Keeping the Lid on Aggregate Spending

How can government maintain a firm “top-down” limit on aggregate spending during the budget preparation process? I suggested in an earlier blog that an answer to this could be found in a budget process based on two key principles. The first is the separation of decisions on net new spending – i.e. new spending initiatives and expenditure cuts – from decisions about baseline funding for continuing programs. The second is the imposition throughout the budget process of the constraint that net new spending must not exceed the aggregate expenditure ceiling minus baseline funding. Let me now expand on this.

As pointed out earlier, the central problem with the crude version of top-down budgeting is that it is quite impractical – other than in the very short term – to exclude bottom-up proposals for new policy from the budget process. However, an entirely top-down determination of the level of funding for ongoing programs is, in principle, possible. Based on this key point, a realistic top-down budget process can work as follows:

* Baseline ceilings providing funding for existing programs are set for each ministry at start of budget process, in an entirely top-down process. These are fixed ceilings in the sense that no additional funding for existing programs would be considered during budget preparation.

* A fixed government-wide new policy envelope is set at the same time, the allocation of the envelope between ministries being then determined during budget preparation.

* The sum of the ministry baseline ceilings plus the government-wide new policy envelope would equal the desired aggregate spending limit.

Crucially, to maintain allocative flexibility and prevent the upward pressure of new spending proposals from threatening the aggregate spending limit,

* The new policy envelope would be a net envelope, in the sense that it would limit net new spending – i.e. new spending minus cuts to existing programs.

* The baseline ceilings would be subject to downward revision as result of spending review.

Such a mechanism ensures that the quantum of new spending is, throughout the budget process, directly linked to the quantum of cuts to existing programs. With a new policy envelope applying to net new spending, the only way of adding additional new spending initiatives is by identifying additional offsetting cuts in existing programs.

To make this work, it is necessary for the ministry of finance to maintain throughout the budget process – as it does in Australia and a number of other countries – a running tally of the net impact of new spending and cuts, to ensure that the net new policy limit is not breached.

In this process, ministries would be invited to submit budget “bids” only in respect to new policy (interpreted to include discretionary expansions of existing programs), and not in respect to requirements for existing programs.

The cornerstone this process is the top-down setting of baseline ceilings. For this to happen, the ministry of finance needs to be able to estimate, with confidence, the “unchanged policy” spending levels of ministries and then use these estimates to determine the baseline ceilings. This requires good quality expenditure forward estimates – which, as I have previously pointed out, are the most fundamental tool of medium-term budgeting.

Of course, this means that the process outlined above can only be applied in unmodified form in countries where the ministry of finance is able to prepare reliable expenditure forward estimates.  Many countries – and not only developing countries – are not able to do that. What to do in such countries is one of a number of issues I will consider in fleshing out the above outline in subsequent blogs.

One Response to “Keeping the Lid on Aggregate Spending”

  1. Joanne says:

    Dear Marc

    I spent a short time working in the New Zealand Treasury so am familiar with the approach that you describe in “Keep the Lid on Aggregate Spending”. However, this approach works well when existing or on-going policy has been fully implemented so that the baseline cost, for instance, of delivering a certain standard of education to all, is known and the policy framework is clear. The annual budget round is therefore centered on new policy proposals and the public policy arguments that support their introduction. However, the first line of defense in The Treasury was to always look first for ways of funding them from within the existing budget baseline.

    As you say, applying the same approach is much harder in less developed countries where the stock of existing policies is often unknown, too numerous and only partly funded. Hence the existing baseline is not yet at a level sufficient to ensure quality public services and is very often scattered and divided thinly across numerous ministries, departments and other public bodies and their associated budget lines (concentrated on wages and allowances and departmental costs, electricity, vehicles and fuel). Adding donors tends to complicate matters further.

    As such, I look forward to the follow-up article on “what to do in such countries”.