The short answer

How OzPolitics reviews a policy proposal

We test the same six things every time: mechanism, legal authority, cost, delivery capacity, supporting evidence and material uncertainty. Political ownership does not change the test.

This is a method for assessing public claims, not a verdict on a party or a direction on how to vote. Conclusions should change when the underlying law, data or implementation evidence changes.

The useful question is not only “what is the rule?” but also “who administers it, which document controls it, and when might it change?” That distinction prevents an accurate general explanation from becoming wrong advice in a particular election, chamber or policy setting.

Evidence review

Why a standardised review framework matters

Australian political debate is saturated with policy announcements, each presented as self-evidently beneficial. A minister may announce a grant program, an opposition leader may propose a tax change, or a crossbench senator may advocate regulatory reform. Each comes with claims about what it will achieve, how much it will cost, and who will benefit. Without a consistent framework for evaluating these claims, the public is left to assess proposals based on the proponent's credibility or partisan alignment. OzPolitics cuts through that noise by applying the same six tests to every proposal, regardless of which party or individual is advancing it.

These six tests are mechanism, legal authority, cost, delivery capacity, supporting evidence, and material uncertainty. They are drawn from practices in public administration, regulatory impact analysis, and program evaluation as applied by institutions such as the ANAO, the PBO, and the Department of Finance. The framework is not original to OzPolitics; it reflects how serious policy analysis is conducted inside and outside government. What distinguishes our approach is that we apply it transparently, in plain English, and without regard to the political identity of the proposal's author.

Evidence review

Test one: mechanism — how is the policy supposed to work?

Every policy proposal is a claim about cause and effect. It asserts that if the government does X, then Y will happen in the world. The mechanism test asks whether the causal chain is plausible and clearly specified. If a policy proposes to reduce petrol prices by cutting fuel excise, the mechanism is straightforward: a lower tax rate should translate into a lower price at the bowser, assuming the reduction is passed through by retailers. If a policy proposes to increase productivity by funding a new industry research centre, the mechanism is more complex: it requires assumptions about how research translates into commercial application, how long that takes, and whether Australian firms can capture the benefits.

We examine whether the proposal identifies a specific problem and explains how the intervention will address it. We ask whether there is a theory of change that connects the government action to the desired outcome through identifiable intermediate steps. We check whether the policy targets the right point in the causal chain. A proposal that funds more hospital beds to reduce emergency department waiting times, for example, implicitly assumes that bed capacity is the binding constraint rather than staffing, discharge planning, or primary care access. If the mechanism is vague, internally inconsistent, or relies on heroic assumptions, we flag that as a significant weakness.

Evidence review

Test two: legal authority — can the government actually do this?

The Australian Government can only act where it has legal authority to do so. The Commonwealth's powers are enumerated in the Constitution and elaborated in legislation. A policy proposal may be politically attractive but legally impossible without new legislation, or it may fall within an area of state rather than Commonwealth responsibility. We check the Federal Register of Legislation at legislation.gov.au to identify the relevant Acts and regulations that confer or constrain the claimed authority. We examine whether the proposal would require new primary legislation, a regulatory change, an intergovernmental agreement, or a constitutional amendment.

This test also considers whether the proposal is consistent with existing legal obligations, including international treaties to which Australia is a party. A proposal to impose a new trade restriction, for example, must be assessed against Australia's commitments under World Trade Organization agreements and bilateral free trade agreements. We check whether analogous policies have been challenged in the High Court and what the jurisprudence suggests about the constitutional head of power being relied upon. A proposal that lacks a clear legal pathway is not necessarily invalid, but it carries implementation risk that must be acknowledged by its proponents.

Evidence review

Test three: cost — how much will it cost and who pays?

Cost is not just the headline budget number announced in a press release. We examine the full fiscal impact of a proposal over the forward estimates period and beyond, drawing on budget papers published at budget.gov.au, costings produced by the Parliamentary Budget Office at pbo.gov.au, and relevant departmental annual reports. We distinguish between direct spending, tax expenditures, contingent liabilities, and indirect costs imposed on other levels of government or the private sector. A policy that appears inexpensive in budget terms may impose significant compliance costs on business or administrative costs on state governments.

We also consider distributional effects: who bears the cost and who receives the benefit. A tax cut that is presented as benefiting all Australians may disproportionately advantage high-income earners when measured in dollar terms, even if the percentage reduction is uniform. A new user charge for a government service may have a larger impact on low-income households. We look for the distributional analysis in the budget papers, regulatory impact statements, or independent costings, and when none is available we note the gap. The Parliamentary Budget Office's capacity to produce independent costings of policies proposed by non-government parties and members has been an important addition to the transparency infrastructure, and we draw on PBO analyses wherever they exist.

Evidence review

Test four: delivery capacity — can this actually be implemented?

A policy is only as good as its implementation. The delivery capacity test asks whether the government agency or agencies responsible for executing the proposal have the staff, skills, systems, and organisational capability to do so within the proposed timeframe. We draw on reports from the Australian National Audit Office at anao.gov.au, which regularly examines program implementation and identifies systemic weaknesses in departmental capability. ANAO reports frequently document problems with procurement, contract management, IT system delivery, and performance measurement that are relevant to assessing whether a new initiative can be executed competently.

We also consider the track record of comparable programs. If a proposal resembles an existing or past initiative that experienced implementation difficulties, we note those precedents. We check whether the responsible department has been subject to capability reviews or recent budget measures affecting its staffing. A proposal requiring coordination across multiple Commonwealth agencies and state governments carries additional delivery risk. We do not assume good intentions will translate into good outcomes; we look for evidence that the implementing institutions have the resources and expertise required.

Evidence review

Test five: supporting evidence — what is the empirical basis for the claim?

This test examines the quality and relevance of the evidence cited in support of a policy proposal. We distinguish between different types of evidence: randomised controlled trials, quasi-experimental studies, longitudinal surveys, administrative data, expert consensus, international comparisons, and anecdotal examples. We assess whether the cited evidence actually supports the claimed mechanism and whether there is countervailing evidence the proposal ignores. We give weight to evidence published by the Australian Bureau of Statistics at abs.gov.au, by independent statutory bodies such as the Productivity Commission, and by peer-reviewed academic research.

We are particularly attentive to the difference between correlation and causation. A claim that a particular policy in another country caused a particular outcome requires evidence of a causal relationship, not merely an observation that the two coincided. We check whether pilot programs or trials have been conducted and what their results showed. We note when a proposal relies on modelling that has not been publicly released or independently verified. The absence of evidence is not necessarily fatal to a proposal, particularly if it addresses a genuinely novel problem, but we require proponents to acknowledge uncertainty rather than assert certainty where none exists.

Evidence review

Test six: material uncertainty — what could go wrong?

Every policy operates in an uncertain world. The material uncertainty test asks what the proposal assumes about future conditions and how sensitive its outcomes are to changes in those conditions. A budget measure that projects savings from reduced welfare payments, for example, depends on assumptions about employment, wage growth, and eligibility patterns that may not materialise. A climate policy that projects emissions reductions depends on assumptions about technology costs, consumer behaviour, and international carbon markets. We identify the key parameters on which the policy's success depends and assess how realistic the assumptions about those parameters are.

We also consider political and institutional uncertainty. A policy announced by a government may be reversed by a successor. Funding promised in a budget may not survive future fiscal consolidation. A regulatory scheme may be challenged in court or rendered ineffective by avoidance behaviour. We do not predict which of these scenarios will occur, but we flag that they are material to the policy's prospects. The purpose of this test is not to generate pessimism but to encourage honest acknowledgment of risk. A policy proposal that is presented as a guaranteed success is less credible than one that identifies the conditions under which it will work and the circumstances under which it will not.

Common questions

Before you rely on the answer

Why does OzPolitics apply the same tests to every policy regardless of which party proposes it?

The credibility of a policy does not depend on who announces it. A proposal's mechanism, legal basis, cost, feasibility, evidence, and uncertainty are independent of the political identity of its author. Applying the same framework to every proposal ensures consistency, prevents bias, and allows readers to compare policies on their merits. This approach is modelled on the practices of independent institutions such as the Parliamentary Budget Office and the Australian National Audit Office, which assess proposals and programs without regard to political ownership.

How does OzPolitics handle policies that have not yet been costed by the Parliamentary Budget Office?

When a policy has not been independently costed, we note that and assess the proponent's own estimate against available data. We check comparable programs for which costings exist, examine relevant budget line items at budget.gov.au, and identify the assumptions most affecting the cost estimate. The absence of a PBO costing does not prevent analysis, but it carries higher uncertainty. We encourage readers to press proponents to submit policies for independent costing.

Do the six tests apply differently to spending proposals versus regulatory proposals?

The framework applies to both, but the emphasis shifts. For a spending proposal, cost and delivery capacity are typically the most heavily weighted tests. For a regulatory proposal, legal authority and mechanism receive more attention because the policy operates by changing behaviour through rules rather than by disbursing money. The evidence test is equally important for both categories, as is material uncertainty. A ban on a product, for example, requires evidence that the product causes harm, a legal basis for the prohibition, an assessment of substitution effects, and consideration of enforcement costs.

Where does OzPolitics get the data it uses for policy analysis?

We draw primarily on publicly available official sources. These include the Australian Bureau of Statistics (abs.gov.au) for economic and social data, the Federal Register of Legislation (legislation.gov.au) for legal instruments, budget.gov.au for fiscal information, the Parliamentary Budget Office (pbo.gov.au) for independent costings, the Australian National Audit Office (anao.gov.au) for program evaluations, and departmental annual reports and regulatory impact statements. We also reference peer-reviewed academic research, Productivity Commission reports, and data published by state and territory governments. We do not use proprietary data that readers cannot independently verify.

Source spine

Primary material used for this guide

Review trigger: Creation of new independent statutory bodies with policy evaluation functions comparable to the PBO or ANAO. Major reform of the federal budget process or regulatory impact analysis requirements that changes the availability or format of policy documentation. Significant High Court rulings that alter the constitutional division of powers in a way that affects the legal authority test for Commonwealth policy.

Archive note: This article describes OzPolitics internal methodology and is not itself derived from a single government source. The six-test framework is synthesised from principles employed by the ANAO, PBO, the Department of Finance regulatory impact analysis guidance, and the Office of Best Practice Regulation. The source links point to the key institutions whose outputs inform our reviews. The methodology may be refined as these institutional practices evolve.

Primary links are provided without affiliate or tracking parameters. Confirm that the source still applies to the bill, sitting date, jurisdiction or reporting period before relying on it.