The Contingency Reserve

Budget Explainer No. 2, 9 December 2021*

A explainer piece that examines the definition and use of the Contingency Reserve and its major components.

Download the full printable version: PDF icon (712 KB)
Data: MS Excel (35 KB)


The Contingency Reserve is an allowance within the government's budget forecasts for items that either cannot or should not be allocated to specific programs at the time of publication. While the Contingency Reserve includes a variety of provisions, the main components are the Conservative Bias Allowance, which is an allowance for the historical tendency for expense estimates to be under-forecast, and Decisions Taken But Not Yet Announced, which is the value of new policy decisions that have not yet been disclosed. By including these items on an aggregate basis, rather than omitting them, the Contingency Reserve improves the accuracy of aggregate budget estimates at the time of their publication.

The Conservative Bias Allowance (CBA) was introduced due to an observed tendency of budgets to underestimate future expenses, even after allowing for changes in policy and economic forecasts. It is therefore designed to account for all the changes in estimates of expenses that are as a result of 'program-specific' factors. The proportional size of the CBA is consistent with average historical variations in program-specific factors, confirming the CBA improves the accuracy of budget estimates in practice.

Policies that have been included in the budget aggregates but not yet announced are known as Decisions Taken But Not Yet Announced (DTBNYA). DTBNYA amounts have been disclosed consistently since 2004-05, with their total size increasing significantly over the years since then. Policies may not be announced, or their amounts left undisclosed, for a variety of reasons. Recent increases in DTBNYA are likely due to a combination of isolated factors rather than representing a long-term trend.

What is the Contingency Reserve?

The Contingency Reserve is an allowance within the Commonwealth budget for events that the government reasonably expects to eventuate, but cannot allocate to specific programs or detail. By including these items on an aggregate basis, rather than omitting them, the Contingency Reserve improves the accuracy of aggregate budget estimates at the time of their publication. 

Since the Contingency Reserve was introduced in 1987, it has included a variety of items which generally fall within the following groupings:

  • The Conservative Bias Allowance (CBA), which makes provision for the historical tendency for the estimates of expenses for existing government policy to be under-forecast, and hence revised upwards later. The CBA is discussed further in this explainer.
  • Expense and non-tax revenue policy decisions that were not announced in the budget. These are classified as Decisions Taken But Not Yet Announced (DTBNYA). The Contingency Reserve contains the cumulative impact of funding for outstanding and new DTBNYA. These policies are usually announced and published in a future budget update.1 Budget aggregates can also include DTBNYA that relate to taxation, but these are included elsewhere in the budget as opposed to the Contingency Reserve. DTBNYA are explored in more detail later in this explainer.
  • Items that are too sensitive to be disclosed separately, such as those that are commercial‑in‑confidence or affect national security. In some cases, a description of the item is disclosed in the budget while the funding allocated to it is not, and in other cases no information is disclosed. The 2021-22 Budget, for instance, included a provision for the manufacturing of messenger RNA vaccines in Australia but funding was not disclosed as the exact price negotiated with pharmaceutical companies would be commercial-in-confidence.2
  • An aggregate provision for underspends in the budget year, reflecting the tendency for budgeted expenses for some entities or functions not to be met.
  • Information that is received too late during the preparation of the budget to change estimates of individual items, such as policy decisions or changes to the expected economic outlook.3
  • Provisions for items that have a significant amount of uncertainty at the time of publication, such as programs still under negotiation with State and Territory governments, or for the establishment of new agencies whose legislation has not yet passed Parliament.

The figure below shows how the components of the Contingency Reserve and DTBNYA are related, highlighting that not all DTBNYA are included in the Contingency Reserve.

The description of the Contingency Reserve in the budget will occasionally provide information on other specific items included. A good example is the 2009-10 Budget, where the Contingency Reserve amounts were relatively large, and included: provisions for future equity investments in the National Broadband Network (subject to an implementation study); estimates of revenue and expenses related to the then proposed Carbon Pollution Reduction Scheme (where the timing was dependent on the actions of other major economies); and provisions for future increases in Australia's Official Development Assistance yet to be allocated to specific aid programs.4 See Appendix A for a complete list of all items that have been included in the description of the Contingency Reserve throughout history.

Information on the Contingency Reserve can be found in Budget Paper 1, near the end of the expenses and net capital investment statement.5 In general, this shows the total provisions under the Contingency Reserve and the amount allocated to the CBA. Other provisions within the Contingency Reserve are rarely explicitly identified or quantified, and so can only be examined on an aggregate basis.6 A description of the changes in the Contingency Reserve over time for a given financial year is provided in Box 1.7

  • Box 1: Contingency Reserve amounts over time

    At the 2008-09 Budget, the last year of the forward estimates was 2011-12, which included an amount in the Contingency Reserve of $12 billion. At the following Budget this 2011-12 amount was reduced to $7 billion, and then to $3 billion at the 2010‑11 Budget

    Amounts allocated to the Contingency Reserve in a given financial year tend to decrease in subsequent budgets. This is because provisions are removed as the expected events occur or are otherwise resolved. For example, decisions are announced, which moves amounts out of the Contingency Reserve, and the CBA decreases as the estimated financial year becomes nearer. As the Contingency Reserve only includes provisions for the future, it is not included in the Final Budget Outcome, effectively reducing the value to zero so there are no historical 'actual' amounts.

The Contingency Reserve can also include provisions for items expected to affect non-tax revenue. These could include DTBNYA, for example measures to decrease the size of government investment funds. It does not include DTBNYA expected to impact tax revenue alone. These measures are instead incorporated into the revenue forecasts in Budget Paper 1.

The Contingency Reserve is not a general reserve of money that can be used to fund future policies.8 That is, the Contingency Reserve cannot include money to be simply put away 'for a rainy day' or for aspirational policy goals. If it operated like this, the budget estimates would not be an accurate reflection of total expected spending and revenue, and policies would more regularly be announced without changing the baseline totals.

Other countries also have contingency reserves which differ in their disclosed use and size. For example, the government of South Africa's budget Contingency Reserve provides for unforeseen and unavoidable expenditure, such as natural disasters, while Bosnia's Contingency Reserve provides for shortfalls in revenue.9 Other functions of contingency reserves may include financing of foreign investment projects and unfunded mandates, used as loan guarantees, support for infrastructure spending, and other appropriations. The size of contingency reserves also varies across countries, ranging from less than 1 per cent (the United Kingdom, the Netherlands) to up to 3 per cent (Sweden, Russia) of total government expenditure.10

History of the Contingency Reserve

The Contingency Reserve has been included in budget forecasts since the 1987-88 Budget.11,12 The Contingency Reserve was included to provide more accurate estimates of future outlays, as this was the first budget to include estimates beyond the next year. The budget figures for the second, third and fourth years beyond the year ahead are now known as the 'forward estimates'. Figure 1 shows the breakdown between the CBA, and other provisions included in the Contingency Reserve. The CBA is typically the largest component of the Contingency Reserve, and other provisions generally include the cumulative amount of all DTBNYA measures from previous budget updates that have not yet been allocated and still have an impact on the budget estimates. The Contingency Reserve has almost always increased expense estimates.

Figure 1: Disaggregation of the Contingency Reserve (CR), by budget

Source: Commonwealth budgets and PBO analysis.
Note: Impact is presented on an expense basis and is aggregated over the forward estimates for each budget. For example, '2011-12' refers to the total amounts allocated to the Contingency Reserve at the 2011-12 Budget, for the budget year and forward estimates years of 2011-12 to 2014-15. Provisions that would increase expenses or decrease revenue are recorded as positive, while provisions that would decrease expenses or increase revenue are reported as negative. CBA amounts are PBO calculations, which may differ very slightly from those published in budgets.

What is the Conservative Bias Allowance?

The CBA is an allowance for the historical tendency to underestimate expenses over the forward estimates. The amount of this provision is calculated as a certain percentage of total expenses, excluding GST payments to the States (the 'CBA percentage'). The current CBA percentage is zero in the Budget year, half a per cent in the first year of the forward estimates, 1 per cent in the second year, and 2 per cent in the final year. 13 An example of a calculation of the CBA is shown in Box 2.

  • Box 2: Conservative Bias Allowance calculation
      Consider the estimates for the year 2019-20 in the 2018-19 Budget. Total expenses were estimated to be $504,171 million and GST payments to the states were estimated to be $69,790 million. The CBA percentage for this year was 0.5 per cent, so the total amount provisioned in the CBA may be estimated as:
    0.5% × (504,171 – 69,790) =
    $2,172 million.

The CBA does not account for all possible reasons why expense outcomes may differ from their budgeted forecasts. The budget specifies the following five reasons why forecasts for expenses vary from one update to the next:

  1. new policy decisions
  2. revisions to economic parameters, such as economic growth, prices and unemployment
  3. revisions to other parameters, specific to particular programs
  4. changes to the amount of interest the government is expected to pay on its borrowings
  5. other variations not captured otherwise.

The CBA percentage is intended to compensate for only the third of these factors, referred to in the budget as 'program specific parameter variations'.15 Box 3 provides an example to clarify the differences between these types of variations.

  • Box 3: Sources of variations in forecasts for expenses

    Expenses forecasts are revised between budget forecast updates for a number of reasons.

    If, for example, the government decided to change the payment rate of the Age Pension, this change to previously forecast Age Pension spending would be captured under 'effect of policy decisions'.

    If the Consumer Price Index (CPI) forecast was revised upwards, then the payment rate for the Age Pension, partly indexed to the CPI, would also be expected to increase. This increase to previously forecast Age Pension spending would fall under 'total economic parameter variations'.

    If the expected proportion of Age Pension recipients who are single was revised upwards due to new demographic information, more people would be expected to receive a higher rate of payment. This increase to Age Pension forecasts would be captured under 'program specific parameter variations', as the proportion would be estimated specifically for Age Pension forecasts, rather than as part of the general economic parameters for the budget.

The CBA percentage is based on recent historical trends in revisions to forecasts of expenses and is therefore revised from time to time. Appendix B shows the history of the CBA percentage.

Table 1: Sample reconciliation of expense estimates

Source: 2021-22 Budget, p 160.
Note: Numbers may not sum due to rounding. Program specific parameter variations are highlighted for emphasis. The effect of policy decisions excludes secondary impacts on public debt interest and offsets from the Contingency Reserve for decisions taken that were included in previous budget updates within DTBNYA.

The source of 'program specific' variations is primarily 'demand-driven' programs–that is, programs such as the Age Pension where the required amount of funding is not capped, but depends on the number of users. Estimates of future spending are therefore more uncertain than for programs with capped funding.16,17 As a given year moves through the forward estimates in later updates, the amount of uncertainty in aggregate spending estimates decreases and this is reflected in a decrease to the amount allocated to the CBA. These decreases are referred to as 'drawdowns' and are explored in more detail in Box 4.

  • Box 4: Drawdowns of the CBA

    The 2020-21 MYEFO included a CBA for the 2021-22 year of $1.3 billion. The 2021-22 Budget referred to a 'drawdown' of the 2021-22 CBA of $1.3 billion, reducing the amount to zero as the CBA for the budget year is equal to 0% of expenses, reflecting the fact that the total expenses for the budget year are usually very close to the final outcome. This language can be confusing because the CBA does not represent an actual fund and a 'drawdown' does not represent a saving to the budget.

Does the Conservative Bias Allowance improve the accuracy of budget estimates?

Our high level analysis confirms that there is a historical tendency to underestimate aggregate expenses and suggests that the CBA improves the accuracy of aggregate expense estimates.

The effectiveness of the CBA can be determined by comparing the sum of 'program specific variations' for a given year to the CBA percentage in the first budget update where that year is in the forward estimates. For example, the CBA for 2020-21 in the 2017-18 Budget would ideally equal the sum of all the 'program specific variations' for 2020-21, from the 2017-18 Budget through to the 2020‑21 Budget. If the CBA was above this average, aggregate expenses would be over-estimated, while if the CBA was smaller, aggregate expenses would be under-estimated.

While some individual years vary significantly, the long‑run average of the sum of program-specific variations is 2 per cent, the same size as the current CBA.

  • The fact that the long-run average of program-specific variations is positive confirms that there is a historical tendency to underestimate aggregate expenses.
  • The consistency between the size of this tendency and the CBA percentage in this high level test means that the CBA does improve the accuracy of aggregate expense estimates.

See Appendix B for a more detailed breakdown of the CBA, including comparisons of program specific variations for each year of the forward estimates, and a complete history of the CBA.

What are 'Decisions Taken But Not Yet Announced'?

Most budget updates include some new policies, the details and costs of which have been calculated but have not been announced. The fiscal impacts of all these unannounced policies are aggregated, and presented as a single line 'Decisions Taken But Not Yet Announced' in Budget Paper 2, in the tables showing the revenue, expense and capital measures.18

The total published DTBNYA amount includes new policies not yet announced as well as policies that have been announced but whose costs cannot be published. In the latter case, short summaries of policies, known as 'measure descriptions' are still published for these policies, but the amounts are replaced by 'nfp' (not for publication).

The financial impacts of unannounced and not for publication policies are generally included in the Contingency Reserve. An important exception to this is tax policies, where the impacts are explicitly incorporated into the estimates for the relevant taxes, rather than within the Contingency Reserve.

Both unannounced tax and non-tax revenue policies are included in the aggregate figure for revenue DTBNYA in Budget Paper 2, but only the unannounced non-tax revenue policies are included in the Contingency Reserve.

There are several reasons why policies may not be announced, or amounts cannot be published.

  • Information may be commercial-in-confidence or its disclosure could prejudice national security. For example, the cost of the 2021-22 Budget measure, COVID-19 Vaccine Manufacturing Capabilities, was listed as 'not for publication (nfp) due to commercial in confidence sensitivities'.19 Similarly, the fiscal impact of new listings to the Pharmaceutical Benefits Scheme is usually reported as not for publication.20
  • Some policies may have been finalised very late in the budget process.
  • The government, who choose the timing of their policy announcements, may include some policies within DTBNYA in order to announce them later.

If a subsequent budget update announces a policy which has previously been included within DTBNYA then the measure tables show no impact on budget estimates, because the budget estimates have already included the impact. This is often reflected with a note such as 'Funding for this measure has already been provided for by the Government'. The description of the policy in Budget Paper 2 may also include the cost of the policy which has already been accounted for.21 An example of this is shown in Box 5.

  • Box 5: Example of a policy previously included as a DTBNYA

    This is a simple example of the publication of a policy which was announced in one update (the 2019‑20 Budget) but whose financial impact had already been included in a previous update. The cost of the policy for this update is zero, but the description reveals the full cost of the policy ($5.0 million), which was already incorporated in a previous budget update.

When a previously unannounced policy is announced, the financial impact of that policy is moved from the Contingency Reserve to the relevant program or non-tax revenue item, unless it is a tax revenue item in which case it was already included in tax estimates.

Policies included in DTBNYA are not necessarily announced by the next budget update, so unannounced policies may remain within the Contingency Reserve for multiple updates. The DTBNYA lines in the tables in Budget Paper 2 include only new decisions since the last update, but the cumulative impact of all these policies which remain unannounced is in the remaining aggregate provisions of the Contingency Reserve.22 This means that the Contingency Reserve may include many unannounced policies, from multiple updates, at one time.23

While measures under DTBNYA can be classified as a revenue, expense, or capital measure, the expenses DTBNYA is usually larger than the revenue DTBNYA, with an average size of $200 million for expenses and $100 million for revenue between the 2003-04 MYEFO and 2021-22 Budget.24

How have Decisions Taken But Not Yet Announced changed over time?

Up until 2004-05, DTBNYA was rarely disclosed in the budget papers, and when it was disclosed it was usually very small. Since then, the size of new DTBNYA has increased significantly, both in absolute terms and relative to the rest of the budget, as shown in Figure 2.

The analysis of the largest amounts of DTBNYA in history in the next section indicates that the recent rises may be due to isolated instances and may not represent a long-term trend.

Figure 2: Size of new DTBNYA over the forward estimates at each budget update
($ billion and as a percentage of total expenses)
2003-04 to 2021-22

Source: Commonwealth budgets and PBO analysis.
Note: Values for both DTBNYA and total expenses are the total impact over the forward estimates for a given budget update. DTBNYA only includes new policies at each update, as included in Budget Paper 2, and not the total cumulative impact of all unannounced policies.

Examples of Decisions Taken But Not Yet Announced

To illustrate the varying uses, budget treatments and reporting for DTBNYA, this section of the explainer discusses the largest historical cases. There are nine budget updates which have included a DTBNYA of at least $1 billion in a single year. Given the increase in DTBNYA in recent years, as well as the impacts of economic growth and inflation, these cases largely come from the last five years. These DTBNYA examples include expenses, which are budgeted through the Contingency Reserve, as well as revenue, of which tax measures are included in revenue forecasts as opposed to the Contingency Reserve. Some of these examples may include the financial impacts of measures that are not for publication, some of which are not held in the Contingency Reserve.

  • 1. 2021-22 Budget25

    The largest DTBNYA in a single year was included in the latest Budget, and represented $3.8 billion of payments for 2021-22, contributing to a total amount of almost $10 billion over the forward estimates.

    While these amounts are likely to include several individual items, a significant component is likely to be related to COVID-19 vaccines and other medical goods.26 As these relate to commercial‑in‑confidence arrangements, the precise amounts provisioned for in the Budget were not made public. More details of some of these measures were included in the Budget, but the specific amounts allocated to each are not for publication.27 It has also been confirmed that a provision of $600 million was included as a DTBNYA for a new power plant at Kurri Kurri.28

  • 2. 2018-19 MYEFO29

    The 2018-19 MYEFO also included a DTBNYA amount of $3.8 billion in 2020-21, this time for revenue. Most of this was due to an unannounced tax policy, and therefore did not affect the Contingency Reserve. It was instead explicitly factored into the tax forecasts at the 2018-19 MYEFO.

    Careful readers of the MYEFO may have noticed that income tax withholding was revised down significantly, despite positive parameter and other variations, potentially indicating sizeable tax cuts being factored into the estimates.30 The following budget announced the government's policy, Lower taxes for hard-working Australians, and confirmed that the MYEFO had included a 'provision for the impact of this measure' of a similar size to the DTBNYA. The impact of the announced policy at the 2019-20 Budget was therefore relatively small, since most of the policy had already been allowed for at the previous MYEFO.

  • 3. 2020-21 Budget31

    The 2020-21 Budget, delayed five months due to the pandemic, also included a large DTBNYA peaking in 2021-22. This was likely for similar reasons to the 2021-22 Budget.

  • 4. 2019-20 Budget32

    The 2019-20 Budget, which led into the 2019-20 federal election, included a large negative DTBNYA in the final year of the forward estimates, following much smaller amounts for the preceding years.

    The overall size of the Contingency Reserve suggested that the net impact of all the non‑CBA provisions was small. The 2019-20 Budget reported that the CBA explained all but $500 million of the Contingency Reserve for 2022-23.33

    Some of this may be due to reversals of policies introduced but not announced at previous budget updates.34 In this case, DTBNYA amounts reducing expenses by $2.7 billion in 2022-23 could have been the reversal of previous measures which increased expenses. With 2022‑23 being outside of the forward estimates before the 2019-20 Budget, the value of the previously unannounced decisions would be unknown.

  • 5. 2016-17 Budget35

    Similar to the 2019-20 Budget, the 2016-17 Budget also included a reduction of expense estimates in DTBNYA in the final year of the forward estimates.36 Just as in the 2019‑20 Budget, the value of the previously unannounced decisions that were reversed would be unknown.

  • 6. 2007-08 MYEFO37

    A $1.2 billion amount for DTBNYA appeared in the 2007-08 MYEFO, which led into the 2007 federal election.

    Unlike the cases for the 2016-17 Budget and the 2019-20 Budget, the DTBNYA were not only large in the final year. The 2007 PEFO, released following the dissolution of parliament, revealed that the amounts were almost entirely for the government’s policy, The Utilities Allowance — increase and extend eligibility to include all recipients of Disability Support Pension, Service Pension, and Carer Payment.38

  • 7. 2018-19 Budget39

    An increase in revenue of around $1.2 billion for the budget year was included as DTBNYA at the 2018-19 Budget, decreasing in size over the forward estimates.

    Components of the DTBNYA which were tax revenue would have been included in the forecasts for the relevant taxes, but components of non-tax revenue would have been included in the Contingency Reserve.

    The Budget also included a $1.1 billion increase in expenses in DTBNYA, affecting 2020-21, with much smaller amounts for all other years.

    The nature of these amounts and their relationship, if any, is unknown.

  • 8. 2014-15 Budget40

    At the 2014-15 Budget, DTBNYA reduced expenses by increasing amounts across the forward estimates, reaching nearly $1.1 billion for the 2017-18 year.

    The subsequent 2014-15 MYEFO announced several policies which had been included as DTBNYA at the 2014-15 Budget, but they were all relatively small, leaving the amounts shown here largely unexplained. By the 2015-16 Budget, the Contingency Reserve amounts were almost zero (excluding the CBA), indicating the absence of remaining DTBNYA amounts. This may indicate that amounts for DTBNYA at the 2014-15 Budget were largely due to a decision to not proceed with some policies.

Appendix

  • Appendix A: Components listed in the Contingency Reserve at each budget

    The list below includes all the items that have been specifically mentioned as being in the Contingency Reserve in a budget. It does not include items revealed to be in the Contingency Reserve outside of budget updates, for example in media releases or during Senate Estimates. Items are presented in chronological order of their appearance. Unless otherwise stated, items have appeared in multiple updates. Some of these items, such as asset sale related expenses, have been moved in and out of the Contingency Reserve at various points.

    • Wage and salary increases for departments and Commonwealth statutory authorities.
    • Salaries costs of former Public Service Board staff to be redeployed as a result of the restructuring of that organisation (1987-88 Budget only).
    • Carryover of unspent administrative expenses appropriations.
    • Savings resulting from the streamlining of the travel arrangements for the Public Service (1987‑88 Budget only).
    • Possible increased payments by departments and budget financed Commonwealth statutory authorities to superannuation schemes for employees who are not members of the Commonwealth Superannuation Scheme under the guidelines of the National Wage Case (1988‑89 Budget only).
    • Expected underspends by departments of amounts appropriated to them to pay for design and construction services provided by the Department of Administrative Services.
    • Property expenses, including medium and minor works, new leases and fitout and rent revisions, which relate to the delivery of existing programs.
    • The Conservative Bias Allowance.
    • Expected savings from maintaining the efficiency dividend on departmental running costs.
    • Parameter revisions received late in the budget process.
    • Payment of customs duty on imports by Commonwealth departments (1990-91 Budget only).
    • A limited number of possible net additions to outlays reflecting extensions of existing programs.
    • Additional supplementation for the wage components of some programs this year, predominantly in the health and education areas.
    • Commercial-in-confidence items that cannot be disclosed separately.
    • Possible inaccuracies in the year just completed, due to the timing of the release of the Budget.
    • Adjustment of running costs for the outcome of the administrative expenses deflator.
    • Impact of changed arrangements for the management of agency property costs (1994‑95 Budget only).
    • Carryovers for Defence relating to potential contract delays (1994-95 Budget only).
    • The effect of an initiative to improve the management of cash between the Commonwealth, its statutory authorities and the States (1995-96 Budget only).
    • The avoidable direct costs of the sale of Telstra, and the environment package which are contingent on the approval of the Telstra (Dilution of Public Ownership) Bill 1996 (1996‑97 Budget only).
    • Maintaining pensions at 25% of total male Average Weekly Earnings.
    • Established tendency for estimates of some programs to be overstated in the budget year.
    • Funding for the proposed Federation Fund Trust Account.
    • Increased rate of the Superannuation Guarantee Charge (1998-99 Budget only).
    • Difference between proposed Australian Health Care Agreements being negotiated with States and Territories and the amounts in the forward estimates (1998-99 Budget only).
    • National security-in-confidence items that cannot be disclosed separately.
    • Minor decisions made late in the Budget process.
    • Government's decision to extend the operation of the Natural Heritage Trust.
    • Expected changes to funding for the Department of Immigration and Multicultural Affairs and Indigenous Affairs as a result of a review yet to be conducted (2003-04 Budget only).
    • Financial impact of health insurance premium growth on the Private Health Insurance Rebate.
    • Impact of foreign exchange rate movements (2003-04 Budget only).
    • Asset sale related expenses.
    • Provision for events and pressures that are reasonably expected to affect the budget estimates.
    • Continuation of the National Action Plan on Salinity and Water Quality (2006-07 Budget only).
    • The difference between the amount of Official Development Assistance already committed and the government's target levels.
    • Ongoing costs associated with the Northern Territory Emergency Response (2008‑09 Budget only).
    • Establishment of the National Broadband Network (2008-09 Budget only).
    • Anticipated expenditure from the Building Australia Fund, Health and Hospitals Fund and the Education Investment Fund (2008-09 Budget only).
    • Continuation of drought relief spending (2009-10 Budget only).
    • Updated estimates of the Carbon Pollution Reduction Scheme (2009-10 Budget only) – these provisions were then explicitly removed in the 2010-11 Budget due to a government decision to not legislate this scheme until after the end of the Kyoto Protocol.
    • Future equity investments into the National Broadband Network.
    • The second tranche of the Nation Building program (2011-12 Budget only).
    • Election commitments made at the 2010-11 MYEFO, which were then removed at the 2011‑12 Budget.
    • The Stronger Futures in the Northern Territory package which was still under negotiation with the Northern Territory government at the time of publication (2012-13 Budget only).
    • Commitment to pay the Commonwealth's share of costs in response to the pay equity claim on behalf of certain workers in the Social and Community Services sector (2012-13 Budget only).
    • Paid Parental Leave Scheme (2014-15 Budget only).
    • Company tax rate reduction (2014-15 Budget only).
    • Programmes that are yet to be renegotiated with State and Territory governments.
    • Decisions taken but not yet announced by the government.
    • Continuation of some expiring National Partnerships and possible by-election and redistribution costs for the Australian Electoral Commission.
    • Funding for Investing in ChildCare – ICT system, Better Management of Biosecurity Risks – advanced analytical capability, Public Sector Transformation and the Efficiency Dividend and New Veteran Public Hospital Pricing Agreement (2016-17 Budget only).
    • Increases in new medicine listings (2018-19 Budget only).
    • Funding for the Indigenous Recognition Referendum, Commonwealth Integrity Commission, Future Drought Fund – establishment and Establish the Emergency Response Fund (2019‑20 Budget only).
  • Appendix B: History and effectiveness of the Conservative Bias Allowance

    History of the Conservative Bias Allowance

    In the 1987-88 Budget, the Commonwealth Government introduced the Conservative Bias Allowance (CBA) to improve the accuracy of expenses forecast over the forward estimates, by providing an allowance for potential variations in program estimates due to 'program specific' factors.41 The CBA is not intended to capture variations in economic parameters (for example, changes in the Consumer Price Index) or policy decisions (for example, a measure to change the rate of the Age Pension). Program specific parameter variations may include, for example, an increase in the expected proportion of Age Pension recipients who are single, leading to higher average costs per recipient than estimated (see Box 3 of this explainer).

    The CBA percentage is based on historical program-specific parameter variations that occur as a given year progresses from being the third forward estimates year (FE3) to being the budget year. As these variations in expenses tend to be positive (upwards revisions) in the forward years, the CBA has been set at a positive rate throughout its history (Table B1 and Figure B1). The CBA amount is calculated as a percentage of total general government expenses, less GST payments to states and territories. The CBA percentage was initially set at 0.75 per cent of outlays in FE3. It was gradually increased to 4 per cent by the 2005-06 Budget, before being revised down to 2 per cent by 2009-10. The CBA percentage has remained stable from 2009-10 to the time of publishing this explainer, in 2021-22.

    Table B1: History of changes to the CBA

    Figure B1: CBA percentage, by budget

    Source: Commonwealth budgets.
    Note: FE stands for forward estimates – FE1 is the first year of the forward estimates period, FE2 the second year and FE3 the third year.

    The effectiveness of the Conservative Bias Allowance

    To investigate whether the CBA improves the accuracy of budget estimates, the CBA allocated in FE3 can be compared with the total reported program specific parameter variations for a given year over the period it is included in the budget, from when that year is FE3 to its final outcome.42 While these might not align for a given year, if the long-run average of program specific parameter variations is similar to the CBA percentage, then the CBA is improving the accuracy of budget estimates.43

    For example, in the 2014-15 Budget, the year 2017-18 had a CBA of 2 per cent of total expenses ($8.2 billion). For the CBA to exactly account for reported variations in program estimates, the sum of all the 'program specific parameter variations' for 2017-18, through each budget update from the 2014‑15 MYEFO to the 2018-19 Budget, should equal $8.2 billion. However, the sum of these was $5.1 billion, or 1.2 per cent of the original expenses estimate for 2017-18, suggesting that the CBA slightly overestimated expenses for this year.

    Figure B2 compares the relationship between the CBA reported in FE3 and the total reported program specific parameter variations between FE3 and the budget year. Over the last three decades, there has been a tendency to revise expenses upwards, except between 2005-06 and 2008‑09 and more recently from 2018-19 onwards although the recent revisions are likely due to the COVID-19 pandemic. The long‑run average of 2 per cent is the same as the level factored into the CBA, confirming that the CBA accounts for the tendency of budgets to underestimate expenses.

    Figure B2: Program specific parameter variations compared to the CBA percentage

    Source: Commonwealth budgets and PBO analysis.
    Note: Amounts are expressed as a percentage of total government expenses minus GST payments to the states and territories. This includes all reconciliations in budget updates from 1993-94 onwards, except where reconciliations were not published such as at the 1999-00 Budget and the 2010 Pre-election fiscal outlook. The long-run average excludes variations from 2019-20 and onwards as they may have been impacted by the COVID-19 pandemic.

  • Notes

    * This explainer was authored by Travis Ey, Harry Withers, Anderson Cheung, and John Clark, with the benefit of comments and contributions from Kathryn Smith and Stein Helgeby. The contents of the explainer are the sole responsibility of the Parliamentary Budget Office.

    1. A 'budget update' is a release of one of the documents specified in the Charter of Budget Honesty Act 1998: the Budget, the Mid-Year Economic and Fiscal Outlook (MYEFO) and the Pre-Election Fiscal Outlook (PEFO). Governments sometimes release other updates in between these, but any measures announced are always repeated at the following budget update.
    2. https://www.financeminister.gov.au/transcript/2021/05/12/sky-news-live-first-edition-peter-stefanovic.
    3. For example, the 2004-05 Budget (p. 6-31) included an allowance for the impact of the Australian Industrial Relations Commission's Living Wage Case Decision, released less than a week before the budget.
    4. 2009-10 Budget Paper 1, p. 6-54.
    5. In the 2021-22 Budget, this is Statement 6 of Budget Paper 1, where the Contingency Reserve is considered as a separate 'function'. In the forecast financial statements (Statement 10) the Contingency Reserve is allocated to various line items according to the details of the individual components, with the CBA allocated entirely to 'Current transfers: personal benefits'. The CBA is included in both the fiscal balance and underlying cash balance, while other provisions in the Contingency Reserve are allocated depending on their exact nature.
    6. While Commonwealth budgets usually include lists of items that might be included in the Contingency Reserve, they are not required to provide details on what has been included in the Contingency Reserve. The Charter of Budget Honesty Act 1998 does not provide guidance on the use of the Contingency Reserve.
    7. Box 1 states that there are no 'actual' amounts for the Contingency Reserve. For several years up to 2008-09, however, the Contingency Reserve included expenses related to asset sales, which resulted in small 'actuals', and a larger amount of $589 million for 2006-07. Since 2009-10, the Contingency Reserve actual has been zero.
    8. https://www.finance.gov.au/about-us/glossary/pgpa/term-contingency-reserve.
    9. International Monetary Fund (2008), Fiscal Risks—Sources, Disclosure, and Management, p. 27, available here https://www.imf.org/external/np/pp/eng/2008/052108.pdf.
    10. International Monetary Fund (2018), Austria Fiscal Transparency Evaluation, p. 57, available here https://www.imf.org/-/media/Files/Publications/CR/2018/cr18193.ashx.
    11. The 1984-85 Budget did not mention a 'Contingency Reserve' but included a 'Provision for Contingent Manpower Costs' of $60m in the forecasts for defence expenses only (Budget Paper 1, p. 82).
    12. The 1987-88 Budget (Paper 1, p. 294) was the first update where the Contingency Reserve was explicitly identified, noting that the then $150 million Contingency Reserve was a reasonable provision to cover the uncertainty around estimates of increases in public sector wages and other allowances.
    13. Different percentages are also used for the MYEFO, however these are simply adjustments to reconcile time passing. For example, the first forward estimate year at the 2020-21 MYEFO (2021-22) used a CBA percentage of 0.25 per cent - as roughly half a year had passed, the CBA needed to be reduced by an equivalent amount.
    14. The CBA is typically estimated from preliminary budget estimates, so there can be small discrepancies when recreating the CBA using final aggregates. The CBA amounts are rounded to the nearest $100 million.
    15. Answer to Question on Notice, reference F6, Review of the conservative bias allowance, p. 2, available here Finance and Deregulation portfolio – Parliament of Australia (aph.gov.au).
    16. 'Demand-driven' programs have their funding determined by the number of users, and have no fixed limits on the number of people who can access the program. For more information, see the PBO's Online Budget Glossary.
    17. Answer to Question on Notice, reference F6, Review of the conservative bias allowance, p. 2, available here Finance and Deregulation portfolio – Parliament of Australia (aph.gov.au).
    18. In the 2021-22 Budget, these are Table 1 (p. 5) for receipts and Table 2 (p. 50) for payments. DTBNYA is usually the last item in the tables.
    19. 2021-22 Budget Paper 2, p. 134.
    20. These measures for new listings may have both a revenue and expense component, depending on the nature of the contract. For example, if the government purchases medicine at a discount from the market price, then this may be reported in the budget as a purchase at the market price, recorded as an expense, and a rebate from the seller, recorded as revenue.
    21. In some cases, the policy may have been partly included in a previous update as DTBNYA, or partly announced in a previous update, in which case the amounts may not be zero.
    22. Except for policies which only affect tax revenue, which are never allocated to the Contingency Reserve.
    23. For example, the 2019-20 Budget announced 37 measures which had been partly or wholly included in DTBNYA in previous budget updates. www.aph.gov.au/-/media/Estimates/fpa/bud1920/tabled_docs/finance/TabledDoc15_Finance_DTBNYA.pdf.
    24. While accrual terms are used here, some budgets, including the 2021-22 Budget, instead use cash definitions. For more information on cash and accrual accounting, see the PBO's Online Budget Glossary.
    25. 2021-22 Budget Paper 2, p. 50.
    26. Finance Minister Simon Birmingham reported that the budget documents "haven't detailed how much money is budgeted [for mRNA vaccines] because there are two very sensitive commercial negotiations that have to be pursued there. […] But it is baked into the Budget bottom line. It's part of the contingency reserve in the Budget." https://www.financeminister.gov.au/transcript/2021/05/12/sky-news-live-first-edition-peter-stefanovic.
    27. For example, investment in manufacturing of mRNA vaccines is on p. 134 of Budget Paper 2 in the 2021‑22 Budget.
    28. Confirmed by the Department of Industry, Science, Energy and Resources through a question on notice, available at https://www.aph.gov.au/api/qon/downloadestimatesquestions/EstimatesQuestion-CommitteeId8-EstimatesRoundId11-PortfolioId34-QuestionNumber1.
    29. 2018-19 MYEFO, p. 111.
    30. Forecasts for total growth in Compensation of Employees were unchanged or slightly higher at the MYEFO (p. 46) compared to the previous Budget (p. 5-9), while growth in Gross income tax withholding was significantly reduced in the same years (p. 5-17 of the 2018-19 Budget compared to p. 44 of the MYEFO). The MYEFO reported that "excluding policy decisions, individuals taxes have been revised up by $4.1 billion (1.9 per cent) in 2018-19 and $13.8 billion over the four years to 2021-22." (p. 46)
    31. 2020-21 Budget Paper 2, p. 45.
    32. 2019-20 Budget Paper 2, p. 45.
    33. 2019-20 Budget Paper 1, Statement 5, p. 5-43.
    34. The 2019 Pre-Election Fiscal Outlook, published two weeks after the Budget, noted that "Consistent with past practice, the DTBNYA lines in Budget Paper 2: Budget Measures, included the reversal of some DTBNYA items from previous budget rounds."
    35. 2016-17 Budget Paper 2, p. 61.
    36. The size of these provisions is also very similar, assuming that the previous year's estimate carried through to the final year. If the policies comprising the DTBNYA of $502 million in 2018-19 were to continue to 2019‑20 at a similar size, in both budgets the additional DTBNYA in the final year reduced expenses by around $2.5 billion.
    37. 2007-08 MYEFO, p. 74.
    38. 2007 Pre-Election Fiscal Outlook, p. 6.
    39. 2018-19 Budget Paper 2, pp. 6 and 68.
    40. 2014-15 Budget Paper 2, p. 47.
    41. In earlier budgets, a general provision for wage increases served a similar function to, but was not specifically, the CBA. For example, the 1986-87 Budget included $125 million for the provision for contingency salary and related costs (non-Defence) - see 1986-87 Budget, p. 75.
    42. As the budget is generally released in May, the outcome year is an estimate and not the final outcome.
    43. Program specific parameter variations are published in reconciliation of expenses tables in budget updates (see Table 1 of this report).