Revisiting NSF’s Founding Compromise
To meet the National Science Foundation’s growing challenges, the National Science Board’s role should be refocused on providing science advice.
The National Science Foundation (NSF) and National Science Board (NSB) were founded in 1950 with the passage of the National Science Foundation Act of 1950 (the “Organic Act”) and a small $37 million appropriation from Congress. Since that time, NSF funding has grown significantly to its current level of $8.6 billion. Bipartisan legislation in the House and Senate scientific committees are now converging to rapidly increase the budget, expanding NSF’s role within the federal government. Moreover, this year—for the first time in 30 years—NSF established a new directorate: the Directorate for Technology, Innovation, and Partnerships. Under this regime, NSF will have “a primary role in strengthening the US research enterprise so the United States can better compete against China and the rest of the world.” To accomplish this, Congress has proposed a total of $78 billion to be appropriated to NSF for the next five years, which could result in an agency annual budget of $17.9 billion in 2026—approximately twice its current budget.
With this possible influx of money and expectations, it’s time to revisit how NSF and NSB work together. Over the past seven decades, these two entities have created a national ecosystem for scientific innovation that has enabled generations of discoveries and discoverers—though NSF is far more visible to the public. While this visibility might suggest that NSF is “in charge” of NSB, it is important to note that the Organic Act did not create a traditional board structure. Instead, the NSF director and NSB were designated as co-equals, sharing the responsibilities of managing the awarding of grants and other funding vehicles to promote research and education in science and engineering. In carrying out this responsibility, NSB provided strategic direction for the agency and approval of programs and awards, while the director managed the day-to-day operations of the agency and carried out the board’s policies. NSB also plays a role in providing scientific advice to the president and Congress.
NSF and NSB are linked by the NSF director, who serves as chief executive of NSF, a voting member of NSB, and the chair of the NSB Executive Committee. The NSF director is appointed by the president and confirmed by the US Senate. However, each NSB member is also presidentially appointed. Typical federal agencies, by contrast, have an accountability chain that leads from the agency head to the president, often through a cabinet secretary. As NSF historian J. Merton England wrote in A Patron for Pure Science, “NSF’s two-headed structure was an unusual one for a federal agency.” Indeed, this structure is unique among federal science agencies and even major philanthropic funding organizations outside the government.
As NSF grows, its administrative structure needs to be aligned with its mission, particularly with respect to new “big science” endeavors. To the best of our knowledge, the subject of NSF’s dual power structure has not been addressed in the policy literature since the 1967 policy paper “National Science Board: Its Place in National Policy” by National Academy of Engineering president Eric Walker. However, the relationship between NSF and NSB has evolved significantly since that time. The dual structure introduces complexities and inefficiencies, as was evidenced by a recent crisis in one of NSF’s large facility projects, the National Ecological Observatory Network (NEON).
To live up to ever-growing expectations for the agency and an increase in big science projects like NEON, NSF’s governance structure must be simplified. We recommend streamlining NSF’s reporting structure to better match that of other large federal agencies. Reducing NSB’s day-to-day management responsibilities would create an opportunity for its scientific advice functions to be brought to the fore, strengthening the board’s ability to influence national policy.
NSF and NSB: a unique structure born out of compromise
To the casual observer, it might appear that Congress had always intended to create NSF as an agency whose authority would be shared by two entities. However, like most legislation, NSF’s Organic Act reflected a compromise of two visions of how the agency should be run.
After Vannevar Bush’s Science, the Endless Frontier put forward a strong argument for an independent federal agency that would support US basic research, Senators Warren G. Magnuson and Harley M. Kilgore held hearings in October 1945 on proposed legislation based on the report’s recommendation to create a “National Research Foundation.” There was widespread support for the establishment of “a single federal agency that would award grants, contracts, and fellowships to advance knowledge … in all fields of the natural sciences.” However, policymakers were divided on how to structure the leadership of the new foundation. Kilgore envisioned a director selected by and accountable to the president. Bush advocated for a foundation that would employ a part-time board appointed by the president, which would be responsible for hiring and firing a director, more in the model of a corporate board of directors.
Several attempts to pass the legislation died in the House or the Senate before a 1947 version of the bill was finally agreed upon. President Truman, who felt that the board should not select the director, vetoed the bill. Eventually a compromise was reached where the NSF director and NSB would both be appointed by and accountable directly to the president and would share executive authority for the future NSF.
The initial instantiation of NSF was much different than the current foundation, particularly with respect to the delegation of authority between the director and NSB. In the 1950 Organic Act, the director received the power to award scholarships and graduate fellowships, as well as carry out the general authority of the foundation. However, that power was restricted so that “no final action shall be taken by the Director in the exercise of any power granted … unless, in each instance, the NSB has reviewed and approved the action proposed to be taken.” Thus, NSB was given the ultimate power of review over all actions that the director was charged with carrying out.
This compromise led to criticisms that the agency’s structure made the director more accountable to NSB than to the executive branch. In addition, this structure put a huge burden on NSB to approve all the awarded grants. The burden increased as NSF, and the number of grants, grew. By 1960, the grants program had increased from around $1 million in awards in 1952 to $57.8 million. A reorganization plan enacted by Congress in 1962 and 1965 addressed these issues by transferring some of the powers to the director.
In 1966, the House Committee on Science and Astronautics undertook a comprehensive review that resulted in a recommendation that would lead to major structural changes to NSF. A 1968 authorization bill subsequently aimed to improve “the organization and the operation of the foundation” to prepare for the agency’s growing importance and size. Under the 1968 authorization bill, many parts of the Organic Act were changed, including the power dynamic between the director and board, with the director gaining more power. Section 5 (b) of the 1968 bill states, “except as otherwise specifically provided in this Act (1) the Director shall exercise all of the authority granted to the Foundation”—and makes all the director’s decisions “final and binding.”
NSF should have, at that point, vested the power of the agency in the director, following the model of most other independent agencies. However, NSB retained two remaining oversight functions which are still in place today: the authority to establish the policies of the foundation, and the authority to approve very large awards such as those for major instrumentation.
These two functions are critical to the character and culture of NSF. Effectively, NSB controls not only who can be awarded grants and under what conditions, but also the construction of NSF’s largest infrastructure assets, such as South Pole Station and the large telescopes. While the director retains day-to-day authority for running the foundation and serves as its public face, NSB approval is needed for many substantive activities and agency policies, including NSF’s unique “no cost overrun” policy, which creates a framework to contain costs on the agency’s largest infrastructure projects.
NSF’s organizational structure has no analogue to other US agencies that award grants in support of research activities. NSF and NASA alone stand out from other agencies involved in science as their leadership interacts directly with the White House instead of through a cabinet secretary. NASA’s administrator is “responsible for the exercise of all powers and the discharge of all duties of the Administration and shall have authority and control over all personnel” and, as with the NSF director, does not report to a cabinet secretary but instead reports directly to the president.
Similarly, there is no analogue to NSF’s structure in the major philanthropic science funding agencies outside the government. For example, Howard Hughes Medical Institute, the Bill & Melinda Gates Foundation, and the Chan Zuckerberg Initiative are each incorporated under nonprofit legal status and have a board of directors that hires and fires the chief executive officer (usually a nonvoting member of the board). In all of these cases, the governance is constrained by both federal and state statutes in which the board holds fiduciary responsibility. These organizations play a major role in the US science research and development ecosystem, funding both individual principal investigators and institutions. Overall these institutions have a linear reporting structure with the board having decisionmaking power and fiduciary responsibility, allowing for clarity and the ability to manage issues when and if they arise.
The shortcomings in NSF’s governance structure are highlighted by the recent crisis at NEON. NEON’s 2015–2018 upheaval provides evidence for how the complex relationship between NSF and NSB played out in practice. One of us (Olds) served as assistant director for biological sciences during that period and was an active participant in those interactions.
With a total construction cost of close to $500 million, NEON represented a novel approach to continental-scale macrobiology. It consisted of a distributed network of terrestrial and aquatic biosphere sensors supplemented by robust human-based sampling protocols that extended from Hawaii to Puerto Rico, including the continental United States and Alaska. In addition to the costs associated with construction, over the course of its planned 30 years of operation, NEON will cost an additional $2 billion to operate. Now fully commissioned and operated by Battelle, the observatory produces a multidimensional time series of close to 200 data products that provide a unique synoptic survey of the US biosphere.
The construction of the NEON project was a major challenge for NSF because it had no precedent. Unlike oceanographic research vessels or astronomical telescopes, there was no history of design and construction to fall back on. When NEON project construction commenced, NSB provided approval for it to move forward, as required for all large infrastructure projects proposed by NSF. In 2015, in the middle of the construction, it became clear to NSF’s Directorate for Biological Sciences leadership that NEON, Inc.—the nonprofit corporation established by NSF to design and build the observatory—would be unable to successfully complete its work on time and on budget. Multiple emergency site visits by NSF staff revealed that the project was on a trajectory for an $80 million cost overrun, which would violate Section 1.4.6 of NSF’s Major Facilities Guide.
In response to the crisis, both NSF management and NSB stood up independent ad hoc processes including a “director’s watch list” and an NSB Task Force on NEON Performance and Plans (NPP). These processes involved internal deliberation, hiring outside consultants to determine the root causes of the emergency and chart an effective long-term response, and the production of multiple reports, including a commissioned report from the National Academy of Public Administration (NAPA). In addition to the NSF and NSB responses, Allison Lerner, the agency’s inspector general, commenced her own investigation. As a result of this activity, the House Science Committee increased its oversight over the project.
The sheer scale of the response drew the attention of the scientific news media, which created further challenges for the project in the scientific community. For example, 16 current and past presidents of the Ecological Society of America felt compelled to write a joint commentary supporting the continuance of the project.
NSF’s consultants and external oversight bodies pointed to the agency’s process deficiencies in large facility management. Key recommendations of the NAPA report were to “publish a joint NSF-NSB duties and responsibilities document institutionalizing roles and addressing key working relationships” and to form a Federal Advisory Committee Act (FACA) advisory committee “for the Director to use as a sounding board for objective insight on large research projects.” NSF’s most significant response to the NEON crisis has been the formation of such a FACA committee, the Business Operations and Advisory Committee, and the appointment of a chief officer for research facilities, who sits in the Office of the Director.
Crucially, NSB’s Task Force on NEON Performance and Plans recognized the inherent difficulties of a science board having the expertise to provide effective oversight for a large-scale infrastructure project such as NEON. “NPP found that NSB needs to attend to both the science and management aspects of facility projects when making award decisions. The board’s ongoing efforts to add members with management and business expertise, as recommended by the NAPA report, will help address this finding.”
The central problem, as elucidated in the quote above, is that members of NSB are selected for their scientific skills and insight, not to provide the business and management oversight required for large projects. This is in keeping with the original statute, which states that board members “shall be eminent in the fields of the basic, medical, or social sciences, engineering, agriculture, education, research management or public affairs” and “shall be selected solely on the basis of established records of distinguished service.”
NEON’s challenges do not stand alone. Other examples of NSF large facilities that also experienced crises include the Oceans Observatory Initiative and the Deep Underground Science and Engineering Laboratory, which was canceled. As NSF looks toward greater growth in the future and embarks on new large-scale projects, these issues will grow, placing an increasing burden on NSF’s ability to respond to the challenges that arise as it supports the next generation of science and engineering discoveries.
Streamlining NSF and strengthening NSB’s science advice role
We recommend that NSB step back from its NSF role, allowing NSF to adopt a more normative federal agency structure with a unitary executive as its director. As mentioned earlier, NSF has recently created a Business Operations and Advisory Committee to provide advice on issues related to performance of the foundation’s business operations. We believe that NSF should also charter a new NSF Director’s Advisory Committee that includes a subset of NSB members as well as individuals whose expertise lies in managing large facilities. This new committee, reporting to the director, could provide the combination of management, business, and scientific advice that is necessary for the oversight of large-scale scientific projects.
Stepping back from its NSF role will allow NSB to enhance its second statutory role: providing science advice to the president and Congress. Throughout its history, NSB membership has included a diverse set of America’s top scientists and engineers from across sectors. Already, NSB provides advice and insight that transcends an exclusive focus on NSF. For instance, the board’s Vision 2030 report laid out a national agenda for US innovation over the next decade, and its biennial Science and Engineering Indicators constitute a definitive data source on the US science and engineering enterprise. However, NSB lacks a line of direct access to the president, in the way that the President’s Council of Advisors on Science and Technology (PCAST) has through the White House Office of Science and Technology Policy.
At a time when numerous ideas for scientific management and advice are being proposed and considered, NSB has a key advantage over PCAST in having been statutorily authorized by Congress. NSB thus retains both independence and institutional stability. Whereas PCAST, and its earlier iteration as the President’s Science Advisory Committee, has sometimes ceased to function for years at a time, NSB members, with staggered 6-year terms, provide an important continuity between administrations. One option would be for Congress to leverage the strength of both NSB and PCAST in the creation of a new board, the National Science and Technology Board (NSTB), that would supersede both of them. The NSTB would have the advantage of being written into statute with members serving staggered 6-year terms.
As NSF continues to grow and support more large projects, a unitary governance structure will help to provide clarity in decisionmaking. NSB has the scientific experience and expertise that, if given the opportunity, could provide a structure for useful advice to the president and Congress in a way that has not been possible with the current governance model.