Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

Comment on Presidential Intelligence

Regulating the secret executive actions required to defend the state within a government legitimized by the informed consent of the electorate has challenged American leaders since the Revolution. The unprecedented recent growth of the U.S. intelligence establishment, a dizzyingly complex global threat environment, and the transformative impact of digital technology make this a fruitful time to reexamine current approaches. To Professor Samuel Rascoff’s careful research and serious policy recommendations, I will add comments informed by practical experience as an intelligence collector, operations manager, and policy official.

I. Bottom Line, Up Front

This is an opportune moment to consider how our government evaluates risky intelligence collection operations. Outsiders can contribute to that examination, but judgments regarding the adequacy of current practices must be based on more than an incomplete public record shaped by unlawful disclosures. More rigorous scrutiny of certain proposed collection operations using existing processes, or recommended new processes, may yield better decisions but “more” does not always equal “better” in intelligence oversight. Adopting Rascoff’s recommendation to add more presidentially appointed officials to the management ranks of our intelligence agencies would be a step in the wrong direction.

II. Presidential Intelligence in Context

Sensational reporting of disclosures by former National Security Agency (NSA) contractor Edward Snowden of previously secret intelligence activities1×1. See, e.g., Glenn Greenwald, NSA Collecting Phone Records of Millions of Verizon Customers Daily, The Guardian (June 6, 2013, 6:05 AM), http://www.theguardian.com/world/2013/jun/06/nsa-phone-records-verizon-court-order [http://perma.cc/LV7C-NHZQ]. has energized public and scholarly interest in the supervision and oversight of America’s large, highly capable intelligence enterprise. Rascoff’s Presidential Intelligence2×2. Samuel J. Rascoff, Presidential Intelligence, 129 Harv. L. Rev. 633 (2016). contributes painstaking research and thoughtful analysis to this growing body of scholarship. Rascoff offers constructive recommendations for new mechanisms to ensure U.S. intelligence enhances our security without imperiling core principles of democratic governance or other important national interests.

At the outset, Rascoff usefully unpacks the term “oversight,” which is overbroad and dangerously imprecise when used in the context of U.S. intelligence.3×3. Id. at 646–59. He also acknowledges that multiple and redundant mechanisms — notably involving all three branches of our government — already exist to monitor our intelligence agencies for the purpose of ensuring they adhere to the Constitution and other applicable laws.4×4. See id. at 635–36. Rascoff’s more narrow aim is to promote “effective intelligence collection while seeking to minimize a wide range of costs, including diplomatic blowback, economic harm to American firms, and intrusiveness that threatens privacy rights” (emphasis added).5×5. Id. at 633. Rascoff distinguishes actions undertaken by U.S. intelligence agencies to collect information from activities pursuant to the other principal missions assigned to these agencies (for example, to evaluate and present that information to policymakers, and to conduct covert actions abroad). This distinction matters because supervision and oversight of the type he favors — directly by the President and his staff — is relatively less robust in the collection discipline.

Drawing on the literature and practice of federal administrative law over the last generation, including its centralization in the White House (dubbed “presidential administration”6×6. See generally Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245 (2001). ), Rascoff coins the term “presidential intelligence” to describe a trend he discerns from recent actions taken by the Obama Administration in intelligence policy.7×7. Rascoff, supra note 2, at 669–74. This trend, in Rascoff’s view, is positive and should be further institutionalized.

Rascoff first recommends a more formal process for interagency review of proposed sensitive intelligence collection operations by the National Security Council (NSC) staff before approval (or, presumably, disapproval) by the President. A written record of the main attributes of an approved collection operation (modeled on a covert action “finding”) should, according to Rascoff, be prepared and also used to notify the Congress of the President’s decision.8×8. Id. at 706–07. Second, Rascoff would improve the President’s access to information on intelligence activities and the intelligence agencies’ overall responsiveness to his direction by increasing the number of presidentially appointed, Senate-confirmed officials serving in the U.S. intelligence community (IC).9×9. Id. at 712–14.

III. IC and White House Review of Collection Activities in Practice

Rascoff’s foundational claim — that the U.S. IC has historically operated outside of presidential control, whether describing collection operations or any other intelligence activity — is not consistent with my experience. To the contrary, outside of the President’s immediate staff within the Executive Office of the President, no element of the U.S. government is more directly and immediately responsive to a President than the IC. The Central Intelligence Agency (CIA) has long nurtured its brand as “The President’s Agency.” Direction from the President — conveyed in person or through senior NSC staff — regarding analysis, collection, or covert action is routinely implemented within hours. Rather than resisting White House engagement on sensitive intelligence activities, modern IC managers actively seek out the “top cover” afforded by a presidential decision, or even by his awareness of a planned action.10×10. See Joby Warrick, CIA Tactics Endorsed in Secret Memos, Wash. Post (Oct. 15, 2008), http://www.washingtonpost.com/wp-dyn/content/article/2008/10/14/AR2008101403331.html [http://perma.cc/X8J5-JXFH], for reporting on efforts by CIA in 2003 and 2004 to gain “top cover” from the White House in the form of a signed memorandum approving the plan for interrogating high-value al Qaeda detainees: “[a]lthough Justice Department lawyers, beginning in 2002, had signed off on the agency’s interrogation methods, senior CIA officials were troubled that White House policymakers had never endorsed the program in writing.” In John Rizzo’s book Company Man: Thirty Years of Controversy and Crisis in the CIA (2014), CIA’s former Acting General Counsel described steps he took to expand awareness of specific interrogation techniques to the broadest possible circle of government national security lawyers so “[w]e would all be in this together, for better or worse,” id. at 189. The days of freewheeling U.S. intelligence operators and “plausible deniability” for the Presidents they serve are a distant memory.

At this point, it may be useful to introduce a general note of caution on the reliability of available research data in the intelligence field, as it may partially explain why my friend Professor Rascoff and I disagree on certain matters.

Scholarly research and debate on contemporary intelligence topics is inevitably handicapped by the secrecy that necessarily surrounds this broad class of government activity. The evidence available for research on current intelligence operations in an unclassified environment is always incomplete, usually deployed in service of a particular policy objective, and routinely distorted by biased commentary from both the critics and the defenders of the government’s actions.

This is certainly the case with the materials describing NSA’s electronic surveillance activities illegally revealed by Snowden and, to a greater or lesser degree, each of the other collection operations cited in Presidential Intelligence. The handful of publicly known (and arguably underconsidered) U.S. collection operations represents a statistically insignificant fraction of the total number of ongoing U.S. operations to gather national security information. There is simply no way for an outsider to know what share of current IC collection operations were prudently undertaken and how many others may have escaped rigorous scrutiny and could eventually come to grief. This is not a reason to abstain from studying and seeking to draw lessons from current intelligence controversies, but it is a caution against making sweeping judgments and proposing disruptive changes based on limited evidence.

Rascoff’s conclusion that the President and his senior staff engage intelligence most regularly in the context of analytic and briefing support or while steering covert actions is valid, insightful, and justifies the close examination of the processes by which equally sensitive collection activities are evaluated. However, it would be mistaken to conclude that, just because collection operations command less of the President’s time, the White House lacks adequate structures, procedures, and expertise to oversee collection — or that such mechanisms did not exist before the current administration put them into place.

Interagency committees to oversee and advise the President on all aspects of intelligence have existed with different names, memberships, and mandates within the NSC virtually since its creation in 1947. In recent decades, under Presidents of both political parties, the NSC staff has included a directorate responsible for intelligence matters led by a senior director who is also commissioned as a special assistant to the President and therefore part of the sitting administration. The NSC’s senior director for intelligence also serves as the chair of a standing interagency committee that coordinates relevant policies and specific intelligence and counterintelligence programs.11×11. See, for example, National Security Presidential Directive-1, Memorandum from the President on the Organization of the National Security Council System (Feb. 13, 2001), http://fas.org/irp/offdocs/nspd/nspd-1.pdf [http://perma.cc/VN6B-LYX6], issued during the administration of George W. Bush, which, among other things, established a Policy Coordination Committee for Intelligence and Counterintelligence.
For a more general discussion of NSC-led interagency policy committees, see Alan G. Whittaker et al., Nat’l Def. Univ., U.S. Dep’t of Def., The National Security Policy Process: The National Security Council and the Interagency System (2011), http://www.virginia.edu/cnsl/pdf/national-security-policy-process-2011.pdf [http://perma.cc/XB5D-VYPP].
These have historically included both proposed and ongoing sensitive collection operations. The findings and recommendations of this committee (presently designated an Interagency Policy Committee, and previously the Policy Coordinating Committee on Intelligence and Counterintelligence) are conveyed to more senior NSC bodies (for example, the “deputies” and “principals” committees) or directly to the National Security Advisor who may initiate further interagency coordination, consult with other principals, raise the decision to the President, or approve the proposed intelligence activity on the President’s behalf. National Security Advisors have historically employed all of these means to evaluate the potential benefits and risks of proposed intelligence operations with the result that some proposals were approved and others disapproved or significantly modified.

During recent administrations, the NSC intelligence directorate also received and reviewed periodic schedules for overseas collection by the Department of Defense under its Sensitive Reconnaissance Operations program, the Department of Justice for collections under the Foreign Intelligence Surveillance Act with possible foreign policy implications, as well as other IC collection programs. The NSC staff reviewed these “scheduled” operations based on the supporting paperwork and in-person briefings, and ultimately provided a recommendation to the National Security Advisor, including on the question of whether the President should be made aware of a planned activity. Through their actions, the President and National Security Advisor signal to the NSC staff — and IC leaders — how deeply they wish to be engaged on sensitive intelligence gathering activities.

Finally, a practice developed over successive administrations of weekly West Wing meetings between the National Security Advisor and the head of the IC (now the Director of National Intelligence, but previously the Director of Central Intelligence).12×12. Nat’l Comm’n on Terrorist Attacks upon the U.S., The 9/11 Commission Report 263 (2004), http://govinfo.library.unt.edu/911/report/911Report.pdf [http://perma.cc/4Z7F-3YXY]. These meetings often included their respective deputies, other relevant agency heads, and senior staff who developed the agenda and documented decisions taken. They offered a regular opportunity for the leaders of the IC to frame and consult on proposed activities, and also to coordinate the flow of information and decisions to the President.

Rascoff notes President Barack Obama’s January 2014 promise to strengthen executive branch oversight of intelligence activities, including through an annual review of “priorities and sensitive targets” by his senior national security team.13×13. Rascoff, supra note 2, at 670 (quoting Press Release, Office of the Press Sec’y, Remarks by the President on Review of Signals Intelligence (Jan. 17, 2014), https://www.whitehouse.gov/the-press-office/2014/01/17/remarks-president-review-signals-intelligence [http://perma.cc/P97T-DNQ6]). It is not yet apparent what new structures or procedures may have been established based on this pledge or their impact on longstanding White House processes to review proposed sensitive collection operations.

IV. Expanding the White House Role in Collection

Notwithstanding the existing and promised new mechanisms for White House supervision of IC collection, Rascoff recommends a still more formal, rigorous, and accountable review process. For example, he usefully highlights the need to expand participation in such reviews to ensure that potential economic and commercial implications are appropriately considered alongside traditional legal, diplomatic, and security interests.14×14. Id. at 671–73. Fresh thinking on how to improve these review mechanisms should be welcomed.

Decisions regarding new collection programs, like all other White House actions, are already extensively documented. However, to the extent that a standardized written template or “collection finding” may make the review process more thorough, efficient, and accountable, this recommendation also deserves careful consideration.

There are, however, a number of serious practical challenges to instituting the regime for review and approval of sensitive collection activities proposed by Rascoff. Several of these are acknowledged.15×15. See id. at 708–12.

First, no President can afford to invest the same amount of time and attention in sensitive intelligence collection activities as he does in his daily briefings and the handful of high-interest covert-action programs. Depending on how “sensitive” is defined and where the threshold for NSC review is set, the number of operations requiring White House attention could easily number in the hundreds — if not thousands. Every clandestine collection activity involves risk, and many also involve sophisticated technologies. Moreover, the equation used to weigh potential gains against risks can change daily over the course of an operation in a fluid overseas environment. Significantly increasing the number of intelligence operations reviewed at the White House would therefore require many more highly qualified staff at the NSC, the Office of the Director of National Intelligence, and other agencies regularly participating in such reviews. The IC’s speed and agility would inevitably suffer.

Second, centralized White House supervision of human intelligence (HUMINT) operations conducted by CIA and other agencies would present unique challenges. Each week, U.S. intelligence officers deployed globally undertake countless operational acts in the course of recruiting and handling human sources. A single miscue can compromise an agent. HUMINT “flaps” invariably incur a diplomatic penalty, often a high one. In truly extraordinary cases, the White House may play a role in approving or supervising a HUMINT operation, but in day-to-day practice there is no alternative to relying on CIA’s chain of command (as informed by the views of U.S. ambassadors overseas) to make sound judgments in pursuing human intelligence.

Finally, if more intelligence collection activities were referred to the White House for approval, it would be important to consider what standard of review the NSC should apply. A textbook formula would weigh the anticipated information gains against the foreseeable risks, but both are often difficult to measure in advance. Rascoff cites approvingly the “front page rule,” a time-honored maxim within national security circles that has recently been advanced as an appropriate standard for intelligence managers evaluating proposed new operations.16×16. Id. at 678; see also Jack Goldsmith, My Speech at ODNI Legal Conference: “Toward Greater Transparency of National Security Legal Work,” Lawfare (May 12, 2015, 8:30 AM), https://www.lawfareblog.com/my-speech-odni-legal-conference-toward-greater-transparency-national-security-legal-work [http://perma.cc/8AXJ-YSHF]. Intelligence managers and their political masters would certainly be well advised to pause and take account of how a secret intelligence or military activity might be described in the New York Times (or even Der Spiegel) if it were disclosed. The “front page rule,” however, does not on closer examination offer a particularly useful standard for decisions by officials actually entrusted with protecting the Nation. Presidents have often been required to make decisions in secret that would not have been ratified by the editors and readers of the Times at the time they were made. The job of a U.S. President would certainly be considerably easier if national security decisions could be grounded exclusively on an assessment of prevailing media and public preferences.

Whatever new structures and processes may be put into place at the White House to review and approve sensitive intelligence activities, it is important to remember that humans will be making these judgments based on incomplete information and often under extreme time pressure. Thoughtful design of such review mechanisms can minimize but never remove the risk from intelligence operations. There will be intelligence flaps in the future.

V. Presidential Control Through Appointments

Borrowing from administrative law, Rascoff also recommends strengthening presidential control over IC collection activities through the appointment of more loyalists to senior intelligence posts.17×17. Rascoff, supra note 2, at 712–16. Rascoff argues that this step would improve the flow of information from the IC to the White House and also ensure the President’s priorities are factored more routinely into decisionmaking about collection and other intelligence activities.

There is, however, little evidence that the White House suffers from a lack of information from the IC. In fact, the opposite is more often the case. The President and his senior national security team are inundated with written intelligence products (both analytic assessments and operational updates) and they meet daily with intelligence officials in different settings. The government’s principal foreign affairs agencies, the Department of State and the Department of Defense, recognize and occasionally resent the IC’s broad access and influence with White House officials.

As Rascoff acknowledges, introducing a large number of nonprofessionals with partisan credentials would, at a minimum, create the appearance of politicizing the functioning of U.S. intelligence.18×18. See id. at 715. The IC’s most valuable asset is, in fact, a carefully cultivated reputation for providing objective, nonpartisan information and judgments. The presence within intelligence agencies of larger numbers of nonprofessionals who owe their current jobs (and future career prospects) to the electoral success of a political party is, at best, a source of bias and, at worst, an invitation to corruption.

There is still another unique aspect of intelligence practice that argues against replicating this feature of administrative law in the U.S. IC: the need for a clear, unambiguous, and accountable chain of command for operations. In this regard, intelligence operations are more similar to military missions than regulatory actions. Successful overseas intelligence operations — that can easily number in the hundreds each day — depend critically on a clear chain of command at all levels. In the most significant cases, the President may give an order to his intelligence chief who, in turn, will direct the operation and report the results. A proliferation of channels purporting to convey direction from the White House to IC agencies or field information from the IC to the White House would needlessly complicate an operation and potentially place lives and other important government interests in jeopardy. The maintenance of multiple and redundant communication channels to and from the White House by political appointees may be beneficial in administering domestic programs, but it represents a potential hazard in the national security field.

U.S. intelligence is singularly focused on the information and policy support needs of the President. If a President does not routinely receive the information, advice, and support he needs from the U.S. IC, he should replace its leader. A larger cadre of political appointees in our intelligence agencies would not benefit the IC or, ultimately, the President who appointed them.


* The Author is the Director of the Intelligence Studies Project at the University of Texas at Austin. He served as a Special Assistant to the President and the Senior Director for Intelligence Programs and Reform on the National Security Council staff from 2005 to 2009.
This Comment was reviewed and approved by the Publications Review Board of the Central Intelligence Agency (CIA) because of the Author’s prior employment with the CIA. This does not constitute an official release of CIA information. All statements of fact, opinion, or analysis expressed are those of the Author and do not reflect the official positions or views of the CIA or any other U.S. Government agency. Nothing in the contents should be construed as asserting or implying U.S. Government authentication of information or CIA endorsement of the Author’s views. This material has been reviewed solely for classification.