Notes

^1.

See Aboody and Lev (1998) for a description of the issues surrounding the capitalization of software.

^2.

We will deal with depreciation rates and estimates of capital stock in a subsequent paper.

^3.

Further work will extend our investigations into areas where measurement problems are greater.

^4.

See Baldwin, Beckstead and Gellatly (2005) for arguments on what needs to be included to fully capture science-based innovation expenditures.

^5.

The 1993 SNA manual indicated that it considered R&D to be an investment but did not recommend that it be capitalized because of controversy around how narrowly it should be defined. The revised manual has now accepted that it should be capitalized, but included as a Satellite Account.

^6.

Peleg (2003) notes that the tax code in the United Kingdom specifically excludes from its definition of R&D expenditures on commercial development that do not involve scientific or technological investigation or that occur after the resolution of uncertainties.

^7.

None of these expenditures are included in the official R&D statistics that consider only work performed in Canada—see Baldwin, Beckstead and Gellatly (2005).None of these expenditures are included in the official R&D statistics that consider only work performed in Canada—see Baldwin, Beckstead and Gellatly (2005).

^8.

Using the methodology developed in Beckstead and Gellatly (2003b).

^9.

For a discussion of the importance of engineering investments, see Baldwin and Dixon (2008).

^10.

The importance of R&D would be even smaller if exports of R&D were removed from the Own-account R&D expenditures, as is done in some satellite accounts of R&D.