Public expenditure in the developing countries increased its share of GNP by nearly 41% between 1972 and 1986, central governmental expenditures rising from 18.7% of GNP in 1972 to 26.3% in 1986. Much of this was due to rising interest payments on the foreign debt. When one examines the composition of central governmental expenditures, two things stand out. First, the share of expenditure devoted to human development (notably, education and health) declined, especially expenditure on education, and secondly, the proportion of expenditure devoted to the military also fell. These were, of course, falling shares of a total which was itself an increasing proportion of GNP. Thus when these expenditure categories are expressed as proportions not of total central governmental expenditure but of GNP, it transpires, paradoxically, that public expenditure on the military and on education and health rose. It would be highly desirable if developing countries in the 1990s could take advantage of the improved international political climate and curtail military expenditure, thereby releasing additional resources for expenditure on human development and for physical investment. Indeed, real public expenditure on education per student and real public expenditure per capita on health often either stagnated or declined. The central government current deficit, which can be interpreted as negative savings, rose very sharply, from 3.5% of GNP in 1972 to 6.2% in 1986. This is a further indication of the need to reform the public finances through improved tax and expenditure policies in order to make a larger contribution to accelerating growth.
Indeed in many countries, above all in those affected by serious debt-servicing problems, the fiscal deficit of the central government is the major constraint on development. The ease with which several countries have generated trade surpluses indicates that the balance-of-payments constraint is not always as severe as was once thought, and widespread evidence of massive capital flight indicates that in those countries at least a savings constraint is not binding. Fiscal reforms (designed to raise more revenue) and debt relief (which would release public-sector funds for development purposes) often are the main priorities.
****** FROM DUPLICATE Context ****** Governments can promote both economic growth and equity by supplying the physical infrastructure needed for productive private investment and by providing social services to meet the basic needs and improve the productivity of the population. But the high cost of raising revenue means that it is vital to set priorities and attain quality in public spending. Priorities can be set by considering what governments do best and what markets do best. Governments provide "public goods" that benefit all citizens, such as law and order and national defence. They should also be involved in providing goods and services with large external benefits to society, such as primary education, basic health care and immunization programmes. Direct investment or regulation is needed to control monopolies caused by a single source of supply or large returns to scale relative to the size of the market - water supply, sanitation and power, for instance. Government subsidies on goods and services consumed by the poor are sometimes justified but, to contain the cost, they should be accurately targeted. In general these criteria support widespread public provision of infrastructure for transport, communications, power, water supply and irrigation - areas critical for growth in the early stages of development. They also promote support for basic education and health, which has been instrumental in producing higher literacy rates and skill levels, reducing mortality and morbidity and lowering fertility rates. In contrast, these criteria generally do not support direct public production or marketing of industrial or agricultural products or direct public provision of bus transport or housing.
The tasks of containing and allocating public spending call for medium-term plans and shorter term budgets that set clear priorities and ceilings. Developing countries typically face shortages of skills and information, fragile political systems, and unstable macroeconomic conditions that all tend to exacerbate the difficulties of governments normally face in these matters.
Improving the efficiency and effectiveness of public spending requires reform of fiscal planning, budgeting, implementation, and monitoring. Fiscal planning ideally involves formulating a phased investment programme, projecting current spending needs and assessing revenue availability and borrowing requirements for three to five years, all set in the context of a consistent macroeconomic framework. The annual budget would then be a comprehensive one-year slice of the medium-term plan. For plans and budgets to promote effective decisionmaking by individual public agents, the tradeoffs among agencies, programmes and projects must be explicit, and the budget constraint for each agency, once set, must be firm so that an agency may not exceed a budget on its own initiative.
The breakdown of spending by category varies tremendously among countries but generalizations are possible. For example, industrial countries spend much more (as a share of both total spending and GDP) on subsidies and transfers, primarily for health and social security, while developing countries tend to allocate more of their spending to investment.
Although the capacity to carry out medium-term fiscal planning and comprehensive annual budgeting is limited in most developing countries, some have shown it possible to improve the efficiency and effectiveness of public spending. Botswana, for example, has developed procedures to ensure that careful attention is paid to the recurrent cost implications of its investment spending. Chile has used economic analysis - primarily cost-benefit analysis - to screen potential investments thoroughly. Others are trimming government payrolls through hiring freezes, civil service censuses and voluntary retirement schemes; some countries are trying to rationalize the civil service wage structure. Mexico is moving toward targeted food subsidies.
The technical and institutional problems involved in planning, budgeting, implementing and monitoring expenditure are very great. Governments must set priorities if they are to control the total level of spending and allocate it efficiently. These priorities should be based on two considerations: (a) an appreciation of where government involvement is necessary and, conversely, where the private sector can be counted upon to provide the same output as well or better; (b) an understanding of how limited resources can be spent most efficiently and effectively in the areas in which public involvement is called for. Governments should concentrate their spending in areas where their participation is necessary for a well functioning market, for economic growth and for the alleviation of poverty. Government involvement is clearly needed to supply public goods, such as defence and law and order. It is also needed where the private sector would under-supply goods or services that benefit society at large, such as primary education, basic preventive health care, transport infrastructure and agricultural research. Priorities emerge more forcefully when all parties are aware of their specific resource constraints. Setting overall spending limits means balancing needs against the cost of raising revenue. It is best achieved through a coordinated process of medium-term fiscal planning, annual budgeting, and regular monitoring of revenue and expenditure. Binding spending limits should apply not just to finance and planning ministries, but also to sector ministries, subnational levels of government and state-owned enterprises.
Setting priorities is only the first step. All dimensions of investment projects - economic, technical, administrative, and financial - must be appropriately designed and implemented in a policy environment that provides incentives for good performance. Priorities and quality must also be considered in allocating recurrent public spending: adequate spending on operation and maintenance will often be more important than new investment, hiring fewer civil servants and paying them competitive wages will generally be preferable to using the government as the employer of last resort, and subsidies will be more efficient when targeted to the poor rather than dispersed across the entire populace.