• Flat or declining funding in many disciplines makes it harder to justify risky or unorthodox projects.

• The peer review system tends to favor established investigators who use well-known methods.

• Industry, university, and federal laboratories are under pressure to produce short-term results—especially DOD, which once was the nation’s largest source of basic-research funding.

• Increased public scrutiny of government R&D spending makes it harder to justify non-peer-reviewed awards, and peer reviewers tend to place confidence in older, established researchers.

• High-risk, high-potential projects are prone to failure, and government oversight and media and public scrutiny make those projects increasingly untenable to those responsible for the work.

A National Research Council study indicates that the Department of Defense’s budgets for basic research have declined and that “there has been a trend within DOD for reduced attention to unfettered exploration in its basic research program.”³⁸ The Defense Advanced Research Projects Agency (DARPA) was created in part because of this consideration (see Box 6-2).³⁹

Defense Advanced Research Projects Agency managers, unlike program managers at NSF or NIH, for example, were encouraged to fund promising work for long periods in highly flexible programs—in other words, to take risks.⁴⁰ The National Institutes of Health and National Science Foundation recently acknowledged that their peer review systems today tend to screen out risky projects, and both organizations are working to reverse this trend.

In 2004, the National Institutes of Health awarded its first Director’s Pioneer Award to foster high-risk research by investigators in the early to middle stages of their careers. Similarly, in 1990 the National Science Foundation started a program called Small Grants for Exploratory Research (SGER), which allows program officers to make grants without formal external review. Small Grants Exploratory Research awards are for “preliminary work on untested and novel ideas; ventures into emerging research; and potentially transformative ideas.”⁴¹ At $29.5 million, however, the total SGER budget for 2004 was just 0.5% of NSF’s operating budget for