Taxation of entities

TAXATION OF ENTITIES

Taxation of Entities with Neutral Effect on Investment Decisions

Outcomes of Practical Entity Taxation Arrangements Compared to Neutral Ideal

Some Implications for Design of Entity Taxation of 'Permanent' Tax Preferences

TAXATION OF ENTITIES WITH NEUTRAL EFFECT ON INVESTMENT DECISIONS

An entity may hold identical assets and liabilities that produce income to those held by an individual or unincorporated business. Both individual investors and entities also use investment capital (or contributed capital) plus debt funding to acquire assets.

Unlike an unincorporated business, however, an entity is owned by other entities or individuals. Ultimately, at the end of a chain of entities are individual owners (or beneficiaries of the income and/or contributed capital). Examples of entities are companies and trusts.

An entity is a separate body for tax purposes from its ultimate individual owners. Ownership interests in entities (like shares in companies or units in unit trusts) can usually be bought and sold. Thus, these interests have their own year-by-year values for tax purposes separate from the tax values of the assets and liabilities held by the entities themselves.

It is the assets and liabilities of an entity that produce the entity's income, not the entity itself. An entity may be viewed as an 'administrative artefact' over its assets and liabilities that yield income year by year.

The income tax treatment of assets and liabilities held directly by individuals with neutral effect on investment decisions is:

to include all 'benchmark' income (which may be a loss) arising each year from an individual's investment assets and liabilities in the individual's tax assessment for that year,
with 'benchmark' (economic) income comprising the annual net receipts from these assets/liabilities plus their annual change in value.

Thus, economic income from an individual's investments is assessed at the individual's marginal tax rate in the year that it arises. Economic profits in a year attract tax, economic losses in a year attract tax savings (or full loss offset).

Figure 2 illustrates the neutrality property of the benchmark tax base. Pre-tax returns of 10% across alternative investments are reduced to 5.3% for individual taxpayers on a 47% tax rate and to 7% for individual taxpayers on a 30% tax rate.

This neat neutrality property does not change just because an 'administrative artefact' in the form of an entity is placed over particular assets and liabilities. Investment neutrality should be maintained if individual owners of the entity who ultimate feel the effect of the entity's investment decisions have their share of the entity's annual economic income included in their annual tax assessments.

Thus the same benchmark tax base for investment neutrality applies to individuals owning assets and liabilities indirectly via entities as applies to individuals investing directly.

OUTCOMES OF PRACTICAL ENTITY TAXATION ARRANGEMENTS COMPARED TO NEUTRAL IDEAL

How does the benchmark tax outcome compare with the tax outcomes of practical entity taxation arrangements designed in the realistic context of an income tax base containing concessional treatment?

In making this comparison, practical considerations include the following.

Invariably the entity and its individual owners are assessed separately on the income of the entity and at different rates of tax. Comparisons of entity treatment need to take into account the design of the taxation of the entity and of its owners.
The rates of after-tax returns to people investing via an entity depend on:
1. the assets and liabilities held by the entity and their tax treatment in the entity (often the same as for the same types of assets and liabilities held directly by individuals), including the shape of tax concessions or preferences;
2. the extent of retention by the entity of annual income versus annual distributions of income and contributed capital;
3. the income tax treatment of individual owners/beneficiaries of income of the entity, including annual distributions of income;
4. the tax treatment of the buying and selling of interests in the entity (while taxing the buying and selling of entity interests ideally should not change the aggregate tax take between entity start up and liquidation - because the only income produced is from the entity's underlying assets and liabilities - the tax treatment of buying and selling can have major effects on year-by-year returns to people selling their interests);
5. the tax treatment of the liquidation of the entity; and
6. the tax rates of the entity and of individual entity owners/beneficiaries.

Annual income for tax assessment purposes will often differ markedly from the benchmark tax base - for both entities and the direct investor. Therefore, in practice, when assessing entity tax design against a neutral investment ideal there are two forms of neutrality to be considered:

investor neutrality - where the tax outcome is the same for individuals investing directly or indirectly via an entity even though that tax outcome may differ from the neutral investment ideal
investment neutrality - where the tax outcome of investing via an entity is consistent with the neutral investment ideal. Where this neutral outcome is also achieved by the direct investor both investor and investment neutrality are brought together.

Some entity tax systems are designed to tax the income from entities' assets and liabilities once over time. Other entity tax systems - such as the classical taxation of companies in the United States - are designed to tax this income twice, once in the entity and again when the income is distributed to individual shareholders. Such classical treatment with its double taxation of income cannot meet the requirements of either investor or investment neutrality, simply because individual direct investors are only taxed once on the income included in their annual tax assessments from their direct investments and the benchmark treatment also imposes just one layer of tax. Similarly, proposals for a zero company tax rate miss the investment neutrality point - by ignoring the requirement to have economic income taxed at the investor's marginal rate in the year it occurs, not at some later time when the company decides to distribute after retaining that income tax free for years.

Table 1 summarises individual taxpayer treatment and four entity taxation systems that are designed to impose a single layer of tax. For each entity tax system the table summarises its basic design features, its key features relating in particular to tax-preferred income (annual economic income not included in a taxpayer's tax assessment for the year) and plausible practical requirements for achievement of investment and/or investor neutrality under that system. The international dimension is ignored so that the possibility of foreign income and foreign owners/beneficiaries is not part of the analysis. The table assumes in all cases that capital gains taxation (CGT) applies to 100% of the capital gain realised on the sale of interests in entities.

Included also in the table are investor after-tax rates of return for the direct investor, as well as people obtaining returns indirectly via the four entity systems. The same two assets of a hypothetical project underlie these varying returns. Assets identical to these two are held both directly by the individual investor and by the entity under each of the four entity systems:

an appreciating asset (land), acquired in Year 0 for $1000, that increases in value by 10% a year; and
a depreciating asset (building, plant, horticultural plants, etc), again acquired in Year 0 for $1000, that produces net receipts in Year 1. Its value declines at 15% a year (reflecting declining annual net receipts).

In addition, when an entity retains cash from the project's net receipts rather than distributing it, the cash is deposited in a regular bank account which provides compounding interest income at the rate of 10% per annum.

After five years the direct investor sells the hypothetical project's assets and the entities running an identical project dispose of its assets and liquidate. For both individual investors and entities, 100% of the capital gain realised on the sale of the land is included in assessable income and the depreciating asset attracts 30% reducing balance depreciation for tax purposes (accelerated depreciation of double the actual decline in value) with excess depreciation deductions clawed back at the time of sale. Thus, the tax preferences underlying the after-tax investor returns in the table:

arise because annual accrued capital gains of the appreciating asset are not taxed and because annual depreciation deductions for accrued capital losses of the depreciating asset are provided at twice the actual rate of decline in value of the asset;
provide tax savings in the form of timing benefits - the delayed payment of tax - because depreciation balancing adjustments and 100% CGT on realisation have the effect of ensuring that over the five years of the project the actual change in value of each asset is included in tax assessments (even though the change in value each year is not included in that year's assessment as the benchmark base would require); and
may therefore be terms 'temporary' tax preferences, as opposed to 'permanent' tax preferences such as exemptions for certain forms of income like only taxing part of the capital gain realised or not clawing back excess depreciation deductions on the sale of assets.

Because the tax preferences are temporary, on sale of assets and liquidation of entities in Year 5, the same overall amount of tax is paid in relation to the depreciating and appreciating assets by the individual investor and by the entity plus owners/beneficiaries under each of the four entity systems in Table 1. When an entity retains income the bank interest earned on those retentions adds to aggregate tax paid (what individuals do with cash from directly investing or from entity distributions is not included).

Full loss offset is assumed throughout whereby individuals or entities have sufficient other assessable income to absorb any annual tax losses from the project.

For the different entity tax systems, investor (or entity interest holder or owner/beneficiary) after-tax returns are shown in Table 1 assuming both (1) full annual distribution each year of cash from net receipts and (2) no distribution each year until liquidation. In each of the full distribution and retention situations, analyses are done both with sales of entity interests occurring each year and with no sales occurring between capitalisation and liquidation. With no sale of interests, a single return is shown over the five years of the project. With sales of entity interests each year, after-tax returns are summarised in the table for the interest holders selling out each year, having acquired the entity interests in the prior year. A range of returns is shown in the table if these returns vary from year to year.

The returns shown in the table are static in the sense that there is no attempt to take into account the capitalisation effect on the price of entity interests of future taxation events - such as the ultimate removal (via distribution and sale, liquidation or share buy-backs) of double tax that can be imposed on retained income under a full imputation system. Nevertheless, the investor returns in the examples give a good feel for the pressures imposed by different tax treatment on the pre-tax situation. For example, when annual sale of entity interests is simulated, a wide variation in year-by-year returns might suggest poor tax design.

In the absence of tax, the project's two assets (or three assets with retained income) provide a per annum return of 10%. Where the entity tax system imposes tax at the entity level (full imputation and full integration), the entity tax rate is 30%. Thus, were the benchmark tax base to apply at the entity level, after tax the annual return to the entity would be 7.0%.

More importantly, if the design of entity taxation was such that the effect of the benchmark tax base were to apply through to the interest holder level (whether or not the entity was subject to tax), after tax the annual return to interest holders would be 5.3%. Comparison of the after-tax returns in Table 1 with the benchmark 5.3% return should provide an indication of how neutral is the tax base for individual direct investors and how close entity tax design is to achieving investment neutrality in the particular circumstances depicted of income retention/distribution and sale/no sale of entity interests. An indication of how close entity tax design is to achieving investor neutrality is obtained by comparing returns to interest holders with the return to the individual direct investor.

**TABLE 1: ENTITY INCOME TAX SYSTEMS IMPOSING A SINGLE LAYER OF TAX**
System	Design	Key features	Investor returns	Requirements for neutrality
Individual direct investor	Annual taxable income direct to individual.	Attracts full benefit of tax preferences.	5.8%	Achieves investment neutrality if taxable income matches benchmark treatment
Full imputation	All distributed 'income' taxed in hands of individual holders of entity interests (a). Credit for tax paid by the entity on distributed income provided to individual holders of entity interests. Losses held at entity level. No distribution requirements. Tax value (cost base) reduction on entity interests for returns of contributed capital (c).	Timing benefit from low entity tax rate and tax-preferred income if income retained by entity and no sale of entity interests. Sale of interests can produce benchmark outcome as can distributions each year (distributions may effectively incorporate unrealised gains (b) to be taxed as unfranked dividends). NB. While CGT on entity interests does not increase the tax take from entity formation to liquidation, it is an important design features of imputation arrangements, the absence of which can have considerable effect on the pricing of entity interests. To illustrate, with no CGT and no distributions but sales of entity interests each year annual after-tax returns vary from 10.1% to -9.7% compared to close to 5.3% each year with CGT applying. Capitalisation of these effects of no CGT into the price of the entity's interests would be expected (d) but such price effects point to poor tax design.	Distribution: 5.3% (no sale) Distribution of economic income taxed each year at 47% because accrued gains on land included in tax authority's measure of income for entity distribution purposes - with these gains excluded from the measure, the return increases to 5.6% (a). 5.3% each year (sale) Net receipts (distributed) plus change in asset value (CGT) taxed at 47% each year. Retention: 5.75% (no sale) No tax on tax-preferred income until assets sold (as with direct investor) and entity liquidated - but, with no distributions, different cash flows to direct investor. 5.3% to 5.2% per year (sale) Annual income retained and taxed via CGT at 47%, but some double tax on retentions.	Achieves investment neutrality: over the five years with distributions but no sales; and each year with annual sale of entity interests and either full distribution of cash from net receipts each year - the situation depicted in Example 10 in the set of Kyscope Examples - or with no distributions (though some double tax on retention occurs). May broadly achieve investor neutrality with no distributions and no sale of interests each year - lower entity tax rate may be offset by delay in receiving cash. Achievement of investor neutrality and investment neutrality in all circumstances of distribution/retention and sale/no sale of interests by equating the tax base of entities and individuals to the benchmark base is best considered in the context of full integration arrangements (below).
Fixed (unit) trust treatment	Distributed annual taxable income taxed in hands of fixed interest holders (beneficiaries). Losses held at entity level. Taxable income not distributed to interest holders taxed at entity level at individuals' top marginal rate. Tax value (cost base) reduction on entity interests for distributions that are not taxable income (ie untaxed income and returns of contributed capital) (e). Income retains its character on distribution.	Timing benefit from tax-preferred income if interests not sold each year. Distributions in a year beyond taxable income of cash from net receipts - which reduce tax value of entity interests and so are taxed on sale of interests or liquidation - may effectively include unrealised gains if say accelerated depreciation deductions reduce taxable income in that year.	Distribution: 5.8% (no sale) Same return as direct individual investor, though direct investor attracts full loss offset and losses with fixed trust treatment carried forward within entity. 5.3% each year (sale) Economic income taxed each year via CGT and taxable income component of distributions.	Tax value reductions of entity interests for distributed income not in taxable income mean such income is ultimately taxed when interests are sold. Therefore, with full distributions investor neutrality may be achieved if entity interests are not sold each year (though tax losses are carried forward at the entity level and the direct individual investor may achieve full loss offset by applying tax losses to other income) and investment neutrality may be achieved if entity interests are sold each year. Achievement of investor neutrality and investment neutrality in all circumstances of distribution/retention and sale/no sale of interests by equating the tax base of entities and individuals to the benchmark base is best considered in the context of full integration arrangements (below).
Discretionary trust treatment	Tax applied to individual beneficiaries entitled to a share of entity income. Losses held at entity level. Income not distributed to beneficiaries taxed at entity level at individuals' top marginal rate. No sale of, and no tax value adjustments to, beneficial interests in entity - so 'flow-through' to beneficiaries of untaxed income including distributions from unrealised capital gains. Income retains its character on distribution.	Attracts full benefit of tax preferences. Distributions out of unrealised gains not taxed.	Distribution: 5.8% (no sale) Same return as direct individual investor, though different cash flows because direct investor attracts full loss offset and with discretionary trust treatment losses carried forward within the entity.	Achieves investor neutrality. Potential to achieve investment neutrality if taxable income matches benchmark treatment but issues arise when tax losses occur and when economic income is greater than cash distributions - the design of full integration arrangements addresses the latter issue.
Full integration	Annual taxable income of entity (or economic income) taxed in hands of individual holders of entity interests regardless of whether or not the income is actually distributed. Tax may still be paid at entity level with an imputation system applying (or, with entity income being attributed to interest holders, the entity could pay no tax). No distribution requirements. Annual tax value (cost base) change to entity interests for taxable income (or economic income) less entity tax paid less actual distributions (ie retained income) - to avoid double taxation year by year (g).	If economic income is taxed to individual interests (regardless of what entity's taxable income might be): investment neutrality (ie 10% pre-tax to 5.3% post-tax outcome) achieved under all circumstances of: (1) distribution/no sale of interests, either with imputation or no entity tax; (2) distribution/sale, either with imputation or no entity tax; (3) retention/no sale, either with imputation or no entity tax; and (4) retention/sale, either with imputation or no entity tax. If taxable income is taxed to individual interests and taxable income is smaller base than economic income (eg. accrued capital gains excluded): attracts full benefit of temporary tax preferences if interests not sold; and distributions out of unrealised gains not taxed in the absence of sale of interests. Integration of either economic or taxable income means domestic interest holders attract tax on entities' annual income at their tax rates. But no tax at the entity level would see no domestic tax paid by non-resident interest holders on their share of entity income (absent dividend withholding tax). Overlaying integration on full imputation arrangements would address that tax revenue concern.	With taxable income basis of integration and with imputation arrangements applying - ie what might be termed 'practical integration' - returns are as follows. Distribution: 5.8% (no sale) With full loss offset applying, cash flows for holders of entity interests are identical to those of individual direct investors. (f) 5.3% each year (sale) Economic income taxed each year through taxable income component integrated with personal incomes and rest via CGT after tax value of entity interests reduced by distribution and entity interests sold (rather than via unfranked dividends under full imputation). Retention: 5.7% (no sale) Similar return to direct investor but with different flows. 5.3% each year (sale) Economic income taxed each year through taxable income component integrated with personal incomes and rest via CGT.	With integration based on economic income: Achieves investor neutrality under all circumstances of retentions/distributions and sale/no sale of entity interests. As taxable income for individuals approaches the benchmark base (economic income), achievement of both investor neutrality and investment neutrality approached. With integration based on taxable income ('practical integration'): Achieves investor neutrality with no sale of entity interests - because the equivalent of unfranked dividends under full imputation is not taxed on distribution but reduces tax value of entity interests like fixed trust treatment. Achieves investment neutrality with annual sale of entity interests - because annual taxable income taxed to individual interest holders and prior reduction in tax values of entity interests for distributions of untaxed income - or if taxable income matches benchmark treatment. As taxable income for individuals and entities approaches the benchmark base (economic income), achievement of both investor neutrality and investment neutrality approached.

Notes to table:

'Income' as defined by the taxing authorities determines the components of dividends in a distribution - either 'franked' dividends (on which entity tax has been paid) or unfranked dividends ('income' that has not been taxed at the entity level) - as opposed to return of capital. Unless explained otherwise, this 'income' matches economic income and includes accrued capital gains. Outcomes may be affected by how tax authorities determine 'income'. For example, where 100% realisations CGT applies, if accrued gains of appreciating assets like land are not included in tax authorities' measure of income, circumstances of annual distribution but no sale of entity interests will increase the overall return because a lower proportion of distributions will be classed as unfranked dividends in years prior to liquidation.
Distributions out of accrued capital gains may be taxed as unfranked dividends in a year if (1) tax authorities include unrealised gains in 'income' or profit for the purposes of determining the proportions in an entity's annual distribution of contributed capital and unfranked dividends and (2) cash from net receipts in that year are distributed and income tax deductions, like accelerated depreciation deductions, available to the entity mean that some of the distributed net receipts are classed as unfranked dividends.
This tax value adjustment means that if say $1000 capital is contributed (creating a tax value of $1000 for entity interests), $100 profit earned ($50 taxable in the entity and $50 not) and $200 distributed, all the $100 of profit would be taxed in the hands of individual interest holders (with credit for entity tax paid) and the $1000 tax value would be reduced to $900. If the interests were then sold, for $900, there would be no capital gain or loss for tax purposes. A single layer of tax (at individual interest holders' rates at the time of distribution) would have been applied to the $100 of profit. Without the tax value reduction, a $100 capital loss for tax purpose would have been realised. As no tax value adjustment occurs for retained income, in the absence of distribution double taxation would occur if entity interests were sold for $1100 (with offsetting capital losses with buy-backs or on liquidation).
What is capitalised into the price of entity interests is the knowledge that capital gains unrealised in the early years will ultimately be realised and distributed and then taxed as unfranked dividends if not effectively distributed before - see footnote (b).
Tax value reductions for distributions of untaxed profits, in addition to the reductions for returns of capital that are part of full imputation arrangements, mean that ultimately all distributions of profit may be subject to a single layer of tax at interest holders' rates - profits in taxable income when the profits are distributed and untaxed profits (commonly called 'deferred income') when the entity interests are sold or on liquidation. To illustrate, if say $1000 capital is contributed (creating a tax value of $1000 for entity interests), $100 profit earned ($50 in taxable income and $50 not) and $200 distributed, all the $50 of profit would be immediately taxed in the hands of individual interest holders and the $1000 tax value would be reduced to $850. If the interests were then sold, for $900, there would be a capital gain of $50. If all of that realised gain were taxed, again as with imputation, a single layer of tax (at individual interest holders' rates at the time of distribution or at the time of sale/liquidation) would have been applied to the $100 of profit.
This outcome is unchanged by tax authorities excluding accrued capital gains from their income measure for entity distribution purposes. With integration of taxable income, the taxable income that is taxed to individual holders of entity interests already excludes accrued capital gains.
As entity taxable income (economic income or something less) is integrated with interest holders' taxable income, distributions of profit not out of taxable income are not immediately taxed in the hands of interest holders. The fact that any distributions (capital or profits) not out of taxable income result in tax value reductions of entity interests mean that ultimately entity profits should be subject to a single layer of taxation - paralleling the fixed trust treatment. In addition, a tax value increase for the profit component included in taxable income less associated tax paid by the entity means that (unlike full imputation) double tax is avoided in circumstances where profit is retained and entity interests sold (this increase being reversed when those retentions are distributed because the total amount of distributions reduce tax values). With integration incorporating imputation arrangements, retained profits are taxed first in the entity and then to individual interest holders. Even though the interest holders get credit for entity tax paid, with interest holders themselves being taxed on retained income, sound design argues for avoiding double tax when those same interest holders sell their interests while this income remains in the entity.

The different designs of the entity taxation systems in Table 1 perform differently against the criteria of entity investor and investment neutrality. The temporary tax preferences imposed, including 100% realisations CGT applying to the sale of entity interests, mean that both direct and indirect investors are ultimately taxed on all income of the assets of the hypothetical project. Income out of the tax preferences is included in taxable income (via accelerated depreciation 'claw-back' and 100% realisations CGT) when the assets are sold in Year 5. The direct investor is taxed on this tax-preferred income in Year 5. It is the timing of the taxing of this tax-preferred income at the tax rates of entity owners/beneficiaries that determines how each entity tax system performs against the criteria of entity investor neutrality and entity investment neutrality.

Full imputation achieves investment neutrality under all circumstances except no distributions and no sales of entity interests where investor neutrality may be achieved (depending on the entity tax rate and length of delay before asset sales and full distribution occur).
Fixed trust treatment with distribution each year achieves investment neutrality with sales of entity interests and investor neutrality with no sales of entity interests (although cash flow differences arise because losses are carried forward in the entity).
Discretionary trust treatment with distribution each year achieves investor neutrality although cash flow differences arise because losses are carried forward in the entity.
Full integration of economic income achieves investment neutrality under all circumstances of distribution/retention and sale/no sale of entity interests regardless of the tax rate of the entity.
Integration of taxable income achieves investment neutrality with sale of entity interests with or without distributions, as with full imputation, and achieves investor neutrality with no sales of entity interests, similarly to fixed trust treatment but regardless of whether distributions occur or not (if distributions do occur cash flows are the same as for the direct investor).

Overall, by design, full integration of economic income produces entity investment neutrality under all circumstances but can never achieve investor neutrality so long as tax preferences (like excluding accrued gains) are general features of the income tax base. Where taxable income provides temporary tax preferences only, full imputation leans more towards entity investment neutrality and discretionary trust treatment entity investor neutrality. Fixed trust treatment and integration of taxable income both straddle the worlds of entity investor neutrality and entity investment neutrality - though integration of taxable income does this flexibly with retention, as well as distribution, of entity income.

Moreover, integration of taxable income with accompanying tax value adjustments to entity interests would readily accommodate any future movement of taxable income more towards economic income (moves which are becoming more common particularly with financial assets/liabilities). In contrast, wider inclusion of accrued capital gains in taxable income: would ideally require adjustments to the tax value of entity interests under full imputation to remove double taxation on the sale of those interests when taxed income (including accrued gains) is retained; and would require changes to the fixed trust treatment to accommodate the increased possibility of taxable income (including accrued gains) exceeding cash available for distribution.

With integration of taxable income, however, the key practical question is whether the required allocation of retained income to interest holders and adjustments to tax values of entity interests can be made taking into account complex ownership structures and widely held entities whose interests are regularly traded. International attribution rules included in various countries' income tax laws already apply full integration-type arrangements when they attribute retained income of specified offshore entities to local owners to be taxed in the owners' hands. If such arrangements can apply successfully to offshore entities, why not to local entities?

Practical implementation issues associated with integrating taxable income are discussed after analysis of the implications for entity income tax design of 'permanent' tax preferences.

SOME IMPLICATIONS FOR DESIGN OF ENTITY TAXATION OF 'PERMANENT' TAX PREFERENCES

The tax treatment of assets underlying the after-tax rates of return in Table 1 incorporates 'temporary' tax preferences only (100% CGT on realisation and accelerated depreciation). In the hypothetical project underlying the results in Table 1, ultimately the actual change in values of both the appreciating and depreciating assets between acquisition and sale are included in investors' tax assessments.

What are some implications of turning the temporary CGT preference into a 'permanent' tax preference by having, say, only half the capital gains (and losses) realised by individual investors on CGT assets included in their tax assessments? The results of making this change to the analyses in Table 1 are shown in Table 2. This change sees (unless otherwise indicated in the table):

the individual direct investor attracting half CGT treatment on the appreciating asset;
sales of entity interests by the individual interest holders (entity owners/beneficiaries) attracting half CGT treatment; and
where an entity is subject to tax, 100% of the capital gain realised on the sale of the appreciating asset still included in the entity's tax assessment.

The analysis in Table 2 incorporating permanent CGT preferences highlights in particular:

the possibility of perverse effects under full imputation; and
the need for special tax value adjustments on entity interests under fixed trust treatment (and integration of taxable income) if the permanent nature of the 1/2 CGT treatment is to be preserved without making the temporary nature of accelerated depreciation permanent.

**TABLE 2: EFFECTS ON INVESTOR RETURNS OF 1/2 CGT ARRANGEMENTS**
System	Changed tax treatment	Investor returns	Comment
Individual direct investor	No change	7.2%	Reflects timing benefit of accelerated depreciation plus permanent 1/2 CGT tax preference.
Full imputation	No change (a). Again, unless otherwise indicated, income as defined by the taxing authorities for determining unfranked dividend/return of capital components in a distribution matches economic income and therefore includes accrued capital gains.	Distribution: 3.8% (no sale) More than a single layer of 47% tax is paid on income as only half the capital loss on liquidation is allowed for CGT purposes. If the entity's annual income for tax purposes only includes the capital gain on land in the final year the overall return increases to 4.9% (as less unfranked dividends and more return of capital occurs in years prior to liquidation, resulting in a lower capital loss on liquidation). 5.8% to -3.5% per year (sale) Considerable annual variations in annual returns occur because of 1/2 CGT. Retention: 5.75% (no sale) Outcome with 1/2 CGT is the same as with 100% realisations CGT as no CGT gains or losses arise. 7.7% to -2.3% per year (sale) 1/2 CGT provides tax savings on sales of interests before liquidation, offset by only half capital loss on liquidation allowed for CGT purposes. (e)	Perverse outcomes can occur with distributions occurring in the presence of permanent tax preferences that also impact on the sale of entity interests: more than a single layer of tax may be paid on the income from an entity's assets because only 1/2 any capital loss on liquidation is allowed for CGT purposes. (b) (c) The excessive tax stems from the combination of 1/2 CGT applying only at the interest holder level and tax preferences being excluded from the annual profit calculation that determines the amount in a distribution representing unfranked dividends or a return of capital. (d) Large variations in year-by-year returns with sales suggests poor tax design. If the permanent tax preferences did not impact on the sale of entity interests (eg some specified annual receipts exempt from tax), the outcomes would be similar to those in Table 1 - with distributions of the exempt income taxed as unfranked dividends.
Fixed (unit) trust treatment	With the Table 1 adjustment to tax values of entity interests (ie reduction for all distributions that are not taxable income), some of the 'permanent' 1/2 CGT preference is clawed back. For example, with annual distributions and no sales of interests, a return of 6.5% is realised (with returns varying from 7.65% pa to 0.1% pa if entity interests are sold each year). To maintain permanent nature of 1/2 CGT concession adjustments to the tax value of entity interests for untaxed distributions would exclude those distributions out of income subject to the 1/2 CGT preference.	With the exclusion from entity tax value adjustments of untaxed distributions out of 1/2 CGT preference: Distribution: 7.2% (no sale) The same return as the direct individual investor (though cash flows vary because of different loss treatment). 7.65 to 3.7% per year (sale) Annual distributions out of tax preferences (accelerated depreciation and accrued gains) are subject to 1/2 CGT in years prior to liquidation, with the resulting CGT revenue offset by a CGT loss in liquidation year. (f)	Accelerated depreciation restricted to timing benefit but 1/2 CGT remains permanent. Therefore, with full distributions investor neutrality may be achieved (though tax losses are carried forward at the entity level and the direct individual investor achieves full loss offset by applying tax losses to other income). Nevertheless, the sale of entity interests creates large variations in annual returns.
Discretionary trust treatment	No change.	Distribution: 7.2% Same return as direct individual investor, though different cash flows because direct investor attracts full loss offset and tax losses with discretionary trust carried forward.	Achieves investor neutrality. Timing benefit of accelerated depreciation and permanent 1/2 CGT effect realised (no sales to affect these).
Full integration	Integration of (entity) taxable income is applied together with full imputation arrangements. If 100% realisations CGT were applied to the entity and 50% to holders of entity interests (the CGT structure in this table for full imputation), the outcomes with no sales of interests would be identical to those shown in Table 1 (ie with no 1/2 CGT applicable) because it is the entity's taxable income that is integrated with the personal income of individual interest holders. Consequently, 50% realisations CGT is applied at the entity level (forming the basis of the integration), as well as at the interest holder level. (a)	Returns and tax paid are between those of investment neutral arrangements and those that would allow full flow-through of 1/2 CGT preference. That is because the half of realised gains not included in integrated taxable income in a year reduces tax values of entity interests when distributed or on liquidation - so that on sale of entity interests or on liquidation half of the resulting increased capital gain is clawed back (as with fixed trust treatment with no special tax value adjustments). Distribution: 6.5% (no sale) Half CGT preference clawed back. 8.2% to 0.1% (sale) Half CGT preference clawed back at end and 1/2 CGT applied to year-by-year tax value reductions for distributions out of accelerated depreciation and accrued gains. Retention: 6.3% (no sale) Half CGT preference clawed back. 8.2% to 1.7% (sale) Half CGT preference clawed back at end and 1/2 CGT applied to year-by-year tax value reductions for distributions out of accelerated depreciation and accrued gains.	With permanent tax preferences applying that also impact on the sale of entity interests, does not achieve either investor neutrality or investment neutrality. Additional tax value adjustments to entity interests, like those used with the fixed trust treatment, would be required to maintain permanent nature of all the 1/2 CGT tax preference. Nevertheless, as taxable income for individuals and entities approaches the benchmark base (economic income), achievement of both investor neutrality and investment neutrality approached under all circumstances of distribution/retention of income or sale/no sale of entity interests. Large variations in year-by-year returns with sales suggests poor tax design. If the permanent tax preferences did not impact on the sale of entity interests (eg some specified annual receipts exempt from tax), the outcomes would be similar to those in Table 1 - with distributions of the exempt income reducing tax values of entity interests.

Notes to table:

If 100% realisations CGT were applied at interest holder level with 50% realisations CGT applied at entity level (rather than the other way around as for full imputation in these illustrations), after-tax cash flows for interest holders and overall tax paid would be identical to those in Table 1 with 100% realisations CGT applying at both levels. But such a structure would not be logical design.
The excessive tax would be reduced, but not eliminated, were 50% realisations CGT to apply at the entity level as well as the interest holder level. The 50% CGT in these examples applies to capital losses as well as capital gains. That is because the CGT design has net capital gains reduced by half after gross capital gains are reduced by gross capital losses. Outcomes would differ if, instead, gross realised capital gains were reduced by half before applying gross capital losses. Such a change in design applied at both the entity and interest holder levels (again assuming other capital gains are always available to absorb capital losses) would see the excessive tax situation turned into a deficiency of tax. The deficiency of tax situation would be unaltered were 100% realisations CGT to apply at the entity level and the half capital gain/full capital loss design to apply at interest holder level - though larger capital losses would be faced by interest holders on liquidation (or sale of assets followed by cash distribution)
The Australian Government's terms of reference for the Review of Business Taxation (1999) required the Review to examine a proposal to extend full imputation arrangements from companies to also apply to trusts, cooperatives, limited partnerships and life insurers - but set in the context of also considering the benefits of bringing tax value and commercial value closer together (benefits that the analysis in Table 1 suggests would have been related to entity investment neutrality).
With 50% realisations CGT also applying at the entity level, if accrued gains on land were excluded from the tax authority's measure of income for entity distribution purposes the excessive tax would be eliminated. Large variations in year-by-year returns would still occur.
The early tax savings offset (in terms of aggregate tax paid) by higher tax in liquidation year would be expected to be capitalised into the price of entity interests - but such capitalisation effects point to less than ideal tax design.
The same aggregate tax would be paid (with 1/2 CGT preference left intact) but very different year-by-year returns achieved with annual sales of interests (10% to -4.5%) if tax value reductions to entity interests in a year were instead made on the basis of distributions of returns of capital (as under the imputation system) plus prior untaxed income that is subject to balancing adjustments in the year as a result of asset sales or entity liquidation. Such tax value adjustments, necessary to avoid making 'temporary' preferences 'permanent', are complicated and the large variation of after-tax returns with sales of entity interests suggest poor tax design.

CONCLUDING COMMENT

General

Main conclusions from the above discussion are as follows.

Simplicity and entity investor neutrality would be furthered across different entity tax regimes if no permanent tax preferences applied to the tax bases of entities and of their individual owners/beneficiaries. Perverse double tax outcomes would be avoided with full imputation arrangements if permanent tax preferences were eschewed that impacted on the tax treatment of sales of entity interests. Tax concessions would be best structured as temporary (ie providing timing benefits), with balancing adjustment provisions ensuring that over the period that assets/liabilities are held their changes in value are included in the tax base. In the case of Australia in 2009, for example, this would argue for such changes as:
- Removing the 50% exclusion of realised capital gains from taxable income of individuals and some trusts (with consequent changes to superannuation funds).
- Treating all buildings as assets separate from associated land with full balancing adjustments applied to building depreciation on disposal.
Entity investor neutrality would be furthered across different entity arrangements if the same tax base design were to apply to entities, to their individual owners/beneficiaries and to individual direct investors. In the case of Australia in 2009, for example, this would argue for such changes as:
- As already mentioned, removing the 50% exclusion of realised capital gains from taxable income of individuals and some trusts (with consequent changes to superannuation funds). Companies already include 100% of realised capital gains on CGT assets in their taxable income.
- Scrapping the 'simplified' small business regime, which treats eligible small business that opt to enter the regime differently in relation to depreciation, trading stock, etc - a system which, in any case, imposes unnecessary complexity and different tax outcomes around an arbitrary boundary line.
How income is defined by tax authorities for determining the unfranked dividend/return of capital components of distributions purposes is an important design feature of full imputation arrangements - and therefore should be very transparent.
Full integration of taxable income implemented in conjunction with full imputation is a design for the taxation for entities worthy of consideration if consistency of treatment across entities used for various purposes is desired, as well as principle-based arrangements that are sympathetic to both investor and investment neutrality.
- It would be a solid option to be considered by any jurisdiction implementing fresh legislation for the income taxation of entities.

Integrating taxable income

Various inquiries in the past - such as the Carter Report (1966) and the Campbell Report (1981) - have proposed traditional full integration of the company and personal income tax systems. Such traditional full integration arrangements incorporate:

withholding arrangements at the company level to effect prepayments of personal tax on company income; and
taxation at individual shareholder level on any dividends actually paid plus shareholders' share of their companies' undistributed 'profits' (with credits for tax withheld by the company).

Perceived practical difficulties have militated against the implementation of such full integration arrangements, difficulties like determining what a company's undistributed 'profits' are each year (including potentially changes in value of all the company's assets and liabilities) and how to pass out annual estimates of undistributed profits through chains of companies, many of which may be widely held by individual shareholders. The Australian Government (1985) in its Draft White Paper canvassed full integration before singling out full imputation arrangements for further examination. Full integration was seen to involve 'major administrative problems' and 'international implications', most notably large tax revenue costs if refunds were provided to non-resident shareholders for tax withheld by domestic companies.

Commentary by various past inquiries on major administrative problem associated with full integration were made without the benefit of practical experience of the operation of now quite common full imputation systems. The Australian Government (1985) in its Draft White Paper noted that the introduction of imputation 'would provide an appropriate basis for extension to a full integration system were the practical difficulties of that system eventually adequately resolved'. Major administrative systems are required to implement full imputation arrangements and these systems go a long way towards addressing the main administrative challenges posed by full integration. The implementation of full integration therefore becomes more practicable if done in conjunction with full imputation arrangements. Experience in Australia with implementation of fixed trust taxation in conjunction with CGT and associated annual adjustments to trust units' tax values is also helpful. Moreover, full integration would be much more feasible if based around integration of taxable income rather than economic income. Thus, experience with full imputation and fixed trust arrangements plus a shift from integrating economic to taxable income help address the following perceived practical difficulties usually associated with full integration.

Concerns over refunds to non-residents of domestic tax withheld. Under common design of full imputation arrangements, franking credits (and associated franked dividends plus unfranked dividends) flow with biannual dividend advice through chains of domestic companies and out to individual domestic shareholders, as well as non-resident shareholders. Any unfranked dividends going to non-residents attract dividend withholding tax (DWT) - subject to any special arrangements designed to provide DWT relief on foreign income flowing through domestic companies and out to foreign shareholders ('conduit' income). It is then up to the tax authorities in the non-residents' countries to determine what credits they might provide for the imputation franking credits and DWT.
- There is no justification for the country running a full imputation regime to provide non-residents refunds for franking credits just as there is none for a country running a classical system for providing refunds to non-residents for DWT and underlying company tax associated with dividends from that country. The countries of the non-residents can decide the extent of foreign tax crediting or exemption arrangements and even whether to provide refunds of tax for extra foreign taxes paid beyond what is payable in the non-residents' countries - but that is not an issue to concern the country from where the dividends last flowed.
- The same arguments apply equally to full integration arrangements.
Measure of entity income to be taxed to owners/beneficiaries regardless of distribution. Full imputation arrangements might not seem to address a key design feature of traditional full integration arrangements: defining entity 'profits' for the purposes of determining an entity's undistributed profits to be included in beneficiaries/owners tax assessments? But as the snapshots of full imputation arrangements in Tables 1 and 2 have highlighted, implicit (but not always very transparent) in these full imputation arrangements is a definition of profits or income that is used by tax authorities to determine the proportion of unfranked dividends and returns of capital in distributions to shareholders.
- The snapshots give support to enthusiastic tax economists who might argue that the definition of entity profits under both full imputation and full integration should be economic income (and thus including annual changes in value of all assets and liabilities associated with an investment).
- Despite its theoretical appeal, the use of economic income raises practical measurement issues particularly for appreciating assets with no ready market value and raises the same sort of practical political considerations as proposals for CGT to apply to accrued gains of long-term holdings of assets like land and shares.
- Were the taxable income of entities (ideally matching that of individual investors undertaking the same investments) used as the profit measure in a full integration system it would already be part of the fabric of income taxation and, most importantly, it would be transparent.
Advice on taxable income of entity. Full imputation results in the inclusion in company shareholder dividend notices that flow through the entity chain amounts of franked dividends and associated franking credits, as well as unfranked dividends. Under integration of taxable income, those same dividend notices would show taxed income and associated franking credits - whether or not that income was actually distributed - plus the total amount that is actually distributed.
- For domestic owners/beneficiaries, separate identification of unfranked dividends would not be required. They would include the entity's annual taxed income grossed up by associated franking credits in their tax assessments and claim credit off their tax payable for the amount of franking credits. But they would not include unfranked dividends in their assessments - rather, the total amount of the distribution actually made - which might include taxed income, unfranked dividends and a return of capital - would reduce the tax value of their interests in entity (offset by the amount of taxed income plus associated franking credits).
- This would be the same for individuals and for domestic entities further down a chain of domestic entities. Entity tax value reductions, rather than unfranked dividends, would flow down the chain of domestic entities to local individual owners/beneficiaries. The unfranked dividend component of distributions would be treated just like returns of capital are treated under full imputation.
- One possible reason for maintaining the full imputation distinction between franked and unfranked dividends would be to impose DWT on unfranked dividends repatriated to non-resident owners/beneficiaries (requiring special arrangements if DWT relief were to be afforded any 'conduit' foreign income in unfranked dividends).
Required annual tax value changes to entity interests. Under integration of taxable income, final dividend notices could also contain required adjustments to the tax values (cost bases) of entity interests for CGT purposes for the past financial year. Those adjustments would be taken into account in tax assessments for the following financial year in the event of sales of those entity interests in that year.
- We have seen that the required adjustments to tax values of entity interests under full integration of taxable income are: annual taxable income (net receipts plus net changes in tax values of assets/liabilities plus balancing adjustments on disposal) less actual distributions less entity tax paid.
- Inclusion of the size of the required tax value adjustment in final dividend notice would seem to involve similar compliance requirements as does the advice in final distribution notices of Australian unit trusts of required reductions in the cost base of trust units for untaxed distributions (including 'deferred income'). In Australia unit trusts are required to distribute annual taxable income or face tax on that income. (If integration of taxable income were to replace trust treatment, an accompanying simplification would be to have dividend payments replace the traditional trust treatment of having distributed income retain its character.)
- Similarly, companies under full imputation arrangements know the year-by-year taxable income giving rise to tax payments and franking credits. Thus, under integration of taxable income, adjustments for annual taxable income and associated tax payments would be known. Lags in tax return and assessment processes are already dealt with in full imputation and fixed trust arrangements. The handling of issues associated with amendments to initial returns and final assessments would also presumably be similar to the handling of those issues under full imputation and fixed trust arrangements now.
Potential loss in tax revenue. Substituting integration of taxable income for fixed trust treatment may have minimal effect on tax revenue collections depending on any special arrangements under the two systems to maintain the nature of 'permanent' tax preferences (see Table 2). However, substituting integration of taxable income for full imputation arrangements applying to companies would have both positive and negative effects on tax revenue collections.
- Depending on company and personal tax rates, integrating taxable income could result in increased tax collections from individual domestic shareholders being required to include their companies' annual taxable income in their tax assessments even if the companies do not actually distribute that income.
- Compared with full imputation arrangements, integration of taxable income when the company tax base contains significant tax preferences would put much greater emphasis for domestic revenue collection on CGT arrangements.
  - The switch from payment of tax on unfranked dividends by domestic shareholders under full imputation to reduction in the tax value of entity interests under integration of taxable income (as occurs under fixed trust treatment) might not result in significant revenue loss over time if:
    - entity tax preferences were 'temporary' in nature - in which case, tax-preferred income would ultimately be included in entities' taxable income; or
    - no percentage exclusion were allowed for CGT purposes - in which case, CGT revenue would, over time, substitute for regular income tax on unfranked dividends out of tax-preferred income under full imputation.
  - However, if the company tax preferences were 'permanent' and a percentage exclusion were allowed for CGT purposes (as in Table 2), some of the income sheltered by these preferences would not be taxed at a cost to revenue compared to full imputation.
Non-resident owners/beneficiaries. The incorporation of full imputation mechanisms as an integral part of integration of taxable income would ensure that potential tax revenue concerns associated with non-resident beneficiaries/owners were addressed. Tax would still be paid by entities. Franking credit amounts on dividend statements showing how much entity tax underlies a distribution could be used by domestic beneficiaries/owners as tax credits - and would continue to clearly identify how much foreign tax has been paid on dividends received by non-resident beneficiaries/owners for purposes of any crediting arrangements in their countries. With domestic personal tax rates of domestic beneficiaries/owners applying to their share of their entities' annual taxable income, the setting of the entity tax rate would likely focus on how much tax to impose on foreigners investing in, or through, domestic entities. DWT could be made redundant if unfranked dividends were left unidentified. Alternatively, if the abolition of DWT on unfranked dividends out of domestic income raised tax revenue concerns, either:
1. the income tax base could be broadened, recognising that unfranked dividends out of domestic income only arise to the extent taxable domestic income falls short of economic income; and
2. DWT could still be payable on the 'unfranked' dividend component of a distribution to non-resident beneficiaries/owners with that component determined from separate income measurement (retained from imputation arrangements) to distinguish untaxed income from returns of capital. If foreign taxes were not included in franking accounts, special arrangements would then be required if DWT relief were to be given to unfranked foreign 'conduit' income.

Against all of this, dividend statements under full integration of taxable income would:

contain amounts of annual taxed income (regardless of whether distributed or not), associated tax paid (franking credits), unfranked dividends (if a separate measure of income was used to distinguish between untaxed income and returns of capital for DWT purposes) and total distributions actually made;
give domestic beneficiaries/owners the amount of annual taxed income and associated franking credits for inclusion in their tax assessments (along with any net capital gains from entity interests sold during the year) and the size of the adjustment to be made to the tax values of their entity interests - annual franked dividends (annual taxable income less entity tax) less actual distributions; and
give non-resident beneficiaries/owners distribution amounts and associated tax payments for purposes of foreign tax crediting arrangements available in the non-residents' home countries plus unfranked dividend amounts on which DWT was payable (again, if a separate measure of income was used to distinguish between untaxed income and returns of capital). Should a non-residents home country also be integrating taxable income, all the required details for assessment, crediting and tax value adjustments are available (as ideally should be the case now with domestic arrangements that already commonly attribute to domestic owners foreign income of specified types of foreign entities when the income is not actually distributed).

Overall, integration of entity taxable income implemented in conjunction with full imputation arrangements is an amalgam of full imputation and fixed trust arrangements within the concept of integration of entity and personal taxation - accompanied by adjustments to tax values of entity interests to minimise double taxation of retained taxed income. It is a tax design arrangement for entities that has practical and theoretical appeal:

The hard work done by tax authorities in their implementation of full imputation arrangements - as well as fixed trust arrangements in conjunction with CGT - provides a working platform on which to superimpose arrangements for the integration of taxable income.
With taxable income integrated, rather than full economic income, investor neutrality is achievable to the point when entity interests are sold (matching the sale of assets by the alternate direct individual investor). Applied in conjunction with a 100% realisations CGT system, investment neutrality is achievable on an ongoing basis even with significant tax preferences present through the effect of CGT applying to the market value of entity interests when they are sold regularly.
Moreover, as tax preferences are reduced and taxable income approaches economic income, the achievement of both investor and investment neutrality under integration of taxable income is approached under all circumstances of retention/distribution and sale/no sale of entity interests.
With significant tax preferences present, some tax revenue issues associated with superimposing integration of taxable income on full imputation are inevitable. That is because, while annual taxable income is taxed to individual domestic owners/beneficiaries regardless of actual distribution, distributed untaxed income is shifted from regular assessable income to the CGT base.
No major revenue or implementation issues should arise in relation to non-resident beneficiaries/owners - as with full imputation it is up to the non-residents' home countries to determine the extent of foreign tax crediting arrangements. The level of domestic tax to be imposed on foreigners investing in (or through) domestic entities would be a prime consideration in the setting of the domestic entity tax rate.