The Treasury

Global Navigation

Personal tools

1  Fixed nominal baselines

1.1  Background

Under the PFA, government financial reporting moved from an input to an output basis. Previously, appropriations for each department were made based on three input types (personnel, operating costs and capital). The PFA required that appropriations be specified according to the nature of the outputs (i.e. goods or services) produced by each department.

Under the old system, the budget process involved making regular adjustments to personnel costs based on the outcome of public service wage negotiations. Similarly, the other two streams were generally adjusted annually to reflect expected cost movements. Government budgets were made only for the year to come, and no forecasts were presented for spending in subsequent years.

Under the PFA, appropriations for departmental spending were instead based on an output price[2]. As the funding for particular inputs was no longer specified in the annual appropriation, departmental chief executives were free to determine which inputs to purchase in order to provide the contracted output at the specified price. This price was no longer subject to automatic formula-based adjustments based on input cost changes, but instead was notionally benchmarked to a hypothetical market price for the good or service being provided.

The reality however is that few departmental outputs can be benchmarked to an external market price. In the absence of evidence justifying a price increase, there is no mechanism in the budget process to increase an output price from year to year. To the extent that departments are required to manage input cost changes within existing funding levels, the effect is a real saving in departmental spending each year equivalent to the rate of inflation.

The removal of automatic adjustments for departmental outputs is referred to throughout this paper as “fixed nominal baselines”[3].

1.2  Impact on fiscal and budget management

Fixing nominal baselines meant that, for forecast purposes, government spending could be characterised as being split into two tracks: “formula-driven” (i.e. indexation) and “fixed” (i.e. no change to nominal amounts).

Indexation still applies to much non-departmental (or Crown) spending. In particular, welfare benefits are adjusted annually for inflation, superannuation payments are indexed to the average wage, and health and education spending are automatically adjusted for demographic changes.

The nominal value of the majority of all other spending is assumed, for forecast purposes, to remain constant over the forecast period. A specific policy decision of the government is required to change the amount spent on non-indexed outputs. Any such increases must be traded off against other new expenditure priorities, and form part of the annual budget negotiations.

A dual budget process has developed. As part of the Budget Baseline Update, departments can request changes to their three-year forecast budgets (or “baselines”) for formula-driven items, based on changes to demographics, inflation or wage growth as appropriate. Any other spending increases are agreed through the Budget Initiatives process, where increases have to be met from a limited pool of funds allocated for new budget spending, and traded off against spending proposals in all other areas.

In contrast to the previous system the burden of proof is now on the department to demonstrate that it will not be able to deliver services effectively within existing funding levels. In practice this has proved to be a high hurdle. Very few departments have had funding explicitly agreed for price increases, as the general assumption is that departments should be able to make annual efficiency gains that are at least equivalent to the rate of inflation.

While this is the general position, some departments may have found other ways to respond to input pressures. For example, departments may have responded by redefining their core business more narrowly and bidding for funds for projects that would, in the past, have been carried out within baselines. Similarly, in the case of some large departments (e.g. Police), regular increases in baselines have been agreed, but the budget process hasn’t provided the leverage required where output measurement is imprecise, and it is not clear that additional outputs are being produced.

1.3  Benefits of fixed nominal baselines

Fixing nominal baselines has often perceived as a mechanism that reduced growth in public expenditure. However, the impact of introducing the policy needs to be evaluated in the context of other concurrent changes.

Between 1991 and 1993, at the same time as it put a halt to most automatic expenditure increases, the government made cuts to all departmental baselines of between 1% and 5%. In addition, many government agencies have downsized since 1989; and it seems likely that efficiency gains have been made over the period in most areas (information technology being a key driver). However, it is difficult to isolate the impact of removing indexation from that of other public sector reforms. It is also difficult to measure what changes in quality or quantity of outputs have taken place over this period.[4]

In considering whether the policy did serve to reduce expenditure growth, it is useful to consider how budget targets are formulated. Since 1994, operating balance (or budget surplus) targets in the annual budget process have generally been set within the framework of the FRA (discussed in section 1.5). That is, annual expenditure targets are initially determined by a ‘top-down’ process based on long-term fiscal objectives and revenue forecasts, rather than a ‘bottom-up’ process driven by indexation processes. This suggests that any savings made by holding most spending on existing departmental outputs fixed would likely have been “spent” via an increase in the amount of funding available for allocation to new programmes in the budget round.

The real gain obtained by removing indexation therefore relates to an increase in the government’s ability to scrutinise and reprioritise spending on existing outputs. Departments are now required to submit a formal budget bid for any compensation for the loss in real resources due to fixing nominal baselines. This gives Ministers the opportunity to compare the benefits of increased spending on existing programmes with those that might arise from spending on new programmes. Rather than reducing spending in total, fixing nominal baselines therefore facilitates the reallocation of spending from one area to another. This is what makes the policy a key tool in improving the effectiveness of government spending as a whole.

1.4  Risks of fixed nominal baselines

Criticism of the policy has focused on those areas of government spending that may be a low budget priority for Ministers but are nevertheless essential services (e.g. elements of government administration). These areas may find it difficult to maintain their real level of resourcing via budget negotiations and service delivery may be put at risk. In the mid 1990s, in response to concerns that departmental capability in some areas might be suffering through under-funding, an Output Price Review mechanism was put into place. Departments facing input cost pressures were able to request a review of output pricing levels, overseen by an interdepartmental officials committee. In the last five years only five such reviews have been carried out, with mixed results (in one case funding levels dropped).

Another risk is the volatility of the real efficiency gain each year, which varies with the rate of inflation. An alternative would be to index baselines to inflation, and simultaneously extract a predetermined efficiency dividend.

We believe however that the fixed nominal baselines policy has two distinct advantages:

  • simplicity: there is no need to introduce into the process negotiations and decisions around the appropriate indexation levels; and
  • certainty: departments can plan based on known nominal levels of future funding.


  • [2]Output prices are fully costed, contain no hidden subsidies, and where possible are based on market based comparisons. The original input costs also influenced the output price established.
  • [3]The “baseline” is the agreed budget allocation over the forecast period.
  • [4]For a fuller discussion of the expenditure impact of public sector reforms, see Petrie and Webber (2001).
Page top