1 The revenue estimates contained in this document have been prepared by the Review Secretariat with substantial assistance from the Australian Taxation Office (ATO). Industry views expressed in relation to particular revenue estimates presented in A Platform for Consultation have been taken into account when finalising the revenue estimates.
2 The estimates have been prepared against the policy framework set by existing taxation law and the policy measures announced in A New Tax System. The revenue estimates for individual measures need to be considered in the context of both that broader policy framework and the other policy recommendations of the Review. The economic context in which the estimates are set is the same as that used in framing the Budget.
3 Estimates are reported for the period 1999-2000 to 2004-05. Although some individual measures have transitional periods extending well beyond 2004-05, the estimation errors involved in extending the estimates into later years increase very rapidly. The sustainability of the revenue outcomes is discussed in more detail later in this section.
4 The revenue estimates are based on the best available data and methodology. However, it is not realistic to attach a high degree of precision to individual revenue estimates. This is because the data upon which the estimates are based are often seriously inadequate. In addition, it is typically necessary to make judgements regarding behavioural changes by taxpayers in response to a measure. The extent and timing of such responses are often not easily judged.
5 History has shown that estimates relating to new tax measures, such as when the fringe benefits tax and the capital gains tax were introduced, can sometimes be very significantly in error. The estimated revenues for those particular measures at the time they were introduced significantly understated the amount of revenue actually raised. Given the extent of the reforms proposed, there is likely to be a range of unanticipated revenue outcomes -- some positive and some negative. On balance, the Review believes that the revenue estimates presented here are likely to understate the overall positive impact on revenue of the recommended package, possibly to a significant extent.
6 The revenue impact of the Review's recommendations has been measured relative to the revenue base that would otherwise prevail under existing tax law together with the policy proposals outlined in A New Tax System. For example:
where relevant, revenue implications have been estimated using the personal income tax rates that will apply from 1 July 2000;
the estimates for investment and capital gains related measures take into account the impact that indirect tax reform is expected to have on the price of investment goods and the consumer price index; and
where the Review has proposed changes to the entity measures presented in A New Tax System, the revenue impact of those changes has been measured relative to the revenue that would otherwise be raised by the measure as outlined in A New Tax System.
7 The policy setting for an individual measure will also be determined by other Review recommendations. That is, the revenue estimates for an individual Review recommendation are contingent on the policy framework reflected in the range of recommendations proposed by the Review. For example, the revenue implications of removing balancing charge rollovers and removing the ability to assign lease payments for depreciable equipment would differ from those reported if accelerated depreciation were to be retained. In some cases, the revenue outcome from a particular measure may vary substantially on the basis of other policy recommendations.
8 The principal source of data is the ATO Taxation Statistics for the 1996-97 income year. Unpublished ATO data are also used to cost some Review recommendations. Other data sources include the Australian Bureau of Statistics (ABS), the Australian Stock Exchange (ASX), company annual reports, research papers and data supplied by industry bodies.
9 The reliability of the estimates varies in accordance with the availability and quality of data. In some circumstances reliable taxation data are not available because taxpayers are not required to provide the relevant information to the ATO. Reliance must then be had on secondary sources of information or on assumptions in order to derive an estimate of the relevant tax base. The use of secondary source data or assumptions is likely to increase the scope for error in the revenue estimates.
10 The economic environment against which the revenue estimates have been prepared is the same as that used for the purposes of preparing Budget revenue estimates. Although the Budget revenue figures do not extend beyond 2002-03, the same methodology that is used to project the Budget revenue figures to 2002-03 is applied in extrapolating beyond that period.
11 As noted above, the revenue estimates are typically based on taxation and other data that relate to a period several years before the period of revenue estimation. In deriving the revenue estimates, the data were projected forward using available statistical and industry data and Treasury Budget parameters.
12 For example, the revenue estimates for removing accelerated depreciation as a general measure are based on taxation data relating to 1996-97. An estimate of the depreciable assets base in 1999-2000 was derived by increasing the value of the depreciable assets base in 1996-97 in line with ABS data and Treasury Budget parameters for private investment in plant and equipment.
13 To the extent that the Review's recommendations might lead to an increase in economic efficiency or other macroeconomic effects, those effects are not reflected in the parameters used to cost the individual reform measures. Similarly, the macroeconomic effects of an individual measure are not included as part of the revenue estimate for that measure. The effects of the various measures on economic growth and revenue have been taken into account separately in the form of an estimate of the growth dividend. The growth dividend represents the impact on Commonwealth revenue that is expected to arise as a consequence of the Review's recommendations.
14 As noted above, in some cases data are not readily available to determine the size and characteristics of the tax base to which a recommendation applies. In those circumstances, it is necessary to rely on secondary data sources, other indicators of the potential tax base, or educated but somewhat arbitrary assumptions about the relevant tax base. Such assumptions will typically be measure specific.
15 The motivation for most of the reforms proposed by the Review is to change taxpayer behaviour by changing the incentives taxpayers currently face. Consequently these intended changes in behaviour can be an important factor in estimating the revenue impact of particular measures.
16 Little information is typically available about the likely magnitude of taxpayer responses to changed taxation arrangements. The Review has attempted to identify possible behavioural responses and, where they are likely to be significant in terms of revenue, include the likely impact in the revenue estimate. Such estimates are always very difficult and in some cases the Review has identified a response effect but not felt able to make any estimate at all.
17 Recommended reforms may also lead to some unanticipated changes in taxpayer behaviour and obviously these have the potential to impact on revenues in ways that cannot be foreseen.
18 Transitional revenue effects may arise from taxpayers anticipating changes to taxation arrangements. The Review has sought to minimise the extent of such responses in some cases by recommending particular transitional arrangements and timing of implementation. Where it is expected to be significant, the revenue estimates reflect this transitional behaviour by taxpayers.
19 The revenue effect reported for each measure is the net impact of the policy change on revenue, after taking into account behavioural responses by taxpayers and any indirect revenue effects. In some cases the gross or `headline' revenue effect of a measure can be considerably different from the net revenue effect. For example, the net revenue impact of a cut in the company tax rate is considerably lower than the impact on company tax paid. The difference reflects an offsetting increase in tax paid by individual taxpayers, superannuation funds and life insurance companies on distributed earnings because of the reduced quantum of imputation credits.
20 When estimating the net revenue impact of the various base broadening measures, it was assumed that a proportion of the revenue raised from entities would be offset by reduced collections from other taxpayers. The offset reflects the fact that franked and unfranked dividends are both ultimately taxed at the shareholders' marginal tax rate. This is particularly true with the introduction of refunds of excess franking credits.
21 The revenue implications of the measures are generally calculated on an income year basis using available tax data. The income year estimates are converted to a financial year basis using consistent assumptions about the timing of tax payments for different types of taxpayers. The timing adjustments take into account the payment arrangements outlined in A New Tax System. For example, the financial year implications of measures affecting company tax are influenced by whether the companies are large or small and the months in which the companies are required to lodge instalments.
22 The potential revenue impacts for individual measures are reported as point estimates. The point estimates typically reflect what is considered to be the mean of potential revenue outcomes, though in some cases a conservative approach has been taken. The range of potential outcomes surrounding the point estimates can be relatively large for some measures. The discussion above highlighted several aspects of the revenue estimation process that bear on the degree of accuracy of the revenue figures reported for individual measures. Nevertheless, it is likely that, to a reasonable extent, differences between estimated and actual revenue outcomes for one measure will be offset by those for other measures. The extent to which this occurs will influence the margin of error surrounding the overall revenue position reported for the package of reform measures.
23 Table 24.1 presents summary data on the revenue implications of the Review's recommendations.
Table 24.1 Revenue implications of Review's recommendations
(a) The estimates incorporate the impact of base broadening on revenue gained from trusts at the recommended company tax rate; that is, the measure is costed against the Review's recommendations.
24 Over the period 2000-01 to 2004-05 the revenue cost of reducing the company tax rate to 30 per cent rises steadily as a consequence of the assumed maintenance of average economic growth. On the investment side, the gain to revenue accrues quickly, due to the removal of balancing charge rollovers and accelerated depreciation, but then broadly stabilises somewhat below its peak level. As a consequence, the trade-off between reducing the company tax rate and broadening the investment tax base deteriorates over the latter part of the period for which the revenue estimates are reported. The entity reforms proposed by the Review and the simplified tax system for small business also detract from the revenue balance, particularly in the early years.
25 The wedge between the revenue cost of these entity measures and reducing the company tax rate, and the gain to revenue from broadening the investment base, is largely offset in the early years by revenue raised from the integrity measures applied to individuals and entities, the capital gains tax measures and the tax design reforms. The integrity measures provide a substantial ongoing boost to revenue. The revenue gains from the tax design reforms are relatively stable across the estimation period. However, this reflects some phasing of measures and the revenue gains will fall away beyond the estimation period. The fringe benefits tax measures are broadly revenue neutral over the reported estimation period.
26 Toward the end of the period for which the revenue estimates are reported, the estimated revenue deficit for the specific measures recommended by the Review is estimated to be fully offset by increased revenues attributable to the positive impact of the reform package on economic growth and incomes.
27 A key revenue issue raised in submissions received by the Review on A Platform for Consultation was whether the trade-off between lowering the company tax rate and broadening the investment tax base would be sustainable in the longer term. A common concern was that the revenue arising from the removal of accelerated depreciation would be only temporary, whereas lowering the company tax rate would have a permanent, or structural, impact on revenue collections.
28 The view that removing accelerated depreciation would only result in a temporary gain to revenue stems from the fact that allowing an accelerated rate of depreciation only influences the timing of depreciation deductions and not the total value of deductions taken. This point is illustrated in Figure 24.1 for a $1,000 asset with an effective life of 10 years that is depreciated using the prime cost method. The accelerated rate of write-off is assumed to be twice that which would apply under effective life depreciation. Over the first five years, the depreciation deductions claimed are twice as great under accelerated depreciation than under effective life. However, over the subsequent five years the difference in depreciation deductions claimed under the accelerated write-off is gradually eroded and, by the end of year 10, completely offset.
29 Relative to write-off based on the effective life of an asset, accelerated write-off results in less revenue being collected during the early part of an asset's life and more revenue being claimed in later years. Consequently, for an individual asset, removing accelerated depreciation will result in a gain to revenue during the early part of the asset's life and a loss of revenue in later years.
Figure 24.1 Cumulative depreciation deducted under effective life and
30 The example above illustrates the revenue impact of removing accelerated depreciation for a single asset with a 10 year effective life. In practice assets are purchased every year by the population of taxpayers as a whole and so it is important to look at the implications of removing accelerated depreciation for the entire population of depreciable assets. Some of the relevant points can be illustrated by looking at the situation where a $1,000 asset with a 10 year effective life is purchased each year.
31 Figure 24.2 illustrates the revenue implication of removing accelerated depreciation for 10 year assets in the case where one new asset is acquired each year at a constant purchase price of $1,000. The difference in depreciation deducted and, hence, the gain to revenue, from removing accelerated depreciation steadily accumulates over the first five years and then declines to zero by year 10.
32 The total amount of additional depreciation deducted under accelerated write-off increases over the first five years as each successive asset commences write-off. In year five all assets purchased attract greater depreciation deductions than under effective life. After year five the difference in the total amount of additional depreciation deducted under accelerated write-off and effective life declines steadily, reflecting the fact that some assets receive less depreciation under accelerated write-off than under effective life write-off. By the tenth year there is no difference in the amount of depreciation claimed each year under accelerated write-off and effective life write-off for the same level of aggregate spending.
Figure 24.2 Cumulative increase in depreciation deducted under
33 This illustrates that where the capital stock reaches a steady state in terms of both its composition and level there is no sustained revenue gain from removing accelerated depreciation.
34 Similar depreciation profiles to those described above for 10 year assets would exist for assets with other effective lives. The longer the effective life of the asset the longer would be the revenue accumulation phase and the longer the depreciation claw back phase. The relative importance of the depreciation profile for different types of assets will depend upon the total value of investment in those assets and the extent to which rates of depreciation are accelerated under the current regime.
35 Figure 24.3 illustrates estimated profiles of the additional depreciation deducted under the current regime for selected classes of assets with different average effective lives, but assuming that the asset base is constant through time. In each case the difference in depreciation deductions peaks then declines, though for longer lived assets the peaks occur later.
Figure 24.3 Additional depreciation deducted under accelerated depreciation compared with effective life with constant investment
36 Incorporating growth in the price of replacement assets and in the total volume of investment significantly alters the overall revenue profile as illustrated in Figure 24.4. In this case the profiles do not decline to zero, nor do they begin to decline as early or as quickly as in the case with no growth in investment. This outcome arises because a growing capital stock has a higher proportion of assets in the phase where depreciation deductions under accelerated depreciation exceed those under effective life depreciation.
Figure 24.4 Additional depreciation deducted under accelerated depreciation compared with effective life with growth in investment
37 As shown in Figure 24.5, in the absence of growth in the investment base, removing accelerated depreciation would yield only a temporary gain to revenue, though the peak in revenue would be around five years from the time of implementation of the effective life regime and the subsequent decline in revenue only gradual. In contrast, underlying growth in the nominal value of investment in plant and equipment averaging around six per cent per year is sufficient to prevent the revenue profile from declining significantly over the medium to long term.
Figure 24.5 Gain to revenue from the removal of accelerated depreciation with and without growth in investment
38 This section provides tables setting out the revenue impact of all the Review's recommendations. Where measures have a significant revenue impact, it also provides a description of the methodology used in deriving those estimates.
39 The revenue estimates presented in this report were derived in a sequenced manner to enable them to be totalled in such a way as to provide an assessment of the overall revenue implications of the Review's recommendations. An outline of the sequencing of the revenue estimation is provided below.
40 As noted earlier, the benchmark for achieving revenue neutrality is the revenue base that would otherwise prevail under existing tax law together with the policy proposals outlined in A New Tax System. The revenue estimates for the policy proposals in A New Tax System have been revised to reflect the latest available information. Hence, the starting point for estimating the revenue implications of the Review's policy recommendations is the impact of the reduction in the company tax rate on the existing tax base and the business entity measures announced in A New Tax System.
41 The Review's entity recommendations were costed relative to the revenue from the entity measures contained in A New Tax System at the recommended company tax rates.
42 The investment measures are also costed on the basis of the recommended company tax rates. The revenue effect of the removal of accelerated depreciation for all businesses was estimated prior to estimating the effect of removing balancing charge rollovers for all businesses and the lease assignment measures. Hence, the balancing charge and lease assignment costings were estimated on the basis that accelerated depreciation had been removed. The revenue impact of both measures would be significantly larger if accelerated depreciation were to be retained. The low-value asset pool is also costed on the basis that accelerated depreciation had been removed. The small business depreciation measures, including the delayed removal of accelerated depreciation and balancing charge rollovers, are costed on the basis that accelerated depreciation had been removed and the low-value asset pool for all businesses had been implemented.
43 In the case of the capital gains measures, removal of averaging is costed on the basis that indexation had been frozen. The costing of the proportional base reduction measures and scrip-for-scrip rollover relief are on the basis that indexation had been frozen and averaging removed.
44 All other measures are costed independently but on the basis of the recommended company tax rates.
45 Table 24.2 sets out the cost to revenue from reducing the company tax rate, in respect of both the existing company tax base and the measures proposed in A New Tax System.
Table 24.2 Cost to revenue of reducing the company tax rate
(a) The estimates for these measures have been updated for changes in economic parameters and costing methodology. A New Tax System also included revenue estimates for share buy-backs and liquidations. The cost of that measure is not considered to be significant.
(b) The estimates incorporate the impact of base broadening on revenue gained from trusts at the recommended company tax rate; that is, the measure is costed against the Review's recommendations.
46 A New Tax System foreshadowed that the taxable income of trusts would be taxed at the company rate (rather than in the hands of beneficiaries as at present) and that tax-preferred income would be taxable upon distribution. Furthermore, excess franking credits would be refundable to resident individual members and complying superannuation funds.
47 The tax base for this measure was constructed from ATO data for trusts and data on tax-preferred income from listed property trusts.
48 The gain to revenue from taxing trusts as companies arises from changes to the timing of tax collections on the taxable income of trusts and from the taxation of distributions of tax-preferred income. The estimates also include the impact of the trust transitional provisions announced in A New Tax System.
49 Comparing the timing of revenue receipts under the new company tax payment arrangements with receipts under the existing arrangements was the basis for the estimated gain to revenue arising from distributions of the taxable income of trusts to individuals, superannuation funds and companies.
50 The gain to revenue from taxing distributions of tax-preferred income to individuals was calculated using assumptions about the extent of tax-preferred income earned by trusts and likely response effects in respect of the distribution of that income. The tax timing implications of the payment arrangements were also taken into account.
51 Reflected in the estimates is an assumption that a proportion of trusts restructure to partnership or individual arrangements in response to the change in taxation treatment. An allowance was made for refunds of excess imputation credits to individuals receiving trust distributions. Complying superannuation funds were assumed to vary down tax instalments from 2000-01 in response to the increased availability of imputation credits. This costing includes trusts that are defined as CIVs and the impact of extending the same business test to trusts.
52 This measure has not been recommended by the Review. However, the measure requires costing at the recommended company tax rates in order to establish a benchmark for costing the Review's recommended option of taxing inter-entity distributions.
53 The introduction of deferred company tax would have taxed all distributions of tax-preferred income at the entity rate. Distributions of tax-preferred income between entities in non-consolidated groups and from companies to non-residents would have been taxed at the entity rate (currently such distributions to non-residents are exempt although most attract dividend withholding tax). Tax-preferred distributions to residents would have continued to attract the entity tax rate and be fully franked on distribution.
54 ATO data were used to estimate the amount of unfranked dividends paid to resident public companies from resident public and private companies. ATO and ABS data were used to estimate the amount of unfranked dividends paid to residents and non-residents.
55 The estimated gain to revenue would have arisen from the taxation of distributions of tax-preferred income between resident entities that are not part of a consolidated group and a higher rate of taxation on distributions of tax-preferred income to non-residents (currently taxed at a rate of 15 per cent under dividend withholding tax). Deferred company tax on trust distributions was included in the costing for taxing trusts as companies.
56 The estimated gain to revenue from resident taxpayers that are not entities was assumed largely to reflect a bring forward in the timing of taxation. Tax would have been paid at the time of distribution from the originating company rather than following distribution to resident shareholders who currently remit tax on unfranked dividends through provisional tax or on assessment. An allowance was made for refunds to individuals of excess imputation credits arising under this measure.
57 The estimated gain to revenue from non-residents was derived by comparing revenue collections under the existing dividend withholding tax with that which would result at the company tax rate. Response effects were included for an expected lower level of distribution of unfranked dividends to non-resident entities. It was also assumed that there would be a bring forward of unfranked dividends paid to non-resident parent entities prior to the imposition of this measure.
58 A refund of excess imputation credits would be available for resident individuals and complying superannuation funds for distributions received from entities in the new entity tax system.
59 ATO and ABS Household Expenditure Survey data were used to identify resident individual taxpayers that would be likely to receive excess imputation credits and thereby qualify for a refund.
60 The estimated cost to revenue of this measure was derived by taking the difference between the entity tax rate applied on existing franked distributions received by the individuals identified above and their marginal tax rates. In doing so, imputation credits were applied after all other rebates, which were assumed to remain constant. All refunds were assumed to occur on assessment. The impact of allowing refundable imputation credits on distributions arising under the other business measures outlined in A New Tax System were included within the costings for those measures.
61 The following measures for the taxation of life insurers were proposed in A New Tax System.
The current multiple tax rate structure for life insurers undertaking life insurance business to be replaced by a single company tax.
The tax base of life insurers to be broadened to include all their funds management, underwriting and other profit on their life insurance and immediate annuity business.
Bonuses assigned to new life insurance investment policies (excluding risk policies) to be taxed at the marginal tax rates of policyholders regardless of the period of investment.
62 The life insurance estimates are based upon ATO analysis of industry data on fees and charges and its own data on the taxable income of life insurers, assessable life insurance bonus income, and life insurance rebates.
63 The effect of the change to the proposed company tax rate was considered for each class of life insurers' business. Non-complying superannuation and shareholders' class of business were not considered to be significant.
64 The potential tax base attributable to management fees was derived from data on management fees levied as a percentage of the account balance and flat fees charged annually on accounts.
65 In determining the revenue impact of taxing the management fee income of life insurers it was assumed that life insurers will be able to claim additional deductions for assessable income from account establishment fees and management fees. The revenue estimates also recognise that individuals will be entitled to deductions for some fees paid to life insurers.
66 The estimates for the policyholder measures reflect the different tax treatment to be applied to bonuses on policies taken out before and after the commencement of the new regime.
67 To ensure no double counting of revenue, the total net tax collected was derived as the sum of company tax paid by life insurers and net additional tax paid by policyholders, less the amount of revenue presently collected from income assessable to policyholders (net of rebates). Similarly, the revenue gain from taxing the immediate annuity business of life insurers was adjusted to avoid potential double counting with the management and underwriting income measures.
68 During the consultative process the life insurance industry has argued that the revenue estimates published in A Platform for Consultation relating to funds management profit were higher than their expectations. The Review has retained its original basis for the costing of this measure because the industry has been limited in its ability to provide detailed material to support its claims, due to commercial sensitivities.
69 This costing excludes the impact of proposed amendments to the commencement of the new tax regime for life insurers and the application of the rebate rate for existing life insurance investment policyholders. This costing also excludes the proposed transitional measures for the taxation of life insurers' management fee income.
70 The costing for the company rate cut consists of two broad components: the impact on the projected company tax base in the absence of the entity measures announced in A New Tax System, and the impact on the revenue raised from the entity measures in A New Tax System.
71 This item reflects the net impact on revenue from the existing company tax base of reducing the company tax rate from 36 per cent in 1999-2000 to 34 per cent in 2000-01 and 30 per cent in 2001-02.
72 The costing is based on ATO data for company tax paid and imputation credits utilised by residents.
73 The net revenue impact of reducing the company tax rate was estimated by making proportional adjustments to the value of tax paid at the company level and the value of imputation credits utilised by residents and then taking the difference of the two estimates.
74 Reducing the company tax rate would increase the after-tax profits of a company and reduce the amount of franking credits available to be distributed to shareholders. In estimating the net revenue impact, public companies were assumed to distribute the same proportion of after-tax profits before and after the change in the company tax rate. A small increase in retained earnings was assumed for private companies. If public companies were to reduce the proportion of after-tax profits distributed in response to the Review's recommendations the revenue cost of the measure would be greater.
75 This item reflects the impact of reducing the company tax rate on the revenue collected from the measures outlined in A New Tax System.
76 The revenue figures were derived from the models used to produce the costings for the measures in A New Tax System. The revenue estimates reported for these measures do not always vary in proportion to the reduction in the company tax rate.
77 In the case of the costing for taxing trusts as companies an adjustment was also made to reflect a likely reduction in the amount of tax-preferred income distributed from trusts as a result of the investment base broadening measures.
78 In the case of the estimates for life insurers, this reflects the fact that the revenue raised under the proposals reported in A New Tax System is not entirely the result of base broadening.
Table 24.3 Revenue from changes to taxation of investments
79 Table 24.3 provides details of the costing of all investment measures. The removal of accelerated deprecation is to assist funding of the company tax rate reduction. The Review believes this package represents a more efficient and consistent approach to the taxation of investments.
80 Accelerated depreciation is recommended to be removed as of the date of announcement. For businesses with a turnover of less than $1 million, who elect to use the simplified tax system, accelerated depreciation is recommended to be removed as of 1 July 2000 but with the removal to apply to the written down value at that time of all assets acquired after the date of announcement.
81 The revenue estimates are based on ATO data for depreciable assets purchased and ABS investment and capital stock data. The depreciable assets base was adjusted for the impact of indirect tax reform on the price of investment goods and for a bring-forward of some expenditure on depreciable assets in anticipation of the possible removal of accelerated depreciation.
82 The revenue estimates were derived by comparing the difference in the amount of depreciation that would be claimed under the existing depreciation regime with that under an effective life regime.
83 Due to an absence of detailed data on the distribution of assets by effective life, an assumed distribution of assets was derived using the Commissioner's schedule of effective lives of assets. The assumed distribution of assets was benchmarked to have an average effective life equal to the ABS average age of private sector plant and equipment.
84 The proportion of depreciable assets valued using the diminishing value method, rather than the prime cost method, was assumed to be around 70 per cent. If a higher proportion had been assumed the estimated revenue gains from removing accelerated depreciation, over the period reported, would be higher.
85 The impact on revenue of delaying removal for small business was estimated separately by defining a separate asset base for small businesses. The difference in revenue is reported as part of the small business measures.
86 Immediate deductibility for depreciable assets with a value of less than $300 is to be replaced with a 37.5 per cent declining balance or depreciated value write-off rate for all depreciable assets worth less than $1,000 as of 1 July 2000.
87 The revenue estimates are based on ATO data on purchases of depreciable assets and ABS data on plant and equipment.
88 The revenue estimates were derived by estimating the difference between the estimated amount of revenue collected from the removal of the $300 immediate write-off rule and the estimated revenue cost of the proposed $1,000 asset pool.
89 Due to an absence of detailed data on the distribution of assets by effective life, an assumed distribution of assets was derived. It was assumed that low-value depreciable assets have a lower average effective life than higher value assets. Hence, the profile of asset lives was skewed toward shorter effective life assets relative to the assumed profile for all depreciable assets. It was also assumed that the effective lives of assets falling under the current $300 limit was shorter than those valued at less than $1,000. This assumption explains the relatively large transitional gain to revenue in 2001-02.
90 The revenue impact of implementing the $1,000 asset pool is estimated relative to the current expensing of assets acquired for less than $300 and the costing benchmark set by effective life treatment of assets acquired for between $300 and $1,000.
91 Although the costing is based on taxation data, it relies on several key assumptions on the asset life distribution of low cost depreciable assets. The estimates are sensitive to these particular assumptions.
92 For residential and non-residential buildings the existing depreciation arrangements are to be replaced with an effective life regime as of 1 July 2000. The new arrangements will only apply to buildings constructed after 1 July 2000.
93 The revenue estimates are based on ABS data on non-residential and residential new construction and improvements. They were derived by comparing the difference between the amount of revenue forgone had the existing depreciation arrangements for new buildings and structures remained, with the revenue cost of an effective life regime for new buildings and structures.
94 Detailed information on the distribution of depreciable buildings and structures by effective life is not available from ABS statistics. The estimates assume that, relative to the proposed depreciated value/prime cost effective life depreciation regime, existing prime cost depreciation arrangements are generous for residential buildings but less generous for non-residential buildings and structures.
95 Although the estimate is based on ABS data, it relies on several key assumptions on the assumed asset life distribution of depreciable buildings and structures.
96 The balancing charge rollover provisions for plant and equipment are recommended to be abolished as of the date of announcement for businesses with a turnover of $1 million or more. For businesses with a turnover of less than $1 million, balancing charge rollovers are recommended to be removed from 1 July 2000.
97 Revenue estimates are based on ATO data on purchases of net depreciable assets.
98 Removal of balancing charge rollovers for depreciable assets was modelled using four classes of depreciable assets -- with average effective lives of 5, 6, 10 and 13 years. These assets were assumed to be disposed of two years prior to the end of their effective lives. In addition, two categories of passenger motor vehicles were modelled -- those sold after three years and those sold after five years. Companies were assumed to account for most use of balancing charge rollovers.
99 This costing is also based on an assumption that accelerated depreciation is removed for all businesses. Under effective life depreciation, the balancing charge rollover is assumed to be largely irrelevant because the tax written down value of a depreciable asset and its market value, on average across all assets, would be the same. Hence, the estimate only reflects the impact of imposing balancing charges on assets acquired prior to the date of announcement.
100 These revenue estimates are particularly sensitive to the assumptions concerning the timing of disposal of assets and how representative were the asset classes modelled.
101 The impact on revenue of delaying removal for small businesses was estimated separately, taking into account the asset base and effective life profile of assets held by those businesses.
102 Blackhole expenditures are to be either included in the tax value of an asset, immediately deductible or deductible over the relevant life of the asset or a 5-year statutory period -- depending upon the nature of the expense.
103 The costing is based on data from the 1995 Minerals Council of Australia survey of its members relating to non-deductible business expenditure. In estimating the total expenditure base affected by this measure, 75 per cent of non-deductible expenditures are assumed to be blackhole expenditures with the minerals industry data assumed to account for 33 per cent of all expenditure of this type at the trust/company level. An additional component to this costing was included in respect of start up costs for new businesses. Other blackhole expenditures are included in the costing for the taxation of rights.
104 In determining the timing of the revenue consequences of this recommendation, one-third of the total expenditure was assumed to be included in the tax values of assets, one-third would be immediately deductible, and one-third would be written off over a defined or statutory period.
105 These assumptions were necessary due to the lack of hard information about these expenditures. The revenue estimates are sensitive to the assumptions made.
106 Immediate deductibility for removal of overburden is to be replaced with cost absorption as of 1 July 2000.
107 ATO data on mining industry deductions and the 1997 Minerals Industry Survey were the basis for the revenue estimates.
108 The revenue estimates were derived by comparing the difference between the amount of revenue had immediate deductibility remained, with expected revenue collections had three-quarters of existing deductions been amortised over an assumed average life of a mine of 8 to 10 years and the remainder continued to be immediately deductible.
109 Due to an absence of detailed ATO data on the cost of overburden deductions, the estimates are indicative only. The estimate is sensitive to the assumption on the average life of mines. The estimate assumes that deductions are written off using the prime cost method -- the estimates are not sensitive to this assumption.
110 This measure ensures the full consideration from the assignment of leases is included in assessable income from 22 February 1999.
111 The tax base for this measure was derived from ABS data and data from the Australian Equipment Leasing Association. It was assumed that 5 per cent of all finance leases and 30 per cent of leveraged leases are currently assigned.
112 The revenue implications of addressing the assignment of leases were derived by comparing the tax outcomes if leases are assigned with the outcomes if the same leases are not assigned, following the introduction of the measure. Half of the leases are assumed to be assigned to entities paying tax at 15 per cent and the other half are assigned to effectively tax-exempt entities. The revenue implications were modelled on the basis of representative leases with an average effective life of 10 years that are assigned at the end of the sixth year.
Table 24.4 Revenue from changes to taxation of income from entities
113 Table 24.4 sets out the revenue impact of the Review's recommendations in respect of the taxation of income from entities. The first three measures are revenue costs reflecting changes to measures originally announced in A New Tax System. Allowing imputation credits for foreign DWT imposes a significant revenue cost but is an important reform in support of the increasing globalisation of the Australian economy. The thin capitalisation and gearing rules are directed at ensuring that income subject to Australian tax cannot be artificially reduced. The consolidation measures reflect both a more generous treatment of losses bought into consolidated groups and the integrity benefits of the consolidation regime.
114 Widely held trusts that meet the definition of collective investment vehicles (CIVs) are to be excluded from the new entity tax system.
115 Taxable income will flow through the CIV and be taxed in the hands of the member. Distributions of tax-preferred income will not be taxed. The tax value of a member's interest in a CIV will be reduced when a distribution that does not extinguish membership interests is made from other than taxable income or accumulated taxable amounts -- unless the distribution consists of permanent tax-preferred income such as tax-exempt income received by the CIV from a Pooled Development Fund.
116 The revenue estimate consists of two components. The first is the cost of excluding CIVs from the entity tax system. The second is the cost of allowing tax-preferred income of CIVs to be tax free when distributed to unit holders.
117 The revenue gain from taxing trusts like companies exclusive of CIVs was estimated. The difference between the costing of the original proposal for taxing trusts as companies at the recommended company tax rate, and the costing exclusive of CIVs is the cost of excluding CIVs from the new entity tax system.
118 ATO data were used to estimate the proportion of all trust taxable income distributed to individuals from entities likely to meet the CIV criteria. In the absence of ATO data on the amount of tax preferences that are distributed from CIVs, an indicative assessment was made based on information for the listed property trust industry.
119 Additionally, the proportion of distributions sourced from returns of contributed capital and permanent and temporary tax preferences were estimated based on industry information of CIV investment activities. This allowed an estimate to be made of the revenue cost of allowing distributions of tax-preferred income to be tax free in the hands of unit holders.
120 Revenue from CIV distributions to superannuation funds was assumed to be unchanged on the basis that superannuation funds would vary down their tax instalments to account for excess imputation credits received in respect of trust distributions. Hence, the estimate only reflects the impact of the changed treatment on distributions to individuals.
121 This measure will impose tax at the entity tax rate on unfranked dividends paid between non-consolidated entities. The tax, where applicable, is paid by the receiving entity. The payment of unfranked dividends to non-resident shareholders will remain subject to the existing non-resident dividend withholding tax.
122 ATO data were used to estimate the amount of unfranked dividends paid to public companies from both public and private companies (private companies are already subject to tax on unfranked distributions they receive), and the amount of unfranked dividends paid to non-residents. An indicative assessment of the amount of trust tax preferences distributed to other entities was made using ATO trust data and data from the listed property trust industry.
123 The measure is costed relative to the revenue estimated to arise under the deferred company tax measure at the recommended company tax rates. The cost to revenue arises mainly because this measure, unlike the deferred company tax, does not impose additional tax on distributions to non-residents. The remainder of the cost arises from taxing resident taxpayers upon receipt of the unfranked dividend rather than upon distribution from the originating entity.
124 Resident companies and trusts will receive franking credits in respect of foreign dividend withholding tax (DWT) paid on dividends they receive. This measure will allow a credit for foreign DWT to flow to resident shareholders and beneficiaries. Franking credits will be limited to 15 per cent of the gross dividend received.
125 ATO data were used to estimate the amount of foreign source non-portfolio dividends received, and the utilisation of franking credits by companies and trusts in receipt of those dividends. Data were not available for foreign source portfolio dividends received by companies and trusts. However, these are assumed to be relatively small.
126 The cost to revenue of the recommendation will depend upon the extent to which franking credits provided under the measure are used to frank dividends received by resident shareholders and beneficiaries. The proportion of additional franking credits that would be used to frank dividends was estimated by comparing the foreign source dividends received by company groups and the unfranked dividends paid by those groups. The proportion of dividends paid to resident shareholders was assumed to be 70 per cent.
127 Payment of some foreign source dividends is likely to be deferred from 1999-2000 until the introduction of the measure. An adjustment for this response effect was included in the estimate. Foreign subsidiaries of Australian companies may also increase dividends repatriated to Australia as a result of this measure. This latter response effect has not been modelled and, hence, the estimates may understate the cost to revenue.
128 Under the recommendations, foreign-owned groups will be allowed to gear up to a `safe-harbour' debt to equity ratio of 3:1. Above this level, interest expenses will only be deductible to the extent that the gearing level of the group could have been borne by an independent group operating under the same terms and conditions (the `arms-length' test).
129 Interest expenses incurred by foreign owned groups were estimated using ATO tax return data. The IBIS Business Systems data on gearing were used to ascertain the gearing levels of foreign-owned groups and to estimate the proportion of interest expenses that would exceed the `safe-harbour' gearing ratio and the arm's-length test.
130 The potential reduction in interest expenses that would be allowed under this measure was estimated on the basis of the aggregate interest expenses data and average gearing levels. It was not possible to match data on interest expenses claimed by individual companies with the gearing levels reported by IBIS on a group basis.
131 It was assumed that the measure would encourage foreign multinationals to restructure their arrangements in order to decrease their gearing of Australian operations. The revenue estimates include an allowance for the loss of some interest withholding tax revenue as a result of this assumed response.
132 The measure could also cause some foreign controlled companies to seek other methods of profit shifting to reduce their Australian tax liability. This potential response has not been accounted for in the costing.
133 In A New Tax System (page 120), the Government proposed to change the taxation treatment of life insurers commencing in the 2000-01 income year.
134 Many life insurers have substituted accounting periods. Starting the new system from an income year would create competitive advantages and disadvantages -- early balancing companies would have their lead time shortened while late balancing companies would have a longer lead time and a later start date on the new tax basis. The industry therefore suggested a common start date and the measures will commence from 1 July 2000.
135 The revised implementation will apply to each of the measures affecting the taxation treatment of life insurers. The revenue impact of this measure was estimated by changing the date of commencement within the models for costing the life insurer measures proposed in A New Tax System.
136 The costing incorporates the effect of the one year lag in reducing the rebate rate that applies to bonuses paid on existing life insurance policies to the company tax rate.
137 The recommended transitional measure for the taxation of life insurers' management fee income was costed separately.
Transitional taxation of the management income from existing life insurance, annuity and pension business
138 As a transitional measure, only two-thirds of a life insurer's management fees derived from life insurance policies taken out before the date of announcement will be included in the taxable income of life insurers during the five years following commencement of the new regime.
139 The revenue impact of this recommendation was estimated by reducing, by one-third, the management fee tax base used to estimate the revenue impact of the life insurance measures proposed in A New Tax System.
140 Consolidated groups will be able to bring realised carry-forward losses of a subsidiary entity into the group -- subject to limits on the amount that can be brought in and the period over which the losses can be claimed. Similar rules will be applied to unrealised losses brought into a consolidated group.
141 The revenue estimates are based on ATO data on carry-forward losses -- both capital and revenue. They were derived by comparing the potential usage of losses by groups under consolidation with the usage of losses by group companies under the present law, including the loss transfer provisions.
142 The estimates are based on an assumption that 95 per cent of groups will choose to consolidate. Two-thirds of losses are assumed to be continuity of ownership losses and one-third same business test losses. It was also assumed that 10 per cent of non-group losses will be purchased and brought into a consolidated group.
143 The estimates are based on projected growth in the stock of losses adjusted for those losses assumed to be the result of value shifting or loss duplication.
144 The existing grouping provisions are to be repealed and wholly owned groups of entities allowed to elect to be taxed as a single, consolidated entity. Consolidation will prevent value shifting, loss duplication and loss cascading within a group.
145 The estimates are based on ATO data. They reflect the gain to revenue from preventing the use of losses created by value shifting or duplication within company groups.
146 Based on the loss transfer history of company groups it was assumed that 95 per cent of groups will choose to consolidate. It was also assumed that around one quarter of capital losses transferred into companies would be eliminated under consolidation.
147 The estimates include an increase in revenue in the 2000-01 year for the interim loss duplication measures which apply from 22 February 1999. The estimate is calculated as a proportion of the expected gain to revenue from the removal of loss duplication in the first year of consolidation. It assumes opportunities were made of the current deficiency in the law between 22 February 1999 and the date of announcement of the measure.
148 Businesses with a turnover of less than $1 million will be able to elect to use a simplified tax system consisting of:
cash accounting for a broad range of receipts and payments;
a simple treatment of trading stock; and
a simplified depreciation regime for most tangible depreciating assets.
149 Taxpayers will be required to adopt all aspects of the simplified tax system if elected.
150 Table 24.5 sets out the revenue costs of the Review's recommendations in respect of small business. The impact of allowing small business to move to a cash accounting basis is largely a timing effect and this explains why there are only revenue consequences in the first two years of the measure.
Table 24.5 Revenue cost of small business measures
151 The revenue estimates are based on ATO data for closing stock on hand and the closing balances of `debtors' and `creditors' of those businesses with a turnover of less than $1 million.
152 A key assumption required to estimate the revenue impact of this recommendation is the proportion of businesses that are likely to elect to adopt the simplified tax system. Some businesses will benefit financially from the use of the cash accounting method. Others, even though they may not benefit financially, will be attracted to the package by the simplified depreciation provisions and the overall simplicity of the cash accounting system.
153 An analysis of relevant tax data was undertaken to provide a basis for the estimation of the likely participation rate in the simplified tax system. This was applied to determine the overall revenue impact of adopting the cash accounting treatment and simplified treatment of trading stock.
154 The simplified depreciation arrangements for small businesses consist of immediate write-off of assets acquired for less than $1,000, inclusion in a common pool of all depreciable assets acquired for $1,000 or more with an effective life of less than 25 years (including existing assets) -- with a write-off rate of 30 per cent per year (declining balance)-- and effective life treatment for all depreciable assets with an effective life greater than 25 years.
155 Small business also benefits from the delayed removal of accelerated depreciation and balancing charge rollovers. The methodology of estimating the cost of the deferrals is outlined under the general description for removal of accelerated depreciation and balancing charge rollovers.
156 The revenue estimates for the simplified depreciation regime for small businesses are based on the same framework for estimating the revenue implications of the general depreciation measures. This was done by defining a separate asset base for small businesses with its own assumed distribution of assets by effective life.
157 The costing is based on the same assumption as that used for the proportion of small business adopting the simplified tax system. It includes the revenue impact of allowing assets acquired after 1 July 2000 to be treated in accordance with the small business depreciation provisions, and of including all existing depreciating assets in the asset pool at their written down value as recorded at 30 June 2000.
158 The revenue estimates were calculated relative to the benchmark set by the removal of accelerated depreciation and the general pooling arrangements for assets valued at less than $1,000.
Table 24.6 Revenue impact of integrity measures
159 The existing share value shifting and `asset stripping' rules are to be replaced with general value shifting rules.
160 The main impact of this measure is on trusts. The revenue estimates are based on data obtained from ATO case studies involving trusts that are outside the scope of the current value shifting regime but will be covered by the general value shifting rules.
161 The estimates reflect the increase in capital gains tax collections resulting from the removal of the effects of value shifting arrangements. It was assumed that the revenue arising from ATO case studies represents half of the gain from all trusts that will arise as a result of the introduction of the general value shifting rules.
162 The cost bases of interests in trusts would influence the amount of capital gains assessed under the general value shifting rules. As that information was not available, an assumption was made that, for the ATO case studies, the cost bases of the trust interests were equivalent to half the value that was shifted.
163 The revenue estimate for the 2000-01 year covers the interim value shifting measures (dealing with debt forgiveness) which apply from 22 February 1999. The estimate is based on ATO case studies and assumes an increase in value shifting activity between 22 February 1999 and the date of announcement of the measure.
164 Where there is a change in an entity's majority underlying ownership after the date of announcement, losses realised in respect of assets held at the time of ownership change will be made subject to the same business test to the extent of the net unrealised loss relating to the period before the change. Unrealised losses on inter-entity equity interests will be denied where entities are interposed between the ultimate shareholders and the entity with the loss.
165 The estimates reflect the increase in capital gains tax collections resulting from a reduction in unrealised capital losses currently able to be realised and offset against capital gains.
166 The estimates are based on the assumption that there is a one-to-one ratio between realised and unrealised capital losses. Consequently, the revenue estimates are based on ATO data on realised net capital losses carried forward as at 30 June 1997. As was assumed for the costing of consolidation of losses in non-group companies, it was assumed that two-thirds of unrealised losses are `continuity of ownership' losses and one-third are `same business test' losses. It was also assumed that unrealised losses transferable between entities in the group constitute one quarter of the total pool of unrealised losses.
167 Under this measure, individual taxpayers will not be able to offset losses from non-commercial activities (including hobbies and lifestyle choices) against other income unless certain conditions are met.
168 The tax base for this measure was estimated from ATO tax return data for losses claimed by individuals and partnerships over a period of six years. The potential value of non-commercial losses affected by this measure was estimated by screening out losses from those activities that would meet the $20,000 annual turnover threshold or the relevant total business assets threshold.
169 The revenue base was constructed by applying the average marginal tax rate of the individuals with non-commercial losses to the identified value of losses. Historical taxation data for small businesses were used to develop a representative profile over which carried forward losses were assumed to be utilised. The estimate reported for an individual year reflects the revenue impact of denying non-commercial losses generated in that year, net of the estimated utilisation of losses denied in previous years and carried forward.
170 Under this measure, personal services income derived in an employee-like manner will be taxed in that way regardless of the legal form used to provide the services.
171 The costing uses ABS data, a 1994 report by the National Institute of Labour Studies and fieldwork undertaken by the ATO.
172 An estimate of the number of non-agricultural unincorporated contractors and owner/manager incorporated enterprises was derived from the ABS data. Based on data provided in the National Institute of Labour Studies report it was assumed that approximately 40 per cent of these contractors deliver their services in an `employee-like way' and, hence, would be affected by this measure.
173 Based on ATO field work, it was assumed that around three quarters of the unincorporated contractors defined above currently claim deductions to which they would not otherwise be entitled at an average value of around $3,000 per year.
174 In the case of incorporated enterprises, it was assumed that one quarter claim additional deductions equivalent to those assumed for unincorporated contractors. The remaining three quarters were assumed to claim additional deductions to the value of $6,000 per year, as well as split the earned income with a spouse and use the entity to retain income.
175 The revenue estimates reflect the impact of denying the additional deductions assumed to be claimed and of taxing the assumed income of the above owner/manager entities as if received by a single taxpayer as wages or salary.
176 Table 24.7 sets out the revenue impacts of the Review's recommendations in respect of CGT reforms. The reforms have been designed to be as revenue neutral as possible within the CGT arrangements. This broad revenue neutrality might be taken to indicate that the reforms are doing no more than moving the burden of taxation around. In fact the Review believes that the behavioural responses to its reform will substantially increase the capital gains base. Consequently the tax burden on any individual investment will generally be lower.
Table 24.7 Revenue impact of CGT reforms
(a) Includes effect of all changes on individuals; 50 per cent discount of base, abolition of averaging and freezing of indexation.
(b) Includes effect of freezing of indexation and 33 per cent discount of base.
177 The methodology used to arrive at the above revenue estimates is described later in this section.
178 Table 24.8 sets out the revenue implications of the Review's recommendations in respect of fringe benefits tax. Once again the Review has sought a broadly revenue neutral outcome in respect of these measures.
Table 24.8 Revenue impact of FBT reforms
179 From and including income year 2001-02, business entertainment expenses incurred by taxable and tax-exempt employers in relation to restaurant and catered meals, admissions to functions and the provision of venues and associated costs for business-related functions are recommended to be removed from fringe benefit tax (FBT) coverage. These expenses would also become non-deductible to taxable employers.
180 The existing tax base for this measure was estimated from annual FBT returns on meal entertainment and entertainment fringe benefits, along with a small proportion of expense payments, airline transport, property and other fringe benefits categories.
181 The revenue impact of this estimate reflects the difference in the FBT revenue forgone by exempting business-related entertainment and the revenue gain from removing deductibility for such expenses.
182 This measure will remove from fringe benefits coverage car parking provided `on-premises' or in associated buildings covered under leasing or rental arrangements from and including income year 2001-02.
183 The cost of the measure is the tax revenue forgone as a result of removing on-premises car parking from FBT coverage. The revenue cost of this measure was estimated using FBT returns on car-parking fringe benefits.
184 From and including 2001-02, the value of car fringe benefits are to be assigned to an employee on the basis of 55 per cent presumed private use.
185 The tax base for this measure was estimated from FBT returns on car fringe benefits. A distribution of vehicles was constructed to match 1996-97 data for the statutory method and the existing statutory percentages. The operating cost of motor vehicles was drawn from NRMA Vehicle Operating Costs, June 1998.
186 The new arrangements were assumed to be fully phased-in by 2004-05, when existing leasing arrangements will have been renewed. The revenue estimate represents the net effect of taxing the benefits in the hands of employees under the PAYE system, rather than under the FBT system, and of applying the schedular method for calculating the taxable value of car benefits.
Table 24.9 Revenue impact of high level reform to tax design
187 As part of moving to the cashflow/tax value framework, the 13 month rule for pre-paid expenditure is to be removed. As a consequence, the outstanding value of any prepayment will be brought to account at the end of the year. Taxpayers will be allowed to spread the increased tax liability of applying this measure in 2000-01 over a period of five years.
188 The revenue estimates were derived from data on prepayments contained in the published annual reports of a sample of Australian companies. The value of prepayments derived from the sampled companies was factored up to provide an estimate for all taxpayers, taking into consideration that prepayments up to a period of 12 months will be treated on a cash basis for businesses with a turnover below $1 million. The aggregate value of prepayments was adjusted to reflect the fact that some prepayments would currently be taxable in the hands of the recipient at the time of receipt. In those cases, an asset and liability treatment for prepayments would not result in a net impact on taxation revenue, although the timing of tax paid by both parties would be affected.
189 The estimates are sensitive to the adjustment factors used.
190 The growth dividend reflects the impact on Commonwealth revenue arising from the expected impact of the Review's recommendations on economic growth.
191 As discussed in the Overview, the Review's recommendations are considered likely to result in an increase in national income of around three quarters of one per cent of gross domestic product in the long run. The increase in income was assumed to emerge slowly but at an accelerating pace over a 10 year period. This reflects the time it would take for the allocation of resources within the economy, particularly fixed capital, to adjust fully to the altered incentives arising from such reform.
192 The revenue from the growth dividend was calculated by multiplying projected Commonwealth revenue by the assumed cumulative increase in national income. Only revenue accruing to the Commonwealth government is included in the growth dividend. It was assumed that the associated gain to State revenues would be retained by the States. The Review believes that this estimate is at the low end of the range if all of its recommendations are implemented.
193 In conservatively estimating a delayed emergence of the growth dividend, the Review has been conscious of the paramount importance of underwriting the fiscal integrity of the estimates for its reform package.
194 Major reform of the taxation of capital gains includes
freezing of indexation at the September 1999 level for all taxpayers;
cessation of averaging for individuals; and
allowing inclusion in income of a portion of capital gains of:
195 Where the above proportions are adopted, the gain is calculated against the original cost without taking account of indexation.
196 Data limitations have been a significant problem in estimating the impact of the proposed reforms to capital gains tax. Under the self-assessment system, taxpayers provide minimal information on tax returns. This has compliance cost benefits. But there is a downside for policy analysis. At present, a taxpayer returns little more than the net value of capital gains or losses. There is no information on indexation applied, the underlying value of the assets, the mix of losses and gains, or the period for which the assets were held.
197 The lack of historical information has required assumptions to be made about the value, ownership and composition of the asset stock subject to capital gains tax. These assumptions have been informed by judgments of the Review and available partial information.
198 As an explanatory tool, the Review Secretariat also developed a simple numerical model of the asset stock held by individuals (see box headed `An asset population model').
199 Capital gains tax collections have proven very volatile, particularly in recent years. Revenue appears to be strongly influenced by movements and activity on the ASX, and in other asset markets. In order to take account of this volatility, the estimates in this report are based on a five year moving average of capital gains tax receipts. Though this approach is likely to smooth out irregular bumps in the series, it runs the danger of disguising secular trends and for a relatively `young' tax like the CGT which is growing from a grandfathered base, that risk is significant.
200 For example, preliminary data for the 1997-98 income year indicate strong growth in receipts over the prior year. In part, this is likely to be a bubble caused by the stock market activity that occurred around the Asian financial instability of 1997 and 1998. But there is no way of knowing whether there was also strong growth in the base because of earlier turnover of pre-CGT assets (those held prior to introduction of CGT in 1985) which are now being turned over for a second time and adding to collections. There is also no way of knowing the extent to which realisations by superannuation funds (which have grown strongly in 1997-98) will be sustained because of demographic changes among fund members.
201 The Review's package has several components which all contribute to the total impact upon the taxpayer. In the case of individuals for example, the package has three sub-measures:
freezing of indexation;
removal of averaging; and
the availability of the option to include half of the nominal gain in taxable income instead of calculating the gain from the frozen indexed base.
202 The revenue impact of this combination of measures was estimated in two stages. First, the revenue impact of each component was estimated for the existing base of taxpayers (that is without any behavioural response) to give `static' estimates. The sum of these estimates gave an overall static impact on revenue. Separate static changes were estimated for individuals, superannuation funds and other entities.
203 A similar two-stage process was applied for the revenue estimates for superannuation funds. In that case the static stage applied only to the freezing of indexation and the option of including two thirds of nominal gains.
204 The tax base for each group was then varied in size to reflect the change in realisations behaviour that the Review expects to occur as a consequence of the change in average effective CGT rate.
205 The estimate for other entities included only the static component related to indexation. That static estimate related only to the asset population within entities which is to get no special treatment other than retaining indexation until the end of September 1999. Assets held by trusts other than CIVs at the time of announcement are to be allowed the optional treatment and the revenue impact arising from those assets is included in the individuals and superannuation funds components on the basis that the beneficial owners of those assets are largely individuals or superannuation funds.
206 An amount was also estimated for the impact upon income taxes arising from an expected tendency for some returns to investment to be taken as capital gains rather than as ordinary income. For example, there will be an increased incentive for shareholders to realise capital gains on shares rather than to receive the income as dividends.
207 Table 24.10 outlines the overall impact on CGT revenues and other taxes of the changes to individuals and superannuation fund CGT treatment.
Table 24.10 Revenue impact of CGT restructuring proposal
Note: Assets held by non-CIV trusts at the time of announcement will be allowed the optional capital gains treatment on realisation at the entity level.
208 The main assumptions underlying the estimates were:
future growth in tax base -- 6 per cent per annum long term after a period of growth at 8 per cent to 2002-03;
future inflation -- 2.5 per cent per annum; and
in large part, the assets held in non-CIV trusts prior to announcement will be replaced, when realised, not by assets also held in such trusts, but by assets in the hands of the members of the trusts.
209 The estimates for 2000-01 collection year (1999-2000 income year) are very dependent upon passage of the legislation by about the middle of the year. Were the legislation to be delayed further, there would be a considerable risk to revenue. The Review expects considerable deferral of asset realisations once the decision is announced but provided the legislation is enacted about the middle of the year, that deferral will be caught up within the year. Any greater delay in enactment would cause further deferral into the next income year (2000-01) when personal tax rates will be lower.
210 A number of submissions to the Review argued that revenue estimates ought to take account of an expected increase in realisations of assets if the tax rate on capital gains were reduced. Submissions referred to the experience of the United States in the 1970s and 1980s when capital gains taxes were altered significantly and there were large responses in capital gains tax revenue. A body of economic literature has developed around the issue of the elasticity of realisations to variations in tax rate.
211 By and large, the literature concludes that there is a significant elasticity -- particularly in the short term. The literature is more divided over how large the longer-term elasticity would be, but most authors conclude that there is evidence of some response, even in the longer term. A study commissioned by the Australian Stock Exchange and provided to the Review concluded that translating the US literature into the Australian context would suggest a longer term elasticity of more than minus 0.9.
212 The Review believes that a strong response effect ought to be expected in both the short and longer terms -- especially in the short term. In the first two years of the measure the Review has estimated that, on average, there will be an increase of around 50 per cent on the normal rate of realisations of gains as asset holders who face a lower tax penalty under the proposal realign their portfolios. The realisations profile adopted by the Review corresponds in approximate terms to an elasticity of about minus 1.7 in the first and second years after implementation. The implicit elasticity in the longer term declines to around minus 0.9.
213 The ASX study noted that elasticities are likely to decline as the marginal rate on capital gains declines. As superannuation funds face a relatively low marginal rate on capital gains already and since they will not see a large proportional change in their effective rate, the Review expects only a modest change in realisations behaviour from superannuation funds.
214 Other entities will see a gradual increase in taxation on capital gains after the pegging of cost base indexation at the September quarter level. However, there would be a reduction in tax rate for companies in the early years while the freezing of indexation would have only a progressive effect in raising effective tax rates. The balance of these two influences would probably be a reduction in effective rate in the early years but, the Review has not included any change in realisations behaviour amongst corporate entities.
215 In coming to its estimate of the revenue effects of allowing scrip-for-scrip rollover for takeovers by widely held entities, the Review took account of a submission from the Securities Institute of Australia. The Institute had commissioned Access Economics to produce estimates of the revenue effect of allowing scrip-for-scrip rollover in respect of takeovers by publicly listed companies only. Access Economics had estimated that on average there are around 14 takeovers annually in this category of average value $214 million. It posited, on the basis of the literature, that reducing the immediate tax impost to zero would increase the annual number of scrip-for-scrip takeovers by 70 per cent, though some of that increase would come from a conversion of previously cash takeovers to scrip-for-scrip.
216 The Review's approach to modelling the revenue impact of changeover to a scrip-for-scrip regime was similar to that of Access Economics and the Review has adopted the Access Economics estimate of the potential size of the ongoing takeover market for publicly listed companies. Analysis done on behalf of the Department of Communications, Information Technology and the Arts has also offered guidance on adjusting that estimate to take account of the market for takeovers involving other than public companies where one of the entities is widely held.
217 The Review has added 25 per cent to the estimate of public company to public company takeovers to account for other takeovers involving widely held entities.
218 The revenue impact of rollover relief depends upon three main parameters:
the extent of increased takeover activity induced by the change;
the average increase in company value that occurs because of the takeover; and
the extent of once-off portfolio realignment that occurs at the time of a takeover.
219 In the Review's judgment, there will be a very strong response in takeover activity when scrip-for-scrip rollover is available -- particularly in the first year. The Review estimates that in the first year there will be a 125 per cent increase in takeover activity and in subsequent years there will be an ongoing 70 per cent increase.
220 The average increase in company value used in the Review's estimates derives from the Access Economics parameters. Share values in current takeovers are assumed to increase on average by 31.5 per cent on account of takeover while under a more liberal regime the average uplift will be 5 percentage points less at 26.5 per cent. The latter is lower because the extra takeovers induced by the rollover would not have been commercially sound at a price uplift of 31.5 per cent and shareholders, without rollover, would require something like that to be compensated for their immediate CGT liability. Under the rollover regime the premium can be reduced to a commercially viable level and there would also be some reduction in the premium paid for takeovers that would have occurred without rollover since CGT compensation would not be needed in those cases either.
221 Once-off portfolio realignment in the wake of a takeover is assumed to be 50 per cent of shares. This reflects a judgment about the extent of portfolio realignment that would take place in the wake of a takeover where many shareholders are left with a higher weighting than desired of the acquiring entity's shares in their portfolios and at a time when the value of the shareholding is at what may be seen to be a peak.