The Business Tax Working Group's terms of reference ask it to provide specific analysis of business tax reform options that relieve the taxation on new investment, including impacts on national income. To achieve this, the Working Group requested Treasury undertake modelling to provide an assessment of the potential long-run economic impacts of company tax reform.
The modelling uses a computable general equilibrium approach. Computable general equilibrium models mathematically represent how the economy operates and how the behaviour of firms and households change in response to incentives. They are useful for exploring the economic impacts of business tax because they trace through linkages between factor markets and product markets to allow an estimate of where the costs of the tax system are borne, which can be very different from where taxes are levied. While these models have their limitations, they provide an integrated framework for analysis, based in economic theory and using the best available economic and taxation data.
The analysis is undertaken using the Independent CGE Model, Independent Economics' computable general equilibrium model. It is a comparative static model of the Australian economy. Treasury and Independent Economics worked together to extend and calibrate the model to make it suitable for modelling the business tax system.20 The model has been designed to represent economic effects of the company tax system on: the size of the capital stock in each industry; the mix of capital types; labour force participation; the location of multinational profits; and the location of multinational firm-specific assets, such as intellectual property.
The modelling suggests that a company tax cut from 30 to 29 per cent would increase the level of GDP by 0.2 per cent compared with what would otherwise be the case. This increase in GDP is driven mainly by greater foreign investment flows into Australia to fund additional projects that are made viable by the reduction in the tax rate. Under reasonable assumptions in the model, additional capital investment increases the capital stock by 0.3 per cent.
The modelling also suggests that Australian workers benefit from the company tax cut in the long run. The productivity of labour increases with the increase in the size of the capital stock and this flows through to an increase in after-tax real wages of 0.2 per cent and a small increase in labour supply of around 0.1 per cent. Overall, the modelling shows that cutting the company tax rate can deliver an improvement in consumption by Australian households of around 0.05 per cent.21
As noted above, the Working Group considers that a cut in the company tax rate of two to three percentage points would be needed to drive a significant investment response. The impact of a cut of two (or three) percentage points would be expected to be slightly less than double (or triple) the impact from a one percentage point cut. . The modelling results represent long-run changes to the economy. They provide an analysis of the change in the economy from now to a time in the future when capital and labour markets have fully adjusted to policy changes. A reasonable working assumption may be that half of the change in the economy will occur within approximately seven years, and the adjustment will be largely complete within 20 years.22
Key features of the model include the following.
- Up-to-date database. The model is designed to represent the 2011-12 Australian economy, based on an uprated version of the Australian Bureau of Statistics' 2007-08 input-output tables. The model is calibrated based on the 2007-08 level of the terms of trade.
- Rich industry detail. The model distinguishes 111 industries.
- Sophisticated production processes. Output in the model is produced by labour, land, fixed factors and nine additional types of capital: transport equipment; machinery; information technology; structures; dwellings; transfer costs; mineral exploration; research; and other intellectual property.
- Company tax system. The model reflects many features of the company tax system, including: deductibility of debt; revenue clawback through dividend imputation; depreciation allowances that reflect an historical cost basis and other aspects of tax laws; expensing of certain investments; and foreign tax credit arrangements.
- Fixed factors. The model identifies fixed factors in industries in which Australian Bureau of Statistics data suggest there may be above normal rates of return on capital and where there are economic grounds for believing these may be sustainable: mining, banking and finance, telecommunications, and beverage manufacturing. These are further divided into location-specific, immobile factors (90 per cent) and firm-specific, mobile factors (10 per cent).23
- Foreign marginal investor. The model assumes that the marginal investment is funded by a foreigner, that capital is perfectly mobile between countries, and that investments at the margin are funded through a mix of debt and equity that matches the historical average.
- Profit shifting. Company tax can affect where firms declare their profits for tax purposes. The model assumes a semi-elasticity of the tax base to the statutory tax rate of -0.5. This affects both the revenue take and the firm's cost of capital.24
- Labour force participation. Households choose between employment and leisure, taking account of the after-tax wage available.25
The model also relies on a range of general assumptions, many of which are shared with other long-run computable general equilibrium models. Consumers choose between different purchases to maximise their wellbeing. Firms choose how to produce and how much to produce in order to maximise profits. Wages adjust so that labour markets clear. The capital stock adjusts so that the after-tax rate of return matches the required world rate. Australia is a price taker in import markets and is close to a price taker in most export markets.26
The main welfare measure is full household consumption. This takes account of household consumption and leisure.
Modelling exercises are always a simplification of the real world. They are designed to capture the most important features of the economic response to policy changes in a sufficiently flexible way. Not all features of the decisions affected by tax changes are incorporated. In particular, while the modelling takes account of the historical shares of corporations and unincorporated entities in the economy, and the historical shares of debt and equity financing, it does not model potential changes in the legal structure of business operations or leverage in response to policy changes.
References
Dandie, S. and Mercante, J. 2007, Australian labour supply elasticities: comparison and critical review, Treasury working paper, 2007-04.
De Mooij, R. and Devereux, M. 2011, An applied analysis of ACE and CBIT reforms in the EU, International Tax and Public Finance, vol 18, pp 93–120.
20 Independent Economics designed the overall economic structure of the model; Treasury calibrated the model to match the business tax data and provided a range of parameters, such as the profit shifting elasticity and the share of rents that are firm-specific.
21 Consumption in this context is a measure of the volume of consumption of goods and services as well as leisure time enjoyed by Australian households.
22 This timeline is consistent with the transition path in response to tax changes within the Monash Multi-Regional Forecasting Model, a widely used dynamic model of the Australian economy.
23 Actual rates of return have been estimated for each industry using data on capital stocks and net operating surplus from the Australian Bureau of Statistics. An industry is said to be earning economic rents, attributed to fixed factors, when their estimated actual rate of return is higher than a 'normal' rate for that industry. In modelling the European business tax system, De Mooij and Devereux (2010) assume that 30 per cent of fixed factors are firm-specific. The model assumes a lower share, reflecting the greater role of natural resource wealth in the Australian economy and its lower level of international economic integration.
24 Recent modelling of European business tax (de Mooij and Devereux 2010) assumed a semi-elasticity of -0.73. The model assumes a lower elasticity reflecting the smaller role of multinationals in the Australian economy.
25 The assumed uncompensated elasticity of labour supply of 0.2 is consistent with the range of empirical estimates from Australian studies (Dandie and Mercante 2007).
26 An elasticity of demand of -12 for Australian exports is assumed for most industries, but a lower elasticity of -6 applies in industries where Australia has some market power (some parts of agriculture and mineral commodities) or product differentiation (such as tourism and education).