Bush submitted Science, The Endless Frontier to President Truman in July 1945, weeks after 'As We May Think' appeared in The Atlantic. The report made three arguments: scientific research is essential for national security, economic prosperity, and public health; basic research requires long-term support that industry cannot provide; and government funding must come without government direction—scientists, not bureaucrats, should decide which questions to pursue. Bush proposed an independent federal agency governed by scientists, insulated from political interference, distributing grants through peer review. The National Science Foundation, established in 1950 after five years of legislative negotiation, adopted this model with modifications. The report established the institutional framework within which American science operated for seven decades—a compact in which public funding supported research whose applications were unknown and often decades distant.
The report was political advocacy dressed as policy analysis. Bush's wartime authority—derived from coordinating the research that produced radar, antibiotics, the proximity fuse, and the atomic bomb—gave his recommendations unusual weight. He leveraged this authority to argue that wartime success proved the value of basic research and that the research infrastructure built during the war should become permanent peacetime investment. The argument was contested: legislators worried about funding research with no guaranteed practical application, and Truman's administration favored more direct political control than Bush's scientist-governed agency would permit.
Bush's institutional model reflected Progressive Era faith in expert administration insulated from partisan politics. The NSF would be governed by scientists because scientists possessed the expertise to evaluate research quality—a claim that assumed scientific merit was objective, detectable through peer review, and separable from political or economic interests. This assumption has been challenged across the NSF's seventy-year history, particularly when research directions involve value-laden questions about risk, equity, or environmental impact. Bush's framework provided no mechanism for incorporating non-expert voices into research priority-setting, a limitation contemporary science policy continues to grapple with.
The 'endless frontier' metaphor positioned scientific research as exploration without terminus—a supply of questions that would never exhaust itself. This optimistic framing served Bush's political purpose (justifying permanent funding) while concealing uncomfortable questions about whether all research directions are equally valuable, whether some questions should not be pursued, and who decides. The metaphor's power derived from its resonance with American frontier mythology, but the resonance also imported assumptions about conquest, resource extraction, and the subordination of nature to human purposes that environmentalist critics would interrogate decades later.
Bush's institutional legacy is mixed. The NSF became the world's premier public science funding agency, supporting research that private capital would not—basic physics, pure mathematics, long-horizon biological investigation. But the compact Bush proposed—public funding without political direction—eroded across decades as Congress imposed mission requirements, directed funding toward applied research, and demanded accountability measured in economic returns. The AI transition intensifies these tensions: the most capable AI systems are built by private firms using proprietary data, methods, and infrastructure that public funding cannot match. Whether Bush's model can govern AI development or requires wholesale revision is the institutional question the AI moment forces.
Truman requested the report in November 1944, asking Bush to recommend how wartime research capacity should be redirected to peacetime needs. Bush used the request to advance an agenda he had developed across two decades—the conviction that American research required permanent federal support organized through institutions insulated from partisan politics. The report's argument was prepared before Truman's request; the request provided occasion and authority for proposals Bush had been formulating since the 1930s.
Bush drafted the report with assistance from committees of scientists, university administrators, and foundation officers—an elite network whose consensus Bush presented as objective expertise. The report's recommendations reflected the interests of this network: university-based research, investigator-initiated projects, peer review, and minimal federal oversight. Excluded voices—industrial researchers, engineers, applied scientists, and anyone outside the elite research university system—had no input into the framework that would govern American science for generations. This exclusion shaped the compact's form and limited its adaptability to later challenges.
Basic research as public good. Bush's central economic argument: research producing fundamental knowledge generates returns no single firm can capture, requiring public investment for efficient provision.
Scientist governance of science funding. Peer review and expert panels should allocate resources, insulating research from political interference while maintaining public accountability through institutional oversight.
Long time horizons for basic research. Applications are unpredictable and often decades distant—requiring patient capital that electoral cycles and quarterly earnings cannot provide.
University-based research model. Universities, not government labs or industrial research centers, should be the primary sites of basic research—combining investigation with training and maintaining independence from mission-driven demands.
Permanent federal investment as national strategy. Science is too important for national security, prosperity, and health to depend on discretionary appropriations—the compact requires permanent institutional commitment.
Bush's institutional model faces mounting challenges. The concentration of AI capability in private firms with resources exceeding public agencies inverts the public-private balance Bush assumed. Peer review works well for evaluating disciplinary research; it provides no mechanism for evaluating research trajectories with society-wide implications. The model of scientist self-governance that Bush championed excludes affected communities from decisions about research directions—a gap that environmental justice, bioethics, and AI governance movements have forced into visibility. Whether Bush's framework can accommodate these challenges through modification or requires replacement is the governing institutional question of contemporary science policy.