Definitions and conceptual issues

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

9. There are two fundamental and related issues facing economists and national accountants with respect to measures of research and development. First, what is the content and scope of R&D; and second, at what point does R&D become an asset in the economy. To initiate this discussion, it is useful to consider the description of an asset.

Definitional issues

10. SNA93 defines an asset as entities functioning as a store of value: a) over which ownership rights are enforced individually or collectively and b) from which economic benefits may be derived by the holder or user over a period of time. The characteristics of R&D seem to fit this definition closely in that it acts as a store of value which can be purchased and sold. Ownership rights can be enforced, although at times, it can be provided freely to the public. R&D also acts as an asset in that benefits may derive from it over a period of time.

11. There have been competing visions of, and some degree of controversy with respect to, how R&D and its sub-components should be defined in the national accounts. The most common starting points for these discussions come from SNA93 and the Organization of Economic Cooperation and Development's (OECD) Frascati Manual3 (FM).

SNA93 definition

12. SNA93, which has been the standard for the national accounts since the early 1990s, focuses on R&D with respect to in its role in production. It refers to research and development as "an activity undertaken for the purpose of discovering or developing new products, including improved versions of qualities of existing products, or discovering new or more efficient processes of production".4 It goes on to state that R&D is "undertaken with the objective of improving efficiency or productivity or deriving other future benefits." 5 This is a broad definition in that it could include both the efforts involved in the initial design of a product as well as new production and marketing costs. This concept more closely matches that of innovation. It points to a goal of improving efficiency, productivity or other future benefits. However the SNA93 definition would seem to exclude research done for non-commercial purposes or with no immediate productive use. Further, SNA93 stopped short of including it in the asset boundary and therefore R&D assets remain excluded from most countries' national accounts' statistics.

The Frascati definition and SNA93 Rev1

13. Upon the recommendation of the Canberra Group on the Measurement of Non-financial Assets, the UNSC has decided that the definition set forward by the Frascati Manual (FM) be used as the updated SNA description of R&D in the national accounts. This definition states that:

"Research and experimental development (R&D) comprise creative work undertaken on a systematic basis in order to increase the stock of knowledge, including the knowledge of man, culture and society, and the use of this stock of knowledge to devise new applications."6

14. Research is further classified by the FM into three types. Basic research is work done without any direct application immediately in sight. This work may or may not be patented or used for commercial purposes at a later date. Research done in universities or by government may often be of this nature. Applied research is original work but done for a specific purpose. Because of its focus, this work often ultimately contributes to production. Experimental development is work that draws on previous or existing experience to produce new systems or products or to substantially improve existing processes or products.

15. In some ways the FM definition is broader than the original SNA definition in that all basic research is included. This would include research that focused on scientific novelty with an aim to furthering knowledge. This R&D may be freely available to the public and may have no enforcement of ownership rights.

16. Conversely, the FM view is narrower than the SNA93 in that costs associated with new production (e.g., manufacturing scale up and new marketing initiatives) are not included. Yet the overlap between the two definitions is significant.

17. Notably, the SNA93 Rev1 states that R&D does not extend to including human capital as assets within the System.7 The guidelines recommend that R&D that provides no economic benefit should be excluded from R&D assets, thereby departing from the FM definition. Since much of R&D is produced on own-account (remains within the organization that produced it), it goes on to suggest that R&D could be valued at cost.

The scope of research and development

18. The scope of R&D is a key issue when considering capital formation since the aggregate value of R&D varies considerably depending on where the R&D boundary is drawn.8

19. SNA93 Rev1 initially suggested that freely available R&D not be included as an asset. However, it went on to state that as a result of measurement issues, and the fact that it usually does not comprise a large portion of R&D expenditure, it could be included with little impact on the total. The reasoning for the exclusion of this research was that there was no guarantee of exclusive benefit of this information. The research, once completed, provided no market advantage and therefore was not an asset for any sector.

Research and development definition summary

SNA93 – production related

FM – three types of R&D (basic, applied and experimental development)

SNA Rev1 – adopted FM definition

20. In this study, freely available R&D is included as an asset and is allocated to the balance sheet of the originator until it becomes obsolete. This decision was taken because the originator of the research may have some loose ownership rights of any unrealized economic benefits. In the case of government funded R&D, societal ownership is also possible. Also, the national accounts definition of an asset is linked to ownership rights and not market advantage. Further, it is included because this R&D still provides economic benefits to society. A practical concern also exists as there is no data available that allows the separation of these expenditures from the rest of R&D in the Canadian statistical system. Data is available for basic research but this is not equivalent to freely available R&D.9

Table 1
Scope of research and development in Canada

21. The freely available issue is closely related to that of spillovers. Spillovers arise when the benefits of R&D accrue to units other than the owner of the R&D. This may occur when the R&D is freely available or when a patent matures. Spillover benefits could be realized by users. However this study makes no attempt to measure spillover benefits. Measurable benefits remain with the originator of the R&D.

22. The treatment of unsuccessful R&D can also be debated. For example, should research that registered no productive result be included in the estimates? The fact that the cost of unsuccessful research is incorporated in the cost of products would seem to argue for its inclusion. Further, unsuccessful tests are useful in that they bring the search process closer to success. These tests may also bring other unexpected results since other products are occasionally developed from them. An example of this occurred in northern Canada where while searching for oil, large diamond deposits were found. In this study, unsuccessful research expenditures are capitalized.

23. The SNA Rev1 also indicates that R&D that is not sold and does not provide a future economic benefit should not be included as an asset. However, in light of the discussion of unsuccessful research and considering the societal benefits the R&D may provide, it is difficult to discern which R&D does not provide eventual economic benefits. No attempt was made to discern or measure the R&D that does not have future economic benefits in this study.

Research and development: intangible produced assets

24. There are several different types of assets defined in the SNA93. Distinctions are made between produced and non-produced assets and also tangible and intangible assets. R&D is a produced asset unlike naturally occurring assets such as land and mineral deposits. Since it is not physical in nature, like a machine, R&D is defined as intangible. In the SNA, intangible assets include mineral exploration, computer software, databases and entertainment, literary or artistic originals. These are all distinct from each other and from R&D.

25. The treatment of intangible assets in the SNA93 was incomplete but after the Canberra Group deliberations and in SNA93 Rev1, there is a much better coverage of intangible assets in both production and wealth. Further, the SNA93 Rev1 states that with the inclusion of R&D within the asset boundary, patents and copyrights will no longer be identified specifically, but rather they will be included in R&D assets.

Stock-flow issues specific to research and development

26. The nature of R&D intangible assets brings forward some conceptual issues not common to other categories of capital. While investment flows that contribute to the stock of capital are relatively easily defined, the rate at which R&D assets are used up in production and transactions in existing R&D assets are less straightforward to measure.

27. R&D capital may not be used up gradually over time like machinery and equipment. It can become obsolete very quickly (e.g., some software R&D) or it can have a long service life (e.g., pharmaceutical R&D). This makes it more difficult to consider the nature of the depreciation of the stock in relation to average services lives, and about what should be included in the stock of capital.

28. Sales of R&D also present a unique issue. When R&D assets are sold by a business, there is a question of whether this reduces the stock of that entity's outstanding R&D capital. In one sense, it does not necessarily alter the stock of accumulated knowledge available to that company. However, the company may have sold certain user rights associated with the R&D capital (patents/copyrights) which would prevent them from using the knowledge in production. On the other hand, the company may have simply sold a service flow associated with the original R&D or even an offshoot of the original R&D. On the latter point, a software producer generally does not sell R&D but rather software produced using a stock of accumulated R&D.

29. From an economy-wide point of view, the impact of imports and exports of R&D on the stock of R&D capital needs further consideration. Specifically, does the stock of R&D capital in national wealth increase with business purchases of R&D from abroad and decrease with sales to non-residents? If the trade in R&D is with affiliates, it becomes an even more difficult issue to sort out.

30. Furthermore, an increase in demand for R&D assets may contribute to a rise in the value of R&D capital through a price effect, though this effect might be difficult to capture. The value of business R&D capital would most certainly be reflected in the market value of the firms engaged in R&D activities.

31. In this study, R&D that is sold is considered to be no longer in the stock of the seller but is completely transferred to the purchaser. This is the case for both domestic transactions as well as for exports of R&D. This method was chosen for practical considerations as well as the fact that it matched the U.S. method. This treatment requires further consideration. When R&D is purchased either domestically or from an international source, it is considered as a purchase of fixed capital formation and not as an input (intermediate expense) for production.


3 . The Frascati Manual is a product of the OECD Working Party of National Experts on Science and Technology Indicators (NESTI).

4 . SNA93, Paragraph 6.142.

5 . SNA93, Paragraph 6.163.

6 . Frascati Manual, OECD 2002, paragraph 63.

7 . The issue of asset boundaries is complex and much discussed by national accountants, particularly when dealing with intangible assets. Human capital is one of these issues. In the SNA, human capital is not treated as an asset. "Ownership rights" in connection with people as well as valuation of this capital, pose difficult issues with this matter. Further, human capital does not fit neatly into the SNA framework, as it straddles the line between labour and capital factors of production.

8 . Baldwin et. al. (2005), in their study of productivity, contend that intangible assets should include other innovation costs, such as product design and marketing, thereby providing a more valid comparison between the Canadian and the U.S. economy. The study indicated that with this broader definition, Canada's R&D costs would more than double those based on the FM definition.

9 . Since some data exists for patented research it has been suggested that non-patented research could be used as a proxy for freely available. However, for some industries, non-patented R&D is not freely available.