Marc Robinson
Public Financial Management Results
Go to content

The IMF on Budgetary Institutions

(Originally written in June 2014)

The IMF has just published a useful paper entitled Budgetary Institutions in G-20 Countries: An Update. This paper present numerical ratings of G-20 country budgetary institutions, based on an explicit set of criteria which is a revised version of a rating methodology first presented in 2010.

The idea of formally rating budgetary institutions based on explicit criteria is an interesting one, and the Fund is to be given full marks for the effort. The majority of its criteria are sensible, and some of the new ones it has added make good sense.

Nevertheless, a number of the criteria which have been used for rating fiscal institutions are based on questionable assumptions, including the following:

  • The view that it is necessarily good practice for all government expenditure to be subject to annual budget authorization, rather than authorization by standing legislation (see criterion 9(b)). However, contrary to this, social security and other similar expenditure is generally based on standing legislation in OECD countries. This is because such expenditure is based on entitlement rules which remain largely unchanged from one year to the next, and the standing legislation authorizes not amounts of expenditure, but rather whatever level of expenditure is determined by the existing entitlement rules. To give continuity to the entitlement rules, it is entirely reasonable and appropriate to base such expenditure on standing legislation rather than the annual budget. What is important is not whether such expenditure is incorporated in some artificial way in annual budget authorizations, but that it is fully factored in as part of the fiscal objectives and framework which accompanies the budget.
  • The insistence that top-down budgeting (TDB) necessarily means setting firm sectoral or ministry ceilings at the start of the budget preparation process, prior to any consideration to spending ministry new spending proposals. However, contrary to this, TDB properly conceived means only setting a firm aggregate spending ceiling, and does not demand that the allocation of this aggregate ceiling between sectors and ministries be locked in before the budget preparation process seriously starts. To lock in sector or ministry allocations at that stage runs the risk – in many countries – of seriously aggravating allocative inflexibility and worsening the incrementalist tendencies of traditional budgeting. I have presented a detailed discussion of these points in my 2013 article in the OECD Journal on Budgeting, and also in this blog.
  • A degree of enthusiasm for limiting the magnitude of budget carryovers from one year to the next such that a country which totally prohibits any carryovers would earn top marks on the relevant criterion (criterion 12(b)). This surely is a case of throwing the baby out with the bath water. Don't we all recall that the introduction of carryover provisions was a major reform of the earlier generation of PFM reforms, intended to reduce the incentives for "end of year spend-ups"? Sure it is appropriate to avoid the types of problems which have arisen in the UK and certain other countries with the built up of excessive accumulated carryovers, but this does not mean that banning carryovers altogether is good practice.
  • The proposition that fiscal objectives should necessarily cover the consolidated central and sub-national governments, and that it is best practice that there be "a legal framework for coordination and sharing the burden of fiscal policy between layers of government" (see criteria 8(b), (c) and (d)). This indicates a failure to understand the federal principal. In countries such as Switzerland, Australia and Canada, the national government is by virtue of the nature of the federal system in no position to set fiscal objectives for sub-national governments, let alone order them to comply with these. One might wish that the national government had such a power, but such a wish would represent a desire to change the political system. The focus should be on improved voluntary coordination. We all understand that the European Union is forcing all its member countries to adopt fiscal rules which cover both central and sub-national government, but we should also be aware of the specific challenges of making this work in federal countries such as Germany and Spain.
The inappropriate criteria concerning the coverage of fiscal objectives seems to indicate an inadequate understanding of differences in political systems which is apparent in other criteria which are employed. One such criteria asserts that it is necessarily best practice that programs – the budgetary categories on which a program budgeting system is based – be used as the basis of budgetary appropriation by the parliament. However, this is something which a number of countries with political systems drawn from the British system (especially Australia and Canada) deliberately do not do. Despite the fact that Australia and Canada have program budgeting systems, the parliament does not appropriate the budget in terms of programs, but leaves the executive free to allocate and reallocate monies between programs. This approach – sometimes called a "global appropriations" approach – is not accidental, but reflects the view widely held in these countries that the allocation of resources in the budget is essentially an executive government (as opposed to legislative) prerogative. Or more precisely, that an executive government with a majority in parliament is entitled to allocate resources as it wishes and have its budget endorsed by its parliamentary majority. One might wish that they took a different view, and asked parliament to vote on program allocations, but this would again be asking for a change in those countries' view of the appropriate balance of budgetary power between executive and legislative wings of government.

It is notable in this context that the paper indicates that during its development, a number of countries requested that more consideration be given to differences in "the structure of the political system or the role of the legislature".

While in general well drafted, the paper could also benefit from tidying up on a range of points, such as:

  • The apparent contradiction between at one point calling for aggregate expenditure ceilings to cover "discretionary" expenditure (p. 29) and at another point asking that they cover as much government expenditure as possible (criterion 6(d)).
  • The conflicting definitions of fiscal rules as requiring "lasting" or "permanent" fiscal constraints (Box 3 versus criterion 5(c)). As I pointed out in an article in the OECD Journal some time ago, the original IMF (Kopits-Symansky) definition of fiscal rules as "a permanent constraint on fiscal policy, expressed in terms of a summary indicator of fiscal performance" is too narrow because some fiscal rules intended to apply for a defined period of time, or indefinitely, without necessarily being intended to be permanent.[1] ... It is more appropriate that fiscal rules be defined as "limits on expenditure which are formulated in such a manner as to have continuing application" – a formulation roughly equivalent to the loose notion of a "lasting" fiscal constraint.
Despite these weaknesses, the IMF Budgetary Institutions work represents a major contribution to the ongoing work to strengthen national budgetary institutions so as, amongst other things, to reinforce the capacity of governments to give effect to sound fiscal policy objectives. With some further review of the rating criteria used, the value of the Fund's framework could be significantly enhanced.

[1] I cited as an example of a fiscal rule (more specifically, an expenditure rule) which was intended to have continuing but not permanent application was that enunciated by the Australian government in 2009 as part of its deficit exit strategy from the Global Financial Crisis. Under this strategy, the Australian Government committed to return the budget to surplus by restraining real growth in spending to 2 percent a year once the economy recovered to grow above trend. Once the budget returned to surplus, and while the economy is growing at or above trend, the government committed to retaining a 2 percent annual cap on real spending growth, on average, until surpluses are at least 1 percent of GDP.
Back to content