David Bushby (Tasmania, Liberal Party) Share this | Hansard source

I too rise to contribute to the debate on the Tax Laws Amendment (Research and Development) Bill 2010 and the Income Tax Rates Amendment (Research and Development) Bill 2010. The coalition has always been strongly supportive of increased business investment in research and development, or R&D, and has therefore considered it highly appropriate to have in place a taxation regime that provides strong incentives for Australian businesses to invest their hard-earned dollars to help achieve this outcome.

We also accept the general principle that the current law, as with all laws, may potentially benefit, in terms of achieving increased R&D investment, through a timely revision of its operation and that this could conceivably be achieved through a higher R&D rate for a greater number of Australian businesses. However, sadly, the changes to the R&D taxation regime presented to this place by the government today, in these bills, fails dismally in terms of achieving these desirable investment outcomes. This is because there are many shortcomings in the bills that need to be addressed before investment is likely to follow.

I was fortunate enough to have participated in the Senate Economics Legislation Committee inquiry into these bills—which occurred well over a year and a quarter ago—and the evidence was overwhelming at those hearings that these bills are flawed and should be rejected. Indeed, when a group jointly appears representing all of unions, academia and business, telling the committee that the government has got it wrong, I believe it is time to sit up and take a look, because, when groups of such divergent interests come together to fight something, they may well have a point.

So let me take the opportunity to have a look at some of the issues raised. I think there was universal condemnation of the degree of and quality of the consultation associated with the legislation. Consultation time lines were highly condensed by the government, rendering it difficult for affected stakeholders to fully understand the impact and consequences of the proposed changes, which are complex, before the then final versions of the bills were introduced in the first half of last year. The second exposure draft was released by Treasury on 31 March last year, with submissions from stakeholders due by 19 April last year, a total of only 10 working days. This was not sufficient time to digest 134 pages of legislation and provide substantive com­ments, particularly given that the second exposure draft incorporated a number of entirely new concepts, including a completely new definition for core R&D.

Let me repeat that, because it was a major matter of concern by affected stakeholders. We are talking about what constitutes core R&D, a matter completely central to the whole purpose of taxation deductions for R&D activity, and a completely new and substantially different definition was first posed in the second exposure draft with only 10 working days provided to allow those affected to consider this and many other changes. In addition, the tight time frames meant that Treasury had not completed its redrafting in time for the 31 March release. As a result, the second exposure draft did not include redrafted feedstock provisions, and instead merely stated that 'a feedstock adjustment rule is under consideration'. Stakeholders were not provided with an opportunity to comment on this aspect of the legislation until after it was introduced into parliament, which is an unfortunate and disappointing outcome.

So the government gave itself an absurdly short timetable for community consultation and examination by the parliament of what is a fundamentally new approach to the definition of eligible expenditure. Of course, despite its claims then of high priority, the government has failed to bring these bills on for debate for more than 12 months. Incredibly, it has failed to use this time to further consult and fix the problems inherent in them. The bills before us today are almost completely the same product of that very poor consultation process that occurred last year.

The Gillard government has been hell-bent on commencing this legislation retrospectively, first of all on 1 July 2010 and now, under great duress, on 1 July 2011. This has caused huge confusion for industry and business, who for the last 18 months have been unsure as to how they should undertake R&D activity and, accordingly and logically, have done less than they would have. In other words, R&D has been scaled back. Over and above the consultation issues, which are something that seems, certainly in respect of the bills that come before the Economics Legislation Com­mittee, a common trait with this government, there are a number of issues in relation to the provisions in the bills themselves, issues which might well have been sorted out if consultation had been properly conducted in a realistic and timely manner or if the government had taken advantage of the past 15 or 16 months to listen. These issues are of such importance that many of the submitters, including Michael Johnson Associates, a specialist R&D firm that works solely in the space of assisting businesses with their R&D activities and the taxable nature of them, argued that the package is in such poor shape that it should not proceed to law in its proposed form. This was widely reflected in the committee hearings with, again, just about all submitters calling for a delay in its implementation or its rejection outright.

As mentioned, the lack of consultation on major shifts in policy is a reflection of the overall approach of the current Labor government, which continues to rush legislation into parliament without due consultation and consideration. Further, the committee was not given sufficient time to consider the bills, another trademark of this government's approach to legislation and policymaking.

An additional consequence of the hasty development of the bills is numerous drafting errors, as well as inconsistencies between the then explanatory memorandum and the bills, as identified in several submissions at the time. Like the lack of consultation and rushed introduction of many of the government's bills, the regular drafting errors that keep arising in legislation under this government is becoming a worrying trend, although it is hardly a surprising outcome given there is often such limited time between drafting and introduction.

As mentioned, the bill substantially alters the definitions of 'core' and 'supporting' R&D. By narrowing the definition of what constitutes genuine R&D, the bills will disqualify from assistance many forms of R&D currently undertaken by Australian businesses for which they attract the existing concessional tax treatment. In turn, the overwhelming expectation of those groups who have lodged submissions on the exposure drafts is that the government's changes will reduce the number of firms qualified for the concession. The evidence suggests that there will be particularly grave consequences for firms focused on industrial R&D and other non-lab/whitecoat activities, including those involved in manufacturing, prototyping and process development.

The problem with the new definition of R&D is essentially that it is poorly aligned with the Frascati manual definition, as had previously been utilised by Treasury. In fact, the eligibility of R&D activities that would fall under the third limb of the Frascati definition—experimental development—is in real doubt under these bills. The bills introduce a much narrower definition of a core activity—essentially being experimental work, unknown outcomes and new knowledge—and then provide for a range of choices as to what a supporting activity might be. These provisions are strongly indicative that anything in a production environment is in danger of not being eligible for the R&D tax concession. Essentially, the provisions will render ineligible the R&D activities of a lot of companies, activities they consider to be core, particularly if a lot of what they do in R&D is in a production context. The provisions thereby knock off the development side of research and development, leaving only research as eligible.

The definition of R&D in the Frascati model, as developed under the auspices of the OECD, is:

1) Research and experimental development (R&D) comprise creative work undertaken on a systematic basis in order to increase the stock of knowledge, including knowledge of man, culture and society, and the use of this stock of knowledge to devise new applications.

This is a highly recognised and accepted international definition. It is completely unclear why the new and vastly more problematic definitions—first included in these bills in April 2010—have been adopted.For around 25 years, our R&D tax incentive has been based on the Frascati model, which, as mentioned, was developed under the auspices of the OECD. The second part of the Frascati definition, 'the use of this stock of knowledge to devise new applications', is central to my objections to the new approach proposed by the government. The object clause of the bill states that the object is to 'encourage industry to conduct research and development activities' by:

… providing a tax incentive for industry to conduct, in a scientific way, experimental activities for the purpose of generating new knowledge or information in either a general or applied form …

Critically, this clause omits the second critical element of the Frascati approach—that is, the use of this knowledge to devise new applications. The new definition of core R&D also precludes the critical second element of the Frascati definition. As such, the new definition of core R&D will require taxpayers, in order to claim the concession, to seek new, previously unknown or undiscovered information and carry out scientific experimentation to uncover that new knowledge. Claimants will need to prove in a retrospective assessment that the knowledge did not exist anywhere else, which will create additional administrative and operational burdens. This creates an innovation system which does not encourage industry to pursue innovation and the development of processes and products. As mentioned, the approach outlined in the bill leaves little room for the majority of what business R&D is actually about—what, in the Frascati model, is called 'experimental development'. Experimental development is defined as:

… systematic work, drawing on existing knowledge gained from research and/or practical experience, which is directed to producing new materials, products or devices, to installing new processes, systems and services, or to improving substantially those already produced or installed.

All evidence from stakeholders supports the view that the proposed definition of core R&D is inappropriate and will add additional red tape issues. For example, the requirement that supporting R&D activities must be either directly related to or be conducted for the dominant purpose of supporting core R&D activities will add significantly to the compliance burden. This requirement splits off the development of new knowledge from the development of new or improved product processes, devices, materials and services. If you had a manufacturing process where you were trying to develop a new process, the first 30 per cent of your R&D spend might be on the creation of new knowledge but the remaining 70 per cent would be on the development of a new process incorporating that new knowledge—the development part of R&D. Unfortunately, it is that 70 per cent that will likely lose its eligibility for favourable tax treatment under this legislation. Similarly, the new definition of 'supporting' R&D is:

(1) Supporting R&D activities are activities directly related to core R&D activities.

(2) However, if an activity:

  (a) is an activity referred to in subsection 355-25(2); or

  (b) produces goods or services; or

  (c) is directly related to producing goods or services;

the activity is a supporting R&D activity only if it is undertaken for the dominant purpose of supporting core R&D activities.

This is of considerable concern to companies involved in R&D, which I remind the Senate is a desirable outcome and is ostensibly the reason parliament provides tax incentives for spending in this area in the first place. By redefining supporting activities as proposed, the incentives provided will distort spending in R&D away from industrial R&D programs to laboratory style programs.

Evidence was received from Caltex about how they conduct industrial R&D on a commercial scale, seeking to deliver continuous improvements to processes and outputs, particularly at their Australian refineries. To not do so would render them uncompetitive against their international competitors—bearing in mind that petroleum products are highly mobile. Caltex's research is, by necessity, trialled in live conditions rather than laboratory or theoretical conditions, and they hold grave concerns about the economics of conducting this research if they are unable to access the incentives. The net result would be the failure to adapt, develop and remain competitive. This would lead to less fuel refined in Australia and more imported, with consequential impacts on jobs and fuel security. The fact is that this inexplicable proposed change will fundamentally alter the whole concept of supporting and inducing additional R&D in Australia.

The Australian Industry Group pointed out two major concerns with the definition of supporting R&D. The first of these is that supporting R&D expenditure is only eligible if it is either 'directly related to' core R&D activities or if it is 'undertaken for the dominant purpose of supporting' core R&D activities. This means that businesses would have to be undertaking core R&D before any of their R&D expenditure could qualify as supporting R&D. If a business had no expenditure that qualified as core R&D, it would have no eligible R&D expenditure whatsoever.

The second main issue that AiG raised is that a lot of experimental development is neglected by the dominant purpose test in the definition of supporting R&D activities. In section 355-30 of the bill, a supporting R&D activity is defined as an activity directly related to core R&D activities except if it is an activity that is explicitly excluded, an activity that 'produces goods or services' or an activity that 'is directly related to producing goods or services'. In any of these cases the expenditure needs to be undertaken for 'the dominant purpose of supporting core R&D activities'. As was noted at the hearings:

… it is difficult to think of many supporting activities that don’t fall into one of the three dominant purpose categories given that any activity directly related to production is captured.

The reality is that there are simply too many ways that supporting R&D activities will be excluded from eligibility for the proposed new R&D incentive for business to have any confidence that experimental development will continue to attract a tax incentive. I am of the opinion that the government has erred greatly by proposing the definitions it has and that the definitions of core and supporting R&D should be reconsidered to be more closely aligned to the Frascati model of R&D before these bills can be supported.

But the poor policy decisions reflected in these bills do not stop there. The government has also introduced sweeping changes to the eligibility requirements—changes which again threaten to significantly erode support for R&D investment in Australia. Again, they are also fundamentally inconsistent with the government's stated intent of making R&D tax support arrangements simpler, more predictable and more generous. Instead, they impose a series of barriers upon firms rather than offering encouragement for innovation. Under this proposed legislation, there will be a significant amount of additional planning required by companies so that they can reassess their eligibility under the new definition, on the one hand, and also to predetermine, perhaps throughout the annual life of a project, what activities will now be core and what will be support. This clearly will impose a significant new and challenging burden in predetermining eligibility.

Further complicating matters is a lack of clarity over what an essential aspect of this proposed legislation actually means—that is, the use of the term 'dominant purpose'. The bills define a range of activities as not being core activities. If a company has an activity that comes under that list of exclusions then it will have to jump over another hurdle. They will then have to show that this activity is for the dominant purpose of supporting a core activity. This leads to the need for a company to be able to define what its core activities and its supporting activities are and to justify the fact that an activity has the dominant purpose of supporting a core activity. This will inevitably prove to be relatively complex and deliver uncertainty to companies planning their R&D activities. Also, companies are required to identify and categorise their activities upfront when they are registering their R&D activity to qualify for the tax credit, which imposes a high compliance burden on all businesses.

As noted at the hearing, in a manu­facturing setting a company may come up with a process concept or a new product concept, but in just about all cases it will need to be tested in real life or through a scale-up version. Scale-up is a real challenge to successfully commercialising R&D, so it is important that this stage is recognised in these bills before they can be supported. Evidence also noted that in manufacturing, typically, there will be a batch run of a new concept or product and there will be feedback R&D. Invariably, the first trial will not be the final product. Feedback R&D highlighting shortcomings and failures within the system gives the R&D team the knowledge to further improve and create the product or process they are seeking. It is clear that in a manufacturing setting this would be a common occurrence, and the proposals before us totally fail to recognise this.

As it stands, under the current definition of R&D activities, all activities qualify under the 'systematic, investigative and experi­mental' test—SIE test—or the 'directly related' test. No distinction is made. Under the proposed legislation, the taxpayer needs to split activities into core or supporting and then establish which of the four tests the supporting activities apply to. These decisions will be based on the overall circumstances of the activities and little if any guidance is available yet for companies to assess these issues. The overwhelming evidence is that a 'dominant purpose' test will exclude a large proportion of the production trial activity that is a necessary and legitimate part of the research and development cycle.

If the government's aim is to contain the cost to revenue associated with large and open-ended production trials, the introduction of a cap on the total value of the group's R&D claim would better achieve this objective, whilst also providing clarity and simplicity for claimants. Of course, the government claims that the changes will be revenue neutral. However, the evidence suggests that eligible activity will be greatly curtailed with the end result that there must be savings for government revenue. The bottom line is that the dominant purpose test should be removed as it imposes an unnecessarily high threshold and does not target those few excessive claims which the government purports are occurring. There are better ways to address these, such as a 'purpose directly related to' test, substantial purpose, apportionment of expenditure, dollar capping the extent of production trials on the total value of the R&D claim for companies with group annual revenue exceeding $1 billion or on eligible R&D expenditure, more sympathetic language, specific provisions for specific excesses, time limits for trials, or pre-approvals for projects above certain values.

Further, the object clause needs to be revised in respect of spillover and additionality benefits. The objects clause of the draft legislation is too narrow and restrictive and implicitly or explicitly accords greater emphasis to research rather than development. It will change the emphasis that has been in the objects clause one way or another since the inception of the R&D tax concession in the mid 1980s. The emphasis that has always been central to the objectives of this incentive has focused on increasing investment in both R&D in Australia and helping Australian industry become more internationally competitive, export oriented and innovative. But, as noted at the hearings, in a sense this restricts the eligible research and development to those circumstances where a company could perhaps be asked: would you not have done this without the credit? That is actually not a very sensible position, because the credit should just be a cost-planning issue in a matrix where you make a decision about whether to do the work or not.

The narrow coverage of the objects clause suggests to us that the government intends to pare back the role of the R&D tax incentive to fund, almost exclusively, research. It does not intend to include much of what business R&D is about—namely, the development of existing knowledge to devise new appli­cations. Instead one can only conclude that the government intends that the R&D tax incentive will apply only to activities conducted for the purpose of producing new knowledge. It would be more honest to refer to it as a research tax credit. (Time expired)