rigidity/Flexibility 25 V. Conclusions 31 Main Elements of a Project Preparation System 35 Subjects for Further Consideration 36 Particular Points 37 Tables 1. Projects Under Preparation and Man-years Devoted, by Region, FY 1974 2 2. Number of Months Required to Bring a Project from Bank Conception Date to First Field Appraisal 7 3. The Four Selected Projects, Compared with Other Related Groups: Number of Months Required to Bring a Project from Bank Conception Date to First Field Appraisal 11 Annexes I. Work Effort 1974 II. Questionnaire Regarding 1974 Projects III. The Method for Case Studies IV. Calculation of the Increase in Net Present Value of a Project from Bringing Forward Execution by Three Months SUMMARY The time required to bring projects from initial Bank contact with them to actual effectiveness of a loan or credit in support of them varies enormously between projects and shows no identifiable regional or sectoral pattern. The median time in recent years has been about 3-1/2 years, of which the first two constitute the period before Bank appraisal, here termed preparation period. The Bank normally has no direct responsibility for a project in this phase of its life, but much of the work to be done is deter- mined by its criteria and standards for appraisal, and it has been devoting upwards of 200 man-years per annum to advising and assisting others on prep- aration of specific projects. This study attempts to illuminate a wide- spread concern that the preparation of projects for Bank/IDA financing takes too long and to identify ways in which the Bank might usefully help to speed it up. Beyond a good deal of statistical work (primarily on 1972-74 lend- ing) which was undertaken because this phase of projects' life has not been much analyzed before in the Bank, the study had three principal bases: four projects approved in 1971-73, two with preparation periods close to the Bank- wide norm for "first" projects (in a sector in a country) of 2-1/2 years and two with preparation periods toward the longest (6 and 8 years), which were analyzed in detail on a flow chart method; an opinion survey of 1974 borrowers, to which 52 full replies were received; and a special review of the broader experience to date on this topic of the Operations Evaluation Department and particularly of the 50 projects mainly approved in the 1960s which are also covered in the 'First Annual Review of Project Performance Audit Results' issued in September 1975. Somewhat over 100 projects have therefore been reviewed in varying degrees of depth, and about 500 projects have been covered in various statistical exercises. The study reaches three major negative conclusions, of which the first and most important is that the standards of projects eventually financed by the Bank/IDA do not seem generally to have been too high and that a median preparation period of two years is not unreasonable given the complexity, and importance to eventual project success, of the work to be accomplished in this period. The detailed case studies revealed limited scope for trading off elements in the quality of what was eventually done against potential time savings. The other negative conclusions concern possible causes of delay. First, it was found that Bank procedural delays -- in the sense of slow response to correspondence, dispatch of missions or selection of consultants -- were not matters of any great significance, even though they did sometimes crop up. Second, slow Government decisions in response to institutional or sector-policy propositions of the Bank were a factor of delay in relatively few cases and, where a judgment could be made, such delay appeared usually to have been worthwhile. Of the 20 respondents to the questionnaire survey who felt that the Bank had raised an issue of institutional nature, all but four felt that the effort had been useful, and a few implied that more could have been done. - ii - Nonetheless there does seem to be scope for shortening, as well as improving, the project preparation process. The individual case studies suggest that some 20-30% of preparation time could have been saved in each case without sacrifice in project quality or assumption of unreasonable fore- sight on the part of those involved. About a quarter of the questionnaire respondents felt that major time savings could have been made, and a further 30% referred to minor savings. However, the generally desirably higher standards of preparation now required by the Bank and the greater complexity of some modern projects probably make it very doubtful that a major change, like a reduction in the median from two years to one, would be possible with- out reducing standards excessively. The most that seems feasible on the basis of the work done is that the process might be shortened from the current median of 24 months to some 21, or by overall average of about 10% -- more in some cases and less in others -- by measures of the sort discussed below; but it is not possible to assert this with certainty nor to establish any generally applicable norms or standards of time required, because the work to be done depends too much on each particular local situation. The main scope for savings now, and the main problems that can be identified from the past, are in connection with studies, usually done by consultants, borrower staff or UN Agency personnel, with a view to developing the project idea, and meeting the needs of the Bank, either stated or presumed. All the case studies and sizeable minority shares of the other samples reviewed give evidence of considerable difficulty in getting appropriate studies done -- for instance, in organizing them, in controlling them, in redoing and comple- menting earlier studies, in reconciling work by different parties, in prepara- tion of supplements at a late stage or in requiring studies of little conse- quence. The result is that the project preparation process often appears more roundabout or partially regressive than it need have been. Sometimes the problem appears to have been more the responsibility of the Bank, and sometimes less, and sometimes it arose due to neglect of the Bank's advice, but the emphasis of this study is not on the past but on how the Bank could help to reduce recurrence of the problems. Since so much of the work is done partly to meet the requirements of the Bank, the Bank would seem to have a special responsibility, to which it has not perhaps always attributed sufficient priority, to be as clear as possible at an early stage as to what its requirements will be, as regards acceptable standards of design, and of study and documentation. Issues which it should have been possible to foresee have sometimes arisen late -- requiring further study and delay or preventing useful improvement of the project. The case studies make it clear that specific guidelines for studies desired would have been useful, and a number of the respondents to the questionnaire survey raise the same subject, but one of the difficulties is that such guidelines, to be most useful, need specific adaptation to each situation and the issues it presents. Much therefore depends on the ability of the visiting Bank mission to foresee the issues which will be significant and on the time that it has to go over the guidelines in detail with the borrower to make sure he understands them. - iii - The Bank can play a very valuable advisory role in helping borrowers to plan project preparation, to ensure the necessary and economical logical progression, starting with a wide range of alternative solutions to a properly defined problem and leading, by process of gradual elimination of non-optimal solutions and increasing detailing of the superior one, to an implementable project. There is evidence of excessive repetition of work in the preparation of some projects, and of a tendency, for lack of prior planning, for questions of a 'feasibility' nature to be raised at the time of detailed engineering. Occasionally project preparation has been held up by a long drawn out dispute on a relatively minor technical point which could have been resolved more quickly. A slightly worrying aspect of the questionnaire survey is that, despite the value attached by the borrowers to some of the studies specifically suggested by the Bank, none of the special economic comparison studies requested by the Bank, often at a late stage, were felt to have contributed to the project, suggesting a possibly excessive concern by the Bank with risk. A particular danger of lack of progression to preparation studies appears to arise when different agencies become involved over time, each having some tendency to go back to basics. The possibility of important new issues emerging can never be eliminated, however expert the initial identification of work to be done, but a firmly scheduled plan can help all parties to avoid returning to ques- tions already settled unless new information with major implications comes to light. While the advantages of early establishment of project units or com- mittees to carry out project preparation and lead into implementation in many circumstances seem to be widely accepted now, and also illustrated by these studies, the evidence collected suggests that it might be wise for the Bank to go further, to extend its systematic consideration of institutional capa- bilities for project planning to appraisal reports for all sectors and to consider the possibility of providing financial and technical assistance to Project Planning Units as such, especially for agriculture. Even though its efforts have not always been successful, and perhaps less successful overall than in other dimensions of institution-building, the Bank's emphasis on the growth of transport planning institutions has clearly been a significant contribution. The overall evidence is that projects with the shortest prep- aration periods, especially when extended through loan effectiveness, are those prepared by Governments or borrowers themselves, and it is generally supposed that local preparation also facilitates effective implementation. Among the four case studies the most effective and expeditious project prep- aration was by the two entities which were strong enough to take main responsibility themselves and to use individual foreign advisors -- a pattern which also seems to have proved the most satisfactory to respondents to the questionnaire survey. In another of the case studies significant delays arose in trying to reconcile the work of the consultants, who had worked rather independently, with that of the prospective borrower, raising a ques- tion whether it would not have been preferable for the consultant to be more directly responsible to the borrower instead of to the Bank as Executing Agent for UNDP, and also the issue of whether it might not have been pref- erable in the longer run to start with more limited studies and prospective project scope, as suggested by a UNDP advisor at the time. About 40% of the projects covered in the survey of 1974 borrowers were prepared with no more - iv - local participation than supply of data to the outsiders doing the work, but the borrowers considered that little more local participation would have been possible, illustrating the fact that the greatest difficulty the Bank faces in project preparation is the frequent lack of a local institutional base and underlining the importance of efforts to fill the gap. A possibility that would seem to warrant serious study is greater differentiation in the depth to which key appraisal issues are pursued in different projects. This could have a significant effect on preparation time and requirements for some projects. The issue is illustrated by the case study project just mentioned, where the Bank may have been too ambitious in what it hoped, but has not yet been able, to achieve with a first project alone. Simpler projects under certain circumstances might make it easier to make smaller, more timely commitments for parts of programs, or particular experimental solutions in agriculture, with less risk, less possibility of delays due to technical disputes and better control over the performance of the borrower. It is raised in a different form by some of the cases in the other samples, which suggest that full revisions of long-term sectoral or program projections may not be necessary for appraising repeater projects where sectoral issues and the strategy for dealing with them remain essen- tially unchanged. Improvements in the Bank's participation in project preparation are not a matter of programming and monitoring, but of ensuring that committed and experienced staff have adequate time for the job and providing them some more support, for instance in the form of comprehensive guidelines for use by borrowers and more possibility of helping local planning units, such as the new Project Preparation facility should open. The Bank appears to have been aware of projects early enough in their lives and procedural delays have not been a major factor, but a more intensive early involvement seems to be needed to effect significant improvement in the project preparation process. Preparation is perhaps the most entrepreneurial and promotional type of work in the Bank, and effective fulfillment of the advisory work on guidelines and preparation planning described above depends directly on the ability of the individual staff members to perceive well the issues and to help work out a plan to solve them. The most important operational suggestion of this report is there- fore that the Bank ensure, very early in the project preparation period, that experienced staff work over the proposal with the borrower, advise on prepara- tion and reach early agreement on what is to be done, by whom, and by when. This early involvement should reduce the need for later missions, as well as delays in project preparation. One possible way to achieve this objective has been outlined in detail for the Bank management to consider. Its main features, together with other suggestions referring to subjects for further study and a few specific lessons from the experiences reviewed, are summarized in the conclusions of this report. MANAGEMENT POLICY REVIEW: DELAYS IN PROJECT PREPARATION I. THE PROBLEM 1.01 The origin of this study was a concern, within the Bank as well as outside, that it takes too long to bring projects from the stage of initial discussion with the Bank to actual financing, and that the Bank may be excessively slow or unduly perfectionist in this important period of project processing, particularly the portion prior to field appraisal. Member countries have been concerned about the long lead-time with which they have to present projects to the Bank, and other potential joint lenders with the Bank have sometimes expressed the same worry. Within the organization, the large amount of individual project slippage from one year's projected lending to another and the size of the 'gross' lending programs that have been necessary as a result to support the fast increase in lending consummated in the last years have been continual sources of anxiety since the initiation of more systematic programming in the late 1960s. 1.02 Over the years, moreover, the Bank has devoted a growing portion of its own resources to assisting in the preparation of projects for financing from its funds, to the point that project preparation (including identifica- tion, where separate from other work) now accounts for about 21% of total operational effort financed out of the Bank/IDA administrative budget, compared with something in the neighborhood of 19% five years ago and less in earlier years; project appraisal accounts for only a few percentage points more, while economic and sector reporting now accounts for significantly less, as project supervision still does. Quite apart from the long-standing and expanding practice of financing project preparation (by consultants) through loans and credits, many initiatives have been taken to help member countries with project preparation. The Bank became an Executing Agent for UNSF/UNDP as soon as these funds were started at the end of the 1950s; 1964 saw the establishment of the first two 'Cooperative Programs' (with FAO and UNESCO); and in the middle 1960s the Bank set up its major regional offices in Nairobi, for East Africa, and Abidjan, for West Africa. Those early steps have been followed by many others, including creation of other field offices, notably the one in Jakarta set up in 1968, with particular emphasis on project prep- aration, and establishment of Cooperative Programs with WHO and UNIDO, in 1972 and 1974 respectively. Pre-appraisal missions sent out from head- quarters alone increased from just over 400 in FY1972 to over 500 in FY1974. 1.03 The extent of Bank involvement in project preparation varies enormously, from almost nil for a straightforward project with an accustomed Bank borrower to virtually total responsibility in a few cases of new borrowers with projects involving little engineering work. For this reason overall averages have limited meaning, but even these show great variation among the Bank's Regions, as the following table indicates: - 2 - Table 1 Projects Under Preparation and Man-years Devoted,La by Region, FY1974 Preparation Man-years as % All Man- No. of Man- years Spent Projects Man-years years on Projects Under Devoted to Per Through Board Preparation Preparation- Project Approval Eastern Africa 74 27.8 0.38 45 Western Africa 63 26.7 0.42 42 East Asia and Pacific 60 36.3 0.61 59 South Asia 64 26.3 0.41 48 EMENA 127 23.9 0.19 32 Latin America & Caribbean 135 34.0 0.25 44 Central Operating Projects Departments 89 29.3 0.33 35 BANKWIDE TOTAL 612 204.3 0.33 43 /a Including all Cooperative Program (CP) operational manpower, accounting for 85.6 man-years in total. This is an exaggeration insofar as a portion of CP staff time is devoted to other work, but it offsets the failure of the figures to capture effort devoted to project preparation under other heads such as project supervision and sector work. Source: Annex I. Since projects appear to stay in pre-appraisal status in the Bank's operations for about two years on average, the Bank's present average expenditure on pre- appraisal work for a project for which a loan is made may be about two-thirds of a man-year, compared with rather less than double this for appraisal. 1.04 Quite apart from any question of Bank involvement, project prepara- tion, as conceived in this report, is a process with strong internal logic and sequence that apply to almost any large investment of resources, although the Bank and other UN agencies have been important propounders of them. Moreover they take time to fulfill. In its essential elements, the process is perhaps best seen as two diverging movements over the course of time: a narrowing down, from an initially broad range of alternative solutions to a problem, each surrounded by uncertainty, to a specific solution, sufficiently defined in respect of physical and institutional components, costs, benefits and risks to be ready for execution, and a simultaneous broadening out over time of the detail and precision to which the work has to go and hence of the costs involved. Starting broad means clear definition of the problem sought - 3 - to be solved in such a way as not to preclude solutions that may be more economic, while narrowing down over time avoids expenditure on unnecessary detailing of solutions that may finally be rejected. To help keep the logical progression to the work, project preparation is generally divided into a number of stages, each terminating with a decision prerequisite to proceeding with the next stage. The number of stages appropriate to a particular project depends mainly on the complexity of the problems (and the uncertainties surrounding the possible solutions), the costs of the studies involved and the size of the eventual investment foreseen.l/ Although every project intrinsically contributes to broader sector development, an additional dimension to preparation of a project specifically for Bank financing is often study of a broader sectoral policy issue on which the Bank hopes to reach agreement with the borrower in connection with the granting of a loan. 1.05 The view appears to be widespread in the Bank that the main problem countries face in project preparation is shortage of necessary human and financial resources and also that the Bank's ratio of pre-appraisal to appraisal effort should be further increased, particularly for the more novel types of project.2 Stress is being given to the more active and creative role that it is felt the Bank can play early in project preparation, before design has become too firm, than at appraisal. Initiatives are underway to increase emphasis on the Bank's work in project preparation and to systematize the work more.-V At the same time, in a modern, and different, variant of a worry that has traditionally restrained the Bank's involvement in project preparation -- that it may be pre-committing itself or losing its appraisal objectivity -- concern is being voiced that a large Bank role in project preparation can lead to projects being considered by the borrower as "the Bank's" schemes rather than his own; and the need to ensure full borrower commitment is being stressed. 1/ The standard stages are five -- Identification, Pre-feasibility study (or preliminary sorting and selection of alternatives), Feasibility Study (selecting the best solution, perhaps with variants), Preliminary Engineer- ing and Detailed Engineering (or their equivalents on the institutional, architectural, agricultural, curricular, etc. sides) -- although some of them may often be combined. An excellent discussion of the process of project preparation, and the rationale of staging, is provided in "Notas sobre Formulacion de Projectos", Cuadernos del Instituto Latinoamericano de Planificacion Economica y Social Serie II No. 12 (Santiago de Chile 1970). 2/ See, for example, IBRD Report No. 588, "Rural Development and Bank Policies: A Progress Report," dated December 2, 1974. No such increase, on a Bank- wide basis, is however reflected in current projections of the Regional offices. But several Regions have indicated an intention to supervise project preparation more closely, and one suggests preparing project prep- aration reports in the form of appraisal reports. 3/ Through the Country Program Papers; a new system of Project Briefs that is being introduced on an experimental basis; and a possible greatly elaborated manpower planning and programming system for "project generation" proposed in a draft report, "Cooperation between FAO and the World Bank," dated August 8, 1975. 1.06 Time required and quality achieved, in project preparation, are obviously intimately related, and this report is intimately concerned with both. The questions posed at the outset of this chapter, and which have been the focus of the study because of the borrowers' great concern about them, are hard to answer not only because there is the presumed trade-off that has constantly to be kept in mind -- between speeding project preparation and increasing project quality (both in the sense of strengthening the physical and institutional specifics of the project eventually agreed and in the sense that the project preparation experience can itself help to improve the quality of the agency which will implement the project) -- but also because this trade- off is difficult, although essential, to analyze by the technique of evaluation of the past: it is too early to be categorical about the ultimate quality of projects prepared under current policies and too late to be very definitive about the preparation difficulties encountered in projects whose operational quality is already reasonably well-established. The answers given here are in fact principally founded on an amalgam of three different bases: first, thorough case studies, extensively discussed with Bank and borrower personnel concerned, of the preparation phase of four projects approved for Bank/IDA support in 1971-73 and selected mainly for diversity of type and what was then thought to be exceptionally long preparation, so as to throw up clearly the problems; second, an opinion survey (see Annex II), carried out by the good offices of the Executive Directors and designed to indicate the extent of current relevance of problems tentatively identified, of most of the Bank's borrowers in CY 1974, with 52 full replies received, giving a reasonably repre- sentative sample of projects financed, of both traditional and new types, and of the Bank/IDA borrowers in that year except in respect of coverage of Africa, which is rather weak; and third, special review of the broader experience to date of the Operations Evaluation Department with Bank/IDA projects and particularly with the 50 projects, for which loans/credits were made principally in 1964-69, discussed in the First Annual Review of Project Performance Audit Results.l/ In addition, since the general picture of how long it takes to prepare projects for Bank financing has not been at all clear, a good deal of statistical work was done, particularly on 1972-74 lending. 1.07 As with other studies of the Management Policy Review type, this review was not designed primarily to give a carefully balanced comprehensive exposition of what the Bank has done, though it does attempt to be properly balanced in the more general statements made, but rather, as the title of the report implies, to focus on trying to identify problems and their extent, whatever their source, and remedies that the Bank might help to bring. This means that, while almost all the borrowers with whom contacts have been made readily admit that they could have helped to avoid some of the delays encountered, the country contexts are taken here as largely given and solu- tions sought as to what the Bank can do either to help change that context or to act within it. The following chapters therefore give much more prominence to projects with problems in respect of project preparation, in these several samples, and to what the Bank might do about such problems, than would be the case if it was sought only to give a balanced overall picture. The reader should be careful not to draw inferences about the overall pattern, for all Bank/IDA lending, from those discussions which intentionally focus specifically on the problem areas. 1/ Board Document SecM75-692 dated September 26, 1975. -5- 1.08 This report proceeds with two short chapters -- first (Chapter II), to present the main facts that emerge about project preparation, the time it takes and borrowers' views about it, and second (Chapter III), to analyze the reasons why it has sometimes been so delayed -- and then two longer chapters focusing more directly on the specific problems that emerge (Chapter IV) and what the Bank might consider doing about them (Chapter V). II. THE FACTS 2.01 One of the principal difficulties of systematically analyzing the preparation period in the project cycle, particularly on a Bank-wide basis, is the enormous variation among projects in the nature and extent of Bank Group involvement at this stage. While later events in the cycle such as field appraisal or loan negotiations can be dated quite precisely on a historical basis, this is far more difficult for earlier stages. For the purposes of the study it was decided to create a concept called 'Bank con- ception date' (as opposed to 'Country conception date'), defined as the date when the Bank put on record that the project was being considered for finance and began to follow it up in a serious and continual way. The date of con- ception, so defined, may coincide with the recorded date of Bank project identification, i.e., when the Bank entered the project's name for finance in the official Project Timetable. It may however precede that date considerably if much project preparation work was necessary before formal identification. In the empirical work underlying the study the 'Bank conception dates' have been chosen conservatively; the continuous record of the Bank Group's interest in a project has not been pursued further back than the earliest date in the official archives, i.e., in the first Project Negotiations File or in the Project Timetables, whichever is the sooner. In fact about 70% have been selected from the Negotiations Files and 30% from the Timetables. 2.02 The median time between Bank conception, so defined, and loan/credit effectiveness is just over 3-1/2 years. This conclusion is based on analysis of data for some 300 projects approved in CYs 1972-73 and, separately, 150 projects approved in CY 1974. The analyses also show that the medians for each region or each major sector all fall within 10% of this figure, the sole exception being that for irrigation projects which is notably, nearly 30%, longer than the Bank-wide median.L/ The only other distinction which is note- worthy at this level is that between first projects (in a sector in a country) and follow-on projects, each of which account for about half the total numbers of projects approved in recent years: median time between Bank conception and loan/credit effectiveness is just over three years for follow-on projects but nearly 3-3/4 years for first projects. 1/ Even whether this is a finding that would apply to all years can also be doubted. The seven irrigation projects covered in the questionnaire survey show average and median project processing time shorter than many of the other sectors represented. - 6 - 2.03= Of this 3-1/2-year project processing time through loan/credit effectiveness, about 2 years constitute the period preceding first field appraisal: 2-1/2 years for first projects and a little less than 2 years for follow-on projects. This period, which is the focus of interest in this report, shows somewhat greater variability between regions, the medians for Western Africa and East Asia and Pacific being a little longer, as Table 2 shows. But the more noteworthy variation is rather within each region or sector. While there are projects, in almost every group, which go from Bank conception to field appraisal in less than six months and others which take upwards of six years, this variability is indicated mainly by the wide ranges between upper and lower quartiles. Thus, while 50% of projects approved in 1972-73 passed through the stage field appraisal to Board approval in 7-1/2 or 12 months and became effective in 3-1/2 to 7 months after Board approval, their time in pre- appraisal status ranged between 1-1/4 and 3 years; and in many groups, as Table 2 shows the range between upper and lower quartiles was even greater, the former value being often nearly three times the latter. 2.04 In principle, Bank conception date might be assumed to occur some- what after, or at earliest simultaneous with, Country conception date. The answers to the questionnaire survey show that this is indeed often the case but that for as many as 40% of the projects Bank conception preceded Country conception, as now seen by the borrowers. This relates mainly to projects in some of the newer areas of Bank activity such as population, tourism, sites and services and some types of agricultural project. It may reflect the impact of special Bank thrust in these areas. 2.05 Despite the Bank's apparent initiative, however, this does not prevent some of these projects falling in the category of those for which the respective borrowers were, in retrospect, most insistent that preparation time and difficulties could have been substantially cut. Nor do the answers to the questionnaire survey show any correlation between length of time actually taken to bring the project to financing and depth of feeling that economies would have been possible. In other words, some borrowers whose projects had run 4 or 5 years from Bank and/or Country conception date to Bank field appraisal appeared to consider that this time was entirely necessary while others were seriously exercised about projects whose preparation time had fallen substan- tially below the medians presented above. 2.06 Whether the time elapsing after Bank conception date should be con- sidered excessive must of course depend in part on the stage in the life of the project itself at which the Bank becomes involved. The questionnaire answers for the 1974 projects suggest that Bank conception was generally early, before feasibility studies or preliminary engineering had been done, and that the main product of the pre-appraisal period, in terms of studies, was indeed feasibility report/preliminary engineering. In a few cases detailed engineer- ing was also completed by the time of Bank field appraisal but most respondents indicated detailed engineering being carried out partly before and partly after loan approval, while for education and irrigation projects, as well as one or two transport projects (excluding highways), detailed engineering and design was undertaken only after loan approval. Table 2 Number of Months Required to Bring a Project Frnm Bank Conception Date to First Field Appraisal (Based on all projects approved in ralendar Years 1972-73) Median Conception Minimum First Median Third Maximum to Appraisal, Mean Project (0%) Quartile (25%) (50%) Quartile (75%) Project (100.) to Median Conception to Effectiveness (Z) All Projects Approved CY72-73 (293) 31 2 15 24 37 130 56 All First Projects (149) 37 4 18 30 43 130 67 All Follow-on Projects (144) 25 2 14 21 34 120 5i ENTIRE REGIONS Eastern Africa (43) 29 4 12 24 35 118 60 Western Africa (43) 36 2 18 27 43 130 59 East Asia Pacific (50) 35 6 21 29 37 109 66 South Asia (31) 30 2 15 21 34 108 57 ENERA (63) 27 6 14 23 34 112 59 LAC (63) 32 2 15 24 43 113 56 ENT1RE SECTORS Agriculture (83): 33 3 16 28 36 130 60 Irrigation (13) 43 14 24 36 66 95 65 Livestock & General Agric. (48) 33 3 15 27 37 130 59 Livestock & General Credit (22) 26 4 15 21 32 79 50 DFCs (25) 31 2 11 26 49 118 59 Education (33) 34 2 18 25 41 113 56 Iudustry (13) 25 2 17 22 31 61 62 Public Utilities (57): 34 2 17 24 39 104 56 Power (31) 28 2 16 24 35 86 62 Transportation (74): 27 4 11 22 37 120 58 Highways (41) 27 4 11 23 38 120 59 Ports (15) 32 8 12 31 35 103 72 a/ Numbers in brackets show the se of each group. - 8 - 2.07 Comparison with the relatively few projects studied by the Operations Evaluation Department from the early 1960s (in terms of loan approval), and before; suggests that the time from Bank conception to field appraisal and loan/credit approval is now somewhat longer than in the past,1/ without any particular explanation merely in terms of 'new borrowers' or 'new types of project' (some of which appear to have relatively short processing times) and that more thorough engineering work is often onow completed before appraisal and loan approval. Certainly it appears to be the case, roughly consistent with the current Bank principle that detailed engine ring for highway projects should preferably be at least 80% completed before appraisal, that more thorough feasibility work and a larger proportion of the detailed engineering are now done for modern highway projects before appraisal and loan approval. For irrigation projects, much wider-ranging feasibility studies, covering agricultural and on-farm aspects more thoroughly than the heavily engineering- oriented studies of the past, are now generally done prior to appraisal. 2.08 The broad pattern of main responsibility for project preparation appears to have remained fairly constant over time, but the Bank, both directly and through its Cooperative Programs with the other UN Agencies, seems to have been taking a more active secondary or tertiary role. It appears that about 50% of projects now being financed by the Bank and IDA are, as in the past, prepared principally by the borrowers themselves, 40% principally by consultant firms (nowadays financed in roughly equal proportions by four sources -- the countries themselves, UNDP, IBRD/IDA loans and credits and bilateral assistance), and 10% principally by other agencies, including the Bank itself and the Cooper- ative Programs, with some tendency for the share accounted for by consultant firms to drop, in favor of the last category. But by 1974 the Bank was signif- icantly involved in one way or another (e.g., through being Executing Agency for UNDP or through preparation being financed under previous loans or credits, as well as in other more direct ways at the earliest stages of preparation of at least 40% of projects for which loans/credits were signed. 2.09 It might be expected that such early Bank involvement would speed up project preparation, but the evidence on this is not entirely clear. Review 1/ Study of the files for the 44 projects approved in CY 1965 (excluding those in courLtries such as Italy and Japan) also shows that median periods Bank conception to first appraisal and Bank conception to loan effectiveness have risen from about 8.5 and 22.5 months, respectively, at that time to 24.0 and 42.0 months for projects approved in the last few years. However, an indeterminate but probably large part of the difference between the 8.5 and 24.0 months is a statistical mirage resulting from the introduction since 1965 of the Bank's Country Program Paper and Project Timetable systems, causing projects to be recorded earlier for potential financing. One estimate we have made of the adjustment necessary for purposes of compara- bility (adding 12 months of 'programing foresight' to the times recorded for the very large number of 1965 projects with 'preparation periods' of only a few months) would raise the 1965 median to about 18 months, so that the total period from (hypothetical) Bank conception to loan effectiveness would have risen from 32 months for projects approved in CY 1965 to about 42 months for the projects approved in the last years. - 9 - of a fairly large sample of projects (60 in all) for which loans and credits were approved and signed in the first half of CY 1974 shows that slippage from original schedule was substantially less through field appraisal in cases where the Bank was involved early in preparation than in others, but that this advantage was largely, though not entirely, lost by the time of loan approval; and it is uncertain how much such lesser slippage may simply reflect improved ability to forecast accurately in a case where the Bank is more involved. Overall time elapsing from Bank conception to loan effectiveness, at a median of 42 months for the 24 cases in which the Bank was involved early, was longer than for the 22 projects prepared by Governments without significant early Bank involvement (median of 37 months) but shorter than for the 14 projects prepared by consultants acting under terms of reference provided by Governments or other agencies (median of 51 months), but the last figure is partly affected by the great technical complexity of many of the projects included. These alternative means of preparation are of course often not available for any given project -- the responsible Government agencies may for instance be too weak, so that outside help is essential -- but the best it seems possible to conclude at present is that, as of now, early Bank involvement in preparation may speed up the process, whoever takes the main responsibility, but not by very much, perhaps a month or so over the full cycle to loan effectiveness. 2.10 No independent test of operational quality of project preparation (other than the important one that, as amended and elaborated -- sometimes substantially -- at loan appraisal, it was considered adequate to justify loan approval) can yet be applied to projects approved in recent years, but it is worth noting some of the major implications for project preparation of the recent Project Performance Audit Review. It showed a good performance on the part of the large majority of projects and found very little evidence of proj- ects having been built to excessive design standards. On the other hand, in a few of the mther numerous cases where the Bank had valuably helped to avoid excessive standards or scale of project, these changes had come late, not indicative of a very efficient project preparation process; while the Bank had been a major proponent of strengthened local planning and project prepa- ration capability and criteria in many of its borrowers' agencies, this tended to be among the less successful of its institution-building efforts, apparently because some of the borrowers did not give it great priority; and, in the relatively few cases of projects that did show disappointing economic results compared with potential, it was clear that the failure had been much more often in project preparation than in project appraisal. Longer, broader views and economic action plans had been needed and these were not of the nature that they could be prepared by brief appraisal missions. Engineering preparation, on the other hand, had generally been notably successful and the small, specific weaknesses on this side -- such as insufficient soil studies for highways or inadequate hydrological work -- were of the sort to be remedied by some of the developments in Bank project preparation standards referred to above. 2.11 In sum, then, it appears that project preparation has been generally adequate in quality for the large majority of Bank/IDA projects and has been improving over time, but it has taken an increasing amount of calendar time -- in excess of two years for half the projects approved by the Bank in the last few years -- and a significant minority, about a quarter, of 1974 borrowers - 10 - surveyed considered that there was significant scope for the preparation of their projects to have been more economic and expeditious. Another 30% identify minor improvements that could have been made. Their suggestions are discussed in the following chapters. But the particular needs to be accom- plished at this vitally important stage in the life of a project are too variable and specific to each situation for it to be possible to establish on the basis of experience, any simple norms, broadly relevant, for the time that should be taken. III. THE REASONS 3.01 The four detailed case studies, which have been carried out on the basis of a flow chart-systems analysis technique (see Annex III) to try to cope with the multiplicity and interrelationship of factors involved in reality, illustrate the kind of reasons why project preparation sometimes gets delayed and their relative importance; the larger samples studied provide a check as to whether the factors have the same relative importance more broadly, but they do not permit such careful analysis of what difference long preparation makes, in terms of improvement of project quality or delay in realization of ultimate project benefits. 3.02 The four cases were originally selected, early in 1974, to cover a wide geographical/sectoral screen and, in particular, to represent cases of exceptionally prolonged preparation. Two were in Asia and one each in Europe and Latin America; loans or credits were approved in 1971-73 and discussions had begun in 1966-69 except in one case that went back to 1961. They are all first projects for the Bank in a sector in a country. But when the compre- hensive statistical studies mentioned in the last chapter were completed, it became apparent that two of the chosen projects -- one in tourism and one in industry (fertilizer) -- actually had preparation periods, from Bank conception date to first appraisal, quite close to the 2-1/2-year Bank-wide median for first projects. The other two projects -- a multipurpose (water and power) scheme and a railway program -- were, on the other hand, in the upper ranges of preparation time, although not the longest, as comparison of Table 2 above and Table 3 below indicates. Bank Procedural Delays 3.03 One possible factor in explaining long preparation times which needed investigation was essentially internal Bank procedural delays, such as slow response, poor scheduling or lack of parallelism in work. The flow charts did bring out a few instances of these, but they did not amount to much overall. For example, the multipurpose project, in the early years of its preparation, suffered from a sharp reversal over a one-year period of the Bank's official position, apparently due to poor interdepartmental coor- dination in the Bank; the conclusion of a Bank technical report favoring immediate preparation of detailed engineering was replaced with a recommenda- tion for further and broader preliminary studies. When preparation was greatly broadened, a few years later under UNDP/FAO, to include a basin-wide Master Plan before returning to a feasibility study for the project itself, the Bank Table 3 The Four Selected Projects, Compared with Other Related Groups-a/ Number of Months Required to bring a Project from Bank Conception Date to First Field Appraisal Median Conception Minimum First Third Maximum to Appraisal Project Quartile Median Quartile Project to Median Conception Mean (O) (25%) (SO%) (75%) (1007) to Effectiveness (%) The Tourism Project 32 57 h/ Corresponding Region First Projects (26)- 40 8 16 30 63 113 68 The Fertilizer Project 34 62 Industry First Projects (8) 24 12 17 25 32 34 68 The Multipurpose Project 98 78 Irrigation First Projects (9) 52 14 34 36 80 95 62 Corresponding Region First Projects (11) 38 8 12 30 39 108 79 The Railway Project .74 80 Corresponding Region Virst Projects (41) 28 6 14 27 34 112 63 All Railway Projects (12) 27 7 15 19 27 74 63 a/ The Groups are all taken from the main data base, for CYs 1972-73, although the multipurpose project was approved in CY 1970 and the tourism project in CY 1971. b/ Numbers in brackets show the size of ead group. - 12 - appears to have been insufficiently in touch to send a review mission, with FAO, at the crucial point between the two stages, although it appears always to have been envisaged that the eventual project would be put up for Bank/IDA financing. On the fertilizer project, nearly half a year was lost at the start of Bank involvement while it awaited a letter to confirm a top-level agreement in principle between the Government and the Bank before studying the preliminary feasibility report that had been sent. A few months were again lost later on, due to staff shortage or poor scheduling, when the Bank was unable to deal quickly with revised technical proposals from the borrower, and some months toward the end of the preparation process by making bidding and appraisal less parallel than they might have been (and would now be). Some- what similar small examples arise in the railways project, with the Bank pro- ceeding in a rather slow, deliberate sequence on the recruitment of consul- tants under a UNDP project, instead of preparing the lists ahead of time for this long-envisaged study; waiting too long for the refinement of the con- sultants' eventual report before initiating action; and requesting comments by mail instead of having joint working sessions. Elimination of all delays of this type might perhaps have saved some two-three years out of the aggre- gate 20-year preparation period for the four projects, but that is on a gross basis, not allowing for other delaying factors occurring simultaneously and the net effect of their elimination would have been useful but very much smaller. Government Decision-makin& Delays 3.04 Another factor which can sometimes be important in explaining long project preparation time is Government difficulty in reaching major decisions, for instance about the Bank's propositions as to institutional arrangements. Preparation of the tourism project, for example, suffered some delay due to Government indecision as to which among its agencies should be responsible for the airport; the original agency involved was replaced and then brought back and, although basic design work continued, this occasioned some delay due to difference of opinion among the agencies on the best site. Institutional issues, mainly related to the extent of responsibility that should be borne by the contractors as compared with the public sector agency in charge of the project, also arose in the preparation of the fertilizer project. But the only significant example of delay of this type, among the four cases, con- cerns the establishment, in connection with the railway project, of a separate agency specifically to deal with general transport coordination; having once been established after much delay, it was then abolished, only to be recreated as a sine qua non of the eventual Bank railway loan, but so far, after its first two or three years of renewed life, it cannot be considered to have had significant effect on policies or investment decisions. The argument over this matter ran over some four or five years of the preparation period, and no doubt an earlier resolution of this problem would have expedited the movement towards the Bank's loan, but there were also considerable simultaneous diffi- culties in preparing the railway project which the Bank was actually to finance. - 13 - Delays in Getting Appropriate Studies Done 3.05 By far the overriding reason for the prolonged preparation for these four projects does indeed seem to be sheer difficulty in getting the project preparation studies done and having them well oriented to be sure a capacious category with many variations but one that it is hard to break down in any generalizable way at this stage. Lengthy periods involved in organizing studies, redoing or complementing previous ones, in some cases undertaking studies which proved to be of little consequence for the project, reconciling work by different sources, seem quite often to have been on the critical path of project preparation. Entirely leaving aside the time that was spent doing studies that made a significant contribution to formulation of the project as eventually undertaken and a reasonable period for organizing the execution of these studies, a gross period of something like 10 years seems to have been lost for reasons of this sort. The problem affected all four projects in varying degrees. 3.06 After giving its revised opinion on the studies needed for the multi- purpose project, the Bank offered to finance them, agreed the terms of reference with the Government, and made preliminary arrangements with a firm of consul- tants. Then it hesitated, on 'country' grounds of poor public financial per- formance and a nationalization issue, and eventually withdrew its offer. The Government approached UNDP for financing and agreed with FAO, the Executing Agent, that a large study, beginning with a basin Master Plan and only then going back to a feasibility study for the project, should be done. The Bank advised against preparation of the Master Plan, as not essential for the project, but became actively involved again only six years after the original agreement for the study that was to have been financed by the Bank;l/ at the request of the Government, the Bank sent a joint Bank-FAO Cooperative Program mission to review the work of the UNDP study, then coming to an end. Certain revisions, and complementary studies on the agricultural and institutional aspects, were carried out, and, due to UNDP/FAO refusal to have remaining funds spent on engineering work, detailed engineering was separately contracted with a con- sultant firm. Nonetheless the Bank appraisal mission which came one year later found the project so inadequately prepared -- particularly in respect of land and soil studies,2/ an issue not raised by the Bank-FA0 CP mission -- that it considered it necessary completely to revise it, postponing any irrigation distribution works until further studies had-been done, this time under the 1971 Bank financing for construction of the bulk supply system. These studies 1/ Modern procedures for Bank expression of 'Special Interest' in a pre- investment study financed by UNDP but executed by another agency, to help keep the work in line with what the Bank would require, did not apparently exist at that time. 2/ This problem, and its appearance so late in the preparation of the projects, resulted partly from a more general difference of opinion between the Bank and FAO, which was prepared to accept lower 'recon- naissance' level standards of land and soil studies as a basis for recommending financing. - 14 - were delayed in execution, gave rise to substantial debate within the Govern- ment and, despite a number of interim missions, were found by the Bank, when the critical time for decision-making came, to be based on excessive design standards and theoretical figures on water requirements, problems not pre- viously identified. It appears now that the main project works, even though runing more than a year late, will be completed several years before the irrigation distribution system has been sufficiently upgraded and extended to make great use of much of the water. Yet a retrospective review of the numer- ous changes through which project design has been since the Bank's first involvement, now 14 years ago, suggests that only those introduced in the first year or two of this period may have made a material difference to the project's basic economics, and there is strong engineering opinion that essential distri- bution works could and indeed should have been undertaken simultaneously with the main scheme, without further major studies. From a modern perspective the Bank now feels that the main error was not to put sufficient emphasis on the agricultural and water-use side in original planning of project preparation, but the project as prepared prior to appraisal would probably have been acceptable by appraisal criteria of the early 1960s, and one of the ways the scheme has been so delayed appears to have been changing Bank preoccupations about irrigation schemes at the rather widely spaced occasions of its involve- ment. 3.07 Considerable difficulties also arose in getting the preparatory studies done for the railway project, although there is less doubt about the necessity of the various studies undertaken -- except possibly, retrospectively. in view of the rather limited success to date of the Bank's 10-year stress on transport coordination, of the studies on this subject, whose execution was made a precondition of the loan. Again some confusion was introduced by a change, due to a Bank policy change at the time, from the possibility of Bank financing for the studies to the need to seek UNDP assistance. It was diffi- cult at short notice to find a place in the UNDP program for such studies. Three years after the Bank's original proposal, the major studies finally got underway, in this case with the Bank as Executing Agent and following terms of reference drafted by the Bank. The consultant apparently worked, for effi- ciency, rather independent of the Railway authority, which went on preparing its own plans. Many areas of the consultant's assignment could only be covered with recommendations for further studies, it took two years from the consul- tants' interim report to reconcile, with Bank technical assistance, their proposals and those of the authority on the basic investment program, and a further year to produce a finally agreed plan of action. The borrower has had great difficulty in keeping to the plan of action, particularly in respect of tariff and financial performance, and he considers the Bank somewhat insensitive to the socio-economic realities of his situation, but he has certainly benefited from the heavy technical assistance effort that has been made by the Bank staff itself over the years. 3.08 The fertilizer and the tourism projects, both of which are being executed fairly close to schedule (though with delays on some parts of the latter, largely due to a difficult land acquisition issue), contrast with the other two in that main responsibility for their preparation lay with nationals of the countries involved, but they too suffered from considerable - 15 - difficulties in preparation of essential studies. Much of the problem seems to have been lack of clarity on the part of the prospective borrowers as to what was required by the Bank. The Bank was continually disappointed with the documentation and information provided in response to its questions. For the fertilizer project the Bank failed to take up the suggestion of its local representative that what was really needed was an outline of the kind of feasibility study desired, and about a year-and-a-half was spent on a techni- cal argument about gasification processes which had finally to be arbitrated by outside professors, who found no very significant difference between the two and a slight advantage to the one favored by the borrower; the saving grace of this delay was that it held up action to a time when shortage of electricity in the region where the plant would be located was a much clearer problem, and the Bank then contributed usefully to the borrower's redesign of the scheme in light of this constraint. The tourism infrastructure project suffered from both the Bank and the implementing agency being new to the field and the lack of agreed logical order to the numerous studies that had to be made for the many different elements of the project; not until two years after the first major technical mission on the project did the Bank finally define precisely what information it needed in connection with the important airport component of the scheme, and Bank technical specialists for some other sub-components were brought in only at the time of appraisal, when they raised a few basic problems, which it was really too late to deal with -- at some cost, probably, to the economic quality of the project. Assessment of Aggregate Potential Time Savings 3.09 Except in the case of the multipurpose project, with the seemingly somewhat unfortunate changes insisted upon by the Bank at a very late stage for lack of land and soil studies of acceptable quality, there is not much evidence from these case studies that the project preparation standards required by the Bank have been too high; indeed they may possibly have been on the low side in the tourism project. For this reason possible savings in time that might have been made in the preparation of these projects were not to be found by trading off sacrifices in project quality against the savings in costs and time that might have been made by eliminating particular studies or steps in the preparation process, or by revising project objectives, but rather by seeing how the studies which contributed little to project quality might have been avoided or how necessary studies might have been brought to successful fruition more expeditiously, or how procedural delays might have been eliminated. Moreover, allowance had also to be made for the fact that there were often multiple debates going on together so that elimination of one particular step or study would by no means necessarily save to project processing time the whole of the time required by such a step or study. 3.10 Cautious and detailed calculations applying the principles described in the last paragraph to the flow charts for each of the individual projects and identifying the particular points at which time might have been saved and how much might have been saved nonetheless suggest that quite sizeable amounts of time -- in aggregate, some 20% to 30% of the project prep- aration times shown at the outset of this chapter -- could have been saved in each case, without sacrifice in project quality and without assuming - 16 - unrealistic amounts of foresight on the part of the participants. The main savings would have been in connection with studies. In the case of the multipurpose project the time-saving could have been substantially greater from a purely technical point of view, as implied by the discussion of it above, but the difficulty here is that, due to the discontinuity of Bank lending to the country involved, the question of feasible savings in time required to consummate the loan and start the project enters quickly the realm of unconstructive political speculation. The monetary values involved in these potential time savings can be quite considerable. For instance it has been estimated that the 14-month time-saving calculated to have been possible for the fertilizer project would have enabled net savings on imports of some $6 million, at the appraisal report's projection of world fertilizer prices, and as much as $40 million given the fertilizer prices that have actually eventuated, for a project with a total capital cost of about $80 million. The borrower for the multipurpose project has pointed out that delays in obtaining Bank financing also meant delays in obtaining substantial amounts of complementary bilateral financing. The Evidence of the Questionnaire Survey 3.11 The responses received to the questionnaire survey of 1974 borrowers yield a general picture regarding the reasons for delays in processing of projects for Bank financing which is surprisingly consistent with that revealed by the few detailed case studies undertaken. The most serious single problem, whether measured by the intensity with which it is sometimes expressed or by the proportion of respondents who raise it (over 25%), is lack of knowing early enough (equally in cases where the Bank was actively involved at that stage as in cases where it was not) exactly what the Bank wants in terms of information or project preparation studies, while the second most common problem, partly related and expressed by just under 25% of respondents, is that the Bank some- times required studies which the borrower still considers, after having done them or having started his project, unnecessary, unduly repetitive or too elaborate; an equivalent number considers that the studies required by the Bank in their cases, mainly engineering studies which they would nonetheless not have done had it not been for the Bank's request, and sometimes special studies on particular sectoral issues, were useful. These points are very consistent with the prominence assumed by the question of project preparation studies, and the difficulties of getting them done appropriately, in the case reviews; however, only one or two refer to any difficulty in obtaining financing for studies. 3.12 The questionnaire-answers also reflect the same relative lack of importance of what was termed above 'Bank procedural delays' and 'Country decision-making' in explaining preparation delays. As regards the former, it received little stress: some borrowers felt that more or longer Bank missions could have helped while a few others would have preferred less; one or two stressed the need for Bank missions, whether visiting or locally resident, to have more authority; and one or two emphasized the delay intro- duced by the Bank's procedures for hire of consultants. As regards the latter, perhaps most striking of all, of the 40% of respondents who felt that the Bank had raised an issue of institutional nature, all but four felt that the effort had been useful and helpful -- although in quite a few cases made - 17 - rather later than would have been desirable -- and a few implied that the Bank could have helped further to expedite country decision-making by efforts of an institution-building nature; the two population borrowers both felt that the major delaying factor had been their own Governments' decision-making. The Evidence of the Project Performance Audits 3.13 The Project Performance Audits, providing critical assessments of projects earlier financed, equally reveal, as might be expected, a few delays in project preparation due to Bank procedural matters (e.g., delayed mission arrangements, sometimes due to staff constraints, or slowness in responding to country correspondence) or to difficult country decisions (sometimes totally outside the Bank's sphere, and when concerned with Bank institutional require- ments, found generally worthwhile in retrospect in these cases). 3.14 Again the much more prominent, and frequently encountered, difficul- ties, in those cases that did present project preparation problems at the time or in retrospect, concern studies: poorly controlled consultant studies, excessive studies, more supplementary studies than should have been necessary had requirements been well specified in the first place; and also, it should be stressed, sometimes inadequate studies or studies that were with hindsight less broad than they should have been -- so that better studies might have extended preparation time, although in fact they could generally have been done simultaneously with other steps that were fulfilled. IV. THE ISSUES 4.01 The preceding chapters imply that most projects financed by the Bank/IDA have been adequately prepared in an amount of time that the Borrowers directly familiar with them have not considered greatly excessive, but that there is a minority of projects which have suffered significantly from problems of inadequate or overly slow preparation, that about a quarter of the prepara- tion time could have been saved in the case of the four projects studied in depth without undesirable sacrifice in quality, or undue foresight on the part of those involved, and that there is scope for at least small time-savings in the preparation of many projects; moreover, the principal problems seem to revolve very much around the studies required. The purpose of this chapter is to explore the specific issues and dilemmas that the Bank confronts in this field in more depth. The effort has been to concentrate, as much as possible, on matters which may be relevant to current operational practice, especially as indicated by the survey of recent borrowers, for there is no doubt that the Bank has learned many of the lessons of the older history reviewed; for instance, there is relatively little reflection in the questionnaire answers of the prob- lems of unconstructive inconsistency between Bank departments and between the Bank and other UN agencies that arose in the history of the multipurpose project or of recent delays in getting studies started comparable with those which characterized this project and the railways one too. 4.02 There appear to be three broad sets of issues -- which we have called Defining Requirements, Local Participation and Bank Rigidity/Flexibility -- and this chapter is divided accordingly. But all three areas are intimately interlinked. - 18 - Defining Requirements 4.03 There is little doubt from the evidence gathered that the place where the Bank could contribute most to improving the process of project preparation is better definition, at an early stage, of what it requires by way of studies and data in support of projects to be financed. The conse- quence of sometimes poor definition in the past has been that studies have sometimes had to be redone or extended or have been more expensive than necessary to provide what the Bank really desired, or that the project financed has been less good than it might have been -- and occasionally has had to be substantially revised, and delayed, in execution, not because the external situation changed but simply for this reason. There are, moreover, two aspects to the poor definition: sometimes, insufficient specificity, leading to misunderstanding and disappointment, and sometimes inadequate selection of issues, resulting in important aspects being left uncovered or undercovered. 4.04 There is a fairly common phenomenon of backtracking. For instance, both the Project Performance Audit Review and the Questionnaire Survey indi- cate a number of examples where the Bank found it necessary to ask the consul- tants, after they had completed their studies, for supplementary information -- particularly on economic aspects and on traffic forecasts for transport proj- ects -- which have generally taken at least several months and occasionally as much as a year to produce. One of the shocks to the engineer responsible for the appraisal of the multipurpose project was that, after ten years of active preparation work and eight years of Bank involvement, no geological investigation had been made of one of the dam-sites while the design of two five-mile-long tunnels was based on no more than two drill-holes; further work on these aspects, after field appraisal, in order to firm up cost esti- mates was a significant factor to delay completion of the appraisal report. Reference was already made earlier to the relatively large number of cases revealed by the PPA Review where a major contribution of the Bank, usually at appraisal, had been to cut down project scope and standards, good for the eventual economics of the project, but still raising a question about the efficiency of preparation. The questionnaire answers suggest that a problem encountered in the tourism project studied may also be characteristic of other projects in this relatively new area for the Bank: the Bank raising questions of a 'feasibility' nature after detailed engineering has been completed or questions of an 'identification' nature after the feasibility study has been done; in the project studied it actually proved to be too late to do anything about the nonetheless valid questions raised at a very late stage with regard to the breakwater and the roads. One or two borrowers covered by the questionnaire survey refer to problems of inconsistency between preparation missions (for instance under the Cooperative Programs) and appraisal missions which they think could have been avoided with firmer plan- ning at the outset. And more than half a dozen respondents raise independently, without any particular inducement from the way the questionnaire is worded (see Annex II), the need for early guidance and specific guidelines to make the preparation process more efficient, with less tendency to regression. - 19 - 4.05 It may be thought that the problems arise from the Bank becoming acquainted with the projects too late, so that it had no opportunity to help specify what the project preparation studies should cover. This does not appear to be generally the case even for the older projects, covered in the PPA Review, and less so for the more recent projects, with their apparently longer time from Bank conception date to appraisal. Rather, the difficulty may be that the Bank did not attach as much importance as would have been worthwhile to its early contact with the project.1/ or to reviewing fully the terms of reference for the preparation studies, or it may be of the types illustrated by the tourism project and the fertilizer project studied in depth. In the former case, at a very early stage soon after the first official request for assistance in the sector came from the country, the Bank sent a high-level mission which seems to have identified very well all the issues which were to become key -- such as land acquisition, inter-agency cooperation, economic methodology -- and to have reviewed the need for basic planning parameters. But it seems never to have gotten into enough detail as to just what would have to be done on the key issues,2/ and in preparation of documentation for the numerous sub-components of the project, a process which eventually took not the six months it forecasted but 24 months, marred by repeated misunderstandings as to what data and what degree of engineering detail were required and frequent downgrading of planned appraisal missions to pre-appraisal; it was not until 18 months after the mission's visit that a written project preparation plan was agreed, and then only for the major items. Bank involvement in the fertilizer project also began at quite an early stage of its preparation, but the difficulty there was that the first mission failed to identify, with sufficient clarity, the overriding importance of the problem of the high opportunity cost of electricity which, had it been foreseen, could have directed efforts immediately to appropriate adaptation of the scheme, quite possibly thereby avoiding the diversion into the long argument about alternative gasification processes. 4.06 Good early identification of issues and rather precise agreement on what should be done about them appear to be additionally important because the studies which the history suggests to be lacking or delayed are not general but rather specific and focused in nature. Project evaluations done to date 1/ The Bank may not have been aware of the importance attached to its advice at this stage: for instance, that there are cases where its general comments on what further work needs to be done have sometimes been handed on, just as they stood, as the terms of reference for a further consultant study. 2/ This finding closely parallels the phenomenon encountered in some of the projects studied for the recent report on Delays in Loan and Credit Effectiveness (IBRD Report No. 813 of July 22, 1975) -- insufficient depth of pursuit, at the outset, of certain legal and institutional issues -- which led to the suggestion in that report that procedures for securing selective early involvement of Bank Lawyers in potential problem cases especially where multiple parties, land acquisitions or a new administra- tive unit is involved - 20 - do not indicate great problems in getting done the broad economic and sector studies (by the country or the Bank) which often led to the project idea nor do they raise much doubts about the basic sectoral priority of the projects undertaken. The case studies underlying this report suggest that the Bank has often been wisely flexible here, being quite ready to accept the apparent validity of the tourism project, for example, without demanding a detailed proof in the form of a large study of the whole tourism sector in the country where it is located; much the same goes for the fertilizer project, where the Bank was hesitant, too, after it had been proved incorrect once, to suggest further optimizations other than one which was clearly convincing and quickly received the approval in principle of the borrower; these applications of the Bank's principle of seeking good projects, not necessarily the best, clearly saved valuable time. 4.07 The multipurpose project, by contrast, seems to have suffered pre- cisely from overly prolonged attempts to optimize it -- perhaps a particular danger when primary involvement in its preparation changes, as it did in this case, from agency to agency, with each new one naturally having some tendency to go back to basics and to try to improve on its predecessors. The UNDP- assisted Master Planning diversion, to which this project was subject against the Bank's advice, does not seem to be an isolated case; another irrigation project, covered by the questionnaire survey, met a very similar fate in the same period of the middle 1960s with equally little positive effect on the project originally envisaged, according to the Government responsible for it. More generally, the ex-post project reviews done to date raise considerable doubts about the real value of major one-shot Master Planning exercises, however good they are in themselves. They become out of date too quickly without leaving a local capacity to update, may not be responsive enough to urgent needs to be practical and can have unsuspected side effects like lulling an agency into a false sense of security that it has now discovered the best project in its field and need not bother to prepare others, and of course they tend to get forgotten in detail. The record suggests that the Bank's apparent reticence about this approach has been prudent. 4.08 Quite aside from excessive optimizing there is of course a problem which can never be totally overcome, even with the most expert and thorough initial review of what project preparation needs to be done -- namely, that crucial new issues will emerge only during the course of the project prepara- tion, rather in the way that the electricity shortage affecting the fertilizer plant did. Such issues sometimes present a serious dilemma, which can only be resolved on an ad hoc basis weighing the relative risks and costs involved in either extending studies to cover the issue and delaying the project or, alternatively, keeping to project schedule and initiating separate action on the issue in question; the dilemma needs to be explicitly confronted and not swept under the carpet. Clearly it was correct in the fertilizer case, as the country authorities insisted, to redesign the scheme in light of the new knowledge, but it must be admitted that the practical question which arises retrospectively, in this case as in other similar ones so far encountered, is whether more adequate initial work would not in fact have identified the issue ahead of time. - 21 - 4.09 There does appear to be a real need for very concrete advice from the Bank on the studies needed to bring a project to fruition, taking into account the specific circumstances of the country and building in a certain progression of work to avoid endless optimization. The Project Performance Audit Reviewl/ tended to reach the conclusion that the type of preparation study which had sometimes been most critically lacking in the past was one concerned with complementary development or investments in the region where the project was located -- to increase the impact of a highway or a primary irrigation investment or to adapt better the curricula in lower schools. But the need for such studies, like others, can clearly not be determined on any a priori basis: there are countries where the economic system has been quite strong enough to respond effectively to the opportunities created by such projects, without the need for any elaborate prior planning, and equally clearly there are other cases where this has not been the case and Bank- assisted projects have had consequently limited effects. Moreover there are cases where an equal need may be for guidance on the most specific details of engineering standards -- the appropriate soils studies for highway projects, an aspect that seems quite often to have been underemphasized, or the level of detail to which geological studies should be taken to reach the Bank's normal standard of cost estimates, accurate (excluding inflation considera- tions) to 15-20%.2/ Choice as to what is required for a particular case can only be made by an experienced mission, that has had time to identify the key issues properly; and this is a service which history suggests it would be advisable to provide to any member country for projects planned for Bank/IDA financing if gaps in the operations pipeline are to be avoided. Local Involvement 4.10 One important reason why a number of borrowers, in responding to the questionnaire, urged that they be given more guidance and guidelines as to what the Bank requires was that they felt this would enable them to do much of what the Bank and Cooperative Program missions and Consultants now do for them and also to produce projects better adapted to their particular circumstances. Clearly this has some importance to the Bank's budget,3/ 1/ And the study, "Comparative Evaluation of Selected Highway Projects", IBRD Report No. 349, dated March 6, 1974. 2/ For the exceptional case of tunnels, see Public Utilities Department, "Guidelines for-Estimating Costs of Tunnel Construction", dated January 17, 1974. 3/ The recent review of the FAQ Cooperative Program, in strongly advocating what it terms a 'bus-stopping' approach for IBRD/FAO assistance to project preparation, meaning periodic brief visits to advise and assist a local team taking main responsibility, presents interesting data for projects whose preparation, assisted by the FA0 CP, was completed in CYs 1973 and 1974; while, for projects prepared with "substantial country participation", the average number of missions was 2.2 and the average number of CP man-days was 201, for those prepared with "minimal country participation" the average number of missions was slightly less, at 1.8, but the average number of CP man-days was 316 -- 56% more than in the cases of substantial country parti- cipation. This does not of course mean that substantial country participa- tion is always possible, but it gives an indication of the budgetary savings that can be made when it is. - 22 - as well as to their own expenditures on consultants, and it contrasts with the view sometimes expressed in the Bank that the main problem of project preparation is lack of appropriate staff in the member countries. Indicative of some of the basis for that view, on the other hand, is the fact that a much larger proportion of the borrowers covered, as many as 40%, had had their proj- ects prepared with seemingly no more local participation than supply of data and information to the outsiders doing the studies and felt that it would not have been feasible to improve on this situation -- except for a very few who urged that more use could have been made of local consultants. 4.11 It appears to be generally believed that, in cases where there is scope for choice, heavier reliance on local institutions and staff for project preparation tends to mean longer project preparation time. Some borrowers and some Bank Departments state that they have on occasion expressly chosen this option, the former stressing the long-run training value of such a course and the latter emphasizing consequent more effective project execution. No doubt the trade-off is a real one in many individual cases, at least in the short run. But it must be admitted that it is hard to find any statistical verifi- cation for it and that whatever analyses it has been possible to make of the relationship between preparation time and primary and secondary responsibility for preparation point to the conclusion that, overall, project preparation periods, especially extended through loan effectiveness, are least for projects prepared principally by Governments and their agencies. 4.12 There are of course cases where the involvement of foreign consultants appears actually to have slowed project preparation. A performance audit of a port project in Africa found that a consultant firm, together with the Bank, had delayed the progress of the project by about a year by requiring more detailed investigations and their reflection in the bid documents than seemed to have been necessary in retrospect or justifiable in light of local knowledge at the time. There appear to be one or two similar cases covered in the questionnaire answers. Three respondents state that they had to re-do the work of the foreign consultants. But delays due to the foreign consultants themselves do not appear common. A more significant factor may be the time required to obtain appropriate foreign consultants -- often in the region of 6-12 months. A few of the cases reviewed suggest that the Bank should consider how it could be more helpful to borrowers by providing lists of firms that have worked satisfactorily in that part of the world recently, in advising on firms available in a particular technical field and in pointing out the need for early action. 4.13 But all the evidence, including that for expeditious project prepa- ration and project execution in the long run, points to the importance of building up local capabilities, which has of course long been an important general institution-building objective of the Bank in its own right. The borrowers covered in the detailed case studies who have benefited the most institutionally from the Bank's involvement are almost certainly those for the tourism and fertilizer projects, who were already strong enough that they were able to use foreign advisors, in the tourism case for airport and urban plan- ning, and, in the fertilizer case mainly equipment manufacturers' personnel, very selectively and to have the benefit of on-the-job training in connection with the Bank-assisted projects almost entirely for their own staff. By its - 23 - advice on plant management arrangements and its insistence that process guarantees be obtained from the manufacturers supplying equipment to try to ensure early operation at full capacity, the Bank appears to have made significant institution-building contributions in connection with the fertilizer project. It is noteworthy that the borrowers covered in the questionnaire survey as well as in some of the Project Performance Audits who seem most pleased with their foreign consultants' contribution are those who have been able to hire them individually, or very selectively, for work directly with their own staff or with local consultants. 4.14 But effective use of outsiders in this way, or indeed of the "bus- stopping" technique referred to earlier, requires already a solid institutional base in the country to incorporate and work with the individual specialists and visitors. And it is precisely the absence of such a base, in the form of a project planning unit, which is the greatest problem that the Bank encounters in effective project preparation, especially in some of the poorer countries and in some sectors now receiving priority, outstandingly agriculture. Nor does the problem seem easy to solve. Even in areas where the Bank has long given close attention to the development of a local planning base, notably in transport, some dramatic successes (e.g., the Brazil Highways sector) are matched by cases of very limited achievement, seemingly due to lack of political priority for the effort in some countries, related salary problems and a certain scepticism about the possibility of planning on economic principles. Not a few long-time borrowers of the Bank still seem to rely mainly on foreigners for project preparation. The very fact that many of the studies specifically required by the Bank which the borrowers answering the questionnaire find of limited value are economic studies, while those they clearly value are mainly of an engineering nature, and the large extent to which components have to be included in new loans to cover consultants to prepare further projects, illus- trate the pervasiveness of this problem. 4.15 Some $5 million have been spent over the last eight years on consul- tants for the preparation, and further advice during early execution, of the railways project studied as part of the background for this report. With a heavy additional investment of technical assistance by the Bank staff, the railway authority's investment planning has greatly improved to include con- cepts of systematic comparison of alternatives and establishment of year-by- year priorities, and the transport coordination agency is at least in opera- tion and carrying out large studies. But this is rather little to show for so much effort. One of the problems has been shortage of competent local staff to work with the consultants and another, political indecision, resulting for instance in the transport coordination agency, which the original consultants were supposed to help get started, being finally set up only 18 months after they commenced work. But also the borrower found the consultants responsive only to the Bank and not to their own problems; it was a major undertaking, as mentioned, to eventually reconcile the investment plans of the two. Given the results to date, a question must be raised, even though it cannot be categorically answered, whether it would not have been preferable to step down from the very comprehensive effort insisted upon by the Bank, to start with a few individual studies of particular lines and other proposed invest- ments, as suggested by an individual UNDP consultant whose conclusions were rejected rather out of hand by the Bank but who seemed to have more in mind - 24 - the need to gradually develop local capabilities, while at the same time coping with a few immediate concrete questions. A smaller consultant team working more closely with employees of the railway authority might have had more significant long-term effect. 4.16 This case, with its poor relations between the consultant and the intended borrower and great delay in reaching an investment program acceptable to the latter, does raise a question about some consequences of the change that was made in 1967 to have all contracts, under UNDP studies for which the Bank was Executing Agent, signed between consultant and Bank, instead of between consultant and the local agency whose project was being prepared as had been the practice up to that time. It was believed that the latter arrangement had proved too 'artificial' and confusing for the consultant and that direct responsibility to the Bank would expedite arrangements and strengthen control, without taking additional Bank staff-time since regular supervision of the consultant would still be in the hands of the local agency. Check of a small random sample of pre- and post-1967 contracts for UNDP- financed studies and a canvas among Bank staff suggest that average time to get the consultants started has indeed fallen from about 11 months to 7, but that more Bank staff-time is typically required under the new arrangements, while quality of reports may be somewhat better. Yet if the experience of the case studied (the contract was signed in 1969) is at all representative of the situation under the new arrangements, it does suggest that in those cases where the consultant is expected to make an important training and institution-building contribution, ways should be sought to re-establish a more direct relationship between the consultant and the borrower, perhaps along the lines of the Bank's approach nowadays to the major water resource planning exercises in which it is involved (e.g., in Mexico and Egypt). 4.17 The investigations made also support the validity, now generally accepted, of certain other steps of more general applicability for helping to increase local involvement in project preparation: for instance, early establishment of a committee or unit, if one does not already exist, to carry the project through preparation and into execution. Worthwhile, even if not fully effective, interdepartmental committees were established at an early stage, partly at the Bank's advice, in connection with the tourism project and for transport coordination in the country with the railway project. A population borrower, responding to the questionnaire, finds the main factor which delayed the preparation of the project to have been the lack of a group established by the Government early on to work with the Bank missions and maintain a realistic local emphasis. A rural development borrower emphasizes the advantages that would have accrued from appointment early in project preparation of the project manager. The Project Performance Audits underline the useful role, perhaps special responsibility, that the Bank could have fulfilled, in supervising consultants, generally appointed at its suggestion, in countries without experience in that and with very limited staff; inadequate allowances for inflation in cost estimates in one case and use of excessively high design standards in another are the kind of problems which it is believed such closer supervision could have avoided. Finally, the difficulties encountered in project preparation due to lack of basic data are a sharp reminder of the importance of sometimes neglected data collection systems (hydrological, agricultural, etc.) in broader institution-building efforts. - 25 - 4.18 But the question also arises whether the Bank should not consider assisting more directly in the creation and development of the kind of project planning units that are required for more effective project preparation, especially in fields such as agriculture, rural development and education. Review of a large sample of appraisal reports for projects approved in FY 1974 suggests in fact that there may now be some imbalance in the assistance provided by the Bank Group to the development of project planning institutions, and a degree of inconsistency with the Bank's current objectives. For all 28 projects in the transport and urban development fields the reports assess, usually quite fully, the adequacy and development of the sectoral planning institutions and in as many as 23 cases Bank/IDA financial assistance is being provided, usually in the form of financing for consultants, while loan/credit conditions in these and other cases quite frequently refer to the creation of planning units, preparation of sector plans, establishment of planning criteria, foreign training of planning staff, etc. By contrast, for the 29 projects in agriculture and rural development, only 4 of the reports discuss the sectoral or sub-sectoral planning arrangements (although they generally do discuss other sector-wide aspects such as agricultural credit and research) and only 19 can be considered to include provisions that, in a wide interpretation, may con- tribute to future project planning -- e.g., financing for project evaluation units, usually with responsibility for preparing a follow-on project, use of an intermediary channel (such as an agriculture credit bank) or financing for further feasibility studies by consultants. Agriculture and transport are of course quite different in their institutional set-ups, but the extent of the contrast described is nonetheless striking, especially when set alongside the evidence that, partly thanks to past Bank efforts, some countries are now more capable at project development in the field of transport than any other. The apparent tendency for some agricultural Phase II or Phase III schemes to have been embarked upon a little too early behoves a cautious approach to reliance on project evaluation units for generation of new projects. The readiness of the Bank in recent years to lend for the costs of local consul- tants suggests that it would be no major policy departure to contribute to financing the operations of general units for planning of agricultural projects, a seemingly very high priority area for future development. Bank Rigidity/Flexibility 4.19 A perennially important question in almost any phase or instance of the Bank's operations is "how rigid should the Bank be?", and this review raised various issues essentially illustrating this dilemma and suggesting that delays in project preparation had sometimes arisen from the Bank erring on the side of excessive rigidity -- either being more insistent about partic- ular technical questions and prescriptions than seemed worthwile in retrospect or applying too liberally and literally standard procedures and patterns. The cases where these issues arise are not particularly numerous, but they are sufficiently significant to warrant awareness on the part of the Bank and consideration of any means that might be found to reduce the problems. 4.20 One type of case is the long drawn-out technical argument, with unconstructive results, such as affected so seriously the early work on the fertilizer project included among the case studies. An additional complica- tion in this instance was that it was largely a matter of the difference between processes employed in the equipment of alternative manufacturers, - 26 - with one of whom the prospective borrower had a long relationship, while the Bank appeared to draw its information from the other; neither Bank nor borrower seems to have made much effort to cross-verify the validity of the information it was obtaining from the parties at interest, there was con- siderable mistrust of one another, and it was not until a very late stage that it was agreed to bring in the outside professors to make the best direct comparison they could. Somewhat similar prolonged technical arguments arose in connection with some of the projects covered by the performance audits -- for instance, the argument over the extent of very hard rock that might be encountered in dredging work for the African port project mentioned earlier, finally resolved only by withdrawal of the bidder who was best prepared to cope with such rock (which did not in fact turn up in the actual works), and another one-year argument over the need for an additional thermal unit in a power project in Latin America, finally resolved by deeper Bank understanding of the situation and recognition, in particular, of the low capacity factors of the planned hydroelectric schemes and the consequent need for complementary thermal energy. 4.21 As most delays of this kind occur in small increments of time, there is rarely a Bank perception that the sum of these increments could after a while constitute an unreasonable cost for the additional benefits being sought. One check to the duration of such arguments would probably be to require more systematic comparison, in the Bank, of the value of the potential improvement proposed with the cost of delay in the overall project, with a view to reaching an order-of-magnitude idea of the amount of time it is worth devoting to the matter. But this has certain game-theory problems, of bargaining and counter- bargaining. Moreover, it would not have helped much in the case of the fer- tilizer project, insofar as the annual operating cost savings of $2.5 million believed by the Bank to be obtainable by using the process it favored were sufficient to warrant, on the basis of such a comparison, delay in the whole scheme of about one year at the then price of fertilizer; whereas reality, as far as it is known now, indicates a cost, not a saving, to the Bank's then preferred process and greatly higher costs of delay, due to the rise in fer- tilizer prices. What may better be said, perhaps, for each of the instances mentioned in the last paragraph is that the Bank should have made more inten- sive and objective efforts, earlier, to verify the validity of the case it was to sustain for so long -- for example, by trying to get an early independent cross-check of the data it was being fed by the equipment manufacturer it knew, naturally somewhat biased in favor of its own process. 4.22 But it must also be admitted that it is hard to find from the projects reviewed by the Operations Evaluation Department so far an instance where the Bank has adopted a very strong position on a technical issue, succeeded in convincing the borrower after many months of argument, and been clearly proven right in retrospect; frequency unification in the Mexican power sector might be considered such a case, but then the Bank has by no means adopted an uncompromising attitude on this over the years, and the matter is not really one of detailed technical improvement but of basic organizational structure, of the institutional type on which the Bank has indeed been helpful in a number of countries. If this is the case, then it clearly suggests that the Bank should clearly and firmly express its views on such issues of relatively detailed technical improvement to a project but not - 27 - pursue them, if firmly rejected, for more than quite a limited amount of time, here respecting the responsibility of the prospective borrower to know better his own particular local situation and constraints and reaching its own decision as to whether the project, in the form desired by the borrower, is too risky to warrant Bank Group financing. At all events, it is important that Bank managers ensure that issues between Bank and borrower which are costing much delay and many man-hours get raised early to higher levels of the hierarchy, for more open discussion, rather than leaving them, as appears to have happened on occasion in the past, to one or two individual technical specialists. 4.23 Calling for an economic comparison study such as the Bank requested of the fertilizer plant borrower, but never in fact obtained from him, is one fairly common means to reducing the risk that a proposed solution is not the most economic. A slightly disturbing feature of the results of the questionnaire survey is that such special economic comparison studies, when called for by the.Bank, often at a fairly late stage, generally seem to have been considered in retrospect not to have contributed significantly to the eventual project. The 10-15% of questionnaire answers relevant to this issue and brief review of related Bank files suggest that the studies did no more than confirm, considerably more elaborately and sometimes at significant cost in terms of consultant fees, judgments already reached during the earlier feasibility studies -- for instance on the siting of educational institutions, the chemical composition of complex fertilizers, the choice of airports to be improved. This does not of course mean that the studies were useless; they did help to reduce risks as perceived by the Bank. But specially added studies which never result in a change in the projects to which they refer do engender doubts about their usefulness. The fact that moreover the borrowers did not seem to set much value on the risk reduction implied does raise a question whether the Bank, in the interest of both efficient project prepara- tion and effective building of borrowers' planning capabilities, may not need to be somewhat more selective in the special economic comparison studies it calls for late in the project preparation cycle, making sure that the issue is clearly a practical one which cannot be handled by experienced judgment and back-of-the-envelope calculations alone and really does need a further study to resolve. 4.24 In about 15% of the questionnaire answers, borrowers said that excessive emphasis was given by the Bank to form instead of substance; they felt that the outside help, whether from the Bank or more often from the Cooperative Programs or Consultants, helped not so much to change in any way the substance of the project conceived locally as to dress it in the form desired by the Bank. In the fertilizer project studied in depth, the Bank seems to have been rather more rigid than necessary about the specific form in which the data had to be presented, not making as much use as possible of the information system that the borrower already had. However, one or two borrowers also saw this emphasis on form as being useful, not only to obtain the external financing, but also, with modification to local circumstances, for possible wider use locally. - 28 - 4.25 About 10% of the respondents found excessive the series of missions from the Bank or the Cooperative Programs that they had to receive in connec- tion with the preparation of a project -- they felt some of these missions duplicated one another's work or that of the Bank's economic missions -- and expressed the view that fewer Bank missions would actually have helped project preparation if the first one had more clearly explained what the Bank needed. However, a larger number of the respondents, nearly 20%, clearly considered that either longer or more frequent Bank missions would have helped project preparation. 4.26 As emphasized, these problems of excessive emphasis on form and rather rigid missions schedules were perceived to affect only a very small minority of projects overall. But there is one sector where they seem to have arisen with particular frequency, unfortunately one not covered by any of the detailed case studies. All five borrowers for education projects who responded to the questionnaire referred either to problems of this sort or to studies that were felt to be overly elaborate or partially but excessively repetitive. The Bank has been experimenting with new patterns of preparation of education projects -- such as proceeding straight from an initial identifi- cation mission to a joint preparation/appraisal mission, followed as necessary by a brief further mission to confirm details, and this approach does seem to have expedited project processing in some cases where it has been applied. But it may well be useful to carry out a broader reassessment of the Bank's information requirement for education projects and of how responsibility for meeting them can best be divided between borrowers, UNESCO, and Bank economic and project missions. 4.27 A question of much more general relevance, referred to earlier in connection with the misunderstandings that have sometimes arisen due to lack of clarity on the part of the Bank, is the standard and degree of detail and accuracy to which the Bank requires engineering and similar studies to be brought before appraisal or loan approval or commencement of execution of works. The case of the multipurpose project suggested that the Bank may have been unduly rigid, at appraisal, with regard to the standard of land and soil surveys required before it would consent to finance any work on the distribu- tion system. However, this case seems to be more the exception than the rule. We have not found evidence that the Bank is generally requiring excessively high standards of project preparation, and there are countries where the Bank has clearly made a significant contribution in helping to introduce standards of project preparation which have helped to avoid uneconomic projects and to ensure more efficient execution of worthwhile schemes. 4.28 In some sectors, especially those involving the construction of many standard buildings, the Bank may indeed need to be a little stricter in regard to the degree of advancement of design work reached by the time of loan approval, to avoid the quite frequent phenomenon of long periods of low dis- bursement in the early life of a loan and unnecessary accumulation of commit- ment charges. One of the several important uses of the Bank's Project Preparation facility,!/ supplementing present retroactive financing facilities, will no doubt be to help here. 1/ See Memorandum from the President, "A Facility to Assist in the Prepara- tion of Bank-Financed Projects", distributed as Board Document R75-224 dated November 17, 1975. - 29 - 4.29 One of the main sectors, as mentioned earlier, where the Bank has for many years been trying to build up the level of design reached before loan appraisal/approval is highways, due to poor earlier experience with large cost and time overruns on approved projects and also the relatively low cost of detailed design in this case and importance of it for determining appropriate construction technique in a field which offers considerable flex- ibility. The result is that highway projects are now generally more advanced than other types of projects when loans are approved, and more advanced than they would be at this stage if the finance was being obtained from other sources. The various ex-post studies done to date on highway projects for which detailed engineering was not available before loan approval definitely do not suggest any lower ultimate economic rate of return to such projects; rather, the Bank's practice does seem to have reduced cost and time overruns over appraisal projections, and may have led to selection of more appropriate ultimate design standards, more precisely agreed, which should tend to improve economic performance of the investment; nor is it clear that completion of detailed engineering before loan approval necessarily delays loan approval by a corresponding amount of time, for earlier projects sometimes had an equally long 'preparation period', extended by discussion of other matters, simulta- neous with which the detailed engineering might very well have proceeded. Indeed, where delays in loan approval result from effort to resolve critical project problems, they may result in accelerating actual project implementa- tion. Nonetheless a number of the questionnaire respondents do raise this problem, implying that what they considered Bank rigidity on this matter delayed project progress. Actually, however, the real problem seemed to be less that the Bank's standards were considered too high than that construction of particular roads had to wait after completion of detailed engineering for them, pending the completion of detailed engineering for other roads to be included in the envisaged Bank loan -- and one borrower states, for instance, that this was done in order to put together a large enough project that it would, in the Bank's opinion, attract international contractors, who then in fact failed to materialize. This problem of delays to certain particular works, awaiting readiness of others not technically related but included in the same loan, can sometimes be solved by retroactive or advance financing (as it was in one case covered by the questionnaire survey), by the inclusion in a loan of a flexible component that may be devoted to roads subsequently agreed with the Bank (as in the last Mexican highway loani!) or by adoption of the sector lending approach described in a recent policy paper.17 An alternative solution is of course to make smaller loans, related in a more timely manner to the readiness of particular components of projects to go ahead. However, limits to this are imposed by the costs to the Bank of appraisal processing (which have led to the general standard of loans to a particular borrower not generally more frequently than once every two or three years) and the demands of a project's own integrity. 1/ Loan 968-ME of March 1974. 2/ "Highway Sector Lending", SecM75-487, June 1975. - 30 - 4.30 In the very large majority of cases, pre-loan institutional requirements by the Bank appear to have been soundly conceived and useful -- indeed the Project Performance Audits identify a few cases where more should have been done, even at the cost of some delay. But occasionally the Bank appears to have been too ambitious, apparently thinking to resolve a large number of quite fundamental issues at one swoop, with the aid of a loan large enough to provide much "leverage" but by consequence also generally very delayed. The railways case studied for this review and a steel mill project analyzed earlier both throw considerable doubt on the view that Bank loans should be designed to address as many problems as feasible because "leverage" in practice disappears after a loan is signed. They suggest that the Bank could have helped more effectively in the long run by starting with a smaller earlier loan, with much more limited objectives and for selected works of clear priority whatever the larger longer-term plan might be, thereby working with the borrower to get to know better his problems and constraints. One particular problem of the very delayed approach is that it has sometimes, as in the case of the railways project, failed to recognize the urgency with which prospective borrowing agencies have to try to resolve some immediate problems, so that decisions get taken long before the studies which were supposed to assist them are completed. 4.31 A more general query relates to the depth to which Bank appraisal questions should be pursued. The elaborateness of the answers sought, which effectively determines much of the preparation work necessary as well as the periodicity with which loans can be made, could be differentiated even more than now among different situations and projects. A few of the responses to the questionnaire survey raise this issue in a particular way, which also finds some reflection in the Project Performance Audits: in the case of some 'repeater' loans, it seemed uneconomic to have spent so much time and effort going over the general sector and program background, such as overall educa- tion sector financing, or transport regulation and coordination, or full regional development framework, where the sector strategy which the project supported remained unchanged; but there have also been repeater loans which seem to have needed even more thorough review than they actually received. What therefore seems to be needed is an early determination of the depth to which particular questions should be pursued. 4.32 When, as is now usually the case, loans are seen as one of a series, there may be scope for a variety of approaches to the main issues to be addressed. One, relevant to a situation where the technical and institu- tional base is weak and where the correct solution (e.g., in agriculture or education) is not yet clear, may be to start with a simple, relatively narrowly focused project and appraisal thereof, with a view to increasing the breadth and complexity of preparation and appraisal over time, as the wider relevance of the solution, and its implications, become clearer. Another, for a situation where the program to be followed could be already quite precisely defined or where much freedom of action would be left to a respon- sible intermediary, might be to start with a major preparation and appraisal effort for the first loan, which could then be followed at relatively short intervals, of one or two years, by further financing, with lesser appraisal effort provided the program framework remained largely unchanged and key performance targets were being fulfilled. In those cases where this is not already happening, projects should be designed to support a continuous, progressive, joint institution-building effort by the borrower and the Bank. - 31 - V. CONCLUSIONS 5.01 Project preparation, the phase in the project cycle when a project is moving from basic conception to readiness for appraisal, is a critical stage in the life of a project, which presents the main opportunity for avoid- ing the problems that have been found in projects so far evaluated ex-post and many of those that take the time of supervision missions. A major con- clusion of this study is that the Bank does not appear to have been generally either too slow or too perfectionist at this stage. Although the project preparation period, defined from 'Bank conception' to (first) appraisal is quite long, at a median of two years for all projects and two-and-a-half years for 'first' projects, it is not unreasonably long given the complexities of the logical process of properly defining a problem and then gradually narrowing down alternative solutions to one best particular scheme, defined in sufficient detail to be implementable. The fact that the typical preparation period for projects financed by the Bank Group appears to have been rising may be con- sidered good to the eKtent it reflects earlier Bank involvement in this process, more thorough preparatory studies on aspects earlier neglected, and fuller responsibility of the borrowing countries' agencies, as opposed to short-term consultants or others, in the preparation of projects. 5.02 However, there are ways in which project preparation might be expedited without sacrifice in fundamental project quality. The preparation process has sometimes been more roundabout than would have been necessary due to lack of clarity on the part of the Bank at an early stage as to what information and studies it would require in the particular case, resulting in excessive backtracking and need of supplementary studies, and occasional over- insistence by the Bank on some technical point. There seems to be scope for cutting repetitive work in the preparation of some projects, especially loans in support of an unchanging sector strategy. Although increasing the role of the borrowers' staff in project preparation may often lengthen the process in the short run, the overall evidence is that projects prepared by Government agencies and other borrowers themselves have the shortest preparation periods; but the Bank's efforts to strengthen local project preparation capacity have been mixed and the results have also suffered from lack of political commit- ment to this kind of work in some countries. 5.03 This study suggests that there is scope for cutting project prepara- tion time for some projects. But it is very doubtful that a major change -- like a reduction of the median from two years to one -- would be possible without reducing current Bank project standards excessively. The detailed case studies done for this report, covering two projects with very long pre- paration periods and two with preparation periods close to the median for 'first' projects nowadays, suggest that preparation time might have been cut in each case by about 25%. But such a sharp reduction does not seem feasible overall in light of the greater complexity of projects which the Bank is handling today and the more thorough treatment it affords them, generally usefully. The most that seems to be feasible, based on all parts of this study including the survey of 1974 borrowers, is something more like, say, a 25% reduction for 25% of the borrowers, 10% for 50% and 0% for 25%, implying - 32 - an overall reduction in the median from 24 months to about 21.! The study also suggests that there is scope for improving the quality of project preparation, without increasing the time required. 5.04 But the central conclusion of this report is that the cliche 'Preparation is the borrower's responsibility' needs to be underlined and applied even more rigorously by the Bank. To enable the borrower to fulfill that responsibility, the Bank, concerned like any lender about the risks to which its funds are put, has a corresponding responsibility to inform the borrower as early and as precisely as possible of what standards of design, and also of study and documentation, it will require. For most first-time borrowers and also many older borrowers, the Bank, as a development institu- tion as well as a lender, should actively help the borrower at an early stage to work out a closely cut plan of project preparation, with benefits to both the speed of project preparation and to maximizing the work to be handled by the borrower's own staff whenever borrowers are able to assume this respon- sibility. For the large number of borrowers not yet in a position to carry the main responsibility for preparation of their projects, the Bank should give priority emphasis to their development of units for planning projects, especially in agriculture, and should consider supporting them more than it now does through its loans and credits. 5.05 Increasing the speed and effectiveness of project preparation is not a matter of Bank programing or monitoring. Procedural delays by the Bank, like slowness in sending letters or missions or in selecting consul- tants, do not seem to have been a factor of any significance and, particularly in recent years, the Bank seems to have been involved in projects at a suf- ficiently early stage in their preparation. The amount of work involved in project preparation and the time required vary enormously between projects, generally differing more within any one group, such as a region of the world or an economic sector, than between groups. Standard prescriptions and norms are impossible. The kind of residual role that the Bank has, and needs to have more, as plans for preparation of specific projects are more fully managed and increasingly implemented by the borrowing agencies themselves, cannot be precisely programed. Merely increasing the proportion of Washington paperwork would in fact have more of a negative than a positive effect on project preparation. 5.06 The key element in improvement will rather be a matter of giving experienced and committed Bank staff, with a good deal of delegation of authority, increased time and scope to exercise the essentially entrepreneur- ial and promotional functions involved in this phase of Bank activity more than any other. The case studies underlying this report repeatedly illustrated the crucial importance to the whole project preparation process of how well the key issues were identified by the initial Bank missions and, secondly, of how much time they had to work with the borrower on a practical plan for dealing 1/ For comparison, the analysis in Delays in Loan and Credit Effectiveness, July 22, 1975 suggests that the time between loan signing and loan effectiveness could be cut from a recent average of about five months to about four. - 33 - with the issues; neither of these tasks, consisting mainly of work directly with the borrower in the field, can be effectively handled by direction. Hence the most important conclusion of this study is that, simultaneous with or just after project identification and before feasiblity studies start, experienced Bank staff assist identification of the key issues and help tailor the investigations to the decisions to be made, building in a logical sequence and agreeing quite precisely with the borrower on both the broad design stan- dards and the specific documentation standards aimed at. Only such a custom- made project preparation plan, identifying the specific places where outside help will be required, can hope to cope with the complexity of the process. It is believed that an effort of this quality at this stage in the life of a project would significantly reduce the need for extended subsequent missions, as well as improving the project preparation process; the object is a more timely Bank participation, to reduce the need for heavy direct Bank involve- ment throughout the preparation period, even though periodic advice on changes in plans will still be needed. 5.07 To support the staff carrying out this kind of assignment, to facilitate their work and enable increased delegation of authority to them, as well as to simplify and clarify the project preparation task of prospec- tive borrowers who need little of the assistance described, the Bank should give higher priority than it has in the past to maintaining precise and up- to-date guidelines, addressed to the borrowers, on what data and justifica- tion it requires in support of projects in the various sectors and sub- sectors for which it lends. The visiting Bank staff members would go over the relevant parts of these guidelines in detail with the borrowers and amend and adjust them to suit the particular project in view and cover the issues that they had identified as central to the case. 5.08 In some sectors in some countries, work of the sort described is seriously hampered by the lack of an effective institutional base for project planning, and a necessary component of making the proposed approach fully effective would be gradually filling this gap. The Bank has made significant contributions in many countries to the gradual development of such a base in the transport sector, while in other countries its efforts in this same sector have met with limited success due to relatively low political and budgetary priority assigned to the matter in the countries. In the latter set of circumstances, its sustained support may nonetheless prove equally if not even more important in the long run. It would be highly complementary to the borrower-focused approach to project preparation described, for the Bank to extend its systematic attention to the project planning function more fully to other-sectors in which it is involved and to consider more extensive technical and financial assistance for such sectoral project planning units. 5.09 While it is hard to obtain a comprehensive view of how the Bank's various Regional Offices, in their different circumstances, are today handling their responsibilities in connection with project preparation, it is believed that the main emphases of this report are consistent with current trends in most of them and are in fact already implemented, in most respects, in some cases such as the Bank's RMEA in Nairobi. The Project Preparation facility whose establishment was approved by the Executive Directors in December - 34 - 1975 should provide a useful element of flexibility to facilitate the financing at short notice of studies and start-up expenses in connection with individual projects and could help timely implementation of particular elements of the project preparation plans already mentioned. Another potentially important mechanism, a periodic Project Brief to summarize the status of a project and related work during the preparation period, is being introduced on an experimental basis. If fully shared in substance with the prospective borrowers, as intended, it should help effectively to focus attention and expertise on substantive project issues, or questions arising in connection with Bank policies, at an early date when steps can be taken to deal with them without delaying project preparation and, by providing an element of continuity over the preparation period of a project, it could help maintain progression and discourage the introduction of new, delaying elements at a late stage. In due course, the Bank should update and elaborate its various project preparation guidelines and introduce supporting changes in Bank practices where needed in respect of financial and technical assistance to planning units. 5.10 The OED believes that full implementation of the basic approach to project preparation and Bank assistance thereto sketched in this report should, after an initial transition period, have no significant net costs to the Bank's administrative budget. Although it would help to raise the quality of project preparation as regards both physical components and institutional capabilities to implement them and use them to best advantage, it would also make the process more efficient (less backtracking, sideways movement and supplementary studies, implying less later preparation, pre-appraisal and supplementary appraisal missions and related work) and more dependent on the borrower. It should also help reduce project slippage and thereby the margin needed between gross and net lending programs. The stronger initial missions to help prepare detailed project preparation plans would be able to accomplish their work quickly with borrowers more familiar with the Bank and Bank standards; where they were longer, as they might be for some new borrowers and complex projects, requiring much preparation work, they should bring cor- responding economies in the later stages when the Bank would be going over the preparation work, to appraise and eventually approve a loan. Other sug- gestions made, such as additional assistance to institutionally weak borrowers in the supervision of consultants, are unlikely to have significant cost implications for the Bank as a whole. 5.11 The following pages summarize the three groups of suggestions that emerge from this study: the main operational elements that might go into the comprehensive project preparation system which we consider to be needed; four important subjects which we believe warrant further consideration by Bank management; and, several more particular matters that should be considered. All are intended to help the Bank provide more effective support to borrowers' project preparation efforts. - 35 - I. Main Elements of a Pro_ect Preparation System The project brief system, which has just been introduced in the Bank on an experimental basis, is intended to become the foundation of a system to identify project preparation issues early and ensure timely agreement with the borrower on them and how they should be addressed. It would thus appear to be addressed to the fundamental concerns raised by this study. The system will be reviewed toward the end of 1976 with a view to establishing it, with such modifications as experience suggests, on a Bank- wide basis. The conclusions of this study suggest that whatever system is finally introduced should include the main elements listed below: 1. A project timetable system which ensures that adequate forethought is given to the Bank resources that will be required for project preparation support to borrowers and the appropriate timing of this support. 2. Sectoral project preparation guidelines which provide borrowers with a comprehensive and specific guide, adjusted to each specific case by Bank staff in con- sultation with the borrower, as to what the Bank requires in support of a project, e.g., number and type of studies, data, maps, level of engineering design, etc. 3. Timely review by experienced Bank staff with borrower, well before feasibility studies start, of the main project objectives and the arrangements needed to achieve them, and agreement with the borrower on a project preparation plan which specifies precisely who would be responsible for doing what, and when. Special project inception missions should be organized wherever this review cannot readily be carried out in other ways. 4. Timely review within the Bank of issues which emerge during the implementation of the preparation plan, to ensure that they are resolved without undue delay and that major changes in the project or the preparation plan are made only in response to fundamental questions about the project. Although some of these elements are already in place, the OED has offered detailed suggestions, constituting one approach to the design of a comprehensive project preparation system, to the Central Projects Staff who are responsible for designing and overseeing the administration of such a system. - 36 - II. Subjects for Further Consideration 5. Appraisal Treatment of Borrower Project Planninr Capabilities: This review strongly reemphasizes the importance of borrower planning units and basic data collection mechanisms andfinds wide differences be- tween the extent to which project appraisals in different sectors give attention to these aspects; those for Transport and Population projects are par- ticularly strong and those for Agriculture projects (and, to a lesser extent, some other sectors such as Electric Power and Tourism) are often weak. Of course the various sectors are structured differently in member countries and face different problems, but it would be worth aiming at more consistent attention to this subject and seeing how the strong Bank experience in analysis of planning competence in transportation -- the criteria to be used, aspects to be emphasized, rate of progress to be expected, types of assistance that work best, etc. -- could be carried over to the other sectors where the planning function appears to be presently given less attention (para. 4.18). More generally, where loans/credits include funds for consultant project preparation work, it would seem desirable to have appraisals justify this more fully than is now the case in terms of the specific deficiencies of the borrower's plan-ing unit and the on-the-job training or other assistance the consultant will provide to strengthen it (para. 4.14). 6. Financing_Project Planning Units through Loans/Credits: In view of the importance of faster growth of project planning capability in the agricultural sector and in rural development, it seems to warrant consideration whether the Bank could not usefully extend the consider- able financial assistance it has given to the development of transportation planning institutions (principally through financing consultants and training) to including components in agricultural project loans/credits to cover the operating costs of more general agricultural and rural development project planning units, as it already some- times helps finance evaluation units to prepare specific follow-on schemes to projects earlier financed in these fields. Such Bank support could help give added priority to a type of work that tends to suffer severely in Govern- ments' short-term budget cuts (paras. 4.14 and 4.18). 7. The Depth to which Particular Appraisal Issues Should be Pursued: Since so much of Bank/IDA lending is for series of projects for one entity or program, consider- ation should be given to making an early determination - 37 - of the depth to which particular appraisal questions should be pursued, to avoid unnecessary time and effort in project preparation. One dimension of this question relates to the ambitiousness of objectives for the early loans in a series; under certain circumstances smaller and more timely commitments for simpler projects with more limited objectives may reduce the risk of delay, inade- quate borrower performance and the cost of failing in an experimental approach to particular sector problems. Another dimension relates to avoiding unnecessarily full updatings of the general sector program or institutional context whenever these remain substantially unchanged. Such early determination would help project preparation missions recommend the appropriate depth of preparation work on different aspects of the project (paras. 4.30- 4.32). 8. Data Requirements and Mission Series for Education Projects: Procedures for Bank assistance in prepara- tion of education projects and the data required of prospective borrowers seem to require special review and reassessment since all the education borrowers covered in the questionnaire survey of the Bank's 1974 borrowers appeared to find them partially unsatisfactory and excessively elaborate (para. 4.26). III. Particular Points 9. Continuity in Bank Advice on Project Preparation: Wherever circumstances cause the Bank to reduce severely or interrupt lending operations in a country, an inde- pendent determination should be made as to whether and to what extent there should be a corresponding inter- ruption in Bank advice on project preparation (especially supply and discussion of abovementioned guidelines and assistance in formulation of individual project prep- aration plans), for substantial gaps in project prep- aration advice will most probably imply substantial later gaps between the decision to resume active oper- ations and the ability to do so (para. 4.09). There may, of course, be circumstances in which the potential for constructive response to advice on work needed to pre- pare specific projects is so low that there is no basis for endeavoring to maintain even a minimal continuing Bank involvement in project preparation. 10. Consultants' Responsibility to Borrowers: Where consultants are expected to make an important training and institution-building contribution, ways should be sought to re-establish a more - 38 - direct relationship between the consultant and the borrower than is provided for under the present arrange- ments by which the Bank recruits and manages consultants as Executing Agency for UNDP study contracts (para. 4.16). 11. Special Economic Comparison Studies: It appears that the Bank should be particularly careful and selective in calling late in the preparation cycle for special studies making economic comparisons of alternatives, and ensure that the study is either essential to risk reduction or likely to improve project design significantly (para. 4.23). 12. Operational Manual Statement on Project Preparation: In due course, it would seem desirable for the Bank to prepare, for the information of Bank staff and borrowers, an Operational Manual Statement on Project Preparation and the assistance the Bank can provide for this, integrating the scattered references presently to be found in separate statements dealing with Technical Assistance, Consultants, and UNDP. WORK EFFMOT 1974 (Man-ye.rs) EAST AFRICA UEST AFRICA EAP STHENA I.AC CPS ANK A CPS (fAW) Nork Effp I. Preparation WUrk Efforq (%. of PAR) 27.8 (45.1) 26.7 (42.3) 36.3 (58.8) 26.3 (48.0) 23.9 (31.6) 34.0 (46.0) 29.3 (34.9) 204.3 (42.71 Excludes UNDP but includes; (1) Cooperative Program with UP Agencies; (a) PAO 5.0 3.5 14,0 14.5 4.5 12.4 1.1 55.0 (b) W11 2.0 0.5 - 2.2 1.9 1.4 0.2 8.2 (c) UNIDo . - - - - 2.0 2.0 (d) UNESCO 3L 1.5 2.2 LL .1 4.0 0.1 ZO. (a) SUB-TUAL 10.8 5.5 16.2 18.4 13.5 17.8 3.4 85.6 (1I) Consultants 1.3 2.0 2.0 1.9 2.3 1.9 5.6 17.0 (Lit) Bank Staff (M of PAN) 15.7 (25.5) 19.2 (30.4) 18.1 (29.3) 6.0 (10.9) 8.1 (10.7) 14.3 (18.5) 20.3 (24.2) 101.7 (21.3) Bank Staff Supervising UNDP Projects 1.8 1.9 1.7 1.5 2.3 2.5 2.8 14.5 Total Bank Staff tork Effort (1 of PAlR) 17.5 (28.4) 21.1 (33.4) 19.8 (32.1) 7.5 (13.7) 10.4 (13.7) 16.8 (21.7) 23.1 (27.5) 116.2 (24.3) No. of Projects Under Preparation 74 63 60 64 127 135 89 612 Total Bank Staff Uork Effort per Project .24 .34 .33 .12 .08 .13 .26 .19 Preparation Urk Effort per Project .38 .42 .61 .41 .19 .25 .33 .33 2. Appraisal Vork Pffort (. of PAR) 30.0 (48.7) 31.2 (49.4) 21.7 (35.2) 25.2 (46.0) 44.2 (58.4) 37.5 (48.5) 45.8 (54.5) 235.6 (49.3) vhich includes; (1) Consultants 3.2 5.1 5.1 4.7 5.7 4.9 14.3 43.0 (it) Bank Staff Effort In Appraisal (2 of PAN) 26.8 (43.5) 26.1 (41.4) 16.6 (26.9) 20.5 (37.4) 38.5 (50.9) 32.6 (42.2) 31.5 (37.5) 192.6 (40.3) me. et Projects Uhder Appraisal 39 47 38 31 55 59 58 327 BSk Staff Effort In Appraisal per Project .69 .56 .44 .66 .70 .55 .54 .59 AvatIe Vark Effort per Pmeject .77 .66 .57 .81 .80 .64 .79 .72 3. Negotiation & Board turk Effort (1 of PAN) 3.8 (6.2) 5.2 (8.3) 3.7 (6.0) 3.3 (6.0) 7.6 (10.0) 5.8 (7.5) 8.9 (10.6) 35. (8.0) Nmber of Projects 29 30 28 18 30 ,37 43 22.3 MCD Work Effort per Project .13 .17 .13 .18 .20 '16 .21 .17 4. Prepgr2rhtD Apnratst aml Jantations (PAE) Urk fa. 61.6 63.1 61.7 54.8 75.7 77.3 84.0 478.2 5. Project "ugervi"lg4 20.0 18.4 22.1 17.5 28.6 28.5 27.2 162.3 6. Sector Work 11.5 10.6 8.5 13.8 14.0 15.8 13.0 190.1 7. Economic Mork 6 C P's 13.8 16.0 10.8 11.9 21.1 27.3 8. other Output 101 10.2 15.7 16.6 21.6 21.2 57a1 152.5 9. TOTAL MORK ( of Total urk Effort) 117.0 (11.9) 120.3 (12.2) 118.8 (12.1) 114.6 (11.7) 161.0 (16.4) 170.1 (17.3) 181.3 (18.4) 983.1 (100.0) ANNEX II Page 1 QUESTIONNAIRE REGARDING 1974 PROJECTS 1. Do you recall when the project was first broadly conceived and could you roughly delineate the main stages of project preparation through which it went? When was the project first brought to the attention of the Bank and approximately when did it reach the following stages (please record all applicable): Government approval ( ); Pre-feasibility Study ( ); Feasibility Study ( ); Preliminary Engineering ( ); Detailed Engineering ( )? 2. Of the various studies related to the project that have been undertaken, were there any which were done specifically because Bank Group financial assistance was sought and which would not have been done otherwise; and, if so, which? 3. Considering all studies and preparatory work done, whether or not specifically with a view to Bank Group financing, which, if any, do you feel in the end made little substantive contribution to the formulation of the project (physical and institutional) as it is now being carried out? 4. Do you think that studies could have been commissioned and requisite consultants appointed more rapidly than they actually were, and, if so, how and by how much? 5. Was there any delay in project preparation caused by what seemed to you to be differences in approach or requirements as between the Bank Group and other aid-supplying agencies, UN agencies, foreign consultants, foreign advisors, etc. involved in the preparation or financing of the project? 6. If the project was not prepared entirely by nationals of your country, how much local participation was there in its preparation, and do you consider that more would have been possible and would have led to earlier execution of an equally effective project? If so, how might it have been best assured? 7. Were there any Bank Group requirements for administrative or insti- tutional or technological changes (for instance creation of a new agency or authority or new system of accounting) which in your view were unnecessary, or not justified in terms of the extra time required to comply with them? 8. If you consider justified the changes proposed by the Bank in these or other matters, do you feel that the Bank might have proposed them earlier in the process and might thereby have avoided some or all of the delays? 9. Would earlier, more frequent or longer Bank staff visits have helped, in your opinion, to bring the project more quickly to the stage of financing? ANNEX II Page 2 10. Did the processing of the project suffer delays because of the issue of "land acquisition", how much, and do you see any way that this issue could have been avoided or solved sooner? 11. If for whatever reason the project was actually begun significantly later than you had originally hoped, are there any parts of the eventual scheme that could, and in your opinion should, have been financed and initiated earlier? 12. Finally, could you raise any issues not covered above which you feel are relevant and make any suggestions which you believe would expedite the project preparation process? ANNEX III Page 1 The Method For Case Studies 1. Four projects were selected for detailed scrutiny from among those approved in recent years with preparation periods that seemed, on first inspection, particularly long. It was thought that more lessons about speeding up project preparation could be learned by concentrating mainly on projects with relatively long preparation periods. 2. For each of the four projects studied in depth, key dates were identified (first consideration by the Bank Group, official identification, pre-appraisal and appraisal missions, negotiations, Board approval, loan effectiveness) and a detailed chronology prepared of the main events, decision points, studies made, Bank missions undertaken, etc. together with changes in the project during the project preparation period and beyond. This information was plotted diagramatically on a "flow chart", showing the approximate retrospective critical path analysis for the preparation of each project, from initial consideration of the project by the Bank Group through loan/credit effectiveness. 3. From such a flow chart, the main areas for possible speed-up in the project preparation cycle were classified according to type, including technical, legal, studies (their financing and progress), appointment of consultants, deficient economic policy, shortage of local currency, com- mercial problems (such as low tariffs), joint financing arrangements, institutional change and Bank Group lending programs. 4. In each case an analysis was made of the difference between the project as originally considered and as finally conceived at loan effective- ness, and any differences incrementally related to the various decisions taken according to the critical path analysis. 5. Approximate increases in costs (a) to the country and (b) to the Bank Group, relating to any speeding up in project preparation or changes made to the project, were analyzed, as were consequent bringing forward of attributable project benefits. 6. Areas where savings in time or effort might possibly have been made in the case of each of the four projects, and any significant adverse effect on the ultimate viability of the project, were identified. Specific scenarios of alternative courses of action that might have been taken by borrowers or the Bank Group (but especially the latter) in order to achieve these savings, were developed for each project and discussed. 7. The results for each project were written up in a case report, including presentation of the alternative scenarios and calculation of ANNEX III Page 2 their approximate costs and benefits, and analysis of the issues of po- tentially broader relevance that seemed to emerge. The case studies were discussed with the borrowers and the Bank staff who had been involved with the projects and revised in light of comments received. 8. The above tasks were accomplished by means of the following steps listed in chronological order (although iterations were required): first, review of Bank files and documents (feasibility studies, etc.) available in the Bank; second, discussions with Bank staff who were involved; third, discussion with consultants (when possible); fourth, discussion with borrower and government. ANNEX IV Calculation of the Increase in Net Present Value of a Project from Bringing Forward Execution by Three Months Basic Assumptions Preparation Period of 3 years Implementation Period of 4 years Operating Life of 20 years Size of Investment $10 million Discount Rate of 10% NPV in year Benefits in Increase in NPV in year 0 from of appraisal first year of bringing forward start of imple- Rate of Growth of Benefits (year 3) useful life mentation from year 4 to year 3 (year 8) by one year a/ by three months (all in $ millions) Case of 15% Internal Rate of Return 0% 3.6 2.0 0.27 0.067 5% 3.8 1.3 0.28 0.070 8% 4.3 1.06 0.32 0.080 10% first 10 years and 0% thereafter 5.2 1.2 0.39 0.097 20% first 5 years and 0% thereafter 4.3 1.06 0.32 0.080 Case of 20% Internal Rate of Return 0% 8.0 2.7 0.60 0.150 5% 8.4 1.8 0.63 0.157 8% 9.6 1.5 0.72 0.180 10% first 10 years and 0% thereafter 11.8 1.8 0.88 0.220 20% first 5 years and 0% thereafter 9.6 1.5 0.72 0.180 a/ Increase in NPV from bringing project start forward from year 4 to year 3 (1.10)-2 NPV - (1.10)-3 NPV NPV /(1.10)-2 - (1.1O)-3 7 = NPV L 0.075_/