Forward Estimates: the Most Fundamental Tool of Medium-Term Budgeting

Posted by Marc Robinson

CABRI has just published the proceedings (Download cabri 2007.pdf) of a seminar on medium-term budgeting held in Ghana last December, under the title Are We Asking the Right Questions? Embedding a Medium-Term Perspective in Budgeting (Download cabri_2007_ch1.pdf) . It’s an interesting read.

I was particularly struck by comments in the overview paper—by Alta Fölscher from South Africa—on the fundamental importance of good forward estimates for successful medium-term budgeting. Alta stresses that a fundamental obstacle to the success of MT budgeting in African countries has been that:

the quality of forward estimates is poor. They consist far too frequently of the proposed budget for the first year of a multi-year framework, followed by inflation adjusted projections of cost for the outer year ...they pay little attention to, for example, the likely phasing of policy implementation, changes in demand that will effect spending unevenly or the impact of once-off capital spending on the base-year estimates. ...A key aspect of embedding a medium-term perspective therefore is deciding what the rules are for rolling over and adjusting and determining the forward estimates.

She is spot on – as I’m sure that anyone who has looked at a representative sample of MTEFs from around the world can attest.

This put me in mind of the stimulating blog mini-debate of the week before last (in reaction to my critique of forward-year expenditure ceilings). It underlines reasons why, in my view, we should indeed distinguish clearly between forward estimates and multi-year ceilings for ministries: namely, that a core problem has been the setting of ceilings which are not underpinned by good estimates. The result is arbitrary ceilings which, as Alta says, do not capture underlying expenditure dynamics—and which can as a consequence either underfund or overfund some ministries significantly.

Forward estimates are just that—a projection or estimate of the spending implied over the coming (say) three years by current expenditure policies (including approved new initiatives). Ceilings, on the other hand, imply an intention or a commitment to provide the ministries concerned with specified levels of funding over the next three years. It is possible to have estimates without setting ceilings, and to set ceilings without having estimates.

Having forward estimates without ceilings is not a problem. Australia, for example, produces forward estimates of ministry expenditure while making it clear that the government reserves fully the right to make change expenditure policy in any annual budget. So all the forward estimates tell ministries is how much funding they could expect in the outer years if the government doesn’t change expenditure policy. (Incidentally, this alone is a considerable help to them in reducing budget uncertainty and letting them improve planning.)

It is setting multi-year ceilings without the benefit of good forward estimates which is the problem. To specify what, say, the ministry of education should receive in three years time without having a reasonable understand of how cost-drivers such as trends in the population numbers of school-age children is to risk getting it significantly wrong.

The catch is that developing reasonable forward estimates of expenditure (and revenue, for that matter) is no simple matter, and there are few LICs or even MICs who seem to do it well. This suggests to me that we should be giving a lot more attention in capacity-building to developing the forward estimates skills of ministries of finance.

Comments

[Onanong]

I agree that government can have a good forward estimates without setting sector/ministry/department ceiling.

I wonder why lot of articles and guidelines point out that sector ceiling is one of the most important elements of the MTEF.

What is your opinion about sector ceiling?

Do you have international guidelines to implement forward estimates? I'm really interested in the medium term budgeting issue and would like to study how to implement the forward estimates.

Onanong

[David Webber]

There are two quite separate issues, or needs, here that often get rolled into one (as this discussion starts to do). Medium term fiscal planning and forward estimates of expenditure are two quite separate tasks, performed by different institutions, with different objectives. Although they can be managed in ways that help each other, good integration is rare.

Medium term fiscal planning (i.e. determing the overall level of resources which are likely to be available in coming budget years) is nearly all about fiscal strategy and deficit control. Given the vagaries of revenue forecasting, even the aggregate budget ceilings available in the out-years will only ever be "indicative" (and therefore rightly subject to later revision). The last 2-3 years have seen some especially rough times in this regard for many fiscal forecasters.

Forward estimates of expenditure, on the other hand, are only that: estimates of needs (and in some cases of agency expectations). Practices (i.e. the instructions given to spending agencies) differ considerably on whether and how they should include the costs of new policies and programs that have not yet started.

Forward estimates of expenditure are therefore frequently developed on an inconsistent set of assumptions across line agencies about what should be included, plus how the future costs of existing programs should be calculated. This is one of many reasons that MTEFs remain relatively weak fiscal tools in most countries. Limited capacity for accurate estimation of future costs, especially of capital projects, is another.

A further and more important reason in this context, however, is the inconsistency with which the overall budget ceilings, as indicated by the fiscal planning (MTFF) process, may be distributed across individual spending agency budgets. While it is invariably helpful for spending agencies to have an indicative ceiling to work to for the out-years, the forward estimates exercise is often frustrated and weakened by what turns out to be a poor indication of their resourcing.

These changes in the budget ceilings of individual agencies may be due to unexpected changes in the overall fiscal envelope and/or to (quite legitimate) changes in sectoral spending priorities. Both causes are common and MTEFs will therefore always be a very imperfect exercise.

Best then not to put great reliance on forward estimates and MTEFs, though the discipline of requiring at least 3-year program estimates from the spending agencies is still worthwhile, if only for their internal management purposes.

Finally, the example given of Australia may be a little misleading in this regard. While the forward estimates are quite rigorously prepared, they are done so within the context of central guidance on the likely availability of future budget resources. Nonetheless, this example does show that both processes can help inform each other, if they are both done well.

Of course, even in Australia, the actual availability of funding in those indicative years will always be subject to change. Multi-year appropriations (based on good forward estimates) may provide a mechanism to reduce some of this uncertainty for a selected few expenditure programs, but these must necessarily be limited in their overall fiscal implications.

[Suhas Joshi]

David makes a good point about the Australian experience and the fact that forward estimates of expenditure are frequently developed on an inconsistent set of assumptions across line agencies. Compounded with the lack of capacity in developing country line ministries, this normally leads to line ministries submitting forward estimates based on current years budgets, plus a percentage of what they feel they should seek in the outer years, much to the frustration of Ministries of Finance. As a result, Finance then attempts to “set things right” mainly by distributing ceilings across line ministries, inconsistently, in the absence of complete information, or little central guidance. This further translates into equally inconsistent ceilings at spending unit level, as Line Ministries try to make operational sense of their ceilings. Little is done to explain to these countries the fine Australian example and the background that it took the Australians several attempts to get it right, despite their capacity, and that things did not start of there as quickly as most people expect them to move now in the developing world. As a result, the forward estimates become an exercise in futility- applied in form rather than content and ends up being divorced from the budget exercise.
But t’was a famous victory!

[Udaya Pant]

Forward estimation has a complicated issue to address too, i.e., inflation and exchange rate estimation in the estimates!The variation could go either way; hence it may be worthwhile to adopt a mid-way for such estimation.

In the developing country perspective, the political factors (like changed sectoral outlay priority rightly hinted by David)could be a major issues in unstable political conditions!The importance of the Forward Estimate is, no doubt, cannot be undermind for budgeting and resource organising; more in the 'capital rationing' situations!