Rather than capping tax revenue, the government should reform the system

As the government pulls together the May budget, Finance Minister Mathias Cormann is aiming to cap tax revenue at 23.9% of GDP. At best, this is an arbitrary cap. More importantly, it distracts from real tax reform, which could increase national productivity and income.

There are many potential reforms to the tax system that are revenue neutral, from broadening the tax base to replacing transaction taxes.

The 2016 federal budget projected that federal tax revenue would be 22.2% of GDP in 2016/17, rising to 23.5% by 2019-20. Meanwhile, the government’s Intergenerational Report projects ever-growing government expenditure as a share of GDP thanks to, among other things, an ageing population and further increases in health outlays.

These two trends make it hard to believe that capping tax revenue at 23.9% of GDP could ever bring the budget back to balance. Furthermore, a comparison of Australia’s tax revenue to GDP ratio with other countries with comparable living standards shows Australia as a relatively low taxed country.

Using the latest numbers compiled by the OECD, we find that tax as a share of GDP when you include all levels of government is just 27.8% in Australia. This compares to an average of 34.2% across all OECD countries. Several European countries have a share above 40%, although both the US and Switzerland are relatively rich countries with lower tax as a share of GDP than Australia.

The diversity of tax to GDP ratios across countries indicates the design and operation of taxation and government expenditure programs are more important than the tax share.

Tax distortion

Collecting a dollar of tax costs society much more than a dollar. For income and GST taxes, the distortion caused by levying a tax is estimated by the Treasury and others to exceed twenty cents per dollar raised.

This is because levying a tax distorts peoples’ decision making about employment, investment and saving, choice of business type, and so forth.

For example: a higher income or payroll tax can influence decisions about how much to work; the capital gains tax concessions given to home owners incentivises the reallocation of savings from business investments to housing.

As tax and government expenditure as a share of the economy increase, the costs of the additional tax rise while the benefits of the additional expenditure fall. As the tax rate rises, the tax distortion costs rise at a greater rate than the rise in the tax rate.

The cost of tax distortions to private sector decisions on employment, investment and so forth is a reason to constrain tax as a share of GDP. But if we are trying to enhance the well-being of Australians, setting a ceiling on taxation as a share of GDP seems an issue of second order importance. While the principle is clear, its practical application is dubious.

So at what level should we tax?

Identifying the right level to tax is subject to conflicting political and personal assessments. This is because, beyond funding the government, tax can serve a number of roles.

A progressive income tax – increasing the marginal tax rate as income rises – is one of the instruments used to redistribute income to those on lower incomes.

Some forms of taxation can be used to account for costs to society that may be ignored in market transactions - such as taxing pollution, or to discourage behaviour - like taxing tobacco or gambling.

Tax can also be one of the levers to smooth economic outcomes over the business cycle. Increasing taxes during economic booms and reducing them during recessions as part of a overarching policy to achieve full employment and low inflation, for example.

An idealistic benchmark to maximise social well-being would increase until they reach the point where the benefit to society of additional government spending no longer exceeds the distortion cost of raising the additional tax. There is no evidence to indicate this point is 23.9%.

In fact, the wide range in tax to GDP ratios, from the US at 25.9% to Denmark at 49.6%, illustrates there is no one answer.

Tax reform

Instead of looking at tax to GDP, there are far more important and higher value reforms that could be made to the taxation system.

Over the past several years we have had a number of inquiries and discussions, including the Henry Review and the more recent Re:think discussion paper.

These have highlighted many options to improve the taxation system in roughly revenue neutral reform packages.

One of the ideas raised was to remove special exemptions and deductions while lowering tax rates. It would collect the same amount of tax revenue while reducing distortions and the associated losses of national income.

Another idea is to replace transaction taxes with broad based asset taxes, such as the ACT model of replacing stamp duty with a land tax. We could also better design special purpose indirect taxes, such as with alcohol.

Currently there are different rates for keg beer, other beer, brandy, spirits, along with the wine equalisation tax. These could be replaced with a common flat rate of tax per unit of alcohol by volume across all alcoholic beverages.

More challenging reforms include changes to the tax mix, such as a larger GST, which has relatively small distortion costs, combined with a smaller income tax.

There are many options to reform taxation and derive greater value from the money diverted from private use. Placing an artificial and contrived ceiling on taxation as a share of GDP has to be both a long way down the list of sensible reforms and one that is controversial in practical implementation.