Infrastucture

INFRASTRUCTURE (AND OTHER DEFERRED-INCOME) PROJECTS

Income taxation of infrastructure assets

Loss flow-through for infrastructure companies

Many different types of investments have long lags between initial investment activity and eventual cash flows from production. Examples of such delayed-income projects occur with mining and petroleum activities (often with many years of exploration and development expenditure before the first receipts from production are realised), forestry (with many years between initial planting and ultimate felling and sale), horticulture (including investments in olive trees, apple orchards, macadamia nuts, etc again with long delays between planting and benefiting from the sale of associated produce) and real estate development (often with many years between land acquisition and ultimate sale).

Investment in 'infrastructure' - like aviation facilities, electricity generation, transmission and distribution facilities, roads and railways and other land transport facilities, port facilities, water storage and treatment facilities, sewage treatment facilities, hospitals, prisons and telecommunications networks - also lay claim to long lead times involved during construction and establishment phases before receipts start to flow. Infrastructure assets - for example, buildings, large depreciable assets (like electricity generators and power stations) and 'structural improvements' like land preparation - are also often characterised by long lives or low rates of depreciation.

Governments or tax-exempt government bodies often provide infrastructure directly for a variety of reasons (for example, practical difficulties in charging for use of some assets like certain roads) - though generous income tax treatment, well out of kilter with the taxing of economic income, can add encouragement for tax-exempt government bodies to sell and lease back infrastructure assets to the private sector. Being tax exempt, government bodies cannot benefit directly from income tax concessions like accelerated depreciation and investment allowances but can indirectly benefit through lower lease rental payments made to lessors of infrastructure assets who do benefit from tax concessions attracted by those assets.

The discussion here focusses on infrastructure provided by taxable private sector operators. The income tax issues are analysed, first, by looking at a stylised hypothetical infrastructure investment undertaken by an unincorporated business and, secondly, by extending this discussion to include such investment by incorporated entities - where the issue of loss flow-through to entity owners becomes relevant.

INCOME TAXATION OF INFRASTRUCTURE ASSETS

As noted, many of the assets associated with infrastructure ultimately decline in value. Nevertheless, as the various assets are being brought together in the construction and establishment phases of an infrastructure project, the value of the overall project is invariably increasing. The closer comes the time when people can get their hands on receipts from goods and services produced by the project, the more people are prepared to pay for the project. Risk of things going wrong in the lead up to production is one thing. But, more fundamentally, a stream of future positive cash flows is more valuable the less time there is to wait for the stream to start.

Thus, a typical stylised value profile of an infrastructure project looks just that of any deferred-income asset, such as an apple orchard, with value first increasing during construction and establishment and then decreasing after production of goods or services commences.

Taxing economic income (the 'benchmark' tax base) would require the year-by-year increase in value of the infrastructure project during the construction/establishment phases to be included in annual tax assessments. In addition, once production did eventually start, depreciation would apply to the full, increased value of the project at that time - not just to the various construction and establishment costs. Again, this parallels the taxing of economic income of other deferred-income assets year by year as income arises.

In practice, the early increase in project value is invariably not included in tax assessments. Not surprisingly, given the lack of tax on these early accrued gains, the later depreciation is then not based on the full value of the project at start of production but on actual capital expenditure undertaken. In addition, tax depreciation not starting until production commences helps ensure that depreciation is not allowed when asset value is likely going up. Typically, a net tax benefit might be expected from such tax arrangements with tax value profile remaining above value profile until brought together at project end. That net benefit would be increased the earlier and faster tax write-off is allowed - illustrated, at the extreme, by immediate write-off which is sometimes attracted by particular delayed-income activities, like forestry. The larger the tax concession aimed at a particular activity, the more likely there will be significant second round effects in terms of price and cost increases driving down the pre-tax return from those activities so that after-tax returns line up with other investments not so generously treated.

Stylised infrastructure taxation. Figure 50 shows the value/tax value profile of a stylised infrastructure project. The project is marginal before tax (NPV of zero with discounting at the going 10% interest rate). All capital expenditure ($1000) is made at the start of Year 1 and there is a 5-year delay before positive net receipts start flowing from the investment. The up-front capital expenditure is equity funded so there are interest payments on debt during the 5-year production lag. Value of the project increases by 10% a year in line with the going rate of interest. After the project starts producing net receipts, project value (and the stream of net receipts) falls at 5% a year. For income tax purposes, tax value remains at $1000 until production starts (that is, accrued gains during the production lag are not taxed) and 5% declining balance tax depreciation, initially applied to the $1000 of capital expenditure, starts when production starts.

Consistent with value/tax value profiles of other deferred-income projects, Figure 50 highlights differences in typical income tax treatment with 'benchmark' economic income taxation. The project's pre-tax return is 10% per annum. However, actual value remains above tax value over the 10-year life of the project because tax value remains at the $1000 of construction expenditure until production starts in Year 5. Consequently, after tax the stylised infrastructure project's return is 6.0% (before any consequent price effects) compared to the after-tax 5.3% return of other investments that have a 10% pre-tax return and are taxed year by year on their economic income.

Taxation of deferred-income projects in practice. It is tempting to suggest that the after-tax cash flows of this stylised infrastructure project would be transformed into those mirroring an actual infrastructure project by simply adding together a series of similarly stylised projects, with each project reflecting expenditure during the lead-in years that adds to the actual overall infrastructure asset. However, the tax treatment of the stylised infrastructure project would require all construction and establishment spending during the lead-in years of the actual infrastructure project to add to the tax value of the overall project - with that overall project tax value attracting write-off once production starts. In practice, capital expenditures of deferred-income projects attract a variety of write-off regimes depending on the type of deferred-income project and the type of capital expenditure involved. For example, with a mining or petroleum project the following variety of write-off regimes might apply to expenditure during the lead-in years:

exploration expenditure might be immediately deductible even though it is successful and therefore adds to asset value;
expenditure on items plant and equipment might be written off on the basis of their 'effective lives' once 'installed ready for use';
development expenditure (including associated community and housing costs) might be written off over the 'life of the mine', starting when production commences; and
other expenditures might be written off over fixed numbers of years.

Thus, in practice, particular write-off regimes may give rise to tax losses during the lead-in years of a deferred-income project when the value of the project is actually increasing year by year. Debt funding, with its negative cash flows and possibility of increasing annual value of outstanding debt liability, adds to the possibility of such tax losses.

Full loss offset for infrastructure projects in lead-in years. With the stylised infrastructure project, during the up-front lag in production there are no debt funding costs, as well as no costs that attract early write-off for tax purposes. Therefore, Figure 50 does not put a focus on the issue of full loss offset - whereby tax losses in the lead-in years would be written off immediately against other income (or, theoretically, would attract a payment from revenue authorities equal to the loss times the investor's tax rate). In practice, where tax losses do occur in the early lead-in years, the investor would be able to write them off against assessable income from other activities - or would have to carry them forward to deduct against later positive net receipts from the project itself. Calls for measures to provide immediate tax savings for early tax losses from stand-alone infrastructure projects (including such projects in corporate structures) invoke broader considerations:

Under economic income taxation (or the 'benchmark' tax treatment) only economic losses would attract full loss offset.
As Figure 50 highlights, in the lead-in years, project value is increasing and with it economic income is accruing year by year.
Thus, all infrastructure expenditure in each of those years would be adding value to the overall infrastructure asset with the expectation of providing future benefits on completion of the asset. This means economic income is accruing on all these expenditures during the lead-in years - even if, in practice, some of them attract tax write-off during those years.
Annual debt funding costs (plus any annual increases in value of financial liabilities) would subtract from the annual accrued gains - but would not be expected to be greater than accrued gains building year by year on the infrastructure expenditures. The size of the annual accrued gains would reflect the investor's expectation of at least earning the going interest rate on infrastructure spending applied to create the infrastructure asset.
Usually therefore there would be no economic losses in the early years at all - just net positive economic income which would be taxed under neutral economic income tax arrangements.

Against these considerations, ensuring the achievement of full loss offset in the lead-in years of infrastructure investments might be a more plausible option were annual accrued gains (and losses) included in the income tax base and all infrastructure spending in the lead-in years added to the infrastructure asset for later depreciation (when production commences). The common lack of taxation of the early accrued gains on all infrastructure costs incurred during the lead-in years of deferred-income projects raises the issue of quarantining early tax losses from these projects (including losses caused by debt funding costs) to be carried forward for deduction only against positive income from the associated project. Such a proposition for loss quarantining is not supported by the type of analysis undertaken in relation to negative gearing of rental properties. More generally, simplistic attempts at offsetting one distortion (not taxing accrued gains of all infrastructure spending in the lead-in years) with the imposition of another (denying immediate tax savings for losses on interest costs faced by those infrastructure investors that have income from other sources) are likely to impose many adverse side effects while failing to address the original distortion. The best strategy is to address directly the lack of tax on accrued gains of all infrastructure capital expenditure.

Second-round infrastructure price effects. Even if the rate of declining balance tax depreciation allowed a particular type of deferred-income infrastructure project matches the actual rate of its eventual decline in value, not taxing the up-front accrued gains provides the project an advantage over other investments facing tax on their year-by-year economic income. Particularly if such advantage singled this type of infrastructure projects out from other investment alternatives, second-round price and cost effects could be expected as resources flowed into this infrastructure activity. Figure 51 illustrates the result of modelling the impact of such second round effects on the value/tax value profiles of the stylised infrastructure project in Figure 50.

In Figure 51, the flow of resources into the tax-advantaged infrastructure project has increased construction costs from $1000 to $1139. The pre-tax return is reduced from 10% pa to 8.9% pa as the post-tax return moves to the 5.3% pa equilibrium level of fully-taxed investments elsewhere. To achieve those returns while maintaining a 5% per annum decline in net receipts, the project has to generate slightly higher net receipts each year - which may seem improbable with more resources moving into this type of activity. Modelling the situation where the stylised infrastructure project's construction costs are not affected but net receipts are (through reduced product prices and/or recurrent cost increases), net receipts in the first year of production fall from $220 to $197. The more realistic outcome in terms of effect on construction costs and net receipts may be somewhere in between these two extreme positions.

INFRASTRUCTURE INVESTMENT BY ENTITIES

For infrastructure investments undertaken within private sector entities, the above analysis of the income tax treatment of a separate infrastructure asset needs to be set within the analysis of effects on investment decision-making of different ways of taxing the income of entities, like companies and trusts.

For example, take the situation of a company facing a 30% tax rate which is capitalised to invest in the above stylised infrastructure project (before any price effects) where shareholders on 47% tax rate sell their shares each year and any change in value of their shares over the year is included in full in their income tax assessments. In these circumstances, with share value reflecting the year-by-year increase in value of the infrastructure project in its lead-in years, the shareholders selling out in those years would be taxed on annual accrued capital gains of the project.
Extending this illustration with the company subject to Australia's full imputation system of company tax, if shareholders continued to sell out each year and the company distributes all available cash each year, the annual economic income of the underlying infrastructure project would be assessed to shareholders in the year it occurs - so every year the annual 10% pre-tax return of the stylised infrastructure project results in a 5.3% return to the infrastructure shareholders on a 47% tax rate (the neutral outcome of the benchmark tax treatment).
Alternative situations of different tax treatment and differing approaches to the company's distribution or retention of cash, as well as shareholders' selling out or staying with the company, can be assessed against the analysis in the discussion of taxation of entities. For example, with the infrastructure asset held by a company subject to full imputation arrangements (and with 100% of capital gains or losses from sale of shares and liquidation of the company in Year 10 included in tax assessments of shareholders):
- if the company's shares are sold each year but the company distributes no cash until liquidation, beyond the 5.3% pa return to shareholders in the lead-in years (when the annual accrued capital gains are taxed), the annual return to shareholders steadily falls below 5.3% because of double taxation (in the company and on the sale of shares) of retained income - with that double tax not recouped until liquidation;
- if no shares are sold but annual distributions of available cash are made, the post-tax return to the original shareholders is 5.8% pa, above the 'benchmark' 5.3% because the early accrued gains are not taxed when they arise; and
- if no shares are sold and no distributions made until liquidation, the post-tax return to the original shareholders is 6.0% pa, above the 'benchmark' 5.3% because the early accrued gains are not taxed when they arise and, prior to liquidation, annual taxable income is taxed at the company's 30% tax rate rather than shareholders' 47%.
Sieper (1995) analyses the taxation of deferred-income projects held by companies in differing circumstances such as debt versus equity financing and interest on debt being immediately deductible, deferred or even lost completely. Optimal dividend policy in differing circumstances is also discussed and the analysis is given the rigor of mathematical formulation.

LOSS FLOW-THROUGH FOR INFRASTRUCTURE COMPANIES

Calls are often made for company tax arrangements to be made more flexible to enable tax losses in the negative cash flow lead-in years of companies' infrastructure investments to 'flow through' to shareholders. Analysis of the merit of such calls brings together the issue of loss flow-through with the issues of general availability of full loss offset and the general taxation of accrued capital gains.

Consistent with the discussion of taxation of individual infrastructure projects, very often there would be no economic losses in the early years at all - just net positive economic income which would be taxed under neutral economic income tax arrangements. Again, achievement of full loss offset in the lead-in years of infrastructure investments - in the case of corporate investors, requiring flow-through of tax losses to shareholders - might be a more plausible option were annual accrued gains (and losses) on all construction and establishment expenditure on the infrastructure asset included in the income tax base.

Given the general lack of taxing such accrued gains, it is not surprising that none of the different ways discussed of taxing entities includes flow-through of losses. The design would be different, however, for ideal corporate income taxation which has neutral impact on investment decisions. Assume, in such ideal design, companies' annual income attracts tax at the corporate level (rather than the alternative of integrating company income directly to shareholders' tax assessments each year). In the design, annual accrued gains and losses would be included in companies' tax assessments and companies would attract 'full loss offset' - whereby a company that did not have sufficient other income against which to offset economic losses in a year would receive a tax refund from the tax authorities equal to the loss times the company tax rate. A theoretically equivalent outcome would be to have losses carried forward at, say, the long term bond rate (reduced in proportion to the tax rate) within an income tax design that ensured no risk of losing the compounded forward losses - see the section on risk in Mayo (1984).

Within this ideal corporate income tax system, companies' annual positive income would also be integrated with the tax assessments of individual shareholders, with CGT cost base adjustments, regardless of whether or not the income was actually distributed (or whether, as under the assumption here, the company was first assessed on that income). To illustrate, take a company which holds our stylised infrastructure project and is taxed at 30% on the project's annual economic income in the year it arises. With full integration of company and shareholder assessment (incorporating imputation arrangements to accommodate tax at the company level), 47% tax rate shareholders of the infrastructure company realise the 'neutral' 5.3% pa post-tax return even if no annual distributions are made by the company - in circumstances of either:

Similarly, within this ideal system, where an infrastructure company did incur annual economic losses, those losses would be sheeted home to the ultimate individual shareholders to attract 'full loss offset' there - again with associated adjustments to capital gains tax (CGT) cost bases of their shares.

If a company with an infrastructure loss had already achieved full loss offset by being paid a refund equal to the loss times the company tax rate and if the integration system were accompanied by imputation arrangements, individual shareholders' assessments would include the 'grossed-up' loss but associated tax savings of shareholders would be reduced by an amount equal to the tax refund for the loss. To illustrate, a $100 loss integrated with the tax assessment of an individual shareholder on a 47% tax rate would result in $47 of tax savings for the shareholder from full loss offset at shareholder level. If the company on a 30% tax rate had already received $30 tax refund for the loss, the shareholder would have the $30 added to their tax payable - resulting in a net additional $17 tax saving from the loss. This mirrors the situation where the company makes a $100 loss on one project but is able to offset that against income from another. The loss means the company pays $30 less tax so that with distribution of annual income the individual shareholder receives $100 less income but also $30 less imputation credits and so pays only $17 less tax than without the loss.

In sum, it is in the ideal setting of company income taxation arrangements which have neutral impact on investment decisions that taxation of accrued gains, full (economic) loss offset and loss flow-through come together.