Back to top anchor
Working paper

Budget Management That Counts: Recent Approaches to Budget and Fiscal Management in New Zealand (WP 01/24)



Budget management in New Zealand altered substantially with the implementation of the Public Finance Act 1989 and the Fiscal Responsibility Act 1994. The paper sets out the evolution of fiscal and budget management over the last ten years, in response to the fiscal policy and financial management framework. It focuses on the top-down management of spending aggregates and how this has evolved into the “fiscal provisions” framework. The paper concludes with an illustration of the challenges facing the provisions framework. Work is currently underway to consider the challenges that have arisen through its operation over the last five years. This paper is a companion paper to Treasury Working Paper 01/25 New Zealand’s Fiscal Policy Framework: Experience and Evolution by John Janssen.


The views expressed in this Working Paper are those of the author(s) and do not necessarily reflect the views of the Treasury. The paper is presented not as policy, but with a view to inform and stimulate wider debate.



A series of reforms in the New Zealand public sector, between 1984 and 1994, radically changed how the financial management framework operated. The most visible changes resulted from the Public Finance Act 1989 (PFA) and the Fiscal Responsibility Act 1994 (FRA). The reforms have been widely publicised and are not discussed here.[1]

This paper will review how fiscal and budget management has evolved over the last ten years, in response to the financial management framework. We focus on the top-down management of spending aggregates. Specifically our two key tools for baseline management are:

  • fixed nominal baselines for departmental spending; and
  • the fiscal provisions framework.


  • [1]* We wish to acknowledge the review and comments provided by members of the Budget & Macroeconomic Branch, including Mark Ahern, Andrew Crisp, Kirsten Jensen and John Janssen. We also thank participants at the Australasian Treasury Officers Conference, October 2000, for comments on an earlier version of the paper. This paper is a companion paper to Treasury Working Paper 01/25 by John Janssen.

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).

1.5  Interaction with the long-term fiscal framework

The mid-1990s saw further developments in the fiscal management framework. In 1994 the FRA was passed, requiring the Government to establish long-term objectives and track progress against them. At the same time, New Zealand was moving out of a period of fiscal crisis. Government spending began to increase more rapidly, though still at a rate less than revenue growth.

Budget management processes needed to adapt to the challenges of a surplus environment. A key issue that emerged was the relationship between fixed nominal baselines and the fiscal forecasts. Three-year budget forecasts prepared under GAAP[5] between 1994 and 1996 would include increases in government spending only for those areas that had automatic indexation. All other spending was assumed to remain constant over time. While this assumption could arguably have been justified over the early 1990s during the period of fiscal crisis, by 1994 this was no longer the case.

Because the fiscal forecasts did not allow for increased spending in future Budgets, they understated the likely spending profiles. This resulted in optimistic projections of the progress towards the long-term fiscal objectives, which reduced the discipline on the annual Budget process[6]. This forecast “bias” is illustrated below:

Operating Balance for 1997/98 – Change in Forecasts


Operating Balance for 1997/98 – Change in Forecasts.

There were also pressures from a political perspective to find a better way to represent spending intentions in the forecasts. In 1996, New Zealand moved to a proportional representation voting system and the first Coalition Government was elected. The government formation process threw up two challenges for the coalition:

  • to alleviate any concerns that a coalition government would usher in a period of instability, incoming Ministers needed a way to demonstrate fiscal prudence;
  • there was now an opportunity for portfolio ministers from different coalition parties to bid up spending in their sector, by turning the issue into a coalition dispute.

Both pressures could be managed if some kind of overall spending cap was agreed by both parties. The response was a statement from the new government committing to a $5 billion cap[7] on new spending over a three-year term of government to 1999/2000, on top of changes already included in the forecasts (i.e. on top of the formula-driven items).

Treasury officials were charged with working out how the $5 billion spending cap would work in practice. The cap evolved into a mechanism now known as the “fiscal provisions”.


  • [5]Generally Accepted Accounting Practice – use of accounting standards that in New Zealand are the same for the private and public sector. GAAP is defined in the PFA.
  • [6]The 10-year progress outlooks did include higher spending track scenarios and therefore did provide some measure of a more realistic path towards long-term objectives.
  • [7]Equivalent to 1.6% of GDP on average across three years.

2  Fiscal provisions

2.1  What are the fiscal provisions?

The provisions framework consists of a pre-determined fiscal limit across the parliamentary cycle (three years, generally from November to November), and a set of principles and rules for “counting" against that limit. The limit is a pool of funds available for implementing government policy. The principles and rules establish when the fiscal implications of policy decisions will be “counted” against the limit.

When fiscal forecasts are produced, the provisions sit “on top of” the projected expenditure for existing government programmes and policies. This means they represent the operating balance impact of changes to existing policy, whether this be an increase in funding for an existing policy or funding for new policy. The framework focuses on ministerial decision-making and, therefore, only applies to discrete policy decisions. This means the framework excludes those automatic forecasting or indexation changes that are factored into baselines based on previously agreed parameters, with little ministerial input from year to year.

This means that the provisions framework built onto and extended the existing practice of using fixed nominal baselines for most departmental spending, while allowing forecasting changes (i.e. formula-driven items) to fluctuate with the state of the economy. For example, an increase in benefit payments due to higher unemployment would not be financed from (or “count against”) the provisions. If, on the other hand, the government decides to increase the amount of the benefit payment, the resulting increase in forecast benefit payments would count against the provisions.

Beyond the three-year Parliamentary cycle, forecasts include a proxy for the provisions that is referred to as a “technical”[8] provision, representing potential future policy decisions to be made as part of future Budgets. The technical provision is determined by the Treasury (and agreed to by Ministers) based on an approximate average of the cost of new policy over the last few years.[9]


  • [8]The term “technical” is used to emphasise the fact that as yet there is no political commitment to the projected expenditure target. In the 2001 Budget the “technical” provisions were altered to “indicative” provisions to provide a better indication of potential future spending (outlined in the Fiscal Strategy Report page 23 of B.2 & B.3).
  • [9]In the current 10-year fiscal projections required as part of the Fiscal Strategy Report, the amount set aside beyond the fiscal forecast horizon is termed a “fiscal allowance” as opposed to a fiscal provision. The allowance indicates the fiscal flexibility the Government has for spending and revenue initiatives, contributions to partially pre-fund future New Zealand superannuation pressures, and responding to developments in the economic and fiscal position. It provides a broad indication of flexibility rather than a specific policy commitment.

2.2  How are the provisions operationalised?

New Zealand is now partway through the second set of three-year fiscal provisions. Set out below are the steps we went through to operationalise the framework. Since 1996 these processes have been gradually refined and institutionalised.

Step 1: Agree the amount of the provision

The first step was to advise the incoming government on an appropriate three-year target. Establishing fiscal provisions requires a Government to consider what its long-term objectives are in relation to the operating balance, debt reduction, and future expense pressures. This will determine what resources are available in the short-term.

By way of example, the current Government’s fiscal provisions have been set at $5.9 billion (GST inclusive) over the 1999/2000 – 2002/03 period[10]. The calculation was based on the most recent set of fiscal forecasts at the time (October 1999), and the Government’s potential longer-term parameters.[11]

Step 2: Allocate the provision across three years

The second step was to allocate the total provision over the three-year period. The original $5 billion provision was defined as a cumulative, three-year total (see diagram below), and this framework is still used. The provision needs to be phased across the three years in accordance with the expected profile of decisions. In illustration, the current government’s three-year fiscal provisions are set out below:

Cumulative Provisions – Including the 2000 Fiscal Strategy Report Extract


Cumulative Provisions – Including the 2000 Fiscal Strategy Report Extract.

The expense track includes a fiscal provision of $5.9 billion (GST inclusive) for the next three years[12] to be used on priority policy objectives. The allocation of the provision has been finalised during the Budget 2000 process on a cumulative basis at (GST inclusive) $420 million for the remaining months in 1999/2000, $1.2 billion in 2000/01, $1.8 billion in 2001/02 and $2.4 billion in 2002/03 (0.4%, 1.1%, 1.5% and 2.0% of GDP respectively).

Provisions are cumulative, assuming that a decision taken to increase a department’s baseline in 2000/01 will generally represent a permanent increase. The provisions provide for the annual increase to “roll out” into the following years. Each individual cumulative provision totals to the $5.9 billion total allocation.

The allocation of a budget target over a three-year period represented a step forward for budget management. Previously, the focus had been on the “annual” budget pot available for distribution. A three-year spending programme allows the government to phase in more significant initiatives such as tax cuts, while showing how they could be credibly funded as part of a three-year policy process. It also enables Ministers to deliver significant one-off packages in key portfolios, rather than drip-feeding annual funding increases. For example, Ministers would have the option of deferring a suite of environmental initiatives to a later budget in order to enable significant tax cuts in a current budget.

Step 3: Clarify the “principles” governing impacts on the provision

The third step was to establish the “principles (or rules)” that determine which items would be treated as forecast changes, and which would be treated as specific policy decisions that would “count” against the provisions. Development of the principles turned out to be far from simple, and the principles have evolved considerably since 1996 (Annex One outlines the current guidelines as set down at October 2000).

Compared with the previous system, the key innovation in developing the principles was to include all policy decisions that impacted on the operating balance. In the past, tax policy decisions and policy decisions affecting state-owned enterprises and Crown entity surpluses had tended to be made in isolation of “spending” budget decisions. These have now been explicitly brought into the budgeting framework to enable the Government to consider all policy proposals (that impact on the operating balance) as part of the same decision-making process.

Step 4: Implementation

The final step was to implement the framework through:

  • informing ministers and departments of the allocations and principles (by issuing Treasury circulars);
  • adjusting Cabinet Office procedures by requiring that all Cabinet papers state the impact on the fiscal provisions of specific policy recommendations;
  • establishing Treasury procedures to record decisions impacting on the provisions;
  • adjusting budget processes: the original Budget Baseline Update process has now been redefined to exclude any proposals impacting on the provisions – all such proposals must now be considered as part of the Budget Initiatives round;
  • incorporating reporting against the fiscal provisions in the budget documents. Targets are stated in the Budget Policy Statement; and the impact of policy decisions is documented in each Economic and Fiscal Update (Annex Two provides an example from the 2000 Budget documentation illustrating the level of transparency of disclosure).

In effect the fiscal provisions have set up a second fiscal monitoring process, in parallel with the monitoring of the overall operating balance.

2.3  The provisions tell only part of the story

The fiscal provisions are only one element of the operating balance – they represent the discretionary policy decisions that are most easily controllable by governments in the short term. Other elements – such as benefits and revenue changes, or valuation changes, are less amenable to the direct short-term control of the Government. The provisions framework therefore helps ensure that annual budgeting decisions can focus on shorter-term pressures, while longer term pressures are able to be addressed either by the three-year framework provided by the total provisions target, or by the medium-to-long term targets established in the Budget Policy Statements and Fiscal Strategy Reports required under the FRA

One obvious driver of the operating balance that is beyond the direct short-term control of the government is the impact of the business cycle[13].

A second driver, with the move to accrual accounting, is fluctuation due to changes in liability valuations. At the moment, two significant liabilities (Accident Compensation Corporation outstanding claims obligation and Government Superannuation Fund (GSF) unfunded liability) are valued on a fair value basis and subject to discount rate changes. New Zealand GAAP is heading in the direction of applying the fair value approach to a wider range of assets and liabilities, and therefore increasing the potential impact on the operating balance.

By way of example, for the year ended 30 June 1999, the GSF unfunded liability increased by around $600 million (0.6% of GDP) from that forecast – due largely to a change in how the discount rate was calculated. No change was required to the fiscal provisions, as no long-term threat existed to the achievement of the Government’s objectives (see Annex Three for an illustration of how focusing on fiscal provisions influences the trend of the operating balance – and why short-term fluctuations from “fair value” changes need to be considered from a longer-term perspective)[14].

In setting the three-year target for the operating provisions, the Government faces uncertainty regarding economic performance and likely cost pressures. The Government needs to take into account the fact that fiscal plans may need to change if circumstances turn out different to current forecasts. Key uncertainties that impact on the operating balance include:

  • the economic outlook and its impact on fiscal parameters - economic conditions impact directly on revenue streams, finance costs and beneficiary numbers;
  • changes in the value of Crown assets and liabilities - for example, a decrease in real interest rates increases the value of both the ACC liability and GSF liability, therefore reducing the operating balance.

These impacts, driven by the economic cycle, are beyond the immediate control of the Government. As a general approach, we would recommend that no change be made to the provisions target in response to forecast changes due to the factors described above. This allows the automatic fiscal stabilisers[15] to operate through the cycle.

However, it may be appropriate for the Government to consider adjusting the provisions in the case of an economic shock that is likely to have long-lasting impacts. For example, at the time of the 1998 Asian crisis forecasts and projections showed that progress towards the previous Government’s long-term objectives was significantly reduced. See Janssen (2001) for details.

In response, the previous Government agreed to progressively reduce the three-year provisions target of $5 billion to $4.25 billion[16]. The result was to ensure steadier progress was made over the three-yea period towards the long-term objectives.

Essentially, the provisions provide the Government with an opportunity to credibly demonstrate that it is following through on its fiscal intentions. The operating balance is subject to too many other factors to be a complete measure by itself in the short-term.

2.4  Benefits of the provisions

Progressively through the last three years, the provisions framework has become:

  • the way in which the Government can demonstrate that short-term fiscal policy is consistent with achievement of longer-term objectives;
  • the means of improving budget decision making, to ensure a more strategic three-year focus rather than solely an annual focus; and
  • a tool for Ministers to focus on a more definable and manageable “target” than the operating balance (as discussed in section 2.3 above).

A more realistic total spending profile was (and is)presented for the forecast years. The result has been:

  • fiscal forecasts that provide a better indication of expected progress towards the Government’s long-term objectives (therefore enhancing the credibility of the fiscal forecasts); and
  • that the fiscal provisions have become one of the anchors of fiscal policy by which the Government is judged.

The increased prominence that commentators have placed on the fiscal provisions has increased the focus of the Government on its decision-making, and therefore introduced greater certainty into fiscal planning. The Government can see what it needs to do to meet its fiscal objectives, and knows whether it can meet its policy and fiscal objectives without running out of resources.

Fiscal provision limits support sound day-to-day fiscal management. By defining an overall limit first, the framework has focused decision-making on trade-offs between policy options. All decisions impacting on the operating balance, whether spending, revenue or SOE surplus decisions, are considered in a common framework.

In terms of budget management, ministers and chief executives have clear signals on which to base expectations of new resources in each budget round. Provided that the limit is sufficiently tight, ministers and chief executives will be encouraged to reprioritise within their budgets. This should also help the Government deliver value for money.


  • [10]Equivalent to 1.7% of GDP on average across three years.
  • [11]Long-term fiscal parameters included:
  • [12]The $5.9 billion provision covers three years. However, because the parliamentary term begins in November and the financial year begins in July, the expenditure needs to be allocated across four financial years. The small amount of expenditure allocated in 1999/2000 year provided for new government initiatives implemented before the Government’s first budget took effect from 1 July 2000.
  • [13]Refer Buckle, Kim and Tam (2001) and Tam and Kirkham (2001).
  • [14]In the 2001 Budget, a new fiscal indicator was introduced – OBERAC – Operating Balance Excluding Revaluations and Accounting policy Changes. OBERAC removes revaluation movements and accounting policy changes to provide a measure of underlying financial stewardship. For a fully explanation refer New Zealand Government (2001) pp 76 .
  • [15]In considering the fiscal position, it is important to focus on trends and to abstract from short-term fluctuations about the trends. If government spending and tax plans remain unchanged during an upswing, higher tax revenues and lower spending on unemployment result in improving fiscal surpluses (and vice versa in the case of a downturn). These fluctuations occur because actual output in the economy fluctuates about its productive potential. Allowing these so-called “automatic stabilisers” to operate across the economic cycle takes pressure off monetary policy and reduces fluctuations in employment and output. For these reasons the Government’s general approach is to let these automatic stabilisers operate unhindered.
  • [16]While the previous Government reduced the fiscal provisions, other decisions made at the time had more of an influence over the longer-term – ensuring that the long-term targets could still be attained.

2.5  Challenges that remain

The fiscal provisions framework is proving a useful tool for fiscal policy. However, as with any tool, it evolves as experience develops. Some future challenges are briefly scoped below:

Rolling out provisions beyond the parliamentary term

The fiscal provisions framework is established for the parliamentary term. However the fiscal forecasts do extend beyond the term, more so as the term progresses. As noted in section 2.1 “technical provisions” are included to ensure a realistic “spending” profile is maintained.

A problem occurs when a government makes decisions that impact beyond the horizon of its fiscal provisions. Such spending increases beyond the period of the provisions are not captured by the provisions framework. The technical provisions for future spending beyond the current term of government are not automatically reduced in this case, as no government commitment has been made to maintain those forecast expenditure levels. This is termed the “bow wave” effect (to describe expenditure creep beyond the current budget period). Mechanisms may be required to limit the extent of this problem.

Incorporating capital into the framework

At the moment, a capital provision, which links to the Government’s debt objectives, exists alongside the operating provisions. However, the setting of a capital target has not proved to be a totally convincing framework for assessing or controlling capital spending[17].

Work is currently underway that seeks to improve the capital allocation process.

Institutionalising the fiscal provisions framework

Although the fiscal provisions framework is prominent in the government’s budget and fiscal reporting, it remains an informal control mechanism. Further institutionalisation[18] could help ensure the continuation of the strengthening of the linkage to the longer-term objectives.

For example, in 1999 it was uncertain whether the incoming Labour/Alliance Coalition Government would agree to continue to use the fiscal provisions framework. In the event they found it a useful tool to signal fiscal intentions. However, it will be important to maintain continuity of the principles of operating the framework from one term of government to the next.

However, institutionalisation of informal mechanisms that may need (and likely will need) to change to reflect a changing operating environment could be inappropriate.

Operation of the provisions framework

Two particular issues include:

  • Determining whether something impacts on the provisions can sometimes be a complex exercise. In applying any principle based approach, issues at the margins will always arise – causing differences in interpretation and application of the fiscal provisions framework.
  • Despite the provisions framework being an informal control mechanism, the predetermined provision levels effectively become “locked in” and it is seen as difficult to alter the limits. There may be good reasons why the limit needs more mechanisms to enable review and alteration in response to changing circumstances.

Currently underway is a review to bring together the experience over the last five years with the aim of improving (and simplifying) the decision making processes, addressing some of the anomalies that can occur at the margins of the framework, as well as a more general review of the provisions framework itself.


Continued education on the link to long-term objectives, and the fact it is only one element of the operating balance, is important. Ongoing GAAP developments are likely to see the introduction of greater potential for fluctuations resulting from fair value assessments of assets and liabilities altering through time.


  • [17]Capital charge is at present excluded from the operating provisions framework (as it has no net impact on the operating balance). A possible alternative would be to “count” capital charge costs associated with increases in operating spending against the operating provision. This would act as a proxy for the resulting increase in Crown finance costs, although it would reduce the linkage between the provision and the operating balance.
  • [18]One suggestion that could be considered is whether external auditors should review the impact of policy decisions on the provisions target – enhancing the framework by introducing an external monitoring mechanism.


John Janssen (2001), “New Zealand’s Fiscal Policy Framework: Experience and Evolution”, Treasury Working Paper 01/25, The Treasury, Wellington.

Julie Tam and Heather Kirkham (2001), “Automatic Fiscal Stabilisers: Implications for New Zealand”, Treasury Working Paper 01/10, The Treasury, Wellington.

New Zealand Government (2000), Budget Speech and Fiscal Strategy Report, Wellington, New Zealand, June 15, pp26.

New Zealand Government (1994) Fiscal Responsibility Act, 1994, Wellington, New Zealand.

New Zealand Government (1989), Public Finance Act, 1989, Wellington, New Zealand.

New Zealand Government and The Treasury (2001), Budget Speech, Fiscal Strategy Report, Economic and Fiscal Update, Wellington, May 24, pp23, 76.

Petrie and Webber (2001), “Review of Evidence on Broad Outcome of the Public Sector Management Regime”, Treasury working paper 01/06, The Treasury, Wellington.

Robert A Buckle, Kunhong Kim and Julie Tam (2001), “A Structural VAR Approach to Estimating Budget Balance Targets”, Treasury Working Paper 01/11, The Treasury, Wellington.

The Treasury (2000), Budget Economic and Fiscal Update, The Treasury, Wellington, June 15, pp 35-36.

The Treasury (1996), Putting It Together, The Treasury, Wellington.

Annex 1: Operating provision principles

Determining whether something impacts on the provisions can sometimes be a complex exercise. Currently underway is a review to bring together the experience over the last five years with the aim of improving (and simplifying) the decision making processes, as well as addressing some of the anomalies that can occur at the margins of the framework.

Following is an extract of guidance material available to Treasury analysts, and reflects (in greater detail) what has been published in various Economic and Fiscal Updates since December 1996.

Operating Balance

Principle – All changes in the operating balance resulting from a discrete government decision to introduce a new initiative or change the cost of existing initiatives impact against the $5.9 billion limit, whether they be changes in revenue, expenses or surpluses of SOEs and Crown entities.

Principle – The calculation of impacts is based on the net impact to the operating balance – where this can be calculated. For example, for benefit increases, the impact is net of the income tax component.

Principle – Excluded from the framework are all costs associated with forecasting changes, the recognition of existing contingent liabilities, and costs arising from natural disasters and civil emergencies.

Principle – Impacts are calculated based on the GST inclusive total, to align the framework to the appropriations framework.

What this means in practice is:

  • Gains or losses on sale of assets do not impact, even though they affect the Crown’s operating balance. They do not impact, as they are a direct function of the carrying value in the financial statements, which could vary substantially depending upon factors such as accounting policies, revaluation policies and timing of cyclical revaluations. Accordingly, the gain or loss may be considered subjective (and potentially manipulable).
  • Any action taken to sell assets below market value by the Government will impact against the $5.9 billion fiscal provisions, as this would be a discrete Government decision. The amount that would impact is the difference between market value and the sale price.
  • On the other hand, if a board of an SOE/CE made the decision to sell assets below market value with no government involvement, as opposed to being a government directive, then this would not impact. The government has no discretion in this instance.
  • Devaluations only impact on the Crown operating balance where the impact of the devaluation exceeds the balance in the Crown’s revaluation reserve for that particular class of asset. Revaluations which impact on the operating balance do not count for two reasons:
    • The rationale for not counting gains/losses on sale of fixed assets is that it simply reflects the latest valuation of the assets. The same should apply for any revaluation reserve deficit taken to the operating balance.
    • Where the devaluation results from a specific government decision (e.g. to introduce competition and lower prices), then the change in asset value effectively represents the discounted value of future profits foregone, which will already have impacted on the fiscal provisions.

Policy and Forecasting Changes

Principle – Forecasting changes do not impact on the fiscal provisions. However, there are boundary issues around what is a forecasting change and what is a policy change. For the purposes of counting, a forecasting change is a change where existing policy is such that there will be an automatic change in funding in response to volume change (and government agreement to this change is simply an administrative or rubber stamping exercise).

Principle – Fiscal impacts of implementing existing policy do not impact. However, decisions to change the cost of existing initiatives do impact.

Forecasting Changes

  • How do you distinguish a forecasting change from a policy decision? The basic criteria for a forecasting change is that there is an existing policy or formula that determines that additional expenditure, and Cabinet approval is just an administrative or rubber-stamping exercise. A forecasting change includes the possibility that the baseline will be reduced if volumes go down.
  • Changes to welfare benefit expenses resulting from changes to CPI (which affects CPI indexation) and volume changes (not attributable to a new Government initiative) do not impact. A Government decision to change the benefit rates, on the other hand, does impact.
  • In relation to settlement of Treaty claims, the settlement itself does not impact on the fiscal provisions. Where there is an interest component (because this component is discretionary) it will impact on the fiscal provisions. However, an increase in interest payments due to a delay in signing the deed (which is not a result of a government decision) would not impact on the provisions.

Policy (versus Forecasting) Change

  • Prison musters: an example of the distinction between a policy decision and a forecast change is a decision to increase funding for prison musters. This impacts on the fiscal provisions, as there is no existing policy that states that there will be an increase/decrease in funding if prison muster levels change, i.e. it is not a forecasting change. There is no set formula in place that gets adjusted regularly to account for volume changes.
  • A significant number of government policy decisions that will impact are made in response to changes in demand.

Existing Policy Changes

  • For example, an increase in quotas for New Zealand at the IMF: this would not impact as it is not a new policy decision, rather compliance with existing policy (i.e. membership of the IMF and associated quota contributions).
  • There is sometimes a fine line between implementing current policy and introducing new policy. These situations will need to be assessed on a case-by-case basis in consultation with the Budget Team.

Existing Contingent Liabilities

Principle – The recognition of existing contingent liabilities does not impact on the fiscal provisions.


  • Settlement of existing litigation does not count and, similarly, additional litigation costs associated with defending an existing liability do not impact.

Annex 2: Extract from 2000 Budget Economic and Fiscal Update

Operating provisions for 1999/2000 to 2002/03

In the Budget Policy Statement (BPS), the Government set out a provision for new spending over the four years from 1999/2000 to 2002/03 of $5.9 billion (GST inclusive). The allocations have altered since those outlined in the BPS, as illustrated in the following table.

Table 2.5 – Change in cumulative fiscal provisions
$ million 1999/2000 2000/01 2001/02 2002/03 Total
Indicative fiscal provisions BPS 150 1,200 1,850 2,650 5,850
Fiscal provisions BEFU 420 1,230 1,790 2,425 5,865
Change 270 30 (60) (225) 15

Source: The Treasury

These amounts include a contingency for further initiatives to be developed during 2000/01. The contingency is $180 million a year in 2000/01 and subsequent years.

Since the BPS, the Government has decided to bring forward some spending into 1999/2000 and 2000/01.

To remain within the overall provision of $5.9 billion, the provisions for spending in the 2001 and 2002 Budgets have been reduced to $550 million and $575 million.

Budget 2000 decisions and the fiscal provisions for 1999/2000 to 2002/03

The Government has allocated $3.65 billion of the $5.9 billion as part of the Budget 2000 process. This is detailed in the following table.

Extract from 2000 Budget Economic and Fiscal Update (continued)

Table 2.6 – Budget 2000 policy decisions
2000 Budget initiatives (operating) ($ million, GST inclusive) 1999/00 2000/01 2001/02 2002/03 Total
Revenue initiatives:[19]          
Multi-rate fringe benefit tax - 80 65 65 210
Cancellation of tariff reductions - (35) (80) (85) (200)
Tobacco excise increase (20) (110) (78) (82) (290)
Anti-avoidance measures for 39% tax rate - (10) (20) (20) (50)
Total revenue (20) (75) (113) (122) (330)
Expense initiatives:          
Student loan changes 32 92 103 110 337
Other education 9 108 177 178 472
Increases in NZ Super rates 52 208 212 212 684
Health[20] 32 137 57 88 314
Housing – income-related rents - 55 98 104 257
Industry and economic development - 37 77 116 230
Social services (2) 56 66 71 191
Police 27 54 54 54 189
West Coast package 135 - - - 135
Inland Revenue - 36 36 36 108
Arts, culture and heritage 55 16 16 16 103
Research, science and technology - 30 30 30 90
Cancellation of border charging - 29 29 29 87
Biodiversity strategy - 17 27 38 82
Other environment initiatives 2 21 23 22 68
Maori Affairs 14 19 15 15 63
Foreign Affairs and TradeNZ - 18 16 12 46
Justice 5 22 19 19 65
All other initiatives 49 91 48 22 210
Total expenses 410 1,046 1,103 1,172 3,731
SOE/Crown entity changes          
Housing – changes to Housing NZ 30 79 70 70 249
Total 2000 Budget 420 1,050 1,060 1,120 3,650
Contingency for further initiatives in 2000/01 - 180 180 180 540
Total 2000 Budget including contingency 420 1,230 1,240 1,300 4,190
Provision for 2001 Budget - - 550 550 1,100
Provision for 2002 Budget - - - 575 575
Total fiscal provision 420 1,230 1,790 2,425 5,865


  • [19]A “negative” number represents an initiative that is generating a saving (or an increase in revenue) and therefore reduces the total amount accumulated against the fiscal provisions.
  • [20]This includes a correction to the Health baseline of ($84 million) in 2001/02 and 2002/03. Total new spending excluding this adjustment is $481 million.

Annex 3: graph showing how the Provisions influence the trend of the operating balance

The following graph provides an illustrative view of how a government can influence the long-term track of the operating balance through a focus on the fiscal provisions.

Operating Balance (1999 Budget)


Operating Balance (1999 Budget).

The operating balance increases as the economy expands, providing greater tax revenue, while baselines are fixed.

  • The operating balance track is reduced through the inclusion of the provisions for future initiatives (or “fiscal allowances” in the years beyond the fiscal forecast horizon.
  • The operating balance may fluctuate within the identified band (illustrative only[21]) due to events beyond the immediate control of the Government (such as liability valuation changes).
  • By altering the fiscal parameters for new spending and baselines, the Government can influence the underlying operating balance track.
  • While short-term volatility can detract from the underlying trends (in the operating balance or debt for example), it is the underlying trends that are central to sound fiscal policy action. As noted in the section 2.3, the fiscal provisions should only be altered if progress towards the Government’s medium- to long-term objectives is seriously threatened.


  • [21]Magnitudes (for different probabilities) are provided in Treasury Working Paper 01/11.