further evaluation. They do not reflect any coherent intellectual framework for priority setting.

Several workshop participants perceived a need to derive more information on subjects of current surveys, particularly scientists and engineers working in industry:

• More precise information on employers and their locations (as is available on academic scientists and engineers) would enable analysts to relate individual and firm characteristics. In particular, if SDR respondents employed in companies were asked to indicate the industrial classification of their establishment (plant, laboratory, etc.) this would help overcome the lack of business unit level R&D expenditure data.¹

• More information on what scientists and engineers do in firms would help illuminate the relationship of research to other functions—strategy, finance, production, and marketing—highly relevant to successful innovation. (See below for a suggestion for obtaining even more detailed information on activities.)

• Information on scientists’ and engineers’ outputs and public activity (publications, conference presentations, involvement in consortia or other collaborations, etc.) would help illuminate R&D spillovers among firms, between industries, and across sectors. The 1995 NSCG and SDR contained a module on patenting and publishing (Morgan et al., 2001). Such a module, modified to reflect changing patterns of publishing, patenting, and collaboration, might be included periodically in the SESTAT surveys.

• Data on stock options, which are prevalent in high technology industries, would fill a growing gap in the information on professionals’ nonwage compensation.

Some workshop participants maintained that it is desirable to expand the NSF definition of the S&E workforce and the information obtained about certain categories of scientists and engineers, although the costs of such steps would have to be considered. In particular,