PML on tax reform

The following guest post is by regular ALS reader P.M.Lawrence.

I have now put a second submission into the Henry Tax Review, for the main part. In the faint hope of constructive criticism, here is the body of this one with my contact details removed and with links to the sources for the appendices:-

Submission to the Henry Tax Review


Incentives to increase employment and GDP that operate through the tax system, and scope for simplifying the revenue base and reducing its incidental burdens.


Recommendations (1a.), (1b.) and (1c.) cover a Pigovian virtual wage subsidy integrated with the tax system, to improve both employment and GDP without the problems usual with wage subsidies, Negative Income Tax, etc.

Recommendation (2.) covers reducing incidental effects of GST on investment.

Recommendation (3.) covers a lower compliance alternative GST approach, suitable for certain businesses.

Recommendation (4.) covers a lower compliance and more globally competitive alternative corporate tax approach.       



Even in normal economic conditions Australia, like many developed countries, has material levels of unemployment and underemployment. In poor conditions like those currently obtaining, not only do these deteriorate but also GDP suffers. Many approaches to address these issues have been suggested and/or adopted, most of which are beyond the scope of this Tax Review. However, during the past fifteen years or so, a number of workers in different countries have independently arrived at proposals which integrate older ideas of wage subsidies with the tax system. Notable among these workers are Professor Kim Swales of the University of Strathclyde and his colleagues, and Nobel winner Professor Edmund S. Phelps, McVickar Professor of Political Economy at Columbia University. I myself have done a game theoretic analysis of aspects of these proposals.

In the course of considering the Australian tax system in the light of this work, with a view to adapting it for use here, I observed a number of situations where other improvements could be made in passing. I bring these out in this submission as convenient.


(1a.) GST offsets should be provided, at a basis level per full time employee per annum (net after Payroll Tax etc., and pro rata for part timers, but not offsetting for overtime). The offsets should apply to all employees, new or old, employed by others or self-employed, quite without regard to whether they had been long-term unemployed, briefly unemployed or had been employed for a long time – and the offsets should apply indefinitely, with no cut-off. This basis level should be set to match typical benefit levels for the unemployed, plus their marginal costs. This is currently above $10,000 per person per annum.

(1b.) While the Tax File Number (TFN) system offers one way to implement these offsets for employees, they should instead be implemented by giving eligible people vouchers for each hour to be offset, to be passed on to employers (giving them out quarterly in advance, say). This particularly allows family members to distribute paid and unpaid work more conveniently within households.

(1c.) To maintain short term revenue neutrality and long term budget neutrality, ideally the base rate of GST should be raised to restore the shortfall, as in Professor Swales’s proposal (see appendix A). However, this appears to conflict with part of the terms of reference of the Tax Review: “…the review will reflect the Government’s policy not to increase the rate or broaden the base of the GST…”. Therefore some combination of the following should be done:-

– Other taxes and sources of revenue may be increased to make up the difference. As the GST is remitted to the states, this may be done at state level.

– In current economic circumstances, the Government may decide that some level of deficit financing is warranted anyway. It may be convenient not to restore some or all of the shortfall, at least in the short term.

– On a broad construction of “not increasing the rate of the GST”, it may only be required not to increase the GST paid, net after the offsets. On this view, the percentage rate itself may be increased as it is only an intermediate figure for purposes of calculation. The rate could be increased to a certain point if all payers were to avoid higher payments, or to levels sufficient to restore the shortfall completely if aggregate payments were not to increase. Note that Professor Phelps’s variant of the proposal has tax offsets that are more specifically targeted to the lower paid, so reducing the shortfall to be made up (see appendix B). As against that, it requires more policing.


Consider how unemployment benefits are funded at present. In the short term, when a firm makes a single hire or fire decision, it faces the costs and benefits to itself of that one employee more or fewer. However, it does not face the corresponding reductions or gains in the direct costs and on costs of that employee’s unemployment benefits since those are passed to the broader tax base (countries without this sort of support for the unemployed face Vagrancy Costs instead). This creates an externality favouring lower employment levels. This incentive applies to each actual or potential employee at each instant, so it continues to pressure employers regardless of whether employees go from one employer to another – the usual case in normal economic circumstances beyond the short term. That is, this effect relates to the immediate impact and not the ultimate incidence of the costs and benefits as would normally be the case.

Wage subsidies and similar methods offer a Pigovian approach to redressing this externality. In general, such systems face problems in a number of areas. For instance, the “Five Economists’ Plan” for a Negative Income Tax provides a subsidy with a point of impact on employees. This leads to problems from high effective marginal rates of (personal) tax, problems of wage stickiness delaying declines in real wages sufficient to price more people into work, political problems from those declines, and funding problems during the delay. Paying wage subsidies directly to employers has huge and continuing funding problems if done broadly enough to be effective, and may face delays by encouraging higher real wages.

All of these problems are engineered out by providing the subsidies as offsets to a broad based carrying tax with its point of impact on employers (in this case GST, but Payroll Tax would also serve). The essential feature is that, in a precise technical sense, these offsets are not an actual subsidy but a virtual subsidy. That is, while they provide the same incentives as an actual subsidy, there is no cash flow or funds flow from the government to employers. Instead the offsets are calculated and applied at an intermediate stage and there is still a net cash flow to the government with no churning. This has the following implications:-

– As the point of impact is on employers, there are no issues arising from effective marginal rates of tax.

– As other parts of the tax system can be adjusted when this is brought in, e.g. by raising the calculation GST rate gross before the offsets, the tax system can be kept revenue neutral in the short term while keeping much the same incidence.

– As employee wages are not affected, there are no delays from waiting until wage rates adjust.

– As the offsets are matched to the direct costs and on costs of unemployment benefits, the system is budget neutral in both the short and long term; any improvements in employment that reduce tax revenues automatically reduce the burden of funding unemployment benefits, and contrariwise any increase in unemployment automatically increases the revenue base to compensate. This means that the system can be maintained indefinitely even if other factors slow employment gains.

The externality is driven by game theoretic effects at the point of impact, much like the “Tragedy of the Commons”, so it is not picked up by analysis of incidence. On the other hand, it neither discredits nor challenges analyses of, say, Payroll Tax that show that its incidence does not harm employment. In fact, these offsets could be applied to Payroll Tax itself rather than to GST, creating a Negative Payroll Tax analogous to Negative Income Tax. What matters for this effect is the detailed structure at the point of impact, and as currently structured Payroll Taxes do no harm. The two areas are distinct, and there is no inconsistency between them (unless, of course, states counterproductively chose to restructure their Payroll Taxes to claw back the offsets).

It is worth noting that this system of tax offsets would also offer an automatic stabiliser effect for the economy, by acting equivalently to Pigou’s Real Balance Effect. If implemented through anonymous transferrable vouchers, over time those vouchers would in effect become money providing an income stream equivalent to one from corresponding personal investments. Indeed, if the role of supplying the vouchers were eventually transferred to a suitable Sovereign Wealth Fund, e.g. the Future Fund or the fund proposed at option (3b.) of my submission on the Retirement Income System (see appendix C), the Pigovian subsidy would be transformed into a much more self managing Coasian solution to the unemployment benefit funding externality.

The reasoning above derives from game theoretic analysis, qualitative apart from making tax offset levels match unemployment benefit outgoings. Professor Swales analysed the situation quantitatively with economic modelling (see appendix A). His work indicates that, in the UK situation under a wide range of parameters, not only employment but also GDP would improve, as one would expect from reducing or eliminating an externality: “The proportionate expansion in total output lies within the range 0.8% and 2.8% and the increase in employment in the range 1.45% to 4.0%” (see “The Employment Effect of Subsidies”, Report to the Directorate General Employment, Industrial Relations and Social Affairs, Commission of the European Communities, SOC 94 100018 05A01, available at and following). Even if there were no market imperfection, Professor Phelps believes that wage subsidies of this sort would be justifiable on equity grounds: “There are cases… where it would be wise to deviate from the free market, and low-wage employment subsidies is one of those cases” (see his interview with Challenge of July-August 1997 at, on his related book “Rewarding Work: How to Restore Participation and Self-Support to Free Enterprise” (Cambridge: Harvard University Press, 1997)). This may justify setting the offsets as high as desired minimum wage levels, which would both remove the need to enforce those and eliminate any additional disincentive to employment minimum wage levels may present from effectively acting as a tax on employment.

(2.) GST is often described as a consumption tax, from its economic incidence on the final consumer as opposed to its impact or legal incidence on various intermediate stage producers. However, as many intermediate GST payments have to be made before final sale, either directly to the ATO or embedded indirectly in payments to suppliers, a simplistic GST has incidence on working capital and on goods and services that enter into capital goods as well. In many countries, adjustments are made to reduce or eliminate this. Australia uses a “provisional tax” approach for many taxes, to obtain payments as soon as possible, which reinforces this incidence on investment in the case of GST (although some adjustments are made). To reduce or eliminate this, provisional payment of GST should be dropped wherever it occurs, payment in arrears should be permitted where practical, and interest should be paid on any GST prepayments that still occur. For goods and services forming fixed capital, this interest may be folded into depreciation allowances.

(3.) Firms operating in a single whole or main business area should be offered the option of calculating their GST obligations based on a deemed proportion of their turnover and giving their customers GST credits based on a deemed proportion of their prices, with those proportions determined by their areas of business.


The burden of GST compliance (e.g. providing BAS information) is uneven between different sorts of business. For some, business requirements need that level of management accounting anyway and there is no great additional burden, but for others that level of detail is unnecessary or not cost effective and the business has to implement further systems and carry the excess of their costs over their benefits. Where a business operates in a single whole or main area, its GST obligations and credits correlate highly with its turnover multiplied by a factor that depends on its particular area of activity. This is a consequence of the Du Pont measure of Return On Investment (ROI) that separates functional components of ROI by breaking it down as (returns/sales) multiplied by (sales/investment). Subject to only comparatively few firms being involved so that Goodhart’s Law does not distort it, ABS statistics provide enough information to determine the turnover and price proportions for particular business areas. However, as many businesses have already incurred the costs of GST compliance systems, this should only be an option rather than a required further change.

(4.) A class of company should be set up with exemption from corporate taxation but making equivalent issues of shares to the government instead. These shares should be held at arm’s length from consolidated revenue and direct governmental control by selling them to one or another Sovereign Wealth Fund, e.g. the Future Fund or the fund proposed at option (3b.) of my submission on the Retirement Income System, which would specifically apply its revenue to consolidated revenue in lieu of separate taxes and charges (see appendix C). Dividend franking should be continued on the basis of this share revenue, or not, according to other policies about franking. Existing companies should be offered the option to reorganise in this form.


Corporate taxation represents an area of competition between national economies within the wider global economy. To an individual firm, there are issues of locating operations in different jurisdictions because the taxes themselves vary and because compliance costs vary. When a firm relocates outside Australia, the effect on taxes here is much greater than the benefits to the firm itself as the firm only gains from the differences in its burdens.

In earlier times a different approach from corporate taxation was sometimes used: rather than levying taxes as such, countries required corporations to provide them with a corresponding issue of shares (often preference shares), which became part of a revenue yielding portfolio or “domain”. Corporations gained from a reduced compliance burden, as it was no longer necessary to calculate tax obligations separately from and in addition to calculating obligations to shareholders. In current circumstances of globalisation, there would be additional advantages to Australia: corporations would no longer have tax incentives to relocate elsewhere, as their obligations would follow them; and, where operational reasons led to a move, the revenue would still return to Australia as invisible earnings.

Unfortunately, abuses and fiscal irresponsibility led this method to fall into disrepute. Some governments did not maintain their portfolios properly, making poor investments of their own (as the Khedive of Egypt did, realising his shares of the Suez Canal Company and applying the proceeds to developing the Egyptian cotton industry) or even worse treating them as income, realising them and spending the proceeds. Later governments then often ignored their predecessors’ share acquisitions, particularly if they had been dissipated misguidedly or even stolen by earlier kleptocrats, and simply observed a situation where corporations were not being taxed or otherwise providing revenue streams. This led to them instituting taxes or even nationalising corporations (as in the case of the Suez Canal Company).

The current situation is that both governments and corporations prefer a system of corporate taxation, the former because they have greater certainty and control over revenue and the latter because they are aware of the political risk (sovereign risk) of issuing shares and later facing taxes anyway. However, the advent and reasons for sovereign wealth funds offer this approach new scope. Since other policy objectives require these funds anyway, shares can be placed with them instead of funding them from other resources. They provide an arm’s length place for shares, reducing or eliminating political risk. And their revenue can be returned to consolidated revenue and/or can reduce the demands on consolidated revenue.

3 thoughts on “PML on tax reform

Comments are closed.