Counting the cost: Treasury lifts the veil on how tax policy proposals are costed and managed

By Angus Isherwood, Joe Sothcott & Robyn Walker


For those with an interest in tax policy, one important area that has long remained shrouded in mystery is the fiscal costings attached to tax policy proposals. This is surprising, given that a proposal’s fiscal impact is often a decisive factor in whether it proceeds, and the costings produced by Inland Revenue and, to a lesser extent, the Treasury provide the Government’s authoritative view of that impact. This matters in practice because the fiscal treatment of a proposal can determine whether an otherwise sound remedial, integrity or simplification measure is able to proceed. In a welcome development, the Treasury has lifted the veil with the August release of its Guidance Document on the Fiscal Management of Tax Policy Proposals.

This issue of Tax Alert considers what the guidance covers, what it reveals about how tax policy is made in New Zealand (including the role of the Budget operating allowance and the Tax Policy Scorecard) and where the process might go from here.

Fiscal Management Approach and the Allowances

The Guidance Document examines how the fiscal management approach (FMA) applies to tax policy. The FMA comprises the Government’s budgeting conventions, which are intended to ensure policy decisions are consistent with the fiscal strategy of the Government of the day.

It identifies three mechanisms for managing the fiscal impacts of tax policy decisions (collectively referred to as the “allowances”):

1. Budget Operating Allowance: Funding set aside in each Budget for new policy initiatives and changes to existing policy resulting from Ministerial decisions.
2. Between-Budget Contingency: The portion of the operating allowance reserved for policy changes requiring funding between Budgets.
3. Tax Policy Scorecard: The default mechanism for managing tax policy proposals that increase or decrease tax revenue. The Scorecard is measured over a five-year period and must remain between zero and $200 million. It is intended for policies that improve the tax system, rather than proposals designed to advance broader economic or social policy objectives, or more structural changes to the tax system.

The Guidance Document explains that Ministers ultimately decide how a proposal’s fiscal impacts are managed. The default is to use one of the allowances, although impacts may flow directly through to the Government’s fiscal indicators. This typically occurs where the impact is indirect, Ministers have limited discretion in the decision-making process, or the change is not attributable to a policy decision, such as revisions to tax revenue forecasts.

Fiscal Costings

Officials are required to estimate the direct fiscal impact of a policy using available information and data sources. The starting point is to establish the counterfactual, being the revenue expected to be collected if no change were made to existing policy and legislative settings.

This requires considering whether the subject matter of the proposal is already reflected in tax revenue forecast baselines. Adjustments may be needed following a change in interpretation, new information or expected enforcement activity, but generally not for little-known issues, matters subject to the Commissioner’s care and management powers, or where a policy change is responding to an adverse event from which the Government was not expecting to receive revenue (for example, depreciation recovery arising from insurance proceeds).

Once the counterfactual has been established, there are three approaches to fiscal costings:

1. Zero-cost: No fiscal impact is attributed where the fiscal effect is expected to be nil, or where direct increases and decreases in revenue offset one another.
2. Notional costing: If the expected fiscal impact is non-zero but below $0.2 million per annum and insufficient information is available, a positive or negative impact of $0.2 million per annum is assigned. This equates to $1 million over the five-year forecast period.
3. Detailed costing: Where the fiscal impact is expected to exceed $0.2 million per annum, a detailed fiscal analysis is required.

Detailed costings incorporate first-order consequential effects on other tax types. For example, changes to FBT can affect corporate income tax collections. Material and quantifiable behavioural responses are also taken into account. The Guidance Document notes that the costing for the 39% trustee tax rate increase reflected possible changes in company profit retention, trust dividend income and beneficiary distributions.

Where broader secondary economic effects can be reliably quantified, officials may recommend disclosing a net fiscal impact that incorporates broader economic effects alongside the direct fiscal impact. Investment Boost is one example where this approach was adopted.

However, this is not the default position. Treasury generally favours disclosure of direct fiscal impacts only, on the basis that wider economic effects are more appropriately captured through aggregate economic and fiscal forecasts. The Guidance Document notes that international practice in this area continues to evolve.

The Guidance Document also confirms that stakeholders can contribute to fiscal impact analysis through the Generic Tax Policy Process. When the Government releases its Tax and Social Policy Work Programme, officials generally provide Ministers with broad indications of whether proposals are fiscally positive, negative or neutral, and their likely scale. As policy design develops, stakeholder input can help improve the accuracy of detailed costings and identify design issues that might otherwise go unrecognised. However, the scope for consultation may be constrained where Budget secrecy applies.

Fiscal Impacts Arising from a Change in Interpretation

The Guidance Document also considers situations where the courts or Inland Revenue adopt a statutory interpretation that is inconsistent with broader policy intent, current Government policy, or established taxpayer practice.

A change in interpretation can affect tax revenue forecasts where the interpretation differs from prevailing taxpayer practice. The Guidance Document states that these impacts generally flow directly through to the Government’s fiscal indicators, rather than being charged against an allowance, provided they are considered reasonably probable (that is, more likely than not to occur).

Factors relevant to assessing reasonable probability include whether taxpayers can avoid the tax outcome through alternative structures or arrangements, whether the Commissioner is likely to exercise their care and management powers, and whether Ministers have signalled an intention to amend the legislation.

Where Ministers decide to change the law, officials will generally present the fiscal impact of restoring the position to the original policy intent, existing Government policy, or prevailing taxpayer practice. As this involves an exercise of Ministerial discretion, the Guidance Document states that the fiscal impact will usually be managed through one of the allowances.

This creates a divergence between the treatment of the fiscal impact arising from the change in interpretation and the fiscal impact of subsequently reversing that outcome through legislation. The result is that the fiscal costings can be prohibitive in restoring the law to the previous interpretation, practice, or policy intent, which is often criticized as a weakness in the current system.

The Guidance Document does acknowledge that, on occasion, officials may recommend that the fiscal impact of the legislative amendment also flow directly through to the fiscal indicators. Officials note that this generally arises where Ministers have very limited practical discretion other than to restore the previous position.

The Guidance Document also refers to an “up-down” or “down-up” approach where a change in interpretation has not yet been incorporated into forecast baselines. In these circumstances, the fiscal effects of both changing and not changing the law can be netted against one another. For transparency purposes, officials present both fiscal outcomes to Ministers, highlighting the discretionary nature of the policy decision.

Non-compliance and care and management

The Guidance Document also considers situations involving significant taxpayer non-compliance.

Where the Commissioner applies the care and management provisions to an issue pending a legislative amendment, a forecast adjustment will generally not be required. Where compliance is partial or mixed, officials estimate the expected level of compliance and incorporate that assumption into the proposal’s fiscal costing. While this approach is sensible in principle, fiscal costings may overstate existing compliance in some areas and, in turn, inflate the estimated cost of a proposed policy change. Recent proposals to modernise the FBT rules provide one example; earlier efforts to simplify the tax treatment of donated trading stock provide another.

Irregular impacts over time

The Guidance Document also addresses situations where fiscal impacts are unevenly distributed over time. Treasury may recommend using an extended assessment period where the timing of costs and benefits differs materially outside the standard five-year forecast period and doing so improves transparency.

Deloitte comment

In Deloitte’s view, the Guidance Document is a welcome first step. While it usefully outlines the system at a high level, practical examples showing how particular fiscal costings have been calculated would provide much-needed clarity. Key assumptions, counterfactuals and any behavioural adjustments underpinning material costings should also be made more transparently available. Greater transparency would help address the concern among tax policy observers that costings can appear arbitrary or insufficiently grounded.

More broadly, the Guidance Document should be a launchpad for improving the FMA, not the final word. Although Ministers ultimately decide how fiscal impacts are treated, officials’ recommendations under the FMA will generally carry significant weight. It is therefore important that responsibility for improving the system is not avoided. In particular, Deloitte would support sensible changes that produce symmetrical, net-nil fiscal treatment where legislation restores a previously widely accepted position. The Guidance Document also notes that treatment of broader “second-round” effects is also evolving internationally, and New Zealand should help lead that development so fiscal costings present a more complete picture of the wider impacts of tax policy changes. Consideration should also be given to whether the default notional costing of $0.2 million per annum remains appropriate for policies whose fiscal impacts are difficult to quantify.

Overall, the Guidance Document provides a useful foundation for understanding how the fiscal impacts of tax policy proposals are assessed and managed. The next step should be greater transparency, supported by practical examples and continued refinement of the FMA, to promote more consistent, balanced and well-informed tax policy decisions. If you have any questions, please reach out to your usual Deloitte adviser.