Historical Tension Between Openness and Secrecy

Concern for governmental openness is not unique to modern democracies, but appears to be a part of the roots of civilization. If a government wants its citizens to comply with its policies, it must somehow communication those policies to the public at large. In one of the earliest public promulgations of government policies, Mesopotamian King Hammurabi (1795–1750 bce) had the legal code of Babylonia inscribed on a stone monument so that his people would know his laws. In South Asia, Emperor Ashoka (304–232 bce) inscribed his edicts on stone pillars scattered throughout his realm. Critical to the development of Western science was the Greek ethos of dialectic and demonstration as a part of an open debate of policy issues. According to G. E. R. Lloyd, for Greek “cosmological and natural scientific debates . . . in some cases the contest was adjudicated by a lay public.”^{Footnote 2} Even for societies such as the Pythagoreans, whose membership and rituals were secret, the mathematical form of the knowledge that they pursued was openly available.^{Footnote 3} Although Greek technologies were under the purview of the artisanal and, therefore, lower class, there was no concerted effort to keep technological knowledge secret.

Over time, the leverage granted to societies by military and economic advantage led to a greater consideration for secrecy in technology.^{Footnote 4} Yet counterpoised against this secrecy is the idea that “the rise of democracy focuses not only on the state’s instrumental strategies, but also on the emergence of a particular kind of knowledgeable citizen, the liberal individual who is capable of attesting, as an informed and reasoning witness, to the legitimacy of the state’s technological actions.”^{Footnote 5} This tension between the openness of a liberal democracy and the protective instincts of national (both military and economic) security is a conflict between fundamental Western values that are played out through the restriction of technological information, through control of words and manipulation of narrative.

In Francis Bacon’s New Atlantis, the scientists, the “merchants of light,” took an oath of secrecy to withhold certain discoveries—some being released to the government and not society at large, and others not being revealed outside of the scientists at Solomon’s House. Science and technology are not only commodities but also instruments of power, acquiring political influence in their own right. Not all technology, however, is created with equal import. Many governments have in place a detailed hierarchy of criteria that determine which technological discoveries and applications are withheld from the pubic and which are not. For those that are withheld, there is a further limitation on distribution based upon the judged trustworthiness of individuals and having a “need to know.” The case of Robert Oppenheimer, the father of nuclear weapons, losing his security clearance is a famous instance of being judged not personally trustworthy. Not only is newly developed government-funded technology frequently not freely disseminated, but increasingly since World War II, the social environment—goals, resource allocation, facilities, organizational structures, and approved applications—surrounding the development of certain categories of technological artifacts exists out of the view of the public, much of the technological community at large, and even many members of Congress. Moreover, in industry, many leading-edge technologies are developed under proprietary secrecy in order to protect an economic advantage. Jointly, the imperatives of the twentieth-century Cold War, the War on Terror, and modern capitalism have entrenched a system of secrecy that has become embedded in both American economic and governmental institutions.

Executive Branch Authority for Control of Information

The U.S. government has held and protected secrets, technological and otherwise, since its colonial founding. Authority for maintaining military and state secrets, however, has no clear basis in the U.S. Constitution or in statute. As Robert M. Pallitto and William G. Weaver point out in Presidential Secrecy and the Law, “The state secrets privilege is universally recognized in law as an absolute privilege and mainly relies on practicality more than constitutional principle for its justification. . . . The ultimate reason to uphold its use is on the practical grounds that it is necessary for the survival of the state.”^{Footnote 6} Regardless of the political party occupying the White House, the constitutional authority of the presidency itself has been cited as providing the classification power over information for the executive branch, with the primary policy mechanism for control typically being an executive order by the president. While the exact wording varies from one administration to the next, they typically site the constitutional authority of the president followed by U.S. laws. Presidents Barack Obama and George W. Bush, who certainly might be expected to differ philosophically on the limits of presidential power, cited identical authority: “By the authority vested in me as President by the Constitution and the laws of the United States of America.”^{Footnote 7} Demonstrating the more typical variation from administration to administration, the Dwight Eisenhower Executive Order authority cited: “By virtue of the authority vested in me by the Constitution and statutes of the United States, and as President of the United States, and deeming such action necessary in the best interests of the national security”—what might be termed as “I exist, therefore I classify.”^{Footnote 8}

The president can also issue decision documents generated from the National Security Council (NSC), which controls activities of the executive branch, but which potentially impacts the entire country. The exact title of the document varies from administration to administration.^{Footnote 9} As with executive orders, the president is under no obligation to consult, or even notify, the Congress or the public of the contents or even the existence of a National Security Order or Directive. The NSC has claimed that even a list of such documents would be classified, with the Congressional Research Service (CRS) describing them as “secret policy instruments, maintained in a security classification status” that “clearly pose a problem for a free and open society and bring the U.S. and all of us very close to one of the most dangerous condition [sic] of authoritarian or totalitarian government, rule by secret law.”^{Footnote 10} While this may be a bit of hyperbole by the CRS, the exaggeration is one of degree and not of kind.

Initially promulgated in 1951 by Executive Order 10290 in the Truman administration, government postwar classification control was originally limited to information in which the government held a proprietary interest. However, in 1982 Executive Order 12356 in the Reagan administration extended government classification control to any information that “is owned by, produced by, produced for, or is under the control of the United States Government. . . . E.O. 12356 gives unprecedented authority to government officials to intrude upon academic research by imposing classification restrictions after projects have been undertaken.”^{Footnote 11}

The ultimate arbiter of the power of the presidency is of course the Supreme Court. The inherent power of the president vis-à-vis the Congress to restrict access to information has been repeatedly upheld in U.S. Supreme Court decisions, such as recognizing in 1936 in United States v. Curtiss-Wright Export Corp. the “plenary and exclusive power of the President as the sole organ of the federal government in the field of international relations—a power which does not require as a basis for its exercise an act of Congress.”^{Footnote 12} In Department of Navy v. Egan, a frequently cited 1988 Court opinion, it was held that: “The President, after all, is the ‘Commander in Chief of the Army and Navy of the United States’ (U.S. Constitution, Article II, Section 2). His authority to classify and control access to information bearing on national security and to determine whether an individual is sufficiently trustworthy to occupy a position in the Executive Branch that will give that person access to such information flows primarily from this constitutional investment of power in the President and exists quite apart from any explicit congressional grant.”^{Footnote 13}

The power of the president specifically to control congressional access to classified information, even to the exclusion of the Congress itself, was evident in 1998 in Barlow v. United States. Barlow, an employee of the Department of Defense, believed that the president was lying to the Congress about Pakistan’s violation of nonproliferation agreements so that the president would not have to “de-certify” Pakistan and bar it from receiving U.S. financial aid. After attempting to contact appropriately cleared Members of Congress, Barlow was fired. In Barlow v. United States, “the U.S. Court of Claims supported the administration’s contention that the state’s secrecy privilege could be invoked to prevent the merits of Barlow’s case from being discussed, effectively upholding the administration’s right to fire an employee for attempting to notify the congress of intentionally misleading information.”^{Footnote 14}

Ironically, as national security restrictions on access to technological information were tightening, in the 1960s the Congress began passing a series of “sunshine” laws, including the Freedom of Information Act, the Presidential Records Act, and the Federal Advisory Committee Act. However, these laws typically included a provision similar to section 3 of the 1946 Administrative Procedure Act, which exempted “any function of the United States requiring secrecy in the public interest.”^{Footnote 15}

The Absence of Policy Theory for Secret Technologies

The function of a theory for a policy of technological secrecy would be to permit development of overarching guidance to the president, the Congress, and the public in reconciling the competing interests of openness, national security, economic advantage, and political prerogatives. However, if there were a theory for explaining a policy of technological secrecy, according to the logic for keeping secrets, the policy itself would need to be secret to keep a perceived enemy from gaining insight as to why certain technologies were worthy of being rendered inaccessible to the public and others were not. As Bruce Berkowitz, formerly on the professional staff of the Senate Select Committee on Intelligence, correctly cast it, the characteristics of these competing interests are sufficiently vague as to make it difficult, if not impossible, to craft such a policy document.

Although a presidential executive order may attempt to standardize the definition of secrecy across the executive agencies, it does not apply to the Congress. Furthermore, each of the major government organizations—the Central Intelligence Agency, the Department of Defense, and the Department of Energy—dealing with classified technologies have retained policy control over what technologies get classified, the structure of its particular administrative process, and the process by which an individual is granted access to that restricted technology. While there may be some intellectual novelty in the epistemological blurring between a physical technological artifact and an associated social value such as secrecy, the loss of such a distinction compounds attempts to define and implement technology policies. When technological artifacts are hidden through classification, social uncertainty is intentionally introduced through this purposeful ambiguity.

Executive Control and Technological Free Speech

The First Amendment to the U.S. Constitution prohibits the Congress from making any law abridging the freedom of speech or of the press. However, U.S. laws prohibit, with criminal penalties for violation, the free and open discussion of selected technological information. Moreover, simply having access to such information can impose restrictions on an individual’s free right of association and freedom to travel to particular foreign countries. The government also reserves the right to impose disclosure restrictions on technology developed in the United States even if the information was developed without the aid of the federal government. Finally, it is even possible that an individual who developed a particular technology could well be barred from further work in that same domain for failure to meet the government’s dual criteria of trustworthiness and need-to-know. “This is a far cry,” said Sheila Jasanoff, “from Ezrahi’s transparent and vulnerable liberal democratic state, which is destined always to deploy science and technology in full view of its citizens, and therefore must continually construct demonstrations—in war and peace—to persuade citizens that it is acting for their collective benefit.”^{Footnote 17}

Shrouding certain technologies isolates bits of information and fragments knowledge. This, at its crux, is the dilemma: What is the price in terms of technological productivity and public freedom compared with the dictates of national security? The federal government funds much of the advanced technology developed in the United States. Article I, Section 9 of the Constitution requires: “No Money shall be drawn from the Treasury, but in Consequence of Appropriations made by Law; and a regular Statement and Account of the Receipts and Expenditures of all public Money shall be published from time to time.”^{Footnote 18} However, given the demanding technological requirements of the federal national security apparatus and the size of its budget, much of this innovative boundary of human knowledge is hidden from the public and some of it is even not disclosed to the Congress at large. Moreover, many Members of Congress have little or no knowledge of these sensitive technologies they are funding, and, except in a disguised manner, the Congress does not openly publish the public money expended on them. As U.S. Supreme Court Justice Robert Jackson succinctly opined in a dissenting opinion in Terminiello, 337 U.S. 1, 37 (1949), the liberty of the citizen does not require that we “convert the constitutional Bill of Rights into a suicide pact.”^{Footnote 19}

Seeds of Legislative Secrecy

“There is,” according to David Guston, “a school of thought favoring ‘congressional dominance,’ which argues that Congress possesses sufficient tools—budgets, new authorizing legislation, confirmations, and oversight—for liberating information and otherwise bending the bureaucracy to its political will.”^{Footnote 20} Such a model interweaves three distinct elements: a Western liberal democratic legislature that fights to empower its citizens; a positive value associated with the sharing of governmental information; and a bureaucracy that impedes the free flow of information. The problem with such an approach is that it assumes that there is a single Congress at work. At any given time, however, there may be several “different” Congresses, at least one of which supports a policy of technological concealment.

The pedigree of the U.S. Congress itself is one of secrecy. The Continental Congress not only met in closed sessions, but it also required its members to sign a formal secrecy agreement acknowledging that any violation could result in expulsion from the Congress. The Resolution of Secrecy Adopted by the Continental Congress, November 9, 1775, imposed a complete prohibition on public discussion of matters that a majority vote determined to be secret, and also barred public comments on matter that was under debate by the Continental Congress:

The secrecy surrounding the proceedings of the colonial Continental Congress continued into the early legislative institutions of the United States—not surprising since the individual personalities were largely the same. The Constitutional Convention met in secret session, and produced Article 1, Section 5 of the U.S. Constitution, which provides that “Each House shall keep a Journal of its Proceedings, and from time to time publish the same, excepting such Parts as may in their Judgment require Secrecy.”^{Footnote 22} Clearly, the founding intent was to permit, when deemed necessary, exclusion of the public from congressional policy deliberations.

A session of either the House or the Senate that excludes the public and the press may be conducted upon the request of a member of that body. In the Senate, closed sessions are most typically used during discussions of national security issues or during impeachment trials. As Mildred Amer showed, until 1794 the Senate met in secret session under a philosophy that reflected a conviction that its “advise and consent” to the executive branch was best performed out of view of the public. From its creation through the War of 1812, the House of Representatives also frequently held closed sessions, with additional secret meetings in 1825 and 1830. Between 1830 and 2004, the House met in secret session only in 1979, 1980, and 1983. The Senate from 1929 to 2004 conducted fifty-four proceedings in secret, generally for reasons of national security, but with six sessions related to the impeachment trial of President Bill Clinton. The 1988 deliberations concerning the Intermediate Range Nuclear Force Missiles Treaty were conducted in secret.^{Footnote 23} Given the decreasing frequency of closed sessions, it would appear that the modern Congress has moved toward more open deliberations. However, during this same period, the details of the day-to-day congressional workload have been increasingly conducted via the committee structure, much of the work of which is conducted in closed committee meetings, especially those dealing with national security matters.

The competition across democratic openness, military secrecy, and bureaucratic control is reflected in the various interest groups that try to impose their view of reality. National security appears to have captured the discursive high ground for controlling technology policy. Legislation governing technology policy invariably includes an exclusionary provision excepting national security domains, with science and technology being subordinated “to the purposes of military power.”^{Footnote 24} The boundaries of this intrinsic policy conflict became clearly visible in the mid-1980s over congressional funding for the B-2 bomber.

B-2 Advanced Technology (Stealth) Bomber

In the early 1980s, President Ronald Reagan revealed that the United States was completing the development of a B-2 stealth bomber and that the Congress would be asked to provide funding to begin its production and deployment. One of the historically most visible congressional policy conflicts, it revolved around the competing interests of protection of a technological advantage over potentially hostile nations; the right of the Congress to demand information from the executive branch; and barring the official dissemination of technological secrets to some Members of Congress while simultaneously providing it to unelected, but influential, committee professional staff.

As is typical for a sensitive advanced technology, the details of the B-2 Bomber program had been provided in classified communications to the Members and appropriately cleared staff of the House and Senate Committees on Armed Services as well as the Defense Subcommittees of the House and Senate Appropriations Committees. In the face of protests in 1986 by congressmen, particularly Representative Mike Synar (D-Okla.), who did not serve on these powerful oversight committees, Defense Secretary Casper Weinberger commented: “We have made a great deal of information available in closed hearings to the appropriate committees. . . . But I’m certainly not prepared to discuss the stealth program any further than we have in the presentations, simply because we don’t think the Soviets have any need or right to know.”^{Footnote 25} Weinberger’s comment is interesting for what it reveals about how information on secret technologies is, or more properly is not, disseminated within the Congress. First, Weinberger implied that a policy of providing information on a classified technology to a Member of Congress not serving on a particular committee was tantamount to providing the information to the Soviet Union. Second, there is nothing in the U.S. Constitution or in statutes that differentiates between congressmen with regard to their right or eligibility to receive classified information. Finally, Weinberger openly acknowledged that the information of interest had already been provided to certain congressional committees that in turn had refused to share it with nonmembers of those committees—obviously privileging those committees’ continuing ability to be deemed partners worthy of receiving sensitive information from the executive branch rather than accommodating the needs of a fellow Member of Congress.

The policy for access to classified information not physically controlled by the executive branch is not clear for a Member of Congress who does not sit on one of the privileged committees. The rules vary from committee to committee. For example, the House Permanent Select Committee on Intelligence has historically granted access to classified information if a nonmember successfully negotiates “a six-stage process, including a written request and recorded votes by a quorum of committee membership. . . . Thus, it is possible for a non-member to be denied attendance at its executive sessions or access to its classified holdings.”^{Footnote 26} The House Armed Services Committee rules have been much less constraining and allow access to classified information by any member of the committee or any other Member of the House of Representatives who has requested an opportunity to review such material.^{Footnote 27} Members must, however, already suspect that a secret technology exists before they can know to ask for details. The current process could be termed an exclusionary chain of providing technological secrets to only a few Members of Congress as a proxy for telling all Members of Congress, which is of course itself a proxy for telling the American public at large. It is also of interest that elected Members of Congress were denied access to information on the B-2 bomber while selected staffs—the experts—of certain committees were given access. Although these staffs are employees of the Congress, they have, in many ways, become an empowered bureaucracy in their own right. In an important way, the elected Members of Congress could be viewed as a constitutional impediment to the power of an unelected congressional staff.

In The New Politics of Science, David Dickson related how the Congress reacted to the 1970s “grass-roots democracy” movement in the United States by passing a series of laws “aimed at limiting the undesirable consequences of science and technology . . . implicitly endorsing the legitimacy of public demands for direct social control over the way science and technology was produced and applied.”^{Footnote 28} What Dickson failed to recognize is that a law is a policy crafted to address a particular perceived need—in this case a social demand for more openness and control over science and technology. There are subject-matter boundaries, however, beyond which a law typically does not go, and future contingencies beyond which it cannot see. The openness laws passed by the Congress in the 1970s did not apply to technologies that were considered militarily sensitive. Consequently, the expansion during the 1970s of the openness initiatives specifically excluded military secrets from their scope. What Representative Synar found out was that the expansion of openness did not empower a duly elected Member of Congress to learn the technological secrets of the B-2 without making an extraordinary effort. If a technological community is defined to include those people who determine which projects are worthy of funding, it is clear that for classified technologies some Members of Congress are a part of the technological community and others are not.

Despite these barriers, the informationally disadvantaged Members are expected to perform the constitutional duties of their offices by voting on authorization and appropriations for technologies about which they know essentially nothing. Synar believed that the secrecy surrounding the B-2 bomber prevented at least some Members of Congress from exercising these oversight responsibilities: “I had to jump through hoops to get the special clearance, and it’s left a very bad taste in my mouth. . . . The Air Force has done basically everything they can to keep me from learning about this program. . . . I think it’s important that all Members of Congress review what could be the most expensive weapons system in the history of this country.”^{Footnote 29} In this instance, it is clear that the tension between national security and openness, between the executive and the legislative, was resolved in favor of national security and the executive.

Although Synar was ostensibly attacking the need to develop a particular technology, the B-2 bomber, he was in reality simultaneously challenging three of the privileged interests in the United States: technological progress, bureaucratic control, and military secrecy. Each of these three elements—the benefits of technology, the prerogatives of bureaucracy (both legislative and executive), and the military propensity for secrecy—had developed its own set of political authorities. Synar attempted to change the established political boundaries exposing “the inherent tensions and contradictions in the structure of the democratic political system.”^{Footnote 30} Synar was unable to challenge successfully the idea that a more modern technology is a better technology without also simultaneously challenging the accepted authorities of congressional bureaucracy and military secrecy.

Congressional Clearances

The B-2 debate also highlighted the issue of control of access to sensitive technological information. By virtue of their offices, the president, vice president, and Members of Congress have traditionally been entitled to classified information without being required to hold a formal security clearance. All others, including executive branch employees, congressional staff, contractors, and outside experts serving on advisory panels, are required to undergo some combination of background investigation, polygraph examination, financial disclosure, and urinalysis screening.

Until recently, while congressional committee staffs were required to take a secrecy oath, Members of Congress were not. Under the presumption that election by the American people and the traditional congressional oath of office to support and defend the Constitution of the United States against all enemies, foreign or domestic, was sufficient, Members were not required to take an oath. However, beginning in 1995, the House instituted a secrecy oath for all Members, officers, and employees requiring that they “solemnly swear (or affirm) that I will not disclose any classified information received in the course of my service with the House of Representatives, except as authorized by the House of Representatives or in accordance with its Rules.”^{Footnote 31} The next steps, requiring that a Member of Congress be given access to classified information only after the granting of a security clearance based upon investigation, polygraph, financial disclosure, and urinalysis has yet to take place. Opponents of these additional measures contend that the proposal raises constitutional issues of congressional independence in the event that an executive branch agency conducts the background investigation and has access to the personal information it generates, and whether an executive branch official should determine if an elected legislator is entitled to receive certain governmental information. Nevertheless, it is not always clear from whom the information is being withheld. The penchant for bureaucratic control can even result in the exclusion of the president himself. In his speech on governmental secrecy to the American Association for the Advancement of Science, the late Daniel Patrick Moynihan described project Venona, which in 1946 broke the Soviet communications codes. “The Soviets knew this. As did the United States Army and the FBI. Further, they knew that the Soviets knew and that we knew they knew. Only the American people did not know. Including it would appear, the American president. This on the direct orders of Omar Nelson Bradley, Chairman of the Joint Chiefs of Staff, recorded in an FBI memorandum dated October 18, 1949. If the president was to know, the General would inform him. Which evidently he did not.”^{Footnote 32}

Enforcing Congressional Policies on Classified Technologies

Despite the difficulty that individual Members of Congress experience in obtaining information from the executive branch on classified technologies, that does not mean that the Congress as an institution is without effective oversight tools. Since the primary constitutional control mechanism of the Congress is the “power of the purse,” a close examination of the annual Department of Defense (DOD) appropriations acts—the spending bills for the Department of Defense—provides clues that the Congress exercises a more in-depth oversight of classified technology programs than might otherwise be realized. Since the House and Senate Appropriations Committees must pass on all defense and intelligence funding requests, they have an effective means of enforcing their policies through punitive, but selective, deletion of funds. Buried in the fine print of the annual DOD appropriations act is a section called General Provisions. As an example of the reach of the Appropriations Committees, there is typically a provision such as Section 8007 of H.R. 4576, the DOD Appropriations Act for 2001, which states: “Funds appropriated by this Act may not be used to initiate a special access program without prior notification 30 calendar days in session to the congressional defense committees.”^{Footnote 33} Although this does not preclude the Secretary of Defense from initiating a highly classified technology program, it does ensure that the committees are promptly informed. Once that is done, a committee can request detailed information through a briefing or formal hearing. Although, as discussed in this article with regard to the B-2 bomber program, such information may well not be made available outside of the committee membership and a few selected professional committee staff members, a path for oversight is created.

Presuming that the committees with oversight responsibilities for classified technologies have been made aware of the existence of an advanced technology project, the logical question is: So what? How can the Congress provide effective policy control over an executive branch project when the Congress cannot describe the program, or in many instances cannot even acknowledge its existence in public? Once again a close reading of the annual DOD appropriations act provides a clue. For example, Section 8084 of H.R. 4650, the DOD Appropriations Act for fiscal year 1995, reads:

This provision incorporates by reference a “classified Annex” that is made available, as part of the presentment of the bill, to the president. Incorporation by reference is a well-established legislative tool and is not reserved simply for classified oversight. For example, in a 2001 ruling in Hershey Foods Corp. v. United States Department of Agriculture, the United States District Court for the District of Columbia upheld the use of incorporation by reference by the Congress.^{Footnote 35}

While it is impossible for the public or, for that matter, apparently many Members of Congress to know the details of the classified Annex, the Department of Defense complains regularly about the Appropriations Committees’ micromanagement of the department’s unclassified projects. It is likely that oversight of classified technology is not significantly less intrusive than its unclassified counterpart simply because it is not public.

In-House “Outside” Experts

Despite the seeming advantage that the executive branch may have in exercising policy control over classified technology projects, the Congress is not without means to influence them. The central question is not whether such oversight is provided, but rather whether it is appropriate in a liberal democracy for a subset—the members of the defense and intelligence committees—of the Congress to exercise legislative power that the Constitution reserves for the Congress as a whole. One of the few mechanisms available to all Members of Congress for applying independent technical judgment on classified technology projects is the cadre of trained engineers, scientists and, accountants of the Government Accountability Office (GAO), one of the few organizations subordinate to the Congress. A 1921 statute authorized the GAO to review executive branch documents as a part of its audits and program evaluations. The Congress has relied heavily on the GAO to provide trusted and independent expertise on science and technology issues, especially in confronting the technical expertise of the executive branch and private interest groups. However, “the Bush Administration challenged the authority of GAO on constitutional grounds, arguing that the Comptroller General, who is the head of GAO, had no ‘standing’ to enforce GAO’s right to federal records. The Bush administration prevailed at the district court level and GAO decided not to appeal, significantly weakening the authority of GAO.”^{Footnote 36} Thus, the Congress will have to place increased reliance for technological information on the executive branch and industry—the very parties that wish to influence the policies being generated by the Congress.

The Office of Technology Assessment (OTA) was a short-lived attempt to create a standing institution that could provide independent technical expertise to the Congress. Created on October 13, 1972, by Public Law 92-484, OTA had mixed success.^{Footnote 37} Its strength was in providing information for nonnational security-related science and technology, since it was never granted full access to the most highly classified technologies of the defense and intelligence communities. Consequently, it could not take a full part in providing expert advice in the area in which it perhaps could have provided the most contribution. It was disestablished in 1995 as a part of the Republican-led reductions to what had been a growing congressional staff.

The GAO, the OTA, and information-gathering organizations such as the Congressional Research Service (CRS) are not, however, the only mechanisms available to the Congress to support the development and oversight of technology policy. One of the organizations that may be the least known but perhaps has been the most able to penetrate the technological secrets of the country is the Surveys and Investigations (S&I) Staff of the House Committee on Appropriations. Created during World War II to help the House Appropriations Committee provide improved oversight over the vast increase in federal spending, from 1943 to 1946 the S&I Staff operated under a House resolution and Committee rules. Thereafter, its authority was codified in section 202(B) of the Legislative Reorganization Act of 1946 (Public law 601, 79th Congress).^{Footnote 38} For the S&I Staff to investigate a particular matter requires the concurrence of both the chairman and ranking minority party member from the House Appropriations Subcommittee of jurisdiction—for example, the Department of Defense, NASA, or the Department of Energy—as well as the concurrence of both the chairman of the full Appropriations Committee and the minority party counterpart. The dual-approval process significantly helps to ensure that the work of the S&I staff is largely insulated from partisan politics. Once an investigative order has been generated, the chief of the S&I Staff has an independent hand to recruit on a contract basis the necessary scientific, engineering, contracting, and accounting expertise. In this instance, it is difficult for both the executive branch and federally funded contactors to deny access to requested information even in the most sensitive technological areas, since the House Appropriations Committee has the power to recommend to the entire Congress termination of funding for any project that does not cooperate with the investigation. As a result, the S&I staff has routinely had access to sensitive technology projects for which access has been denied to organizations such as OTA and GAO. The findings, however, of the S&I Staff are available only to Members of the House Appropriations Committee unless the leadership of the committee approves broader distribution. The work of the GAO and OTA are available to all Members.