The use of research portfolios in science policy

Any research funding and/or research-performing organization in the public, private, and non-profit sectors needs to adopt a portfolio-wide perspective to R&D management to better align research project investments with the organization’s overall strategic goals. Private sector firms have increasingly done so utilizing powerful new methodological tools and large amounts of data becoming availab-le. In contrast, with relatively few exceptions, public R&D management still tends to base selection processes on the excellence of individual projects according to peers rather than considering the merits of the whole portfolio. There are good reasons for additional caution which, besides the usual inertia and the resistance by scientists trusting the peer review process, include multiple objectives of public programs, long-term accrual of results and associated uncertainties, and difficulty to monetize or value. This report argues it is high time for public R&D management to move forward. Portfolio analysis should not be applied similarly across the board. It will serve different purposes for different types of public R&D programs depending on risk/uncertainty, data avai-lability, and target clarity (ability to define unambiguous program goals). Not all methodologies will be appropriate to all programs. Nonetheless, the toolkit, data depositories, and computing capability have expanded tremendously during the past couple of decades to render such experi-mentation possible and absolutely necessary.