【正文】 . The existence of separate budgets through extra budgetary funds or significant use of revenue budgeted and accounted for on a basis, diminishes the disciplining effect of an aggregate expenditure limit. The expenditure limit is also affected by the government’s assessment of the need for countercyclical fiscal policy. Recognizing that in certain circumstances it may be necessary to accelerate the return to equilibrium GDPlevels through expansionary or contractionary fiscal policies, arguments for temporary deviations from the longterm budget balance position have to be taken into account when determining the level of total expenditure. Primarily this involves estimating the effect of the automatic stabilizers on the expenditure side of the budget. In a particularly severe downturn, it may also be necessary to plement the automatic stabilizers with discretionary fiscal policy. A policy stage where the government discusses and establishes sector allocations— preferably in a mediumterm context— provides structure and stability to the budget process. Considering the heterogeneity of constitutional arrangements, public sector management traditions and political conditions, it is difficult to generalize procedures that are appropriation all circumstances. Some elements of an allocative phase appear to be important, however. The allocation of the budget to key sectors is essentially an expression of political priorities. The amount of resources allocated to various objectives such as education, defense, regional development, infrastructure, unemployment support, or social welfare, is the characterizing difference between political parties or factions within parties. Determining the allocation of the budget to overall sectors is, from this perspective, a stage that should attract significant attention, as it is at this time that policies are set. However, a new round of budget preparation does not give unlimited freedom to reassess government policies. For a given year, most expenditure is fixed by legally binding contracts, civil service employment legislation and annual agreements on salary increases, explicit mitments to specific areas or entitlement legislation. Radical changes are rarely possible from one year to another. Budget policy can be redirected, but noticeable results may take longer than one year. When taking into account the inertia of current policies, a significant part of setting sector allocations bees a politically uneventful process of correctly estimating expenditure for policy decisions that are already being implemented. The contentious issue is, therefore, to determine how to distribute any additional room for expenditure— or if so demanded, where to make expenditure cuts— which is a significantly more manageable undertaking than making a full reassessment of government policies. In addition, the discretion wielded by the government is often limited by policy mitments made earlier through coalition agreements, party manifestos, political programs, or by statements by the president, prime minister or other representatives of the government. The allocative phase is therefore, to some extent, a process of consolidating policy mitments, and translating them into a coherent expenditure plan. This view of the process highlights the importance of having a capacity to make accurate expenditure projections. Unless the allocation to sectors is based on a qualified assessment of the cost of continuing current policies and the cost of new initiatives, its usefulness bees highly questionable. Ins