THE IMPACT OF THE 150 PER CENT TAX CONCESSION FOR INDUSTRIAL RESEARCH AND DEVELOPMENT IN AUSTRALIA — A PRELIMINARY ASSESSMENT

NOTES AND REFERENCES

1. The rationale for government intervention is based on correcting or offsetting those defects in the market mechanism which tend to push IR&D expenditures below the social optimum. For a useful summary of the main arguments, see Australian Science and Technology Council (ASTEC), Improving Australia's Competitiveness Through Industrial Research & Development, Canberra, AGPS, September 1987, pp. 79–88.

2. For elaboration, see Barry Bozeman and Albert N. Link, ‘Tax incentives for R&D: a critical evaluation’, Research Policy, 13, 1, 1984, pp. 21–31; Richard Slitor, ‘The tax treatment of research and innovation investment’, American Economic Review, 56, 1966, pp. 217–31.

3. D. McFetridge and J. Warda, Canadian R&D Incentives: Their Adequacy and Impact, Canadian Tax Foundation, Toronto, February 1983; Edwin Mansfield and Lorne Switzer, ‘The effects of R&D tax credits and allowances in Canada’, Research Policy, 14, 1985, pp. 97–107; Edwin Mansfield, ‘The R&D tax credit and other technology policy issues’, American Economic Review, Papers and Proceedings, 76, 2, 1986, pp. 190–93; Edwin Mansfield, ‘Public policy toward industrial innovation: an international study of direct tax incentives for R&D’ in R. Hayes et al. (eds), The Uneasy Alliance: Managing the Productivity — Technology Dilemma, Harvard Business School, Boston, 1985.

4. Australian Manufacturing Council, Future Directions for Australian Manufacturing Industry, AGPS, Canberra, 1986.

5. For a discussion of the implications of these statistics, see ASTEC, op. cit., pp. 11–17.

6. There is a strong positive correlation between the share in manufactured exports taken by high R&D intensity industries and the R&D intensity of the industries. This correlation was statistically significant when tested for eleven OECD countries for the period 1970–1980, OECD, Science and Technology Indicators, Paris, May 1985, II, Annex, Graph 3.

7. Technology-based industries are of strategic importance to international competitiveness. By generating products which can transform production structures, they can lead to cost savings and productivity improvements in many industries simultaneously. For elaboration, see ASTEC, op. cit., para. 1.12.

8. With respect to the proportion of exports by high intensity industries in total manufactured exports, Australia in 1981 ranked lowest of eleven OECD countries studied. Australia's figure of 3.7 per cent was well below that for other small economies such as Switzerland (23 per cent), Sweden (15 per cent), The Netherlands (13 per cent), Belgium (9 per cent), and also well below the OECD average (15 per cent). Australia was the only OECD country to register a decline in this percentage between 1975 and 1981.

9. In 1981 the coverage ratio for Australia was only 0.10, the lowest for fifteen OECD countries measured and much lower than the median figures for major, medium and small economies of 1.21, 0.47 and 0.64 respectively. The ratio of technological payments to business enterprise research and development (BERD) for the same year was 36 per cent, which was the third highest of thirteen countries measured and substantially above the OECD median of 19 per cent. Taking technological receipts as a percentage of BERD, Australia's figure of 3.5 per cent was the lowest of all the countries measured and well below the median figure of 9.9 per cent (OECD), Selected Science of Technology Indicators 1981–1985, Paris, October 1985, Table 48). Although Australia's technological balance of payments ratio rose to 0.22 in 1984–85 and 0.38 in 1986–87, these figures are still low compared with other countries.

10. OECD, op. cit., Table 4.

11. ABS statistics quoted in Scitech, ‘Fillip for Ind. R&D in Revised Figures’, January 1987, p. 9. In 1986–87 and 1987–88, BERD represented 0.46 per cent and 0.45 per cent of GDP respectively.

12. See, for example, Larry Dwyer and Terry Alchin, ‘R&D activity in Australia’, National Australia Bank Monthly Summary, January 1987; Donald F. Lewis and John Mangan, ‘Research and development in Australia: the role of multinational corporations’, Prometheus, 5, 2, 1987, pp. 368–85.

13. The definition of R&D covers such activities as industrial design; engineering design; production engineering; operations research (the improvement of the efficiency of organisations by techniques of numerical analysis); mathematical modelling and analysis; psychological research; computer software development (for sale, rent, license, hire or lease to multiple clients) but not routine computer programming or in-house software; and the design, construction and operation of prototypes. Specifically excluded are activities such as market research, market testing or market development; quality control; prospecting, exploring or drilling for minerals; the making of cosmetic modifications or stylistic changes to products, processes or production methods; management studies and efficiency surveys; research in social sciences, arts or humanities; pre-production activities such as planning or demonstration of commercial inability, tooling-up, trial and production runs. See Industry Research and Development Board, Explanatory Notes. Tax Concession for Research and Development, Department of Industry, Technology and Commerce, Canberra, Form 86/11.8.87.

14. An ASTEC report has claimed that “there has been a definite net increase in the level of IR&D over the last few years as a result of the 150% tax concession … it is expected that further benefits will be realized and expressed through expenditure statistics over the next four years”. ASTEC, 1987, op. cit. para. 2.19. According to a survey produced by the Centre for Technology and Social Change at the University of Wollongong called The 150% Tax Incentive for R&D: As Assessment of its Effects, the scheme has had considerable success. The survey covered 12 large computer and communications firms, 28 small innovative communications companies and 37 large R&D performing firms drawn from membership of the Australian Industry Research Group. The authors conclude that the scheme has not only encouraged an increase in R&D, but has led firms to re-examine their R&D strategies to link them more effectively into their business strategies. See D. Scott-Kemmis, T Darling and R. Johnston, ‘R&D tax concession has positive effect’, Laboratory News, June 1989. A recent study of the employment effects of the tax incentive, however, concludes that there is no clear evidence that the scheme has increased the number of R&D jobs in Australia. See B. Doube and C.D. Beaton, ‘Has the 150% tax incentive scheme created new jobs in R&D’, Search, 20, 3, 1989.

15. For an overview of the legislation, see G. Lehmann, ‘Research and development tax planning’, Australian Accountant, November 1986, pp. 32–6.

16. A. Gosling, ‘The 150% tax concession for research and development’, Australian Technology Magazine, October 1987, pp. 4–5.

17. ibid., p. 17.

18. ibid., p. 17. An alternative way of interpreting the B index is to regard it as the ratio of the after tax cost per dollar of R&D to the after tax cost per dollar of other expenses.

19. More complex expressions can be constructed taking into account the value of any specific allowances or credits for R&D in addition to tax concessions. See McFetridge and Warda, op. cit., p. 71.

20. In most countries current R&D expenditures can be immediately written off against income while capital expenditures must be depreciated, often at different rates. Consequently, tax savings on current R&D expenditure will be greater than these for eligible capital expenditures. In Australia, current expenditure on plant and equipment incurred on or after 1 July 1985 is deductible at the rate of 150 per cent over three years. This means that 50 per cent of the expenditure is deductible in the first year, 50 per cent in the next year and 50 per cent in the year after. A lesser deduction is allowable if the company's ‘aggregate R&D amount’ for any of these years is less than $50,000. Capital expenditure incurred on or after 1 July 1985 on buildings, extensions or improvements wholly attributable to R&D does not attract the 150 per cent deduction, but can be written off over three years.

21. “In general the earlier an expenditure may be written off against taxable income the greater the present value of deduction (hence of the tax saving involved) and the lower the net-of-tax cost of R&D”. McFetridge and Warda, op. cit., p. 12.

22. Table 2 assumes R&D capital expenditure is apportioned equally between equipment and buildings. In the absence of data on the proportions of current and capital expenditure in R&D spending after 1984–85, it is assumed these proportions remained unchanged. A further assumption involved selection of an interest rate in calculating the discounted present value of R&D capital expenditure. Since the interest rate on long term government securities has fluctuated between 10 per cent and 15 per cent during the decade, B indexes were calculated using both figures as discount rates. The B index appears to be not very sensitive to changes in the discount rate. Estimates of B index values were derived by applying the general formula. For example, in 1984–85 (prior to the tax concession), W1 = W2 = W3 = 1; a = .862, b = .069; t = .46. At a discount rate of 10 per cent, PVe = .91. Thus B = (1 – (.862 × .46) – (.069 × .46 × .91) – (0.69 × .46 × .91)/.54 = 1.01. In 1986–87, after introduction of the 150 per cent tax deduction for eligible R&D, W1 = W2 = 1.5, W3 = 1; a = .862, b = .069, c = .069; t = .49. At a discount rate of 15 per cent, PVe = .88, PVb = .88. Thus B = (1 – .862 × 1.5 × .49) – (.069 × 1.5 × .49 × .88) – (.069 × .49 × .88)/.51 = .573.

23. B indexes for 20 countries in 1981 were estimated by McFetridge and Warda, op. cit., Table 5.4 for discount rates of 10 per cent and 15 per cent. Their calculation of a B index for Australia was based on data for 1976–77. Some other estimates of B indexes for selected countries, assuming a 10 per cent discount rate, are as follows: Japan .97, Hong Kong 1.0, South Korea 1.01, Taiwan 1.02, Singapore .41, UK 1.0, US .95, Canada .84, Belgium .97.

24. Top of the list, with a B index of .41, was Singapore because of a 200 per cent deduction on current R&D expenditures.

25. Mansfield, 1986, op. cit. While the Australian government has made considerable effort to define more clearly eligible R&D, the extent to which ‘creative accounting’ has redefined activities in this country remains uncertain.

26. cf. R. E. Seiler, Improving the Effectiveness of Research and Development, McGraw-Hill, New York, 1965; A.M. McCosh and R.D. Maluste, ‘Simulating the forces impinging upon the research budget and organisations of a major firm — a test of a possible method’, R&D Management, 2, 1, 1981, pp. 9–18. Seiler, for example, proposes eight determinants of a company's R&D expenditure — the firm's long term financial policy, stockholder pressure for high current profits, length of life cycle of the firm's products, backlog of research ideas needing exploration, competitors' R&D efforts, availability of resources in the shorter term, capacity to produce or market the results of research effort, and stability of research staff. The links between certain of these determinants and objective financial considerations are quite tenuous.

27. Chor Hin Ong, A.W. Pearson and A. Wilkinson, The Role of Financial Subsidy and Other Government Incentives in Industrial R&D and Innovation: A Survey of the Literature (with Case Studies), a report prepared for the Department of Trade and Industry, R&D Research Unit, Manchester Business School, April 1986. As Bozeman and Link have noted, “Given the various determinants of R&D there is reason to believe that response to R&D incentives will vary considerably according to industry, firm size, current level of R&D intensity, profitability, competitiveness of markets, strategic plans and organisation composition”. B. Bozeman and A.N. Link, Investments in Technology. Corporate Strategies and Public Policy Alternatives, Praeger, New York, 1983, p. 97.

28. Overseas studies show that the use of formal R&D project selection techniques is not as widespread as one might have expected. See J.C. Higgins and K.M. Watts, ‘Some perspectives on the use of management science techniques in R&D management’, R&D Management, 16, 6, 1986. For a perspective on the quality of decision-making by the managers of Australian industry see Larry Dwyer, ‘R&D project assessment as an information and communication process’, Prometheus, 6, 1, 1988, p. 78–93.

29. M. Ishaq Nadiri, ‘Contributions and determinants of research and development expenditures in the US Manufacturing Industries’ in George M. Van Furstenberd (ed.), Capital, Efficiency and Growth, Ballinger, Cambridge, Massachusetts, 1980, pp. 386–8. P. Mohnen, M.I. Nadiri and I. Prucha, ‘R&D production structure and productivity growth in the US, Japanese and German manufacturing sectors: a non-separable dynamic factor demand model’, European Economic Review, 30, 4, 1986, pp. 749–71.

30. Mansfield, 1986, op. cit., p. 191. Commenting on the low elasticity of demand for R&D as revealed by empirical studies, Mansfield and Switzer state “In our interviews many R&D executives said it was low because R&D is named as one component (and often a small one) in a package of inputs that must be considered in relatively fixed proportions to launch an innovation”. Mansfield and Switzer, op. cit., p. 102.

31. Mansfield and Switzer also state that in Canada in 1980 around one-third of the IR&D was performed by firms that did not have sufficient taxable income to use any of their R&D tax credits, ibid.

32. Mansfield, 1986, op. cit., p. 190.

33. The figures for government revenue foregone from the concession come from Taxation Expenditure Statement, AGPS, Canberra, December 1988, p. 23.

34. The required revitallisation of Australian industry depends crucially on its ability to produce quality products which are valued on world markets. For arguments that Australian industry should devote more resources to product innovation rather than process innovation, see Dwyer and Alchin, op. cit.

35. Edwin Mansfield, John Rapoport, Anthony Romeo, Edmond Villari, Samuel Wagner and Frank Husic, The Production and Applications of New Industrial Technology, Norton, New York, 1977.

36. J. Button, ‘R&D incentives were temporary catalysts’, Sydney Morning Herald, 8 June 1988.

37. cf. Barry Bozeman and Albert Linka, ‘Public support for private R&D: the case of the research tax credit’, Journal of Policy Analysis and Management, 4, 3, 1985, pp. 370–82. McFetridge and Warda, op. cit., p. 42–4.

38. In 1985–86, non-dwelling construction accounted for 23 per cent of the net capital stock of private enterprises, with other capital expenditure accounting for the remaining 77 per cent. According to the Australian Bureau of Statistics, “These proportions have been fairly constant over the 20 years covered by these statistics”. Australian Bureau of Statistics, Australian National Accounts. Estimates of Capital Stock 1985–1986, cat. no. 5221.0.

39. It is assumed that the representative firm writes off machinery and equipment over five years at 20 per cent per annum and buildings and structures over forty years at 2½ per cent per annum. Using a discount rate of 10 per cent, the present value of the capital cost allowance on a one dollar expenditure on machinery and equipment is .834, while that on buildings and structures is .270.

40. McFetridge and Warda, op. cit., pp. 84–7.

41. Australian Manufacturing Council, op. cit., p.