|Submission No. 267||Back to full list of submissions|
|Download in either PDF or RTF format|
Tax Reform and Infrastructure
Submission to the Review of Business Taxation
|Agriculture, forestry & fishing||
|Electricity, gas & water supply||
|Accommodation, cafes & restaurants||
|Transport & storage||
|Finance & insurance||
|Property & business services||
|Government administration & defence||
|Health & community services||
|Cultural & recreation||
|Personal & other services||
Source: Australian Bureau of Statistics, Private Sector Construction Industry 1996/97 (Cat. No. 8772.0)
Until recently, the public sector was responsible for much of the nations capital expenditure on infrastructure. The construction of roads, railways, airports, electricity generators and transmission lines, gas pipelines and ports was largely funded by governments which then owned and operated the facilities. Since the late 1980s this has changed markedly. Pressures on Commonwealth and State budgets have led to extensive pruning of outlays and much of this has occurred on the capital side of the budget.
The downward trend in national gross fixed capital investments share of GDP identified in Figure 1.2 therefore disguises a divergent trend between capital formation by the private sector and public sector. These developments are depicted in Figure 1.4, which reveals a slight upward trend in the share of GDP accounted for by private sector capital investment, whilst public investment has declined from around 8 per cent of GDP in the 1960s to less than 4 per cent currently.
A large proportion of gross fixed capital formation by the private sector represents investment in dwellings. In an attempt to focus more closely on what is happening to investment in productive capital (rather than residential housing), Figure 1.4 also shows the trend in private sector non-dwelling capital formation as a share of GDP. This has fluctuated between around 3 and 5 per cent since the 1960s, and currently stands at approximately 4 per cent of GDP.
GROSS FIXED CAPITAL FORMATION BY SECTOR AS A PERCENTAGE OF GDP
Source: Australian National Accounts, National Income and Expenditure.
Note: Derived from data expressed in current price terms.
Most of the longterm decline in public investment has occurred in the provision of infrastructure, as demonstrated in Figures 1.5 and 1.6, which are taken from the National Commission of Audit Report to the Commonwealth Government in 1996.
In its Final Report of September 1995, the Economic Planning Advisory Commission (EPAC) Private Infrastructure Task Force found that the long term decline in the ratio of public investment to GDP is not, in itself, sufficient evidence that current levels of infrastructure are inadequate. Indeed, some of the fall in infrastructure investment was undoubtedly justified, as there was previously obvious over-investment in some areas. The large increase in public expenditure on gas and electricity that occurred in the early 1980s (Figure 1.6) is one obvious example of an area in which there was obvious over-investment, and the subsequent fall in investment in this area may be interpreted as an attempt to correct this imbalance.
In addition, there was a move away from large infrastructure projects built in large part for reasons other than on a purely economic justification (such as the Ord River Scheme). Finally, there was a recognition that, potentially, significant improvements could be made in the productivity of some of the existing infrastructure, which could then yield additional services, as an alternative to building new infrastructure.
PUBLIC EXPENDITURE ON TRANSPORT
Source: National Commission of Audit: Report to the Commonwealth Government, 1996
PUBLIC EXPENDITURE ON COMMUNICATIONS AND UTILITIES
Source: National Commission of Audit: Report to the Commonwealth Government, 1996
Nonetheless, for some categories of infrastructure, and in particular roads, many observers considered it beyond doubt that there had been significant under-investment. Figure 1.5 shows the very large fall in expenditure on roads that has occurred in recent decades. There has also been a trend decline in expenditure on water and sewerage assets.
This had been largely accepted by governments, especially State governments, who are generally keen to see large infrastructure projects in their jurisdictions. However, they have been generally unenthusiastic about building these projects themselves, largely (although not entirely) for budgetary reasons. Instead, as discussed in the following section, governments are now turning to the private sector to build, and in some cases own and operate, the infrastructure projects.
There are a number of reasons why, until recently, governments have believed they should play the major role in the provision of infrastructure. In addition to certain noneconomic considerations (such as a philosophical desire to keep the provision of essential services within the public sector), there are a number of economic arguments for public sector investment in infrastructure:
However, over the past decade or so, the importance attached to the arguments for public provision has declined. Whilst governments have continued to be planners and initiators of infrastructure projects, they have increasingly sought private sector involvement in the provision of finance and the management at some or all project phases. This development was highlighted in the National Commission of Audit Report to the Commonwealth Government in 1996, in which the Commission supported increased private sector involvement and recommended that the role of the government be restricted to regulatory and limited funder/purchaser functions.
There have been several reasons for this move to private provision of infrastructure:
Infrastructure is easy to recognise but difficult to define. Infrastructure investments are different from other forms of business investment for a variety of reasons, including:
These characteristics of infrastructure present some particular aspects that are very significant to private investors. Unless they are addressed adequately by the tax system or some other policy instrument, private investment in infrastructure is likely to be well below the socially optimal level. The three main issues are:
These three issues are considered further below.
2.2 The Tax Disadvantage Problem
A major problem for investors is that infrastructure projects typically do not yield positive cash flows for many years after expenses begin to be incurred. Typically, infrastructure projects have the distinctive characteristic that, for a period of up to 10 years (and in some cases longer) from the beginning of construction, they can yield negative cash flows. This is because there are no revenues during the construction phase, and revenues early in the operations phase are low relative to interest payments (and depreciation).
This is illustrated in Figure 2.1 which shows for a hypothetical road project a series of losses in early years followed by positive returns only after a good deal of elapsed time (12 years of operation). Similar patterns of returns are likely to be exhibited by investors in other infrastructural assets, such as electricity generators.
TAXABLE INCOME AND LOSSES CARRIED FORWARD FOR A ROAD
Source: Macquarie Corporate Finance Ltd
If a single company invested in these assets, the tax problem would be much reduced because losses in the early years could be offset against income earned by the company from its other investments. Investment in infrastructure, however, is typically carried out by consortia of major corporations, which form standalone companies mainly in order to share the substantial capital raisings that are required and, hence, the risk. These special purpose, standalone companies do not have other income against which to offset these losses, and Australian tax laws prohibit the flow through of these losses to company shareholders. The inability to pass through tax losses was one of the major reasons why the Very Fast Train (VFT) project did not proceed at the beginning of the decade.
It is often argued, therefore, that infrastructure projects are placed at a tax disadvantage relative to other projects. This argument, however, is not universally accepted and in particular is disputed by the Commonwealth Treasury. In 1995, the EPAC Private Infrastructure Task Force, relying on a consultants report, argued that in terms of effective tax rates, standalone infrastructure projects are not in fact disadvantaged relative to infrastructure projects where full use can be made of interest deductions, or relative to investment in longlived plant and equipment. For all three types of investment, the effective tax rate was calculated as about 29 per cent. The same view was reiterated by the Treasury in its submission to the Financial System (Wallis) Inquiry.
There a number of points which need to be made in response to this argument. First, the estimates of effective tax rates for infrastructure investment are largely driven by the assumption that capital gains accrue free of tax. While this is true, it is of dubious relevance since, under the Australian tax system, capital gains are in fact taxed on realisation, not accrual. Capital gains accrue tax free on all assets, not just infrastructure projects. Moreover, infrastructure BOOT projects are eventually transferred from the private owners to the public sector free of charge at the end of a specified period. Thus the project owners never realise any capital gains. It is therefore highly questionable to impute to them a theoretical accrued capital gain when calculating effective rates of income taxation.
Secondly, the effective tax rates which the EPAC Task Force drew upon to make its conclusions assumed away any cash flow problems in the early years of a project, when revenues are low (or nil) and costs are high. In reality, cash flow problems may in fact significantly constrain the ability of private investors to build infrastructure, especially large projects.
Thirdly, effective tax rates can be highly sensitive to the assumptions made in their calculations with respect to a large number of variables such as the life of the investment, the rate of depreciation, the interest cost of borrowings, the rate of a companys share turnover, whether the company has sufficient taxable profits to obtain benefits from depreciation allowances in the year that they occur, the rate of inflation and the proportion of a project that is debt financed. It is not at all obvious that the effective tax rates reported by EPAC are not sensitive to small changes in these key parameters.
In short, the argument that infrastructure projects (especially standalone projects) are not tax disadvantaged is questionable. On balance, the cash flow issue has led to two arguments being advanced for special taxation treatment of infrastructure:
A second barrier to the private sector investing in infrastructure is the inability of private investors to recoup the full value of the facility they provide. The widespread acceptance of user charging these days such as for the City Link project in Melbourne and Sydneys toll motorways means that private investors can charge directly for use of the facility. In terms of direct use, the free rider problem has been eliminated.
The difficulty is that indirect free riders cannot be charged for using the facility. The private investor cannot derive, through user charges alone, any compensation for the value the project adds to the local economy. For example, a major road or bridge may add considerably to the productivity of a regional economy by cutting costs of production for users and bringing additional custom to local businesses. Owners of local businesses, as well as governments, may be the major financial beneficiaries of such infrastructure which adds to economic activity in a region, pushing up land values and the tax base. While private owners of infrastructure can charge for direct users of their assets, they cannot capture the value enjoyed by entities which benefit indirectly.
In the past, governments built infrastructure partly because they realised that such activities would ultimately strengthen their own revenue bases and thus repay (at least in part) the cost of the infrastructure. Private investors do not have the means to recapture such external benefits and thus may be forced to charge inappropriately high prices (ie, in excess of marginal costs) to direct users unless government makes a contribution which recognises the collateral benefits to society as a whole, or unless the infrastructure franchises include some other value capture mechanisms.
The value capture problem is a genuine externality which, unless adequately addressed by policy makers, can have a negative effect on the ability of the private sector to provide certain kinds of infrastructure. Arguably, while the tax issue was probably of primary importance, an inability to capture some of these benefits was a major factor in the non-viability of the VFT project.
In combination, the characteristics of infrastructure mean that the returns to private provision are heavily dependent on current and future government actions. In particular, private infrastructure projects can be subject to substantial regulatory (or policy) risks that are not evident in other private sector operations. Regulatory risks can relate to planning requirements, external caps on the pricing of services, environmental requirements and the conditions governing the entry of new competitors. Where these regulatory risks are not managed effectively, investment in infrastructure can be relatively high risk. Moreover, the impact of regulatory risk is exacerbated by the fact that infrastructure assets are typically high cost projects with few alternative uses once the investment is made.
Where governments determine the terms of pricing of infrastructure services (as is becoming common in the case of the energy sectors, for example), then the cashflows of private infrastructure projects are also effectively regulated. The issue in Australia currently is that regulatory systems for infrastructure facilities are still very much in their infancy, which creates much uncertainty for potential investors. From this perspective, it can be argued that investors in Australian infrastructure projects face higher regulatory risks than their counterparts in countries such as the US and UK, where regulation systems are well evolved.
The recent decisions on natural gas transmission access arrangements in Victoria suggest that authorities acknowledge the high regulatory risks faced by investors in Australian infrastructure. In October 1998, the Australian Competition and Consumer Commission (ACCC) and the Victorian Office of the RegulatorGeneral (ORG) released their final decisions relating to the prices that will be set for the next five years for gas transportation in the State. After initially basing the decisions on a real pretax weighted average cost of capital (WACC) of 7 per cent, both the ACCC and ORG subsequently determined that the appropriate WACC to use was 7.75 per cent. One of the reasons for the change was a recognition that there is an element of regulatory risk in Australia. It is an open question whether the WACC levels becoming established in the regulated energy sector compensate adequately for this risk.
The above analysis suggests that there are at least three factors which can make private investment in infrastructure relatively less attractive than investment in other businesses:
Unless the Government provides some offsetting advantage to reduce or eliminate this net disadvantage, private investment in infrastructure is likely to be significantly less than it would be on a level playing field. Clearly, the Review of Business Taxation offers one vehicle by which this issue could be addressed.
As suggested in Chapter Two, the Treasury may not accept the validity of the argument that investment in infrastructure is deterred by the inability of consortia to write-off early losses against other income of the component companies. Nevertheless, the Government does appear to accept the validity of the argument, to the extent at least that they have introduced measures to address it. The Infrastructure Bonds (IB) program and its successor, IBTOS, are examined below.
In response to some of the criticisms of the policy environment for private provision of infrastructure, the Keating Government introduced a new incentive, Develop Australia Bonds, in its 1992 One Nation Statement. Infrastructure Bonds (IB) as they became known, were an indirect means by which (under certain conditions) project owners could access tax deductions on their borrowing costs during the construction and early operations phase of projects. The mechanism was indirect because it worked by making the bonds nonassessable to the lenders (who would be largely indifferent between a taxable bond paying, say, 10 per cent and a nonassessable bond paying 6 per cent) and nondeductible to the borrowers (who had no income against which to deduct the interest costs).
One of the criticisms made in the EPAC Private Infrastructure Task Force Report was that it was not clear that the full benefits of the IB tax concession was being passed on to the borrowers. That criticism was premature, and reflected the (then) immaturity of the market. The progressive reduction in costs to infrastructure borrowers over time is shown in Figure 3.1, reflecting a market better informed about taxexempt bonds, greater retail appetite for them, and consequently reduced costs of underwriting and distribution. At the time of the Treasurers announcement ending the program, the cost of borrowing via Develop Australia Bonds was about 60 per cent of the cost of conventional borrowings. There can be no doubt that infrastructure bonds succeeded in lowering project borrowing costs.
EFFICIENCY OF INFRASTRUCTURE BONDS
Source: Macquarie Corporate Finance Ltd
It must be emphasised that Develop Australia Bonds, while having some unintended and unfortunate consequences as far as tax minimisation schemes were concerned, also brought about significant benefits to both the developers of infrastructure and the community as a whole.
These benefits were recently summarised at the time by Mr George Brouwer, then Chairman and Chief Executive of Invest Australia.
For the developer:
For the community:
Perceived problems with the IB program, however, led the Treasurer to terminate the scheme in February 1997. In making his announcement, the Treasurer identified three major problems with Infrastructure Bonds: the potential cost to revenue, their use in aggressive tax minimisation schemes and the fact that (apart from the projects themselves) the major beneficiaries were high marginal rate taxpayers and financial packagers.
The potential large cost to revenue arose as the financial markets and investors in infrastructure became familiar with the potential of the IB program. It occurred despite the apparent stated cap on revenue costs ($150 million in 199697 and $200 million in 199798) because of a loophole in the Development Allowance Authority Act 1992. Specifically, under the Act, it appears that the Authority was obliged to consider all applications received before the Treasurer directs the Authority not to accept further applications for Infrastructure Borrowing Certificates, in order not to breach the revenue cost cap for the year. The Treasurer gave such a direction on 10 September 1996, but all applications received between 1 July and that date had to be considered on their merits. Between 1 July and 20 August 1996 (Budget night) the DAA in fact received applications for 71 IB projects with estimated borrowings of around $21.6 billion.
The second of the Treasurers objections the use of infrastructure borrowings in aggressive tax minimisation schemes was apparently valid. These schemes were not the ones which have received publicity in the media. (These involved quite innocuous tax arbitrage which was in fact anticipated at the time infrastructure borrowings were introduced.) Rather, they involved complicated structures including numerous related companies (sometimes offshore) which managed to break the symmetry of the taxation arrangements. Other schemes included financing instruments such as zero coupon bonds, which managed to increase the tax benefits by large multiples of the benefits gained from ordinary infrastructure bonds.
The last of the Treasurers claims, that infrastructure bonds provide tax benefits to high marginal tax payers, is irrefutable. These bonds were retailed to investors in packages of at least $500,000 thereby obviating the need for a prospectus which would otherwise be required for an offer to the public. The amount paid to the institutions retailing these schemes was 40 per cent of the deductions sought, thus only investors with a marginal tax rate in excess of 40 per cent could (profitably) participate.
While this may have represented a problem of fairness (low marginal rate taxpayers did not have the opportunity to benefit from the tax treatment of the bonds), it was inevitable given the original desire to reduce borrowing costs to infrastructure projects. These costs could only be reduced inasmuch as lenders were prepared to accept a lower coupon return, which was free of tax. The higher was the marginal tax rate of the lender, the lower the tax free rate at which they were prepared to lend, and hence the lower the borrowing costs for the project.
The industry itself and the finance sector disputes the validity of the view that the IB scheme was fatally flawed. In particular, they feel that the cost to revenue means that the scheme was working as intended: if the Government wished to spend less on supporting infrastructure investment, the cost to revenue cap could have been reduced. Also, the industry claims strongly that most of the rorts had already been removed when the scheme was terminated.
The IB tax concession scheme was replaced by IBTOS, a more limited tax offset scheme that applies to interest derived by lenders to approved infrastructure projects. Division 396 of the Income Tax Assessment Act 1997 provides for a tax offset to be allowed to resident lenders in the first five years of borrowings by the project borrower. The offset is calculated by applying the general company tax rate to the interest that a lender includes in assessable income. The offset may be subject to a maximum limit. Where the lenders interest is subject to a tax offset, the project borrower is denied a deduction in respect of a comparable amount of interest.
Unlike the previous IB scheme, which could be used to finance the construction of a wide range of infrastructure facilities (including land and air transport, gas pipelines, water supply, electricity and sewerage), support under IBTOS is limited to approved road and rail projects only (although nonland projects that applied under the previous scheme are eligible to apply for a tax rebate). There is also a cap on the overall cost to the scheme of $75 million per annum.
There is overwhelming criticism among industry stakeholders about the effectiveness of the IBTOS scheme. For example, in its December 1998 submission to the Review of Business Taxation, the Australian Council for Infrastructure Development (AusCID) described the scheme as "totally inadequate to the task of catalysing private sector investment in public infrastructure". Consultations with financiers of infrastructure developments have been equally damning in their criticism of IBTOS.
The perceived problems of IBTOS are summarised in the following paragraphs.
As discussed above, the tax relief available under the IBTOS is capped at $75 million, which is considered to be unrealistic by industry stakeholders. It is claimed, for example, that one mediumsized rail project (such as the Sydney to Canberra fast train concept) could readily soak up the cap. Overhanging applications from the previous IB scheme could also absorb the budget, thereby making new applications futile. Furthermore, it has been argued that the transactions costs involved in applying for a share of the annual $75 million of tax relief (estimated by one financier at approximately $50,000 per application) are likely to be too high to make it worthwhile.
There is concern that the IBTOS application process is not transparent, leading to great uncertainty about whether a proposed project will qualify for tax relief. The previous IB tax concession was enshrined in legislation and had a transparent process with clear criteria, and where recommendations were made by a body at arms length from Ministers. In contrast, the industry considers the IBTOS to be more of an administered benefit where success or failure can be determined by political currents of which the applicants are quite unaware.
It has been suggested that the residual paranoia from the rorting associated with the previous IB regime means that the taxation authorities adopt a very cautious approach when considering IBTOS applications. The culture of the ATO, and the Treasury, is directed more towards protecting the revenue than assisting business to access the program.
Another criticism of the IBTOS application procedure is that applications can only be made on a twiceyearly basis. This is more restrictive than the previous IB scheme when applications could be made at any time.
The tax benefits afforded by IBTOS (and therefore potential impact on infrastructure investment) varies according to the type of institution undertaking the financing:
A more convincing argument for the lack of attraction of the IBTOS to superannuation funds is that it is rare for superannuation fund investors to invest in unrated senior debt obligations, reflecting the conservatism of most superannuation fund trustees.
The conclusion arising from this analysis is that the IBTOS scheme is only valuable to the individual investor. However, as discussed above, an individual investor will not want to bear the construction risk associated with an infrastructure project, and the ATO has introduced obstacles to prevent individuals from avoiding this risk. It has issued draft taxation determinations that:
The potential impact of these issues are addressed in the following section.
In late November 1998, the Australian Tax Office (ATO) issued three draft taxation determinations relating to the operation of the IBTOS: TD98/D17, D18 and D19. These determinations impose restrictions on IBTOS that have no basis in existing legislation and will lead to outcomes that are manifestly contrary to the intended policy of the incentive scheme.
TD98/D17 and D19 threaten to limit tax deductions for interest paid on funds that have been on-lent to Land Transport Facilities (LTF) at rates lower than the investors funding costs. Such an outcome would go totally against the policy intent of the IBTOS especially given that the primary intent of the IBTOS is to secure lower borrowing costs for transport infrastructure projects (ie, financiers lending at a lower rate than would otherwise apply).
The implication of TD98/D19 is that where investors borrow at a particular interest rate (10 per cent, say) and, due to the availability of the tax rebate, lend at a lower rate (say 7 per cent), it is likely that these investors will have their interest deduction for the 10 per cent interest paid limited to 7 per cent. The effect of this is to penalise the lender and ensure that it increases the rate at which it is willing to lend. In fact, the only level at which the lender does not suffer this penalty is if they lend at 10 per cent. Therefore, it appears that the ATO is suggesting that lenders should not provide any reduction in interest rate and keep all the tax credit to themselves.
The ultimate impact of TD98/D19 will be to discriminate against any LTF investors sourcing funds by way of debt, in favour of LTF investors capable of sourcing funds from equity. In other words, TD98/D19 will effectively limit the market for LTF bonds to superannuation funds, large banks and high net wealth individuals with any benefit generated by the scheme likely to be eliminated by the additional risk premium applied by these investors.
The perversity of the outcome that appears to be suggested by the draft determinations was overcome in the case of the previous Infrastructure Bond Scheme by releasing TD94/80. Although TD94/80 was drafted with specific reference to the Infrastructure Bond Scheme, it did contain the following general principle relating to interest deductibility in the context of a tax incentive scheme involving borrowed funds:
"In order for the interest to be deductible, the investor must have entered into the loan solely for the purpose of funding the investment in or acquisition of the infrastructure borrowings. One indicator that the investor had a purpose other than, or in addition to, funding the infrastructure borrowings would be where the deductions in relation to the infrastructure borrowings are greater than the exempt return on the infrastructure borrowings grossed up by the investors marginal tax rate as it would be but for the infrastructure borrowing investment and any related income and deductions."
It would sensible and important that a similar principle also be applied to investments made under the Tax Offset Scheme. However, TD98/D17 appears expressly directed to denying the operation of such a principle.
On a conceptual level, the determinations fail to recognise that onlending to infrastructure borrowers at a lower interest rate is the primary and critical objective of the Tax Offset Scheme, and as such is not only desirable but necessary. On a policy level, the impact of the determinations could be to limit the range of potential investors in IBTOS borrowing and to increase the benefits generated for the investors, not the projects. Each of these significantly reduces the Schemes competitiveness and efficiency.
Another concern of the infrastructure industry is draft determination TD98/D18, which states that socalled dual funding structures are likely to attract the operation of Part IVA of the Income Tax Assessment Act 1936.
According to TD98/D18, such structures are said to facilitate:
There is little argument that contrived structures involving significant cost to Revenue over and above that anticipated are contrary to intended outcomes and should be disallowed. However, there are objections to the characterisation of all dual funding structures in this manner. Dual funding structures should not be viewed prima facie as a tax avoidance structure. Instead, they should be considered as marketdriven structures motivated by the commercial imperatives of delivering lowest cost finance to infrastructure borrowers. Moreover, given the high degree of discretion available to the Minister under the tax offset legislation, there is little danger that tax aggressive structures involving individual investors will be adopted by project sponsors.
It seems sensible therefore for the ATO to amend TD98/D19 to recognise more fully the commercial merits and intent of dual funding structures and to clarify that Part IVA will not be applied indiscriminately to challenge bona fide dual funding arrangements.
One of the major issues for the Review of Business Taxation is whether the statutory tax rate should be reduced from 36 to 30 per cent. Achieving this within the constraint of revenue neutrality, imposed by the government, would mean that many tax concessions would have to be eliminated. The most important of these is the accelerated depreciation schedule introduced as part of the One Nation economic statement in 1992. Two main points need to be made here.
First, given that one of the main objectives of the present review is to improve the international competitiveness of the corporate tax system, it seems somewhat contradictory to impose the condition of revenue neutrality. If the aggregate tax burden on business remains the same, it is difficult to see how international competitiveness can be significantly improved.
Secondly, the elimination of accelerated depreciation would re-create the situation that existed before 1992 whereby the tax system effectively discriminated against investment in assets with longer lives. This is discussed below.
There has been a longstanding dialogue between business and governments about the tax treatment of longlived assets compared to short and medium lived assets. Before 1992, depreciation allowances were based on effective lives of assets. These were often very similar to physical lives, and herein lay the problem. Many investors argued that the risks in investing in longer term assets are greater than in assets with shorter lives. For the infrastructure sector, the risks of investing in longerlived assets are additional to the other barriers to infrastructure investment identified above.
One substantial risk is that of early obsolescence as new technologies, unforseen when the asset was purchased, become available. Many people would view rapid technological change as affecting shorter lived assets, such as computer equipment, and not being particularly relevant to infrastructure assets such as gas pipelines or electricity generators, for example but in fact this is not the case. The cost of building a gas pipeline has approximately halved in recent years with the introduction of new materials and construction techniques. The efficiency of the latest coal-fired electricity generators is substantially greater than those built ten years ago, while technological advances have, irrespective of any change in relative fuel prices, increased the competitiveness of gas-fired generators.
Another important area of risk for the infrastructure sector is stranded asset risk. Using gas pipelines as an example again, the asset could become stranded if the gas field unexpectedly dried up or could only produce uneconomic quantities of gas. This, however, is an extreme example. An asset such as the Moomba-Sydney pipeline, which was deliberately built with substantial excess capacity to cater for future demand, would become partially stranded if a new line is constructed, as planned, connecting the Gippsland Basin gas reserves with the Sydney market. In the electricity market, assets could become stranded in a number of ways, not least as a result of fiscal measures to address greenhouse gas emissions.
The tax depreciation rates introduced on 27 February 1992 went a long way to providing a level playing field between assets with different lives. One reason for introducing the new schedules, as well as the IB program, was to stimulate investment in infrastructure. As well as acknowledging the need to bring Australian tax treatment of longlived assets more into line with other OECD and Asian countries, the government statement emphasised the need to provide an appropriate environment in which the shift of infrastructure investment from the public to the private sector could go ahead efficiently and effectively.
Since the 1992 accelerated depreciation rates were introduced, a major wave of private investment has taken place in the infrastructure sector. Since 1992, the number of kilometres of gas pipelines has increased by nearly 50 per cent. Significant privately owned motorways have been built in Sydney, and Melbournes City Link Project is one of the worlds largest private infrastructure projects. A substantial number of government-owned infrastructural assets have also been sold to private investors.
Of course, Treasury may argue that this investment in infrastructure has been incentive-driven and therefore represents a misallocation of resources at the expense of more efficient investment in shorter-lived assets. Any detailed examination of the nature of the investment in infrastructure which has taken place since 1992 makes this argument very difficult to sustain.
In these circumstances, there is a high risk that removing accelerated depreciation tax concessions, as part of a package designed to reduce the nominal rate of company tax to 30 per cent, would have the highly undesirable effect of threatening future private investment in infrastructure. On its own, such an action would add one more substantial barrier to infrastructure investment to add to the other disincentives identified above.
It is no understatement to say that Section 51AD of the Income Tax Assessment Act 1936 (s51AD) has been the bane of private sector infrastructure providers in recent years, and represents a major impediment to private sector infrastructure investment in Australia. As the private sectors role in infrastructure provision has grown in importance, so have the problems associated with this legislation. It is reassuring to note that this has already been recognised by the Review of Business Taxation:
"While always criticised for its severe impact, section 51AD has become more problematic because of privatisation and outsourcing of government functions that were not contemplated when it was first conceived."
Review of Business Taxation, A Platform for Consultation, Discussion Paper 2, Volume I, February 1999, p.226.
Specifically, s51AD applies to property predominantly financed by non recourse debt which is leased to, or effectively controlled by, an end user that:
This provision was introduced to counter a form of leveraged leasing used by (taxexempt) State governments in the early 1980s to access the benefits of tax deductions in constructing power generators. The application of s51AD is severe, however, because it disallows all tax deductions relating to the investment, whilst all income remains assessable for tax purposes. Where s51AD is deemed to apply, it effectively destroys the economics of the private sector undertaking infrastructure investment.
A related provision, Division 16D, applies in respect of a qualifying arrangement where s51AD does not apply and where there is use or effective control by an end user who is:
Where Division 16D applies, it denies capital allowances to the owner of the property, and treats lease payments as repayments of principal and payments of interest.
s51AD and Division 16D were introduced in the early 1980s as antiavoidance measures aimed at preventing inappropriate use of tax incentives designed to encourage investment in plant and machinery. The legislation was devised in an era where there was no private ownership of infrastructure in Australia and virtually no private management of this infrastructure. Government was assumed to be the natural owner of infrastructure assets and any other arrangements involving the private sector could automatically be assumed to be shams taxdriven transactions in which the private sector merely pretended to ownership. Furthermore, anti-avoidance laws at the time were generally found to be ineffective.
In the context of private provision of public infrastructure, s51AD and Division 16D are aimed at preventing sale and lease back arrangements whereby the taxexempt government entity sells an infrastructure asset to a private company, so that the company can access tax deductions for depreciation and interest payments, but which leases the facility back so as to effectively maintain public control over the use of the asset.
There is no question that provisions need to be in place to prevent tax abuse. The main concern of infrastructure providers is that, in the light of recent developments, s51D and Division 16D have become inappropriate as measures to prevent tax avoidance. Specifically, s51AD and Division 16D were legislated at a time when the sort of continuum of risk sharing that is becoming increasingly common in the infrastructure sector through BOOTtype arrangements was not envisaged. As a consequence, the current legislation has the potential to deny tax deductions even if the private sector takes on significant risks, thereby jeopardising joint public/private sector infrastructure ventures.
Stated simply, the main problem with s51AD is that its potential scope is extremely broad. Its application is essentially dependent on whether the Commissioner of Taxation considers that a government retains control of the use of the infrastructure asset. Since control need only be potential, more than one party can be deemed to control the use of an asset. Given that governments often retain regulatory, coordination and safety functions in respect of major infrastructure assets, the ATO frequently forms a prima facie view that government control exists.
Virtually no private sector project will proceed without a favourable ruling on s51AD. As discussed above, where s51AD is deemed to apply, all income is taxable yet all deductions are disallowed (even if the private sector takes on considerable risks). Consequently, private sector parties will not proceed with an investment until they have obtained clearance from the ATO. A considerable amount of time and money can be spent structuring arrangements so that a project does not fall foul of s51AD (and so that it is likely to receive a favourable dispensation from the Tax Commissioner in relation to Division 16D deductions). An additional problem is that the operation of these provisions effectively rules out any joint public/private sector infrastructure project.
The excessive time taken to process rulings on the application of s51AD and Division 16D to private infrastructure projects, or on related administrative problems, can also jeopardise investment. ATO rulings can take several months to obtain, by which time finance commitments may be withdrawn and the proposed investment may not proceed. On other occasions, private consortia may decide not to bid at all for proposed infrastructure projects because uncertainty surrounding the applicability of s51AD makes fundraising untenable.
It is common for privately financed infrastructure and major resource projects to use limited recourse debt during and just after construction until refinancing can be based on better business conditions post rampup. Where an asset is financed by limited recourse debt, a taxpayer obtains capital allowance deductions which have previously been permitted if the debt was not fully repaid. However, Division 243 of the proposed Taxation Laws Amendment Bill (No.4) 1998 intended to change these arrangements by clawing back all depreciation deductions in excess of equity contributions plus debt principal repayments whenever limited recourse debt is refinanced or the underlying asset is disposed.
The proposed Taxation Laws Amendment Bill (No.4) 1998 lapsed due to the Federal election. However, Division 243 has been subsequently reintroduced as part of the Taxation Laws Amendment Bill (No.5) 1999. It has been slightly amended with the intention of resolving some of the main concerns raised about the earlier draft. However, it does not seem to have done this.
In addition to the clawing back of depreciation deductions, Division 243 also has the effect of increasing the amount of limited recourse debt that is deemed not to be repaid by:
In drafting Division 243, it appears that the ATO is committed to the view that the availability of capital allowances in project financing should henceforth be linked to the risk exposure associated with the borrowing. Thus, by definition, limited recourse debt is deemed to be less risky to the borrower because it is secured only against project assets and cash flows, and is not linked to the balance sheet of the borrower.
These proposed changes would effectively eliminate the use of limited recourse finance in any meaningful sense, favouring equity and corporate debt as the preferred sources of development capital. Since limited recourse debt is the most common financial structure for infrastructure projects (simply for commercial reasons), this provision is particularly unreasonable in the context of an emerging business sector based on privately financed infrastructure development. Furthermore, as it currently stands, the proposed legislation does not appear to give the Tax Commissioner discretion to apply it only in appropriate circumstances.
It has been suggested that the financing of infrastructure projects has been unintentionally included with the remit of proposed Division 243, reflecting an oversight by drafters of the legislation. Nevertheless, until the situation is clarified, many infrastructure projects are at risk. Schemes acknowledged to be in jeopardy because of the uncertainty surrounding the refinancing of existing limited recourse debt include the proposed Canberra high speed train.
The need in Australia to adopt an internationally competitive tax regime is generally recognised. This is essential both to encourage inward investment and to deter the outflow of domestic capital overseas.
In the context of the infrastructure industry, this issue is important in the context of the ability of domestic companies to compete with foreign competitors when bidding for Australian assets. Under current arrangements, overseas bidders for Australian infrastructure frequently enjoy depreciation benefits in their home tax jurisdictions that they can leverage into higher bids than their Australian rivals.
Evidence of this is provided by the recent purchase of Victorias Westar and Kinetik Energy gas distribution/retail company by Texas Utilities. The US-based company outbidded its Australian rivals by offering a purchase price of $1.617 billion. Financiers are claiming that, purely on the basis of the more favourable tax regime faced by Texas Utilities, it had a $60 million advantage over Australian companies when it prepared its bid.
There is currently a high level of uncertainty surrounding the tax arrangements in relation to private provision of infrastructure. This adds to the sectors costs and can act as a significant deterrent to investment. Uncertainty stems from a number of sources, including:
In theory, the system of rulings should provide greater certainty to taxpayers in calculating their tax liability. Rulings are, in principle at least, designed to provide taxpayers with a definitive statement from the ATO on how the law applies to particular arrangements.
In practice, however, the system of rulings is a major aggravation for infrastructure investors. The often considerable length of time taken to issue many rulings is particularly frustrating and can jeopardise investments (eg, financial commitments may be withdrawn). In addition to concerns over excessive delays, other criticisms of the current system of rulings include their jurisdictional limitations, incomplete coverage, their poor quality and incontestability.
The culture and negative attitude of the ATO in relation to infrastructure projects is not conducive to investment. In particular, the ATO is perceived by the infrastructure industry of:
Whilst it is understandable that the ATO should take all reasonable steps to eliminate tax evasion and close tax loopholes, its methods of doing so are currently sending out negative signals to private sector investors in infrastructure.
There are two broad dimensions to the business tax reform agenda as it affects the infrastructure sector, namely those problems which are essentially technical and the major strategic issue of the competitiveness of private investment in infrastructure.
First, there are the nuts and bolts or technical issues that need to be resolved. Included among these are limited recourse debt, some of the problems surrounding IBTOS and (for all its strategic importance) s51 AD. These issues have been dealt with above and do not need further discussion here except to say, perhaps, that the proposal to address the s51 problems discussed in the Ralph Committees recent discussion paper appears reasonable. The other issues require further evaluation.
The major issue for the infrastructure industries is how the tax review can contribute to providing a more favourable environment for private investment in infrastructure.
Clearly, the 1990s have seen a considerable shift in favour of private provision of infrastructure to the benefit of public sector balance sheets. This has happened despite three major disadvantages faced by private investors namely:
The One Nation statement provided two initiatives that, to some extent at least, offset some of these disadvantages. First, the Infrastructure Bonds scheme was capable of reducing the cost of debt by up to 40 per cent and thereby provided worthwhile compensation for the tax losses problem. That program has now been abolished and replaced by one (IBTOS) which is much less attractive. It is difficult to access, capped at a relatively low cost to revenue and subject to allocation on a discretionary basis. Secondly, accelerated depreciation of plant and equipment was of value to many infrastructure projects and now faces abolition. Both these measures were introduced because the Government believed that without them investment in infrastructure would be inadequate. It is difficult to see why that view would now have changed.
If the IBTOS scheme is maintained it is likely to have a detrimental effect on private investment in infrastructure. Alternatives include:
In practical terms, a return to the IB program is unlikely even if the cost to revenue could be controlled and ability to rort the scheme readily eliminated. Governments rarely re-instate discredited programs, particularly those introduced by the other side. On the other hand, direct subsidies while transparent and relatively simple are rarely favoured these days. A new program, which addressed the shortcomings of IBs and whose cost to revenue was controllable may be the best approach.
One option is an infrastructure voucher system that would work as follows. The Government would designate a certain total value for vouchers to be issued each year (say, $250 million). Eligible infrastructure projects, as determined by an arms length statutory body, would be issued with vouchers which could then be used to obtain a tax refund for the project equal to their face value.
For example, suppose a project wants to borrow $100 million at a cost of 8 per cent per annum, so the annual interest cost is $8 million. If the project consortium could deduct this interest cost, it would yield a tax benefit of $2.88 million (0.36* $8 million) but this is not feasible because it has no income. The Government, however, issues a voucher to the project for $2.88 million, which it redeems with the Australian Tax Office, which would issue a refund for this amount.
This process would be repeated each year for a predetermined number of years i.e. until the time when the project (at its commencement) is forecast to be profitable. The interest payments by the project would, of course, be nondeductible in later years when the project is profitable.
The advantages of this scheme are that:
The issue of whether it is appropriate to trade-off accelerated depreciation for a lower statutory tax rate is difficult. Some industries will benefit, while some will be worse off. In the infrastructure sector, those industries using relatively little plant and equipment, such as road constructors, will be better off while others, such as electricity generators and gas pipeliners are likely to be worse off.
Provided the resulting system is neutral between investment in long or short term assets, it is difficult to object to the lower statutory rate option. This, however, is a substantial proviso. It means that the effective lives of assets with long lives (indeed, probably all assets) would need to be re-estimated by the tax office taking account of the risks involved, such as the risks of early obsolescence or the asset becoming stranded. Essentially, such lives would need to be based on economic rather than physical factors.
An alternative is to continue to allow the owners of assets to nominate the depreciation rate themselves, a process similar to self-assessment in the personal income tax system. While notionally this occurs currently, the criteria underpinning such self-assessment appear to be based on physical life rather than economic life. Since taxpayers would tend to nominate the shortest possible term for writing off the asset, this system would clearly need to be monitored by the ATO, but the Tax Office should be working on the basis of a realistic economic life. The present schedule of prescribed asset lives appears to be based heavily on physical criteria. One option would be for the ATO to review its prescribed asset lives based on consultations with industry on economic life and evaluations of international best practice.
If this system were adopted, the Government may claim that the savings would be less than would have occurred from the simple removal of accelerated depreciation and the tax rate then could not fall as far as to 30 per cent. While revenue collections may fall short in the early years, this is essentially one of timing. The full value of the assets will always be written off in its entirety over some time period, so in the long term the effect on revenue only reflects differences in discount rates due to timing. In particular, if writeoffs occur more quickly, there is pressure on early budgets. While the solution proposed here may not meet the Treasurys definition of revenue neutrality, it would make relatively little difference to revenue collections over the longer term (and may even increase them if investment was stimulated).