B U S I N E S S C O U N C I L O F A U S T R A L I A
A recent survey of business expenditure on research and development (BERD) suggest that tax concessions are an excellent way of assisting R&D.
Some have argued that assistance to R&D should be confined to direct budgetary assistance such as the START grants scheme and/or some more general R&D bounty scheme. There is certainly a case for these schemes within the policy mix, but the BCA’s survey results suggest that tax concessions are an extremely effective policy tool in increasing BERD.
The 1996 changes to the tax concession – in which eligibility was restricted and the rate of assistance was reduced – provide a perfect means of testing the responsiveness of business to tax concessional R&D assistance. If the concession was having a major effect in encouraging increased BERD, then the changes would have a substantial effect on BERD. And if it was ineffectual, this too would be revealed – BERD would not be greatly influenced by a reduction in the tax concession.
The survey which captures a third of Australia’s total BERD suggests that, BERD fell by 3 per cent last financial year and another 5 per cent this financial year. Based on firms’ forecasts of future spending, R&D is still falling. There is evidence, though it is not unambiguous, that further declines will be large.
This means that total BERD is now about one third or $1.5 billion lower than it would have been if past growth had continued. Of course the cuts produce cost savings for both business and government. But over the long term they will cost the Australian economy something like $300 million each year.
Hardest hit was ‘non-core’ or more strategic and speculative R&D.
This study argues that:
Government agencies commission work to understand more fully the effects of the changes;
Greater security of funding be given to R&D assistance;
Reviews of R&D assistance which substantially reduce assistance or conditions of eligibility for assistance be conducted in public by independent and expert bodies such as the Productivity Commission;
Recommendations of the Mortimer Report and Industry Commission be implemented. Both endorsed assistance at or above the level provided by the 150% tax concession and called for this measure to be broadened to allow firms to receive equivalent assistance more directly as grants.
The tax concession is a very efficient – in fact probably ‘super-efficient’ – vehicle for encouraging BERD. That is, R&D decision makers tend to overestimate the benefits of the R&D tax concession to their firm’s shareholders.
More generally the results suggest that action needs to be taken to get BERD back into strong growth. We need action which is positive, strategic and integrated if we are to turn this situation around.
We need a national summit on R&D and innovation. It should focus on Australia's R&D effort in an integrated way and all sectors should be involved – business, government and educational/research agencies.
Australia should have – it must have – an innovation policy which is driven by an integrated and strategic approach to Australia's needs for knowledge into the future to harness the great talents and aspirations of Australians.
There is general agreement in the policy community that business expenditure on research and development (BERD) should be supported by government. This is because new know-how once it is created, benefits everyone. Because of this the firm which invested in developing it cannot capture the full benefits of what they have produced.
But how should BERD be supported? There has been considerable debate in recent years about whether assistance should be through direct outlay measures or through the tax system. Currently BERD is assisted through both mechanisms, with a discretionary grants scheme and a generally available tax concession (providing certain criteria is met).
Although supportive of assistance for BERD in principle, the Commonwealth Treasury is generally antagonistic towards doing this through the tax system. They have good reasons for these views. In principle, assistance should be transparent. Firms should know how much they are getting, and get the same level of assistance whatever their circumstances, and governments should know what this costs. Further support through a tax concession assists different companies in quite different ways depending on their circumstances (see below).
However there are also good reasons for assistance through the tax concession. Firstly, business has worked within this framework for a decade and so there is an argument that things which are working well should not be changed without a compelling reason. Secondly, there is the possibility that the tax concession actually induces managers to think that they are getting more assistance than they actually are.
Under dividend imputation, a good deal of the assistance available through the tax concession is ultimately ‘clawed back’ in the form of higher tax on shareholders. This is because for every dollar the firm saves in tax, it receives that many fewer franking credits to distribute to its shareholders (see below). If managers were fully taking this into account, it is likely that they would consider the tax concession to be of limited assistance.
The recent reduction in the rate of, and tightening of eligibility for the tax concession allows us to test this proposition. If managers were fully discounting the value of the tax concession, the effect of the removal of the concession would be relatively minor. If it is not worth a great deal, reducing it will be of little consequence. It may have been this thinking – the belief that the tax concession had low ‘inducement rates’ (see below) – which led to support for the removal or reduction in the concession.
In the five years to 1995-96 BERD increased each year by an average of 16% or a little over 13 per cent after adjusting for inflation. And BERD increased each year whether measured in terms of the real or nominal amount of money spend or as a proportion of GDP. As a proportion of GDP, R&D lifted sharply over the period, from just over .5% to just under .9%. In addition to the ‘carrot’ of increased R&D assistance, the ‘stick’ of increased competition within markets for goods, services and capital from micro-economic reform saw a dramatic transformation in Australian business’s R&D performance.
Chart 1 Gross Australian Business Expenditure on Research & Development (BERD)
Source: ABS: Research and Experimental Development: Business Enterprises, Australia, 1995-96,
ABS Catalogue No. 8104.0, p. 6.
See Appendix, Table 1
In the 1996 Budget the Commonwealth government narrowed the range of expenditures which received concessional tax treatment, and reduced the extent of the tax concession available from 150 to 125 per cent of eligible expenditure. With a company tax rate of 36% this reduces the immediate tax benefit of the concession from 18 per cent to 9 per cent.1
To date, what information we have on the effects of the changes has been generated by industry and has tended to focus on reduced claims for business research and development (Johnson, 1997). The results from such studies suggest that firms’ claims have fallen considerably.
However the extent to which reduced claims for the tax concession reflected falling total R&D spending has been unclear. Any reported falls in claims might reflect falling R&D spending, or simply the fact that a smaller proportion of firms’ R&D expenditure was claimable within the terms of the tax concession.
Against this background the BCA’s Business Development Task Force decided to embark upon the survey which is reported here. A major theme of the survey was to ensure that results were generated which were capable of distinguishing between total business expenditure on research and development and tax-concessional expenditure.
For this reason firms were asked for information about both total and ‘concessionable’ expenditure. Approximately 150 businesses were surveyed in early March 1998. Replies were received from 66 firms with a total turnover of $125 billion and total R&D expenditure of $1.55 billion. This represents about a third of all business R&D.
The firms cover a good cross-section of industry sectors, of domestic and foreign firm ownership, of rates of intra-firm R&D expenditure growth and intra-firm R&D intensity (proportion of R&D to turnover). The sample is clearly biased towards larger firms. However, there are also a few smaller firms in the sample and these permit a tentative analysis of the extent to which and the direction in which this might bias the overall result.
Research executives in each firm were asked to report their firm’s total R&D expenditure in 1995-96 and 1996-97 along with their anticipated expenditure in the current financial year (1997-98) and in 1998-99. They were also asked to nominate their tax concessional R&D expenditure. The questions replicated those in the regular ABS survey of these topics and firms were encouraged to quote from ABS surveys which they had either returned or which they were preparing.
The major conclusion is that business R&D, which was growing rapidly before 1996, appears to have declined substantially (by about 8 per cent) from the time of the policy changes to the end of the current financial year. The evidence suggests that R&D is continuing to fall and it is possible that a further major decline is in prospect.
Amongst the firms in the survey, BERD is now about one third or $1.5 billion lower than it would have been had it maintained its previous rate of growth of 13 per cent per annum since the policy change.
Firms indicated that they expected R&D expenditure to fall by 23% in the coming financial year. However it should be noted that firms have consistently underestimated their forward R&D expenditure in previous ABS surveys. Whilst their underestimates have been between 16 and 21 per cent – slightly smaller than the 23 per cent fall suggested in this survey – this survey asked firms to look out further than ABS surveys do. The survey’s forward estimates should therefore be treated with caution.