The World Development Report 1983 was prepared by a team led by Pierre Landell-Mills and comprising Ramgopal Agarwala, Richard Heaver, Dominique Lallement, Geoffrey Lamb, Selçuk Özgediz, and Mary Shirley. The Economic Analysis and Projections Department, under the direction of Helen Hughes, prepared the material for the section on global economic prospects and supplied data for the whole Report. The work was carried out under the general direction of Anne O. Krueger. See inside front cover for ordering information.
Governments have varying priorities, but the task for all is to discover the most cost-effective means to achieve their chosen development goals. The current economic crisis, with capital scarcer and more costly than before, has made this task all the more urgent. This quest for efficiency is not ideological; it is as much a preoccupation of the centrally planned states as it is of those with market economies. Major reforms are being attempted in China and Hungary, as well as in Chile and Brazil.
The efficiency of state activities has become increasingly important for economic development as the relative size of the public sector has grown. Most governments have sought to accelerate development through a wide variety of interventions. These have gone beyond providing infrastructure and social services to engaging directly in productive activities—primarily by establishing state enterprises. A myriad of taxes, subsidies, and controls have also been introduced to foster desired economic activities.
The results of this activism have been mixed. Overall, both growth and social development have been impressive; during the past two decades developing countries have made unprecedented gains in per capita income, averaging 3.5 per cent growth per annum. The increase in savings, too, has been impressive. But underlying this achievement is a worrying downward trend. While the average investment rate in developing countries rose from 20 per cent of gross domestic product in 1960 to about 26 per cent in 1980, the average annual GDP growth has declined slightly over the past two decades—from 5.7 per cent in the 1960s to 5.3 per cent in the 1970s. For the future, it will be difficult to raise investment rates further. Discounting more favorable external trends, higher growth must come from more efficient use of resources.
The average growth rates for developing countries conceal great variations in performance among countries, not just because of unequal resource endowments but also because of differences in policies and management. For example, possessing oil wealth has not automatically brought markedly higher growth. During the 1970s the GDP of the middle-income oil exporters grew less than one half percentage point faster than that of the oil importers.
The effectiveness of state activism is severely constrained in many developing countries by weak institutions and management. Some governments have become so overextended that they themselves have become a source of serious inefficiency. Over time, as more experienced and better trained cadres are built up, this critical bottleneck should ease. However, the immediate need is to design strategies that economize on management.
There are three possible responses to this situation. First, decentralize; second, place greater reliance on markets and prices and hence make less use of administrative controls; third, take initiatives to make governmental institutions function better. Much may be gained from reassessing and redirecting governments’ efforts. Bureaucracies need to be reoriented from administering regulations toward managing development, planning agencies from drafting comprehensive blueprints toward more selective interventions backed by better policy analysis, and accountants and auditors from a preoccupation with financial rules to obtaining value for money.
In the light of experience and regardless of ideology, the role of the state merits reassessment to take into account what has been seen to be effective over the years. Ownership—public or private—is not the issue, but rather how to make the best use of available skills and how to design institutional arrangements that encourage efficiency. In capitalist New York, the streets are cleaned by the city administration, while socialist Belgrade has contracted with private enterprise “labor associations” to do this work.
Management skills are scarce, but not uniformly so. In countries where high-level management capacity is particularly limited, lower-level management skills are often underused. Governments tend to establish large organizations that demand scarce high-level skills—for example, replacing small rural traders with large agricultural marketing monopolies or private truckers by state transport companies. Least successful of all are the large state farms that in some countries have replaced individual peasant farmers. Thus, frequently, small can be optimal; it makes enterprises more responsive to client pressure and it limits bureaucratic overheads. On the same principle, where a large organization is deemed necessary, great benefits may be derived from delegating responsibility for day-to-day decisionmaking to the various operating units, provided control of policy is maintained.
Considerable advantages can often be gained for the state by contracting out certain tasks. A good example is road maintenance. In Brazil, private firms have taken over the maintenance of about three quarters of the road network, and there is keen competition for contracts. Available evidence indicates that the contracted work is as much as a third cheaper than work done by the public works department. Specific services may also be provided more efficiently by the private sector; bus transport is an example. Cities that have removed the monopoly previously enjoyed by a public bus company have witnessed an impressive improvement in public transport. In Buenos Aires, some 50 competing bus companies replaced a public company that was losing over US$40 million a year; the service now operates privately with a much higher level of service and without subsidy. In Calcutta, private operators compete successfully with the State Transport Corporation, which needs a $1 million subsidy a month to cover its losses.
With an activist government and scarce skilled manpower, the natural tendency is for decisionmaking functions to be concentrated at the center. Yet this easily leads to inefficiency, as the center becomes heavily overburdened. Decisions are often better and more promptly taken nearer the scene of action, where more accurate and up-to-date information is available and there is greater awareness of the various options. Public enterprises are a good example of this principle. As owner, the government has a duty to maintain control of corporate policy but, for reasons of efficiency, operational control should be fully delegated to the enterprise’s management.
In practice, supervising ministries frequently pay too little attention to issues of policy and, instead, interfere in the enterprise’s day-to-day operational activities, even though they are not well placed to evaluate the situation or appreciate the consequences. For example, they often show great interest in individual hiring decisions but ignore an enterprise’s borrowing activities which can affect a country’s creditworthiness. Serious mismanagement results. Interference from the center undermines the main rationale for the public enterprise, which is to give operational autonomy to a separate legal entity.
Autonomy should help insulate an enterprise’s management from extraneous sectional interests; in turn the enterprise’s management must be made accountable. To place this complex relationship on the correct footing, a public enterprise’s terms of reference should be clearly defined, including objectives, guiding policies, production targets, budgets, investments, and sources of finance. Since the state’s objective is rarely mere profit maximization, it is important to resolve conflicting goals explicitly and to specify the measurable performance indicators other than profitability to be used. One device for achieving this, conceived in France and more recently adopted in Senegal, is the contrat plan, which in essence is an agreement between the government and the enterprise that defines the enterprise’s long-term objectives, medium- and short-term targets, and the specific, detailed means for their achievement. Corporate plans elaborated by enterprises for approval by governments, used for instance in India, serve a similar purpose.
The concept of an agreement or contract can be extended to other situations where central government responsibilities are delegated to other agencies within an administration or devolved entirely to nongovernmental organizations. In this context the contract becomes a coordinating mechanism that seeks to commit the participants to fulfill their respective obligations. In Latin America convenios are widely used for regulating collaboration between agencies.
Decentralization can only be successful if it is accompanied by the transfer of adequate staff and financial resources. Local government authorities are widely found to be weak, precisely because they are denied material support. In the case of public enterprises, an important matter to be settled in a contrat plan is the financial obligations of government to the enterprise, as well as the reverse.
Decentralization and accountability go hand-in-hand for public enterprises. The key to effective accountability is a good information system linked to the well-defined performance indicators already discussed. Private firms achieve accountability by creating “profit centers” with managers being given clearly demarcated responsibilities. Public bureaucracies can adopt a comparable arrangement by fully identifying individual programs in the budget, so that a specific service can be costed and evaluated. Lacking the cost consciousness of a profit making organization, governments have introduced program budgeting slowly, despite considerable discussion of the need for budget reform over the years. The experience of some 12 developing countries that have introduced such a system shows that, for success, a phased approach is best.
External accountability is ostensibly achieved through the political process. This can be greatly assisted by the practice of public bodies openly publishing regular and detailed reports of their expenditures and achievements. Appropriately channeled, client pressures based on such information can have a very salutary effect on public managers. This was demonstrated in Kenya when the Government decided in 1978 to publish school examination results, enabling parents to compare the performance of their children’s school with others. This generated considerable public pressure to improve the quality of teaching in some schools.
A similar approach can be used to enhance internal accountability. The Indonesian family planning program, by publishing the number of new acceptors recruited by different subunits, established a performance norm and thereby encouraged each to strive to meet or better it. The laggards were painfully exposed to view. In this way, a monitoring system—essential if decentralization is not to result in an unacceptable loss of control by the center— can also be used to motivate local managers. A close interconnection can thus exist between decentralization, competition, and accountability, which may be exploited to improve efficiency.
Prices and markets
Prices are a highly effective allocative mechanism, and not only in market economies. Almost all the economic reforms attempted by the centrally planned economies (for example, Hungary and Yugoslavia) place greater reliance on prices, which permits the decentralization of decisionmaking to the level of the enterprise. In China, health care has recently been made subject to a pricing system that allows greater consumer choice. The difficulty lies in ensuring prices that accurately reflect economic costs. In any economy these would need to reflect promptly and accurately a vast amount of rapidly changing information on the supply and demand for goods and services. Competitive markets, which essentially decentralize this process, have the necessary flexibility and responsiveness to do this, while saving scarce administrative capacity. The problem is how to deal with market failures that result, for example, from monopolies and cartels, poor information, economies of scale, and important indirect effects.
Governments are frequently drawn into imposing price and other administrative controls to correct perceived market failures or to achieve specific social objectives. Unfortunately such interventions often misfire. For example, price controls on food crops benefit urban families but penalize the output of the rural population and easily lead to scarcities. Import controls to protect domestic enterprises foster inefficient ones. Cheap credit to stimulate investment favors capital intensity and may result in fewer jobs. And finally, administrative controls almost always open the door to corruption. The welfare cost of market failure must thus be balanced against that of bureaucratic failure.
In many instances, greater reliance on markets, even though they may be imperfect, allows state interventions to be selective, and tailored to the managerial capacity of the public sector. But some market failures are too critical to be ignored, and markets generally do not operate to reduce inequality; here, government intervention would seem appropriate. The challenge is to choose those forms of intervention that minimize the economic costs of interference. Thus, targeted food programs are more efficient than general food subsidies. And competitive exchange rates are better than tariffs and import controls.
The search for profitability in a competitive environment provides private firms with a simple, powerful incentive to satisfy their customers. Where state-owned enterprises operate in competition with the private sector, similar incentives can apply. But alternative incentives are required for monopolies—whether public or private. A close link exists between accountability and incentives. Public exposure, as in the case of the Kenya school examination results, is a form of negative incentive. In some developed countries, consumer councils operate as watchdogs of the public interest, though with only partial success: here again, the key is to establish clearly defined and well-publicized performance criteria. British Rail, for example, posts notices in stations giving the average time of late arrival for all passenger trains during the past month, together with comparative figures for previous months.
Incentives for individual employees are a more complex matter. Public servants are rarely subject to a direct link between performance and reward. Because of the difficulty of establishing measurable performance criteria, pecuniary incentives often produce perverse results. In one country an overnight allowance was given to agricultural extension agents if their work took them more than 25 kilometers from their duty station; as a consequence, the agents rarely visited the farmers living closer.
Governments in practice generally offer a whole variety of “perks” to retain and reward staff-free housing, cars, directorship fees, and the like. It is unusual, however, for these to be tied to performance. Since such advantages are not uniformly available, considerable energy can be expended in wasteful competition for discretionary benefits. Agencies, such as central banks, that command a greater array of such benefits, attract staff, leaving less fortunate but not necessarily less deserving services, such as statistics departments, depleted. These biases can be counterbalanced, but only with determined government leadership. Regular salary reviews are a convenient vehicle for carefully scrutinizing the total compensation package and modifying those allowances that produce perverse behavior.
Several developing countries (including India, Kenya, and the Philippines) are experimenting with staff appraisal systems that strengthen the link between pay increases and individual performance. But even the most efficient administrations have found such systems difficult to implement objectively. For this reason it is as important to lay stress on nonmaterial incentives: an enhanced work environment, greater staff participation in designing work programs, and good career development prospects.
Staff training is frequently seen as a panacea for improving the performance of public bureaucracies. But overall the experience with public administration training has been disappointing. Training programs are widely criticized for being largely irrelevant to needs. Great scope thus exists to upgrade their quality and effectiveness.
Here, too, appropriate incentives are critical. If staff are to be motivated to be trained, it is important to achieve much closer links between the trainers and their clients, as Malaysia’s National Institute of Public Administration has well demonstrated. Training is most successful when it is fully integrated into career development, and the status of training within the bureaucracy is high. In military establishments, where the importance of training is generally well recognized, appointment as head of the staff college is often a key step for those later to be appointed chief of staff. In the civil service, the attitude toward staff training is usually much less positive. The relevance of training can be enhanced by enabling the trainers to engage in research and consultancy activities of value to government departments. This exposes the trainers to current problems and makes them aware of operational needs.
A good training program and a strong system of personnel administration go together. Career paths need to be well defined with schemes of service for each cadre, and the progress of staff kept under review. Bangladesh is currently instituting major reforms in this direction; the Government has established a Training and Career Planning Unit, a Personnel Management Information System, and a National Training Council- all concerned to ensure close links between training and staff appointments and promotions.
Selectivity and flexibility
Selectivity in designing government interventions has already been mentioned. It is as relevant to administrative reform or national economic management as it is to running enterprises or projects. At the most general level, experience with administrative reforms strongly indicates that comprehensive changes are extremely difficult to implement. Bookshelves have been filled with reports of special commissions recommending major civil service reforms, but in most cases the results have been disappointing, encountering, as they must, the solid opposition of entrenched bureaucratic interests. In Pakistan, between 1947 and 1960, 28 separate reports were prepared totaling 3,621 pages, all to no great effect. That experience has been repeated more recently in numbers of countries. More successful are selective, incremental reforms that permit a targeted effort and are able to attract sustained political support, both within and outside the bureaucracy. Recent measures successfully instituted in Brazil to improve the supervision of state enterprises are an example.
Comprehensive national planning has been equally unmanageable. Attempts to produce long-term blueprints for development that specified in detail output targets for all sectors of the economy—inspired in part by the Soviet model—have not been effective. After independence, most developing countries emphasized the role of planning, some using sophisticated macroeconomic models and input-output techniques. Central planners, however, have failed to realize their initial high ambitions. Even India, which pioneered central planning in mixed economies, reverted for a while to annual budgets as the principal planning instrument in the late 1960s. The crisis of planning deepened in the 1970s, as uncertainties and rapid change in the external economic environment made most comprehensive plans obsolete almost as soon as they were published.
Selectivity in the choice of issues and instruments is essential if a planning agency is to achieve good results. Identification of priority targets allows efforts to be concentrated and provides a rallying point for the political leadership. Perhaps the best example of planning is in Japan, where the “visions” of the Ministry of International Trade and Industry have guided the concerted actions of government and the private sector. In the same way, Korea successfully made export promotion the central theme of its plans, while Bangladesh has given clear priority to increasing food production with encouraging results. With less time spent modeling the unknowable, much more effort can and should be given to policy analysis.
National economic management will always be weak unless programs are consistently evaluated, and there is prompt feedback to the planners and policymakers. But, here again, no developing country can afford to monitor all activities all the time. Experience suggests that selective tracking is far more effective. As noted above, government budgets are ideally formulated on a program basis, so that the inputs and outputs can be compared to see whether the country is getting value for money. But instead of trying to check the outcome of every operation routinely each year, governments would use their resources better by analyzing the cost-effectiveness of selected departmental activities. In this way, scarce monitoring and evaluation resources can be concentrated systematically on priority areas of concern. The so-called “Rayner scrutinies” instituted recently in the United Kingdom fall into this category, as do the investigations carried out by India’s Programme Evaluation Office.
Poor information is a major obstacle to better economic management. Governments’ central statistical offices often have an ambitious program of data collection— in many cases too ambitious. Here again, selectivity is one key to efficiency. It may often be better to have less, but more reliable and timely, information. Although basic data collection is a prerequisite for the compilation of essential national income data, many opportunities are missed to use focused sample surveys to collect key information rapidly that is relevant to current priority policy issues. Instead, an inordinate amount of resources tend to be expended on censuses, whose results often take years to be published. Kenya has made particularly good use of limited surveys to keep track of food availability in the rural areas, thus providing early warning of shortages and a good data base for food policy decisions.
Together with selectivity comes a recognition that, in a rapidly changing world, government programs need to be constantly adapted to remain relevant. Bureaucracy tends to be inflexible. Mention has already been made of the weakness of the long-term blueprint approach to national planning. Although it is virtually impossible to foresee how a public investment program will evolve over a period longer than two or three years, some governments persist with a five-year investment planning cycle. A more practical approach (and less wasteful in the use of planning resources) is to formulate a two-to-three-year rolling investment plan consistent with a broadly defined long-term development strategy. This plan can be updated annually and linked directly to the budget process.
The need to allow for flexibility extends to structural adjustment programs. While convincing evidence exists to demonstrate that large income losses arise from distortions in the key prices in the economy— interest rates, wages, exchange rates, and other prices relating to trade, energy, and agricultural crops—experience strongly suggests that the timing, pace, and scope of removing price distortions need to be carefully adapted to each country’s situation. The varying approaches and results achieved by Chile, China, the Republic of Korea, Thailand, and Turkey over the past decade are good illustrations of the fact that no simple blueprint can be applied everywhere. More important, each program needs to be carefully monitored and frequently revised as implementation proceeds.
Flexibility is equally important at the field level (though it is often difficult to achieve with donor-supported projects that have to be spelled out in comprehensive detail before funding can be secured). Most often governments face the difficulty of designing a project with incomplete information. For projects where people are both beneficiaries and participants—such as health, agricultural extension, or squatter upgrading programs—it is particularly important to allow flexibility during implementation. (This point is discussed in more detail by Arturo Israel in his article on management and institutions on p. 15.) By incorporating a learning process into the design, a project can be modified with experience.
Part of the explanation for the cost advantage enjoyed by private road maintenance contractors over public works departments referred to earlier is their flexibility. Unencumbered by public service regulations, they can gear up more easily for peak demand and slim down faster when demand slackens. Since public works departments treat labor and equipment as fixed costs, a cut in the budget falls wholly on variable costs such as fuel. As a result the capacity utilization achieved by contractors is often double that of government operations.
Unfair as it may be, public bureaucracies are commonly viewed as obstructing development rather than contributing to it. Obviously, simple comparisons with the private sector are misleading. But the potential exists for reaping major gains from greater public sector efficiency. A survey of 24 countries in 1977 estimated the deficits of state enterprises at 4 to 5 per cent of GDP. The cost of delays in the implementation of public investment projects alone can easily amount to one percentage point or more of GDP growth. A study of irrigation in Asia has calculated that improvements in water management that are both simple and feasible would yield an additional 20 million tons of rice, enough for 90 million people. And there are many more illustrations that could be given of the potential gains from improved management.
Many lessons may be drawn from the experience of the past three decades on how governments could do a better job. Foremost among these is the need for appropriate incentives and for greater selectivity and flexibility. In addition, in most countries many opportunities exist for the public sector to make greater use of private capacity without abdicating its responsibility for policy. Perhaps most important of all is for government agencies to be given clear objectives and to have continuity in directions and operations; countries that have frequently changed policies, modified institutions, and rotated managers have fared poorly.
The time horizon for institutional development is an extended one—decades rather than years. Yet the very pressures that push governments into an activist role are those that drive political leaders (and donor agencies) to seek quick solutions. Institutional reform can only be successful if pursued with vision, tenacity, and strong political backing.