Faced with the increasing complexity of government activities, many countries have fallen back on the idea of the cash budget. This has the merits of simplicity and comprehensiveness. As used in the United States, it presents total payments by the federal government to and from the public (including other levels of government). It is thus similar to the cash flow account of a modern business. Trust fund expenditures and receipts are included, as well as cash payments and receipts involved in loan transactions. Government business undertakings such as the post office, however, are still included on a net basis.

In the United Kingdom all public expenditure planning is now performed on a cash basis, and many programs are “cash limited,” whatever the level of inflation. This procedure, to which the United Kingdom moved in 1976, is justified on the grounds that such treatment helps to control inflationary pressures and exerts stricter control than, for example, planning in volume terms.

The cash budget suffers from the defect that it is not directly tied to government decision making. Liabilities incurred do not synchronize completely with payments. This is because government expenditures result from appropriations and other forms of commitments; cash expenditures may follow appropriations and other commitments of money only after a considerable lag, notably in the case of construction and procurement. Appropriations relate to actions in the future. Expenditures result from past decisions. Both kinds of information are needed for a complete appraisal.

The U.S. government, in an effort to reduce public confusion over the large variety of budgetary concepts, has adopted a so-called unified budget concept that is more logical than the cash budget but differs from it only in some details that do not materially affect the budget aggregates. The unified budget differs from the traditional administrative budget in two main ways: it includes the receipts and outlays of most funds, and it eliminates interagency transfers.

Traditionally, government expenditures have been considered as inputs rather than outputs. This is because, in the classical 19th-century conception, the well-run government does not produce a marketable output. The program budget derives from this concept; it attempts, however, to classify expenditures in terms of the outputs to which they are devoted. For example, a traditional school budget would categorize expenditures in terms of teachers, books, and buildings; what came out of the process would be left to the reader’s intuition or experience. The program budget, in contrast, attempts to assign expenditures to specific outputs, categorizing them according to numbers of children completing various programs.

In government, budgets have traditionally been constructed according to departments and agencies of government. This may be justified on historical or administrative grounds, but it does not necessarily correspond to the structure of activity. Every country organizes the civilian and military components of its foreign policy in separate departments, but this is frequently a serious obstacle to effective policy making. Again, the requirements of good administration suggest that there should be a single department of agriculture. But that department’s activities impinge on those of others, in both domestic and foreign policy. A budget constructed according to actual programs would cut across departmental boundaries.

Program budgeting is an attempt to apply the economics of choice to public decision making. Its basic assumption is that explicit choice among alternative courses of action leads to better results than do other methods of decision making. At the highest governmental levels difficult choices must be made that involve the use of a portion of the nation’s resources. But the same principles apply to decision making at lower levels. The problem of allocating resources within a specific field, such as health or education, is conceptually similar to that faced in drawing up the national budget.

Program budgeting also takes account of the time dimension in many government programs. New undertakings often take time to come into operation. A typical new program may have to pass through a research and development phase and an investment or construction phase before it reaches the operating phase. Alternative programs may differ considerably in this respect. The process of choosing among alternatives frequently involves trading the present against the future. One alternative may require 10 years before it yields results; another may yield smaller results but more quickly. The kinds of choices made in government often involve alternatives that cannot be measured in terms of market value. For this reason governmental decisions involve much more uncertainty than do most business decisions.

A governmental program must therefore be frequently revised in the light of unfolding circumstances. Indeed, every year should be thought of as the first year of a new program. Pervasive uncertainty also requires a high degree of flexibility and a capacity for program revision. A number of options should be held open, particularly in the development phase. Even though this may appear costly, it is less costly than commitment to a design that proves to be inappropriate because of circumstances that could not be foreseen in the early stages.

In most countries the usual procedure for deciding on government expenditure in a forthcoming year has been to assume that existing expenditure was appropriate and then to decide on incremental expenditure for each program. Such an approach means, however, that the change is likely to increase, rather than decrease, expenditure and that little attention is paid to what the full existing program actually accomplishes.

In the late 1970s many countries recognized that the steady growth in public expenditure was putting a strain on their economies, and they attempted to curtail the growth. The Jimmy Carter administration in the United States, although planning for a steep rise in expenditure as a proportion of gross domestic product (GDP), also attempted to introduce the concept of “zero-base budgeting,” whereby the entire government program, not just its incremental parts, was to be evaluated each year. This idea, which involved considerable changes to existing procedures, was applied to some programs on a selective basis but never had the impact its designers envisaged. A similar attempt was made in the United Kingdom in the introduction of program analysis reviews (PAR), but again attempts to evaluate systematically the whole of government expenditure were unsuccessful. The degree of inertia in the system and the vested interests of existing institutions have proved too entrenched to be overcome by administrative procedure.