Australia's Future Tax System

Final Report: Detailed Analysis

Chapter G: Institutions, governance and administration

G5. Monitoring and reporting on the system

G5–1 Monitoring the performance of the system

Australia needs a tax system that is efficient, equitable, simple, consistent and sustainable in the long term. Monitoring the system is essential to judging its performance against these criteria. Where possible, the performance of specific taxes and transfers should be measured objectively to identify whether they are meeting their policy objectives or not. An objective evidence base can reinforce public and government support for successful economic reforms (Wilkie & Grant 2009), and helps to determine when existing policy settings are no longer appropriate.

The Review has undertaken widespread public consultation in forming its recommendations and has been presented with many conflicting points of view. The judgment of experts about what to tax, and how, departs markedly from current practice in many cases. Moreover, there is no community consensus about existing taxes. Very few taxes received consensus support in submissions; many attracted conflicting recommendations, while a few received both condemnation and praise.

Sometimes, differences in opinion about taxes depend on underlying values or ideology. However, viewpoints are sometimes based on theory or conjecture that can be proved or disproved by appeal to the evidence. Wherever possible, policy-makers should draw on empirical observations in Australian conditions (Leigh 2009).

This becomes more important as governments seek to use tax-like instruments for a range of purposes beyond raising general revenue. For example, quite detailed knowledge is required to set rates of tax to correct market failures. To ensure such taxes actually make society better off overall, they should be calibrated to reflect the latest evidence on marginal social costs and benefits of different activities (see Section E Enhancing social and market outcomes).

Box G5-1: Better evidence is needed to set the rate of alcohol tax

Some taxes, such as congestion taxes and alcohol taxes, are designed to correct spillovers (or 'externalities'), which occur when individual decision-makers fail to take into account the impact of their actions on others. These taxes, while theoretically sound, are hard to implement in practice because objective and reliable information on marginal social costs is often not available.

This makes it difficult to set rates that actually target spillover costs. In the absence of clear evidence, the public may perceive a conflict of interest between setting tax rates to target social problems, and raising revenue. An acute example is the taxation of alcohol (see Section E5 Alcohol taxation).

In submissions and in meetings with the Review, stakeholders stressed the need for data based on an impartial, transparent and open methodology. These data could then form the basis for determining tax rates. For instance, the National Drug Research Institute (NDRI) submitted that:

    The drinking culture and characteristics of countries vary, as does that of different demographic and psychographic groupings within countries. Trends, behaviour and disposable income change over time. It is necessary to develop relevant data and to continually update those data.

    To optimise tax settings designed to influence price points as a contribution to safer and healthier consumption of alcohol, continual empirical scientific study is needed on price elasticities, cross-elasticities and substitution effects in the Australian context (Stockwell & Crosbie 2001). Such research needs to be done independently of vested interests. NDRI urges the Government to provide substantial ongoing funding for such work.

The paucity of information on some specific taxes or transfers means that data designed for other purposes are often used for analytical purposes. For example, the Review has received many submissions from a range of stakeholders about the taxation of alcohol. In making their arguments, many submissions relied on evidence about the social cost of alcohol consumption, drawing on Collins and Lapsley's cost of illness studies (2008b) as well as survey data from the National Health and Medical Research Council. However, the findings in these studies cannot be directly imported into a tax policy context. In his research for this Review, Cnossen (2009) highlighted some of the methodological issues involved.

As a means of increasing the availability of evidence and informing debate, the Review Panel has commissioned a series of papers to explore significant tax and transfer policy issues.

Principles

Monitoring the system is essential to judging its long-term performance. Where possible, governments should objectively measure the performance of specific taxes and transfers to identify whether they are meeting their policy objectives or not. Data on the tax system should be freely available to the public.

Data on the tax system are a public good and should be freely available

Unbiased and systematically collected data on the tax system, based on widely accepted methodology and appropriate for tax policy purposes, are rare and often not available in the public domain. Because such information is a public good (see Section E1 on user charging), and even though society would benefit through improved tax policy based on it, incentives for individuals or businesses to produce it are weak. In addition, the capacity of non-government actors to generate this type of information is limited when government holds much of the data needed to conduct analysis.

Some data on the effect of taxes and transfers are already being produced. In many submissions to the Review, stakeholders have supported arguments with data that range from in-house survey results to more sophisticated economic modelling. The Review has considered this information in coming to its recommendations. However, if the underlying sources and methodology are not transparent, it is difficult to judge between competing claims.

There is a greater risk of biased evidence where the functions of data collection and analysis are not clearly separated. If data are gathered for the purpose of supporting a particular argument or interest group, they are unlikely to persuade those with competing interests.

In some cases, estimates are produced, but only as a snapshot at one particular point in time, and often only on a small or partial scale. For example, while there have been a few studies of compliance costs imposed by the tax system (such as Pope 1994 and Evans et al. 1997), there has been no aggregate study of compliance costs in Australia since 2000 and there is no system-wide study of compliance costs that can monitor, on an ongoing basis, the costs of complexity. Proxies, like the number of tax agents or the length of the income tax laws, are of limited policy use. Well-designed system-wide surveys are expensive, but they would provide valuable information on where simplification would yield the greatest returns.

There is also little published information about the extent of non-compliance with the tax laws. The Australian Taxation Office (ATO) does not currently attempt to derive aggregate estimates of non-compliance for key income and deduction items or publish detailed results of its non-compliance activities.

There is very little information on compliance costs incurred by clients in meeting their obligations in respect of transfer payments — for example, in claiming payments, reporting earnings or changes in income and participating in reviews of their entitlements. However, Centrelink and various policy agencies conduct a range of surveys and analytical exercises that could be extended to provide better, ongoing information about client compliance costs.

The reliability of information could be improved by adopting a consistent methodology, classifications, concepts and data collection approaches that could operate as a common standard for tax data, irrespective of who produces it. This would improve the capacity of business, academics and government agencies to produce evidence to inform the policy process.

The prime consideration for the collection of data is to ensure that collection is not biased. The Australian Bureau of Statistics might be best placed to conduct surveys on compliance costs, as well as the social costs associated with particular activities (for example, the consumption of alcohol, tobacco or transport).

Existing data sources could be better utilised

In the past, Australian tax unit record files have not been generally available for research purposes. However, in 2009 the ATO produced and released a confidentialised 1 per cent sample file containing individual tax return information. Over time, expanded availability of this, and similar data for other taxes, would allow deeper analysis of the individual-level effects of the tax system. These data should only be used for statistical purposes, and should never be provided in a form in which individual privacy could be compromised. The guarantee that data will be kept confidential produces better quality data and higher response rates. Further developing protocols for making such data available for research purposes would unlock valuable public information resources.

Data on the transfer system are more readily available than tax data. Confidentiality and privacy concerns are addressed by providing de-identified data under strict terms relating to its use and safe storage. Research undertaken using social security data has generated many insights that have fed into policy development, including information on the persistence of welfare dependence among specific groups, the 'transmission' of welfare dependence between generations, the responsiveness of different groups to policy changes, and the effectiveness of policies in achieving their intended outcomes. Linked datasets, especially longitudinal datasets, are a powerful tool for evidence based policy. The experience of the Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA) has highlighted the considerable benefits that can be derived from making this type of data available to academic and other researchers.

Findings

Unbiased and systematically collected data based on accepted methodologies and appropriate for policy purposes are rare and often not publicly available.

While society as a whole would benefit through improved tax policy from this information, the incentives for individuals or businesses to produce it are weak. Also, the capacity of non-government actors to conduct analysis of the system is limited when much of the information needed to do this is held by government.

Periodic analysis is needed to judge the system's performance

Recommendation 132:

The government should, every five years, publish a Tax and Transfer Analysis Statement that analyses and reports on the overall performance and impact of the system, including estimates of efficiency costs and distributional impacts.

Recommendation 133:

The Australian and the State governments should systematically collect data on aspects of existing taxes and transfers — including compliance cost data — according to consistent and transparent classifications and concepts, and make this information — including confidentialised tax unit records — freely available for further analysis and research.

Recommendation 134:

The government should support one or more institutions to undertake independent policy research relevant to the Australian tax and transfer system.

Australia needs a tax and transfer system that is efficient and equitable in the long term. To see whether these objectives are being met, it is important to analyse and report on the system's performance objectively and comprehensively. Currently, no such assessment of the system is undertaken.

Instead of assessing the performance of the tax and transfer system as a whole, it has become customary to consider elements of the system separately. As a consequence it is difficult to get a sense of the system's combined performance and effects, and to determine whether the system is making a coherent contribution to our national objectives.

The tax and transfer system is subject to continual change, and operates in a constantly evolving environment. Periodic analysis is needed to ensure it continues to perform as intended, and to identify areas that require attention. In theory, it might be desirable to constantly review and assess the system, but this is not feasible in practice. Instead, the Australian government and the States should periodically report on the performance of the entire system in a 'Tax and Transfer Analysis Statement'. This assessment could be carried out at regular intervals, such as every five years, to balance costs and benefits of the exercise (see Recommendation 132). The Australian government should initiate this change, with the Council of Australian Governments to examine the ways in which the States could follow suit.

Among other things, the Statement could estimate the total and marginal efficiency costs of taxes, tax expenditures and means tests. It could include information about the economic burden and distribution of taxes and transfers, and about tax expenditures not possible to publish in the annual Tax Expenditures Statement (discussed below). The problem of tax evasion and avoidance could be monitored and assessed by publishing detailed estimates of the level of non-compliance. This would inform the policy and administration changes needed to maintain the integrity of the system.

The Tax and Transfer Analysis Statement would contribute to ongoing research and debate about the objectives and performance of the tax and transfer system. Academics, practitioners and the general public should be encouraged to contribute to, and contest, the analysis presented in the Statement. All data used in the analysis and a full description of methodologies should be available to the public and ideally subject to peer review.

Government agencies should ensure that data on all aspects of the tax and transfer system are shared as widely as possible (see Recommendation 133). Provided that confidentiality of individual records is guaranteed, as much data as possible should be freely available to researchers, universities, think-tanks, businesses, other agencies and the public at large. While the analysis of this data would clearly be contestable, it would help provide a common point of reference for discussions of the tax and transfer system.

Another approach would be to establish an academic institution or a partnership between institutions that could draw funding from public and private sources to undertake tax research and evaluation (see Recommendation 134). An arrangement like this exists in the United Kingdom, where an independent business tax research centre is funded by the business community and has access to government data sources.11 Similar bodies operate in Australia, in other areas of research.

On the transfer side of the system, the Department of Families, Housing, Community Services and Indigenous Affairs funds research organisations to provide it with independent, high-quality research services. The research allows social policy to be made on a strong evidence base and permits information, data and analysis to be disseminated for public use. Aside from ongoing formal arrangements such as these, there is also a role for conferences and other forums to facilitate open discussion between public officials and private sector experts.

Any government institution would need to be careful not to 'crowd out' analysis from universities, think tanks and peak bodies. However, it could identify and help fill gaps in analysis, or identify where there are holes in the data available to researchers.


11 The Centre for Business Taxation was established in 2005 as an independent research centre of the University of Oxford.