Warren C. Baum and Stokes M. Tolbert
It has been said that to govern is to choose. In the economic arena, the most basic choice that all societies must face in allocating resources is between current consumption of goods and services and investment in future growth. In developing countries, the choice must be made in the face of the deplorably low, often subsistence, level of consumption of a large proportion of the population and the urgent need to invest as the best hope of achieving higher living standards. When the choice is made in favor of investment, it is imperative that the scarce resources be deployed to obtain the maximum benefit. Packaging these investments into projects through the disciplined project approach developed by the World Bank can be a very effective means toward this end. This special article, which is, with minor modifications, to make it self-contained, the final chapter of a new book by the Bank, highlights some of the principal lessons that can be drawn, for the benefit of developing countries, from the Bank’s more than thirty-five years of experience with the project approach.
Economic development, of which project work is an integral part, is a long, slow, and often painful process of learning from experience. Investing in development through projects is subject to all the vicissitudes and constraints that hamper development generally. Development projects take years to prepare and implement. Through-out this time, project managers must confront and deal with the scarcity of human skills and material resources that are synonymous with underdevelopment, with a chronic shortage of funds, and with shifts in political support. They must operate within often fragile economic structures that are exposed to the worldwide forces of inflation and recession and to the unpredictable forces of nature.
Project work is thus highly demanding, and it sometimes seems to demand most from countries with the least capacity to respond. It is more difficult, but also more important, to do project work well when there is little or no cushion to absorb the effects of unsound policies, weak public administration, or unskilled project management. This is true of all types of projects, but particularly of those intended to alleviate poverty and raise the living standard of large numbers of people.
Fortunately, the rewards of project work are commensurate with the demands. Done well, project work pays high dividends. In the postwar era, it has become one of the most potent instruments for promoting economic growth. By setting investment priorities within a national and sectoral strategy, getting the policies right, and combining investments with technical assistance to strengthen institutions and train people, the project approach can enable countries to move more rapidly along the path of economic development.
Doing project work well is as much an art as a science. Drawing on the wealth of World Bank experience, we can identify many of the ingredients of success as well as many of the pitfalls. Developing countries can profit from this experience and thereby telescope the learning process. Some may be able to achieve in decades the economic progress that took generations in the now developed world.
In reviewing the lessons of the Bank’s experience, we have been struck again and again by how obvious many of them seem to be. They often read like little more than the dictates of common sense. But to say that something is obvious does not mean that it is simple. Perhaps simple to understand—except for a few esoteric subjects such as cost-benefit analysis—but not simple to put into practice. If the lessons of experience have not yet been learned, it is not because of irrationality or obduracy, but rather because a powerful constellation of forces makes things what they are and serves as a formidable obstacle to change. Moreover, economic development is a long-term process; when governments change frequently, the continuity of effort that is a sine qua non of development is soon broken. Even in reasonably stable political conditions, public officials are often and understandably preoccupied with more immediate considerations. They are more likely to suffer the short-term political penalties of policy changes such as raising the prices of basic commodities than to reap any political rewards from their long-term benefits.
To take just one example of the difficulty of applying a simple lesson of experience: it is by now widely accepted that developing countries should use the technology most “appropriate” to their circumstances. These circumstances usually include a surplus of low-cost, unskilled labor and a shortage of capital—facts which suggest that the technologies adopted should be relatively labor-intensive. The point seems obvious; which government would knowingly or willingly espouse an “inappropriate” technology? Yet in countless instances developing countries have adopted or retained technologies clearly unsuitable to their circumstances.
Why does this happen? Foreign consultants or advisers may advocate the technology with which they are most familiar. Local engineers, if educated abroad or the heirs of a colonial legacy, may have acquired a similar bias in favor of advanced technology, or they may simply presume, as do their superiors, that what is most modern is best. Special interest groups may favor a particular technical approach, while those who would benefit most from some other approach may be either unaware of the choice or politically disenfranchised. Deep-seated customs and traditions may favor certain solutions and make others unacceptable. Economic policies that overprice labor (through minimum wage or other legislation) or underprice capital (through subsidized interest rates or an overvalued currency) may send distorted signals to decisionmakers. A simple lack of knowledge or reluctance to experiment may limit the range of choice. Bilateral lending agencies may themselves be part of the problem; when aid is tied to the supply of equipment from the donor country—a policy as widely practiced as it is deplored—freedom to choose an appropriate technology may be compromised. With so many factors at work, it is not surprising that a “simple” lesson—such as selecting an appropriate technology—may prove far from simple to apply.
Trying to summarize the lessons of the World Bank’s experience is a daunting task. We can only highlight the more important lessons, with no expectation of doing justice to the many issues and problems that these lessons raise. The presentation takes up, in turn, national investment management, sector analysis and management, the project cycle, and the various dimensions of project analysis. Many of the lessons of sector analysis are not referred to as such, since they underlie the points discussed under the project cycle and project analysis.
No attempt is made to indicate the relative importance of these lessons, although some are clearly more important than others. Not getting prices right, for example, or not providing adequate recurrent funds to operate and maintain project investments can have more far-reaching consequences than not regularly evaluating completed projects. Since the lessons relate to different aspects of investment planning, sector analysis, and project work, it should be possible, at least in theory, to apply them all. But to many officials in developing countries, they may appear to constitute a formidable agenda. It is for each official, knowing the local circumstances, to establish priorities for action at any time. Progress may often be slow and partial, but responding flexibly to changing circumstances, combining rational economic choice with informed political judgment, avoiding the more egregious mistakes, and maintaining a steady pace are the best way to keep the country on the path to sound economic development.
National investment management
Project work in its broadest sense takes place at three levels:
• At the national level, where national investment plans are formulated, priorities among sectors are established, and the macroeconomic framework of policies for economic growth is put in place
• At the sector level, where priorities for investment within each sector are determined and the issues and problems affecting the development of the sector are addressed
• At the project level, where individual projects are identified, prepared, and implemented and attention is given to their technical, economic, financial, social, institutional, and other dimensions.
Only in the abstract can the project process be described as a sequence of steps proceeding in logical order from the national to the project level. Project work is in fact a continuum; decisions or actions affecting individual projects may take place at each of the three levels simultaneously and in interactive ways.
At the national level, the surge of enthusiasm for comprehensive development planning, which reached a peak in the 1950s, was followed by disenchantment with its limited accomplishments. No clear association could be found between a comprehensive planning effort and sustained economic growth. The elaborate and data-hungry models employed were unable to cope with the complexity and rapid change that characterize the process of development; nor did most developing countries have the administrative capacity to implement comprehensive plans. The technical and administrative deficiencies of comprehensive planning proved to be inherent in the process and, therefore, unlikely to be overcome by more strenuous efforts to improve the planning machinery.
Planning is therefore being directed increasingly toward new goals. Countries covering a wide political and economic spectrum no longer seek to rely on positive commands and negative controls to carry out a long-term plan with input and output targets for all sectors of the economy. Instead, they have turned to a strategy that focuses on two elements: a macroeconomic framework of policies to spur economic growth and elicit desired behavior from both public and private entities; and a public sector investment program that allocates scarce resources to high-priority public needs. The success stories of postwar economic development come from those countries that have managed these two tasks effectively.
The macroeconomic policy framework—comprising fiscal, monetary, exchange rate, wage, and trade policies—affects all aspects of economic behavior, including resource allocation, capital accumulation, balance of payments equilibrium, and, ultimately, the pace of economic growth. No single set of policies, and no specific institutional arrangement for managing those policies, are appropriate to all countries in all circumstances; what is appropriate varies with a country’s natural and human resources, level of development, and development strategy and objectives.
Central to any policy framework, however, is pricing policy. Many of the macroeconomic policy variables—interest rates, wage rates, exchange rates—are prices. Whether determined in free markets, or established by a central planning agency, or modified by government intervention, prices strongly affect a country’s rate and pattern of development. The importance of “getting the prices right” is therefore a fundamental lesson of development experience. The basic pricing rule for obtaining the most efficient use of an economy’s resources is to set the price of each good or service at its marginal cost—a result that a free competitive market would achieve and that public pricing policy should, in principle, emulate. In the real world, however, actual market conditions nearly always diverge from the competitive ideal. Moreover, in pricing public sector goods and services (power, water, rail transport, and so on), public agencies must take into account not only the objective of economic efficiency but also other objectives such as achieving a more equitable distribution of income or generating additional revenue. These considerations may justify departure from a rigorous marginal pricing rule (a point we return to in the discussion on cost recovery below) and may call for carefully designed intervention in the market to achieve both social and economic objectives. It is important that such intervention be devised to make use of market forces rather than to supplant them.
In many developing countries, however, a persistent and often massive distortion of prices has resulted from a multiplicity of government interventions, many of them poorly conceived or mutually incompatible. These distortions have reduced the efficiency of investment and slowed economic growth. The countries that have enjoyed relatively high growth rates are those that have managed to avoid significant price distortions; specifically, they have
• Avoided an overvalued exchange rate
• Kept the rate of effective protection for manufacturing industry relatively low and uniform among products
• Avoided underpricing of agricultural products and other disincentives to farmers
• Kept interest rates and hence the price of capital positive in real terms
• Avoided real wage increases not justified by rising productivity
• Applied cost recovery principles in pricing infrastructure services
• Avoided high inflation.
Many examples from the developing world in recent decades demonstrate how critical the proper management of this interconnected system of prices and incentives is for economic progress. Furthermore, the effects of macroeconomic policies are felt with full force at the project level. Thus, another fundamental lesson is that it is virtually impossible to have good projects in a bad policy environment.
Public investment programs and budgets
In formulating public investment programs, detailed and overly sophisticated forecasting exercises have generally proved counter-productive because of inadequate data and limited understanding of how sectoral investments and outputs are linked. Analytical efforts should therefore concentrate on designing investment programs in key infrastructure sectors, where the market alone cannot guide investment decisions, and on checking the consistency of these programs with the likely requirements of the productive sectors. Governments also need to be selective; they can effectively address only the most important public investment issues at any one time.
Experience has further shown that it is advantageous to combine the programming of public investment with indicative forecasting (not targeting) for the private sector. Emphasis should be placed on coordination and consultation—between departments of government as well as between government and the private sector. Governments also need to develop the capacity to respond quickly to changing events by modifying their policies and programs; this will usually require a much improved data base and information flow as well as centralized responsibility for policy coordination.
Establishing investment priorities among sectors—the attention to be given, for example, to programs in health, education, or housing relative to the productive or infrastructure sectors—is particularly difficult and ultimately entails political choices. Economic analysis can make only a limited contribution to this process, but it may still be crucial in clarifying the costs of alternatives and providing data on which informed judgments can be based. In establishing investment priorities within a sector, however, cost-benefit analysis can be very helpful in improving choices.
Experience has taught a number of other lessons about formulating and implementing public sector investment programs.
• Care should be taken that the investment plan is not too ambitious given the available resources. There is a widespread tendency to underestimate the cost of implementing specific projects and the time required. When too many projects are started at the same time, available skills are dispersed, project implementation is slowed, and economic and financial returns from the investments are reduced.
• New projects should not be started at the expense of adequate funding for those projects already under way. Completion of ongoing projects, if they are still justified given the incremental costs and benefits, should have a high priority for funding, as should the operation and maintenance of completed projects.
• The “free” resources left after the needs of ongoing and completed projects have been met should be calculated. This makes possible rational decisions about how much funding can be devoted to new projects in any budget year.
•Planning agencies need to strengthen their project appraisal capacity and to make greater use of cost-benefit analysis to identify and screen out projects with low rates of return.
• Investment plans need to be flexible so that they can be modified as circumstances change. In addition, a “core program” of investments should be identified so that cuts in programs made necessary by a shortfall in resources can be determined by priorities established in advance. Another technique found useful in many countries is to have a rolling investment program—usually a three-year program—that is updated annually or more frequently if circumstances warrant.
The government budget is the principal administrative instrument for translating public investment programs into concrete action. But the link between the two needs to be made much stronger in most developing countries. Many countries do not have a consolidated public sector budget, which is essential if the government is to obtain a full picture of both resource mobilization and expenditure. When such a budget is prepared, it will reveal whether public funds are being invested in activities or in proportions that are not in accord with the public sector investment program, and whether a substantial redirection of resources is desirable.
In fulfilling its budgetary responsibilities, a government must also concern itself with
• Organization and location of the budget and planning functions (although recruitment of qualified staff and establishment of proper procedures for reviewing investment proposals may be no less important)
• Budget classifications that will facilitate planning and management and improve the integration of budget appropriations with the investment program
• Monitoring and evaluation of the budgetary outcome, with emphasis on the accomplishment of programs and projects in addition to the traditional concern for financial propriety.
Even when investment requirements are appropriately matched with financial resources, two budgetary shortcomings common to many developing countries have adverse effects on projects. First, budgets often fail to provide sufficient recurrent funds to operate and maintain projects once the investment phase is over. Second, tight central budgetary controls, by frequently delaying the provision of funds to project managers, have plagued the implementation of investments. In most countries, there is a clear need to give spending agencies greater operational freedom to carry out their assigned tasks. The corollary—that spending agencies should be held more accountable for results—has encouraged the search for new systems, such as program and performance budgeting, which seek to make evaluation of results an integral part of the budgeting process.
Implementation of such systems, however, has proved difficult; experience suggests that the way to more effective budgetary management in a developing country is probably through careful identification of specific problem areas, followed by adaptive and evolutionary change, rather than through the large-scale import and installation of new systems.
Sector analysis and management
Sector analysis bridges the gap between the macroeconomics of national investment management and the microeconomics of individual projects. It is indispensable for resolving questions of choice, of priority, and of interrelations among projects. Yet it is a neglected activity in many countries. Compared with the elaboration of theories and models for national investment planning or of cost-benefit and other techniques for project analysis, theoretical and practical guidelines for the conduct of sector analysis are much less developed. Nevertheless, many sectoral issues are amenable to analysis, analytical capability is growing with experience, and the number of countries carrying out systematic sector work is steadily increasing.
Sector analysis is clearly useful to external lending agencies—including the World Bank, which has helped to introduce it to many of its member countries—in guiding project selection and design. But it is no less important to officials of ministries and operating agencies in developing countries, who can understandably lose sight of the broad objectives, needs, and priorities of a sector amid the day-to-day tasks of administering regulations and managing crises.
Sector analysis serves several purposes:
• It provides a better understanding of development policies and issues in the sector. This is useful both in enhancing the contribution that the sector can make to the overall strategy for economic development and to the planning of public investment, and in ensuring that the framework for sector policy, no less than that for macroeconomic policy, is conducive to sound project work.
• It makes possible the determination of investment priorities in the sector, which guide the identification and selection of specific projects.
• It evaluates the capacity of the principal institutions in the sector to implement desired policies, programs, and projects.
Most sector analysis falls into one of two broad categories: comprehensive surveys of an entire sector (or important subsector) and special studies of particular topics within a sector. Because the former tend to yield a low return for the heavy investment of manpower and other resources, governments are generally well advised to devote most of their sector work to a program of sharply focused special studies of high-priority issues. Much painful experience argues for keeping the reports on these studies brief and concise, in line with their operational purposes. Good sector analysis requires that qualified staff and other resources be made available for long periods, and this in turn requires that there be sustained commitment from high-level authorities. When these are not forthcoming, the quality of project work also suffers.
Each of the main sectors has distinctive characteristics as well as differing objectives and development issues. No effort is made to summarize them here. At the same time, the sectors share three basic areas for action: policy improvement; institutional strengthening; and better selection, design, analysis, and management of investment projects. The main lessons of experience in these areas are highlighted in the following sections on the project cycle and project analysis.
The project cycle
It is convenient to think of project work as comprising several distinct stages, commonly referred to collectively as the project cycle. The idea of a cycle underscores the point that the stages are closely linked and follow a logical progression, with the later stages providing the basis for a renewal of the cycle. The principal stages of the cycle are the identification of a project; its design, preparation, and appraisal; its implementation; and its evaluation once the investment phase has been completed.
The project cycle begins with the identification of project ideas that appear to represent a high-priority use of a country’s resources to achieve important development objectives. The identification and preliminary screening of ideas is a critical part of the process. Decisions made at this stage, either explicitly or by default, have a far-reaching impact on the final outcome of the project. It is particularly important at the outset to consider as wide a range of alternative approaches as feasible. All too often, project ideas are put forward and accepted without adequately weighing alternative, and possibly cheaper or more effective, means of achieving the same objectives. Whether the result of vested interests, political pressures, or simply lack of information about viable options, the consequences are the same: opportunities lost at this stage can seldom be recaptured. There is no substitute for being right in the first place.
Explicit attention should be paid to defining a project’s objectives at the earliest stage and to ensuring that all of the parties concerned, including external lending agencies, agree on those objectives and on the strategy for achieving them. The intended beneficiaries of the project must be consulted when their participation in designing and implementing the project is important to its ultimate success—as it is, for example, in rural and urban development. Failure to reach an understanding about objectives and to secure a firm commitment from all those concerned has often generated friction later and resulted in poor project implementation.
A project can be considered to have passed the identification stage when
• Major options and alternatives have been identified and some initial choices made
• The principal policy issues affecting project outcome have been identified and appear to be amenable to solution
• The project options selected are likely to be justified, given rough estimates of the expected costs and benefits
• It appears that the project will have adequate support both from the political authorities and from the intended beneficiaries
• The prospects are reasonable that adequate funding will be available from local and, if needed, external sources
• A specific plan for preparation of the project has been established.
At the next stage, that of project preparation, a feasibility study should be undertaken for all but the simplest and most routine projects. As the term implies, its purpose is to establish the feasibility or justification of the project, both as a whole and in its principal dimensions—technical, economic, financial, social, and so forth. Each dimension must be analyzed both separately and in relation to all the others. This is done in a series of approximations that test, for example, different technical approaches for their economic benefit and financial viability. The purpose of the analysis is not to determine whether a particular project idea is good enough to proceed with, but to arrive at the best one possible under the circumstances. Although the feasibility study should be designed with due regard for cost, this is not the place to skimp; if done well, the study is bound to pay for itself many times over through cost savings or increased benefits.
Good project preparation reduces the likelihood of difficulties during implementation by anticipating the problems that may arise and devising measures to deal with them. For most types of projects, detailed engineering should be completed, cost estimates made relatively reliable, institutional arrangements specified, and financial plans worked out before funds—whether from an external lender or from the domestic budget—are committed. One of the common causes of delayed or poor implementation of Bank-assisted projects has been inadequate or incomplete project preparation.
A pervasive shortcoming of preparation work, as already noted, has been to underestimate the cost of implementing projects as well as the time required. This optimism on the part of planners is found in all projects, large and small, complex and simple, and in virtually all countries. The cost overruns that ensue can have adverse effects on both the financial and economic viability of projects. Including physical and price contingencies as an integral part of the cost estimates helps to improve their accuracy.
One of the thorny issues of project design is how to achieve the right balance between simplicity and complexity. Projects that have a few, well-defined objectives and are based on proven technologies or approaches stand a better chance of being implemented successfully than projects that embody many objectives or unproven methods. Given the scarcity of trained staff, managerial skills, and administrative capacity in developing countries, special efforts to simplify objectives and to select technologies and approaches that have a high probability of working in a particular environment have paid off very well.
At times, however, projects suffer from being too restricted in scope. There are countless examples of projects that did not yield their full benefits, or even failed, because the need for complementary investments and activities had not been recognized and met. This is true of many kinds of agricultural projects (in which research, extension, credit, inputs, marketing, processing, and storage may all be necessary elements), of urban and regional development projects, and of highway maintenance projects, among others. Projects can be too simple as well as too complex; indeed, the swing of the pendulum in recent years toward greater complexity reflects the unimpressive results of projects that were too simple.
There is, however, a middle ground. Even when a multicomponent approach, with its inherent complexity, is necessary, it is possible to concentrate efforts and resources on those elements that are essential to achieving the primary objectives and to leave desirable but secondary objectives to subsequent projects or to parallel operations outside the project. Adjusting the scale of a project or the timing of its components can be another way of matching complex project design with limited implementation capacity. Furthermore, the administrative capacity of an agency can be expanded under a project through well-designed technical assistance.
All project identification and preparation work is directed toward facilitating project implementation and helping to ensure its success. In turn, the lessons learned during project implementation are fed back into the planning and design of the next generation of projects, Ideally, if all the preceding work has gone well, project implementation should hold few surprises. But events seldom go according to plan; circumstances change, and implementation therefore becomes a critical stage of project work.
Projects should be designed with a view to how they will be implemented. There must be a close match between a project’s objectives and the capacity of local institutions to implement, operate, and maintain it. Also, projects should be prepared in as much detail as is feasible under the circumstances. Since experience shows that projects are seldom implemented exactly as designed, sufficient flexibility must be preserved so that management can change course, sometimes drastically, during implementation if this becomes necessary.
Particular attention should be paid to the start-up period, before implementation has reached full stride. Use of critical path analysis, monitoring and evaluation techniques, and management information systems can facilitate this and later phases of the implementation process.
How to organize the management of project implementation and how to ensure effective coordination among the various parties can be vexing questions. There has been a tendency, fostered in part by external lending institutions, to assign responsibility for the management of projects in the public sector to special project implementation units established outside the regular line ministries. Such “enclave” units have helped to insulate project implementation from some of the bureaucratic, staffing, and salary weaknesses of the traditional ministries, thereby contributing in some instances to better implementation. But they have been without lasting institutional benefit, since they have operated under conditions that could not be replicated by the line agencies. Project implementation units should be confined to such special cases as projects that embody innovative or very large-scale activities. Even then, arrangements should be made to reincorporate the units as soon as possible into the agencies that bear permanent responsibility for such activities, in accordance with a strategy for strengthening the capacity of those agencies.
Ensuring effective coordination when several agencies or organizations are responsible for different components of a project is inherently difficult. Unless well planned, it can be very time-consuming and unproductive. Some coordinating mechanism must be found; the basic requirements are a clear definition of the responsibilities of each of the participating agencies and adequate incentives for them to cooperate in achieving the project’s objectives. If a coordinating committee is used, its authority needs to be clearly demarcated, and its members should be as close as possible to the operating level. Committees composed of ministers or other high-level officials with many other responsibilities have rarely performed well.
The Bank is continually reviewing its experience with project implementation to determine the principal factors that account for the success or failure of projects. Several factors emerge as essential ingredients of successful implementation.
Probably the most important reason for the success of a project is strong commitment by the government to its objectives. A coalition among the interested agencies and principal actors needs to be formed early in the planning stage and continued throughout implementation. When financial resources allocated to a project dry up, tariffs are not increased on time, or prices and other incentives become distorted, lack of effective support may be the cause. Similarly, mutual commitment and support is needed between the government and any external financing agencies.
Appropriate design and adequate preparation—both already alluded to—are essential. The objectives and design of projects need to be adapted to local political, administrative, economic, and cultural conditions, particularly if success hinges on changing behavior. When technologies and approaches have not been fully worked out, a pilot project to test them can save both time and money.
There is a close correlation between good managers and good projects. Many examples can be cited of projects in serious difficulty that were turned around by the appointment of a particular individual; in others, performance declined markedly when a good manager departed. This does not gainsay the importance of the policy environment, but it does affirm that good managers make a difference.
The underlying reasons for success and for failure are largely mirror images. But some of the specific difficulties that projects typically encounter during implementation deserve special mention.
Managerial problems are the most pervasive. Weak public administration and weak management at the enterprise or project level are often at the root of delays and cost overruns. Such weaknesses are manifested in inadequate planning, delays in land acquisition, protracted bidding and contracting procedures, insufficient project supervision, slow response to changes in the policy environment, and low staff morale and productivity.
Financial difficulties are common. They can arise from an imbalance between the national investment plan and available resources, which leads to underfunding of publicly supported projects across-the-board; from worldwide or domestic inflation; from a government’s reluctance (heightened during inflationary periods) to authorize tariff increases for public services or basic commodities; or from bureaucratic delays in providing funds or even in paying the government’s bills. Whatever the reasons, the consequences are the same: physical delays in project implementation, cost overruns, reductions in the scope of the project when this is possible, and deferred implementation of institutional and policy reforms. Cost overruns are a barometer of the implementation process since practically all of the problems encountered result in additional costs as well as delays.
Technical problems arise continually. These can include difficult or unexpected soil conditions, materials of poor quality, design defects, and unsuitability of imported equipment for local conditions. In agriculture, inadequate technical packages or disappointing results from newly introduced technologies are common. In people-oriented activities such as health delivery systems or education, well-tested approaches to the particular circumstances of a country or region may be lacking. It often takes longer than expected to ease institutional bottlenecks and to develop the services necessary to provide new technologies and to persuade beneficiaries to adopt them. But in most instances, solutions are worked out eventually. Technical problems tend to be less intractable than financial or managerial ones.
Ex post evaluation
The project cycle does not end when implementation is completed and the project goes into operation. There remains yet another stage, that of ex post evaluation, so called to differentiate it from the monitoring and evaluation that should be a regular part of project implementation. Actual results are viewed with the advantage of hindsight. The main purpose is to learn lessons for the design of future projects that supplement those gathered during the course of implementation. A second purpose is to help ensure accountability, whether of international aid agencies to their government shareholders or of national agencies and enterprises to their government and people.
Ex post evaluation should provide a comprehensive and detailed review of the elements of success and failure of the project: what went well and deserves to be repeated; what went wrong and why; how to avoid similar mistakes next time. It normally takes place shortly after the investment phase of the project is completed, at which time the construction costs are known but, for many types of projects, a large portion of the benefits still lie in the future. It may be desirable, therefore, to undertake a second evaluation some years later; such a “second look” can yield fresh insights into projects that undergo significant changes while in operation.
The World Bank has a well-established procedure for evaluating completed projects, based on the principles of close involvement of operational staff, objectivity, full disclosure of results, and comprehensive coverage. The process has resulted in a gold mine of information. Developing countries have been slow to take up the Bank’s example, in part because of shortages of trained staff and the need to give priority to the operational stages of project work. Those countries that do have an ex post evaluation system, however, are discovering that systematic efforts to learn from experience are invaluable for enhancing the developmental impact of project work.
Throughout the several stages of the project cycle, various dimensions of project work are addressed, both separately and in relation to each other, and in varying degrees of detail.
Of these dimensions, technical analysis is perhaps the most familiar. Among the issues of technical design are size, location, timing, and choice of a technology package. We have already alluded to the need to select a technology appropriate to the circumstances of the country and the requirements of the project. The point bears repeating in any list of lessons. “Appropriateness” is a relative concept. It generally means that the technology should be determined in relation to the objectives of the project, to the impact on intended beneficiaries, and to local conditions, including the availability and cost of capital, raw materials, and labor, the size of markets, and the present and potential capacity for planning and implementation.
Numerous tradeoffs are implicit in these considerations; for example, the technology may often have to be reduced in complexity to fit the capabilities of local institutions or the sociocultural traditions of the beneficiaries. The range of choice is very wide, from sophisticated, high technology for long-distance telecommunications even in the least developed countries to a labor-intensive approach for maintenance of rural roads, schools, or tertiary irrigation canals even in middle-income countries. Whatever the choice, it is important that it not be unduly biased by government interventions or price distortions of the kind referred to earlier.
Technical analysis also provides an opportunity to consider how a country can best take advantage of its investment plan, policy framework, and development projects to build a capacity to use science and technology effectively throughout the economy. Both trained people and institutions capable of utilizing their skills are required. In the early stages of a country’s development, the emphasis should be on building local capacity to import technology from abroad knowledgeably and to adapt it to local conditions. At later stages, the emphasis should shift to the development of local technology and its integration with imported technology, and to the encouragement of local research and innovation.
The basic question that economic analysis addresses is how to allocate scarce resources among many competing uses. It seeks to determine not only whether a project can be expected to provide a satisfactory return to the economy, but also whether there is an alternative way of achieving the same objectives that would offer a higher return. The analysis entails a comparison of costs and benefits with and without the project, both discounted to present values, through the use of analytical techniques that by now are highly refined—though no more accurate than the underlying data. Its central concept—that, for purposes of economic analysis, resources should be valued in terms of their opportunity cost to the economy in their best alternative use—is applicable to all economic systems, whether market-oriented, centrally planned, or a combination of the two.
Most lending agencies use cost-benefit analysis to assess the projects they help finance, and many developing countries are applying this analysis in one form or another to projects that they fund from public resources. Private industrial enterprises also use cost-benefit analysis for major investments. If done properly, it can be a powerful tool for guiding investment choices.
Cost-benefit analysis encounters some practical problems, among them: how to define the situation “with” and “without” the project (which should not be confused with the situation “before” and “after” the project); how to handle sunk costs incurred before the analysis (the short answer is to disregard them); how to establish the shadow (that is, opportunity) prices for labor and capital when market prices are distorted for various reasons; and how to deal with nonquantifiable benefits. For most of these problems of definition or measurement, reasonable approaches can be found that are consistent with the reliability of the available data. The results of the analysis can be presented as a measure of net present value, as an internal rate of return, or as a cost-benefit ratio; each has its particular uses.
Not all types of projects are amenable to cost-benefit analysis; it is not customarily used, for example, in evaluating education or health projects. Nor does it have the same meaning in different sectors. In power and water, for example, where prices are publicly administered rather than fixed by the marketplace, the economic rate of return is a minimum estimate, more indicative of the appropriateness of the regulated tariffs than of the real return to the economy. Cost-benefit analysis, therefore, is not very useful in comparing the merits or relative ranking of projects in different sectors. Much depends on the common sense, judgment, and even ingenuity of the analyst. Whim, bias, and intuition will also inevitably play a part. These caveats notwithstanding, economic evaluation can introduce rationality into the decisionmaking process, identify and measure risks, and avoid some of the more serious mistakes that can occur even in the best-laid investment plans.
Issues of financial analysis arise in the course of project design in several ways. The first and most universal concern, applicable to all types of projects, is that there be sufficient funds both to complete the project and to operate and maintain it subsequently. This sounds obvious, but far too many development projects have been launched without adequate consideration of the future availability of funds. This is notably true for education projects, in which recurrent costs for teachers’ salaries may quickly exceed the capital costs of the facilities; or for projects in irrigation, roads, health, water supply, and other infrastructure, in which facilities once completed may deteriorate rapidly for want of adequate maintenance. This reluctance to provide adequate recurrent expenditures reflects in part the greater political appeal of new investments and in part the unwillingness of most external lenders to finance such expenditures.
A second financial concern is to recover an appropriate portion of the costs from the beneficiaries or users. A cost recovery policy has three separate but related objectives:
• Economic efficiency. The resources provided under a project are used most efficiently when they are priced in accordance with their marginal or opportunity cost. The extent to which efficiency prices can be applied, however, varies widely from sector to sector.
• Income distribution. In the interests of equity, the prices or other charges levied to recover costs should take into account differences in income levels (as affected by the project) and in ability to pay.
• Revenue generation. Governments in developing countries, being short of resources, need to generate revenue from projects. In addition, revenue-earning enterprises need to be made or kept financially viable. If rural and urban development, water supply, and other projects are to be replicable on the scale necessary to reach the large numbers of potential beneficiaries, a substantial contribution from the initial beneficiaries will often have to be secured.
Difficult tradeoffs may be required among these sometimes conflicting objectives, and the poorer the country the more painful the choices may be. Cost recovery for basic services can be an explosive political issue. This fact, together with administrative considerations, will determine what can be accomplished and how quickly. Two important rules to follow are that new or increased levies should be related to benefits as they materialize and imposed in small but regular increments, and that similar situations should be treated similarly to avoid the appearance of discrimination.
The financial viability of revenue-earning enterprises—electric and water utilities, public and private industries, railways, telecommunications entities, and so on—can be translated into three subsidiary objectives or tests of performance:
• Will the enterprise have sufficient revenue to earn a reasonable return on its invested capital? As part of the test of “reasonableness,” will it be able to generate enough funds internally from its operations to make a satisfactory contribution to its future capital requirements?
• Will the capital structure of the enterprise enable it to meet all of its capital obligations, including the service of its debt, in a timely manner?
• Will there be adequate liquidity, that is, sufficient working capital to cover all current operational requirements?
Insistence on financial viability is an important means of imposing discipline on an enterprise and encouraging efficient management and use of resources. There is almost always some scope for improving financial performance by raising operating standards and reducing waste. Achieving financial viability will, however, ultimately throw issues of tariff policy into bold relief. In many cases, increases in prices—often substantial—will be necessary to reach the financial objectives. To raise the price of a basic product or service that figures prominently in the cost of living is an act of political courage. But not to raise the price may, in the long run, have even more adverse economic, and therefore political, consequences. During an inflationary period, governments are especially reluctant to increase the cost of basic goods and services. This is understandable, but failure to raise prices may also have an inflationary impact if the government must borrow to cover the deficits.
One means of easing the bureaucratic and political pain of raising the price of public services, as we have noted, is to adjust prices in small increments and at regular intervals. Another is to make price increases automatic or quasi-automatic, as when fuel costs are passed on to users through equivalent increases in electricity prices without the need for government approval each time.
Social analysis deals with the impact of projects on people. Its role is to consider the suitability for the proposed project design for the people it is intended to serve, to make proposals for improving the “fit” between the two, and to fashion strategies for project implementation that can be expected both to win and hold people’s support and to achieve project goals by inducing changes in social attitudes and behavior. Some projects—in such fields as rural development, education, and health—are clearly oriented toward people. Infrastructure projects are less so, but even these may have to deal effectively with people, as workers or consumers, in order to achieve their objectives. A project that runs counter to or ignores the traditions, values, and social organization of the intended beneficiaries, or that is based on objectives which they do not share, has little prospect of success.
Social analysis is a relative newcomer among the dimensions of project analysis. It focuses on four principal areas:
• The sociocultural and demographic characteristics of the project population
• The way in which the project population is organized to carry out productive activities
• The project’s cultural acceptability, including its capacity for both adapting to people’s behavior and perceived needs and for bringing about changes in them
• The strategy necessary to elicit commitment from the project population and to ensure their sustained participation throughout the project cycle.
Predicting social behavior is even more uncertain than forecasting financial or economic behavior. Project planners have often made unduly optimistic assumptions about local people’s interest in and need for a project, the economic and social incentives for them to participate, and the rate at which change in their social condition can be brought about. Social analysts with professional training and broad experience can improve the projections, however; more important, they can make a significant contribution to a project’s success by helping to design effective organizations and approaches to achieve desirable changes in social behavior.
Social analysis of projects has frequently failed to take adequate account of the particular interests and needs of women. Although they constitute half the world’s population, women have reaped far less than their share of the benefits of development. Women tend to have less schooling than men, their mobility is constrained by economic and social considerations, and they often face legal barriers to their ownership of assets or access to credit. Yet women play a key role in the development process, both as producers and consumers; for example, they provide most of the agricultural labor used in production, harvesting, marketing, and storage. Moreover, the central position of women in family life profoundly influences attitudes and decisions on education, nutrition, health, and family size. Thus, even if women never enter the formal labor market, ensuring that they receive an adequate education may be one of the best investments a country can make.
Failure to appreciate the role of women when projects are designed and implemented is not only inequitable but also retards the pace of economic growth. Decisions about which technologies to use and about how to provide a wide range of services must be made with full regard to women’s needs and constraints; otherwise, project benefits will be reduced and the position of women made even worse. In many respects, women are the largest underutilized resource for development.
In recent years, institutional analysis has become one of the important dimensions of project work. The outcome of development projects depends on the quality of the institutions responsible for them. Yet institutions in developing countries typically suffer from an acute shortage of experienced managers and staff, an excess of untrained staff, overstrained services and facilities, low wages and salaries, inadequate data and information systems, and an inimical policy environment. There are no standard solutions and few readily transferable institutional models from either developed or other developing countries. Institutional development must be seen as a lengthy process of experiment and adaptation, subject to many reverses when political or economic fortunes change. Still, when strong institutions have been forged, they have played a key role in development.
The overall policy environment affects the performance not only of projects but also, of the institutions responsible for them. If managers of these institutions are compelled to spend time and effort counteracting the impact of government economic policies (such as a critical shortage of foreign exchange for spare parts owing to an overvalued exchange rate), they cannot build long-term capacity effectively. Institutional problems multiply rapidly when government policies seriously distort the economic environment or when government regulations impose extraneous requirements, such as the employment of extra staff for political or social reasons.
Institutional problems have been compounded as more complex development objectives have been assigned to projects. Executing multicomponent projects or projects focused on alleviating poverty or upgrading social services has placed heavy demands on institutions. Knowledge of the economic, social, and behavioral characteristics of small farmers, urban squatters, or disadvantaged tribal peoples is still rudimentary. Understanding of the management and institutional techniques for dealing with such projects is also at an early stage.
Institutions in the public sector present a particular challenge, both because of their importance and because of the many problems they characteristically confront. A comprehensive reform of the public administration—however badly needed—is a very ambitious undertaking. Few countries, developed or developing, have successfully carried out such a reform in the face of the strong resistance that these efforts generate. While careful diagnosis of the problems of the public service is a necessary first step in establishing an overall strategy for change, implementing the strategy will almost invariably call for a selective approach. The best results have been achieved by concentrating on a few of the more critical problems. The ones to tackle first are those that seriously constrain the performance of the public administration and that are within its capacity to solve; in addition, it should be possible to mobilize and maintain public support for the reform effort.
Parastatals and other forms of state-owned enterprise have proliferated in recent years. Their performance has at best been mixed, and concern has been growing about the need to increase their efficiency, reduce their deficits that burden the public budget, and avoid political interference in their affairs. Given the preponderance of public enterprises in the economies of many developing countries, reforming them has repercussions going far beyond improved project performance. Some governments, in fact, are beginning to devote considerable effort to limiting the spread of public enterprises and to transferring ownership of some of these enterprises to the private sector. The legitimate demands on the public sector are very large, and the capacity for public management is among the scarcest of development resources. This capacity should, therefore, be reserved for use where it is most needed and most likely to be efficacious; governments should exercise great caution in deciding which fields are appropriate for public ownership.
For those agricultural and institutional parastatals carrying out processing, manufacturing, or marketing operations that private producers and merchants can do efficiently, the simple answer may be to abolish them. For other parastatals, reducing the number of staff, rationalizing their structures, and appointing and promoting staff on the basis of merit are among the measures needed. In several countries, policy reforms are being tried that will improve the “rules of the game” under which the enterprises operate. This reform has two objectives: to give public enterprises greater authority and autonomy to carry out their activities in accordance with broad goals and specific performance targets agreed on with the government; and to hold the enterprises accountable for their results, while improving the flow of information to government about their performance.
A widespread weakness of development institutions, particularly but not exclusively in the public sector, is the shortage of qualified managers at the middle and upper echelons and of experienced specialists such as accountants and engineers. This is a basic characteristic of underdevelopment that can change only gradually. In the long run, the educational system in all its forms—from primary education through universities, vocational schools, and management institutes—must provide the solution. But some steps can be taken immediately. Expatriate managers and consultants can be used to free bottlenecks in the short run—as long as care is taken that this does not delay the development of local managerial capacity, but rather fosters it through counterpart and on-the-job training. Whatever the source of management, greater attention to organizational questions and improved management, accounting, and information systems and procedures can enhance effectiveness. Continuity of management is also important; frequent changes for political reasons are highly disruptive.
One fruitful approach to strengthening management in the public service is through improved incentives. Salary and other financial incentives for public officials cannot normally match those in the private sector, but disparities can be narrowed. In particular cases, elements of the financial package other than pay can also be improved without disturbing prevailing standards in the public service. Establishing “super-grades” outside the normal system may be a way to meet special needs, while making lateral entry and exit easier can encourage managers to view public service as a natural stage in a career.
Programs for training managers and staff also deserve much more attention. The potential contribution of such programs is very large, yet experience shows that too often they have little effect. Training programs cannot be added on to projects at the last moment, but must be planned carefully in advance and properly executed. Adequate resources must be provided and skilled trainers recruited. As a general rule, the closer training is to the workplace in which it will be used, the more effective it is likely to be. This argues for on-the-job training when feasible.
It is now widely recognized that environmental analysis is necessary for a country to ensure the sound management and use of its natural resources as an integral part of its strategy for economic growth. Desertification, deforestation, soil erosion, overexploitation of such renewable resources as fisheries, and air and water pollution are lowering the carrying capacity of the environment. Usually the poor are disproportionately affected by environmental degradation. The objective of environmental management should be to achieve a balance between human demands on the natural resource base and the ability of that resource base to meet these demands on a sustainable basis in the interests of future generations as well as those alive today.
With careful environmental management, the pace of economic and social progress need not be slowed. In environmental work, the “ounce of prevention” is almost always more important and less costly than the “pound of cure.” Sometimes remedial action may not be feasible at all. All proposed development projects should be screened to detect those with a potentially harmful impact on the environment. Most environmental problems, if properly anticipated, can be dealt with at relatively small cost—usually less than 5 percent of total project cost. Analysis of the tradeoffs between different design features and their environmental costs and benefits should be a routine part of project work; the analysis is complicated, however, by the fact that the standard time-discounting methodology gives insufficient weight to environmental costs and benefits because of their long-term nature.
Managing the procurement process is an important aspect of project implementation. Delays in acquiring the necessary goods and works are likely to be compounded into further delays and increased costs for the project as a whole. Procurement is therefore a process to be carefully planned, organized, and managed, the more so since the number of ways things can go wrong sometimes appears endless.
Procurement must serve three objectives. The first is to help ensure the efficient execution of the project by acquiring goods and works with the optimal combination of quality, price, and delivery time. The second is to promote such national goals as the development of local industry, the balanced regional development of industry, or the support of small-scale enterprises. The third is to comply with the procurement regulations of any external lending institutions helping to finance the project.
There are a wide variety of types and methods of procurement. Which is the most appropriate depends on the size and nature of the project, the particular goods or works to be procured, and the regulations of lending agencies. For large projects, international competitive bidding is generally the best way of ensuring efficient procurement, safeguarding against waste or corruption, and satisfying the interest of lending agencies that all qualified firms be permitted to bid. Local competitive bidding is more appropriate for small-scale procurement of goods and works in which foreign firms will not be interested; changes in procedures may be necessary to ensure effective competition among local firms.
Use of consultants
Procurement of consulting services raises different issues. There is, first, the question of need. Most of the services for which consultants might be recruited can, in principle, be provided by local staff of the project agency if they are competent, experienced, and available. When such in-house capacity does not exist (as may often be the case) and cannot be put in place by additional training soon enough to meet the demands of the project, the next best choice is the use of local consulting firms. Establishing a local consulting industry is desirable both to provide some competition to stimulate agency staff and to supply more specialized skills or expert services. When local services are not available, firms from other developing countries may offer the dual advantages of better knowledge and understanding of comparable local conditions and relatively low costs. There will, however, remain some assignments—especially for large or complex projects and those requiring highly technical knowledge—for which only expatriate consultants from developed countries are qualified. Joint ventures between foreign and local consulting firms, if entered into voluntarily, are often a good way of developing local capacity.
The consultant-client relationship is a close, personal one; when a satisfactory partnership already exists with a particular firm, it may be eminently sensible to continue it. When no such relationship exists, or when for other reasons it is desirable to invite proposals from a short list of qualified firms, the selection process should ensure that price considerations are subordinated to a concern with quality. The cost of consulting services is usually only a small fraction of total project costs, but the quality of the work performed can have an impact on the final project out of all proportion to the cost.
These, then, are some of the principal lessons that the World Bank has learned in more than thirty-five years of assisting its member countries in managing their investment resources. They are offered not in the belief that they provide final answers to the formidable problems of development, but in the conviction that better national investment planning, macroeconomic and sector policies, and project work can ease the path of development, bring its benefits to people sooner, and distribute them more equitably.