1. Introduction
In his half-century Washington career, J. Edgar Hoover lost few bureaucratic battles. He commanded the respect and, not infrequently, the fear of politicians, generals, journalists, and fellow bureaucrats. He infiltrated the civil rights movement and the Ku Klux Klan, authorized raids and wiretaps, built vast databases of dirt on his enemies, and crafted an indelible public image as a puritanically dedicated crimefighter. The Justice Department's Bureau of Investigation could barely cobble together a reliable team of agents when it hired Hoover during the First World War; by the time death removed Hoover from public service in 1972, this formerly obscure federal agency was infamously powerful. Scholars of American political development (APD) have compiled an increasingly long list of epoch-shaping bureaucrats. By any measure, Hoover must rank among the foremost.
Yet Hoover's career has received surprisingly little attention in studies of the American bureaucracy. To some extent, this relative neglect can be put down to the literature's search for generalizable theories. The overarching question of who or what ultimately controls the bureaucracy is not easily answered through biography. Still, the literature boasts rich accounts of the individuals who founded and drove the modern administrative state—from Dorman B. Eaton and George William Curtis to Gifford Pinchot and Harvey Wiley. Although individuals—especially those bearing the unglamorous label of “administrators”—are constrained by social and political forces, they may be capable of maneuvering and harnessing such forces to attain some degree of autonomy. These autonomous bureaucrats might, in turn, leave lasting imprints on American political development.
To be sure, Hoover's story is exceptionally well documented, despite his own attempts to censor it.Footnote 1 However, this story has been only sparsely connected to wider studies of bureaucratic autonomy in the United States. What, if anything, does Hoover's long and influential career say about the ways in which bureaucrats build, retain, and expand their power? What were the principal sources of Hoover's power? How was this power limited? Was this power distinctively tied to the Federal Bureau of Investigation (FBI), or could Hoover be compared to bureaucrats in other federal agencies?
While biographies of Hoover emphasize his ability to consistently get what he wanted in Washington, this article focuses on a rare instance when Hoover suffered defeat.Footnote 2 In late 1945 and early 1946, he waged a campaign aimed at capitalizing on the FBI's wartime growth. During the Second World War, the Bureau had not only been granted a near-monopoly over domestic policing and surveillance, but it had also received from President Franklin Roosevelt unprecedented license to carry out foreign intelligence operations in Latin America. As Roosevelt's successor, President Harry Truman, began to grapple with the postwar world, Hoover argued that the FBI was perfectly placed and amply experienced to make the Latin American model global. Instead, the new Central Intelligence Agency (CIA) was created. The FBI would be strictly confined to the United States, and Hoover would have to content himself with schemes of petty vengeance against the CIA.
I examine this episode through a variety of sources, including a Foreign Relations of the United States (FRUS) special volume on the Emergence of the Intelligence Establishment, internal CIA and FBI histories, and the recollections of contemporary observers within the Truman administration. I argue that Hoover's attempted overreach was thwarted by three interrelated factors: the resistance of Truman himself; the competing postwar plans advanced by the Joint Chiefs of Staff, the State Department, and the Budget Bureau; and the raw fear of an “American gestapo,” felt especially acutely by Truman.
Because these findings emerge from a single case study, their generalizability cannot be taken for granted. Nevertheless, they offer four important contributions for future scholarship. First, they affirm the importance of the presidency in the national security bureaucracy, particularly in keeping bureaucratic autonomy on a tight, domain-specific leash. Second, they challenge scholarly and public perceptions of Hoover as a bureaucrat with near-limitless power. Third, my analysis of Hoover's competitors suggests that national security bureaucrats may not always profit from war. Wars may fatten their budgets, but also attract rivalling generals, diplomats, and other bureaucrats into national security policymaking. Fourth, by demonstrating the significance of “gestapo” fears after the war, I point to the potential force of raw, recent analogies in constraining bureaucratic autonomy within the national security state.
The article proceeds as follows. I begin by outlining the existing literature on bureaucratic autonomy in the United States, highlighting the relatively sparse attention given to the national security state, the FBI, and Hoover. I then delve into the archival record on the post-WWII national security bureaucracy, describing Hoover's plan for postwar intelligence and his abortive efforts to impress the plan upon the Truman administration. I analyze how and why these efforts failed, then evaluate the extent to which Hoover's defeat reveals broader constraints on bureaucratic autonomy.
2. Bureaucratic autonomy and the national security state
American bureaucrats cannot completely insulate themselves from politics. Presidents regularly shape and control the bureaucracy through the appointment and removal power, while Congress creates, funds, and establishes the administrative procedures for most federal agencies. On paper, at least, bureaucratic power is delegated power, flowing exclusively from the legislative and executive branches of government.Footnote 3
Nevertheless, as Daniel Carpenter has demonstrated in his works on the Food and Drug Administration (FDA) and a variety of other federal agencies, some bureaucrats have mastered the art of carving out their own power bases. Anthony Comstock was a middling postal inspector who, after combining forces with antivice interest groups, led an officially sanctioned moral crusade so influential that he became an “ism.” Gifford Pinchot, while head of the U.S. Department of Agriculture's Forestry Division, established the Society of American Foresters to promote conservation and lobby for the creation of the U.S. Forest Service in 1905. Harvey Wiley mobilized scientific expertise and muckraking journalism to expand the Bureau of Chemistry and eventually drive the creation of the FDA. The FDA has since become an integral component of the American administrative state, often taking on new tasks and wider realms of action before the statute books can catch up.Footnote 4
By cultivating coalitions and public profiles, these bureaucrats developed reputations for fairness, effectiveness, and neutrality. Although they could, in theory, still be defunded, abolished, or fired by politicians on Capitol Hill or in the White House, they became capable of confronting the nation's most feared elected leaders, as demonstrated when Harvey Wiley successfully stamped out unsanitary whiskey distilleries in Speaker “Uncle Joe” Cannon's congressional district.Footnote 5
Other scholars have added to Carpenter's list of savvy bureaucratic operators. Colin Moore has documented how colonial bureaucrats allied with Wall Street financial interests to drive imperial expansion in the late nineteenth century against a heavy tide of congressional skepticism. Andrew Kelly has highlighted how U.S. Geological Survey and U.S. Coast Survey administrators overcame crippling funding constraints by constructing a “complex network of public-private science” centered on major American universities. William Adler has described a less expansive “conditional autonomy” achieved by the Army Corps of Topographical Engineers before the Civil War, which was bolstered by Col. John J. Abert's control over information and intelligence, but critically “lacked significant linkages to outside groups or a network of committed supporters in the private sector.”Footnote 6
An important theme of these works is that autonomy does not mean omnipotence. To be sure, if a bureaucrat successfully confronts a politician—as Harvey Wiley confronted Joe Cannon—this would lend support to any argument for their autonomy. However, the more skillful bureaucrat is adept at selling policy goals to politicians over a long period of time, building a base of meaningful political support that delivers concrete policy outcomes. As Patrick Roberts puts it, “Agencies demonstrate autonomy when they act independently, though not necessarily in defiance, of their political superiors or other agencies for a sustained period.” This concept of autonomy as the power to make policy is relative and domain-specific: There is no magic threshold beyond which autonomy is secured, and autonomy in one policy realm is not necessarily transferable to others.Footnote 7
Because this understanding of autonomy allows for cooperation among bureaucrats and politicians, skeptical scholars have had no difficulty reinterpreting supposedly seminal cases of bureaucratic autonomy as cases of congressional or presidential control.Footnote 8 Yet a more substantial critique of the bureaucratic autonomy literature concerns its focus on some bureaucrats over others. The national security bureaucracy, in particular, has received a surprising dearth of attention. How, if at all, do national security bureaucrats act independently of their political overseers? What are the sources of bureaucratic autonomy within the national security state? How is this autonomy limited?
To the extent that scholars have examined the national security state, they have focused primarily on its origins. The leading work in this area is Amy Zegart's on the CIA, the Joint Chiefs of Staff (JCS), and the National Security Council (NSC). Zegart explains the evolution of these post-WWII institutions through a path-dependent “new institutionalist framework,” locating their overlapping and frequently confusing mandates in the messy politics of the 1947 National Security Act. Zegart includes the FBI in her more recent studies of the intelligence failures leading up to September 11, 2001, while retaining her emphasis on structural “design flaws” traceable to 1947, including interagency rivalry, patchy congressional oversight, and limited integration of intelligence gathering.Footnote 9
Zegart is less concerned with autonomy than she is with effectiveness.Footnote 10 The main puzzle in her analysis is why the national security bureaucracy frequently seems to fail, not how people accrue power within it. Although Michael Hogan's account of the National Security Act is more attentive to the role of key individuals, Hogan primarily confines himself to elected leaders—most notably Harry Truman. In other words, Zegart, Hogan, and other scholars of the 1947 institutions (the CIA, the NSC, and the JCS) provide rich accounts of the national security bureaucracy, but more limited accounts of national security bureaucrats.Footnote 11
Scholars focusing on different time periods have more squarely zeroed in on the question of bureaucratic autonomy. Patrick Roberts, for instance, has suggested that the CIA and FBI possess a distinctive ingredient favoring autonomy, because “they exercise executive power, performing tasks that are so urgent, secretive, or forceful that they cannot be anticipated by law.” This “executive power” lessens the need for the “reputational power” emphasized by Daniel Carpenter. A security agency can, therefore, become relatively autonomous without a glowing public image, especially if its legal mandate is broad. According to Roberts, this fact is reflected in the resilience of the FBI since 9/11.Footnote 12
Unsurprisingly, war has frequently been cited as a driver of bureaucratic autonomy within the national security state. Arguably, as Mark R. Wilson has suggested, this tendency can be traced as far back as the Union Army's Quartermasters General during the Civil War, who were empowered by the necessity of supplying troops during a supreme national crisis. An analogous case is the growth of the Justice Department's Bureau of Investigation during the First World War, when the possibility of sabotage and subversion provided an ideal justification for a more vigorous federal policing apparatus. Meanwhile, critics of the post-1945 national security state have claimed that the Second World War led to the creation of a bureaucracy so unwieldy and unaccountable that it could be labeled a “shadow government.”Footnote 13
Thus, although the bureaucratic autonomy literature has only partially addressed national security policy, several overarching themes have still emerged. First, the national security bureaucracy consists of an often bewildering patchwork of agencies and mandates. Second, the reputational power emphasized by Carpenter and others may not be necessary for national security bureaucrats who wield blunt tools of executive power. Third, wars provide fruitful opportunities for the expansion of bureaucratic autonomy within the national security state.
Yet one common limitation of these studies is their reluctance to develop the portraits and narratives of individual bureaucrats that have so enriched the broader bureaucracy literature. Where, for instance, is the FBI's Hoover, whose career spanned forty-eight years, eight presidents, eighteen attorneys general, and two world wars? Athan Theoharis has raised the question of Hoover's bureaucratic autonomy in his analysis of FBI wiretapping. Although this surveillance tactic was prohibited by Congress under the Federal Communications Act of 1934, President Roosevelt took the bold step, on August 24, 1936, of orally authorizing a clandestine counter-subversive FBI wiretapping program. Attorneys General Jackson and Biddle later provided carefully worded written approvals. Although these approvals limited wiretapping to matters concerning national security, Hoover implemented a far more sweeping program, beginning the immense surveillance operations that would define the FBI in the postwar era. For Theoharis, one of Hoover's greatest skills was the manipulation of presidential directives. The FBI director fastidiously documented all conversations with the commander-in-chief, making sure that any hint of executive branch support could be used to evade congressional scrutiny and stretch the law to its limits. When combined with a publicity-conscious war on organized crime, these tactics ensured that Hoover (in Kathleen Frydl's words) “simultaneously built a formidable organizational reputation while exercising a supple and jealous dominion over it.”Footnote 14
Hoover, on this evidence, possessed some degree of bureaucratic autonomy as early as the 1930s. Earlier—particularly during the First World War and its subsequent Red Scare—he showed flashes of his potential power, but the Bureau of Investigation was sharply reined in by Attorney General Harlan Fiske Stone in 1924. The 1930s, however, offered Hoover the “crime wave” of the Great Depression, during which the FBI (as it was restyled in 1935) was empowered by crime bills aimed at combatting “public enemies” such as John Dillinger, Charles “Pretty Boy” Floyd, Lester “Baby Face” Nelson, Alvin “Old Creepy” Karpis, Clyde Barrow, and Bonnie Parker. Fears of communist and fascist subversion—harbored to varying degrees by the Roosevelt administration and key members of Congress—played further into Hoover's hands as war in Europe again loomed. When the United States finally entered that war, the FBI's resources and duties continued to expand. By 1945, the Bureau could boast nearly 5,000 special agents and over 8,300 supporting staff: five times what it had in 1940, and with a budget three times larger than the prewar level.Footnote 15
Thus, Hoover's ascendancy in Washington has been well documented by a variety of biographers and historians, even though Theoharis and Frydl stand alone in connecting this ascendancy with the broader literature on bureaucratic autonomy. To use the terminology of the literature, it would be reasonable to surmise that Hoover wielded both executive and reputational power as WWII approached. In pursuit of domestic subversives, FDR gave Hoover license to act in evasion, if not defiance, of a congressional ban on wiretapping. Meanwhile, Hoover carefully cultivated his public reputation for incorruptible and relentless crime-fighting. He was firmly in the business of both federal law enforcement and national security, expanding his bureaucratic empire seemingly without limit.
Hoover, in short, represents a compelling, but underexplored, case of bureaucratic autonomy within the national security state. In the next section, I document a rare episode in Hoover's career when the limits of this autonomy were starkly exposed. In histories of both Hoover himself and the national security state, the episode—which saw Hoover fail to impress his plan for post-WWII intelligence upon the Truman administration—has hardly featured prominently. Truman's own firsthand account merely describes “certain objections on behalf of Director J. Edgar Hoover of the FBI”; Hogan's sizeable history of the postwar struggles over intelligence and military unification recounts Hoover's role on a solitary page.Footnote 16
Although Hoover's early marginalization from the decision-making process helps explain why he is almost forgotten in the origin stories of the national security state, the causes of Hoover's marginalization may shed valuable light on the constraints faced by powerful national security bureaucrats. It is tempting to conclude that Hoover simply overreached. He did, after all, propose a single foreign and domestic intelligence agency that looked suspiciously like a gigantic FBI. However, Hoover's defeat was not inevitable. Along with the many facets of his strength and autonomy outlined above, he could make a relatively strong case for a leading FBI role in postwar global intelligence. The Bureau had already completed a test-run during the war through its Special Intelligence Service (SIS) in Latin America, and its primary competitor, the Office of Strategic Services (OSS), had been disbanded after the war. Hoover was well positioned to fill this vacuum, but he utterly failed to do so. Even if—given the scale of his ambitions—his failure was not a total surprise, it nonetheless requires explanation.
3. Tracing Hoover's defeat
The surface-level facts of Hoover's post-WWII defeat are relatively straightforward. In August 1945, he began lobbying for a peacetime “World-Wide Intelligence Service.” His idea, pitched to the attorney general on August 29, was simple: a single intelligence agency with international and domestic responsibilities. This would be efficient, integrated, and effective. While Hoover outwardly maintained that “I am not seeking for the Federal Bureau of Investigation the responsibility for world-wide intelligence system,” he contradicted this façade of modesty by arguing “it is most logical that the system which has worked so successfully in the Western Hemisphere should be extended to a world-wide coverage.” That “Western Hemisphere” system, conveniently enough, was operated by the FBI. Hoover justified this argument for a global FBI by suggesting that “foreign and domestic civil intelligence are inseparable and constitute one field of operation.” He continued: “The German-American Bund and the Italian Fascist organizations in the United States originated and were directed from abroad. The Communist movement originated in Russia but operates in the United States. To follow these organizations access must be had to their origin and headquarters in foreign countries as well as to their activities in the United States.” A single agency with domestic and international purview was necessary in the modern world.Footnote 17
Despite these attempts to market his proposal to the Truman administration, Hoover was on the ropes by January. Increasingly, the administration embraced an alternative vision of postwar intelligence. Instead of a single agency with global and domestic responsibilities, there would be a new agency strictly confined to global intelligence. The FBI would be strictly confined to domestic intelligence. This vision was confirmed in the “Magna Carta” of the post-WWII national security state, the 1947 National Security Act. Unambiguously, the act declared that the new CIA “shall have no police, subpoena, law-enforcement powers, or internal-security functions.” Foreign and domestic intelligence would be discrete.Footnote 18
Why was Hoover marginalized so swiftly and comprehensively? The most obvious answer from the documentary record consists of two words: Harry Truman. Even though President Roosevelt warmly embraced the FBI and Hoover, Truman was wary of the Bureau from the moment he assumed office. In September 1945, with the war over and his administration still in its infancy, Truman told Budget Bureau Director Harold Smith (in Smith's words) “that he thought the FBI should be cut back as soon as possible to at least the prewar level; that he proposed to confine the FBI to the United States; and that he had in mind a quite different plan for intelligence.” Smith's papers also document a meeting in May, when the new president questioned Hoover's much-vaunted Latin American operations: “If we continue our present attitude toward Latin American countries in this respect we will not be in a position to complain very much when they send their intelligence people into the United States.” In July, Truman underlined “that he, himself, had some question, from the standpoint of good neighbor relations, about our having the FBI in South America.”Footnote 19
Although Truman apparently “displayed a most congenial attitude toward the [FBI] and stated that he felt that it was the most efficient organization in Government service” when he met with FBI representatives in October,Footnote 20 Hoover quickly detected his new boss's lack of enthusiasm. As early as September, he was telling Attorney General Tom Clark that “a confidential but thoroughly reliable source” had informed him of the president's desire for the FBI “to act only as a domestic agency” in the postwar world.Footnote 21
Hoover's sources within government were usually well-placed, and his fears were confirmed when Truman issued a directive on January 22, 1946, establishing a National Intelligence Authority (NIA) and Central Intelligence Group (CIG). Truman had already dissolved the OSS through Executive Order 9621 on September 20, splitting its intelligence functions between the War and State Departments. This January directive outlined what would come next. Neither the FBI nor the Department of Justice were mentioned. Instead, the Secretaries of War, State, and Navy would oversee “all federal Foreign intelligence activities.” Paragraph 4 declared that “no police, law enforcement or internal security functions shall be exercised under this Directive,” and paragraph 9 affirmed that “Nothing herein shall be construed to authorize the making of investigations inside the continental limits of the United States and its possessions, except as provided by law and Presidential directives.”Footnote 22
Here, Truman was establishing the precise formula that Hoover opposed: a strict demarcation between foreign and domestic intelligence. The president would take the same formula to Congress when proposing the National Security Act, and Congress would retain it. Truman had decided, early, that he opposed Hoover's plan, and he used the levers of his office to present an alternative. First, he set the tone with an executive order and a presidential directive, both of which reflected the presidency's first-move advantage in national security and foreign affairs.Footnote 23 Second, by making his own proposal public, he clearly signaled his intentions to Congress, the press, and voters. Third, by dissolving the OSS and establishing the NIA and CIG, he unilaterally began the task of bureaucratic reorganization.
Finally, Truman easily rebuffed Hoover's attempts at resistance. After Hoover learned that Colonel Alfred McCormack had been “assigned to the State Department for the purpose of forming a World-wide Intelligence organization,” he confided to Tom Clark, “I am not at all optimistic as to the sympathy which Colonel McCormack may have toward our program.”Footnote 24 Hoover then spent the succeeding months making life difficult for the interim director of Central Intelligence, Sidney Souers, by threatening to rapidly pull all FBI personnel from Latin America before an orderly transition could be facilitated. This caused some discomfort to the president and to Souers, but ultimately did nothing to scupper their postwar plans.Footnote 25
Why was Truman so wary of Hoover and the FBI? Truman's concerns about the FBI in Latin America, and its implications for “good neighbor relations,” suggest that the FBI director may have lost credibility during the Bureau's trial run at overseas intelligence during the war. An internal history of the Bureau's wartime SIS Latin American operations that was declassified in 2004 hints at objections raised by critics. With President Roosevelt's approval, agents were first sent to Peru, Argentina, Brazil, Uruguay, and Venezuela in September 1940. Even in the euphemistic, defensive parlance of internal agency historiography, problems and limitations are freely conceded. “As might be expected due to circumstances over which the Bureau had absolutely no control,” the History of the SIS Division states, “the volume of intelligence information collected from each agent was in the beginning and for some time thereafter quite small and of little real value. The agents were, of course, more or less completely unfamiliar with the countries in which they were trying to operate and usually deficient with regard to the use of the language thereof.”Footnote 26
Although, in the Bureau's own recounting, the efficiency and professionalism of the SIS improved over time, “mistakes and errors” had to be acknowledged. “It was,” for instance, “definitely a mistake to undertake the establishment of intelligence coverage solely on the basis of clandestine operations,” as this inhibited cooperation with the State Department. In addition, “Agents were briefed far too hurriedly and sent out on assignment far too rapidly for proper assimilation and adjustment into the program with resultant ill effects”; they were “younger, and more inexperienced than was desirable”; they “suffered from a lack of adequate supervision, administrative discipline and direct contact with the Bureau”; and “the handling of cover work … was extremely faulty and weak until comparatively late in the SIS program.”Footnote 27 If the faults documented in this internal study were known outside the Bureau, then the consequences could not have been positive for the prestige of Hoover.
However, the FRUS volume on the establishment of the postwar intelligence system displays almost no criticisms of the SIS, even behind Hoover's back. If anything, the record points in the opposite direction. An internal memorandum circulated by the Budget Bureau's Arnold Miles in September 1945 discussed “what funds should be allotted to the FBI for the continuation of their secret activities abroad.” Without qualification, Miles described the SIS as “highly useful especially in connection with the desire to ferret out Nazi, and to some extent Japanese, infiltration into the Western Hemisphere.” He added that “at budget time, the program has been strongly endorsed by those officials in the State Department directly concerned with the use of FBI material,” and observed that “there is no instance known to us, in which the SIS has caused any official embarrassment with the countries involved.” The following year, Acting Secretary of State Dean Acheson also praised “the excellent FBI organization in Latin America” in a memorandum to the members of the NIA, on which Hoover did not sit.Footnote 28
There is further evidence that Hoover was seen as a source of valuable expertise. For instance, Navy Secretary James Forrestal's special assistant conceded to his boss early in the postwar bureaucratic battle that he was not yet willing to assess the peacetime intelligence plans “as I wish to give the matter further consideration and particularly to obtain the views of people with experience in this field such as J. Edgar Hoover.” Likewise, the War Department's committee that aimed at studying and proposing postwar intelligence plans invited Hoover to appear as a witness, seeking out his “views and recommendations” for the grand design of espionage during peacetime.Footnote 29
Truman, then, was not receiving negative assessments of Hoover's wartime conduct. The FBI director had a largely untarnished reputation, especially as his dirtier domestic tricks—including the full scope of wiretapping—remained tightly concealed by his “Official/Confidential” classification system.Footnote 30 Nonetheless, Hoover did have his enemies within the executive branch, particularly the OSS founding father, William “Wild Bill” Donovan. In part, this was a festering feud from their less-than-friendly time as Justice Department colleagues in the early 1920s. Yet the tension also stemmed from Donovan openly plotting to extend something like his OSS into peacetime. A later CIA internal history, which was sympathetic to Donovan, admitted that “Wild Bill” was concerned with “Empire-Building” from the moment President Roosevelt handed him wartime responsibilities, “look[ing] out upon vast stretches of territory” including “information, intelligence, propaganda, morale, espionage, counterespionage, subversion, military operations, strategic planning, and postwar planning.”Footnote 31
None of this was lost on Hoover. Although the FBI director had successfully prevented any OSS encroachment on the FBI's turf in Latin America during the war, he showed signs of postwar anxiety in a memo to Attorney General Clark on September 6, 1945. In typically Hooverian, gossip-mongering language, he warned, “I have received information today from two independent outside sources relating to General William Donovan's plans for the perpetration [sic] of his dynasty within the Office of Strategic Services by the continuation of his agency under another name in the world-wide intelligence field, which I thought would be of interest to you.”Footnote 32
To some extent, though Hoover's fears were not groundless, they may have been exaggerated. The OSS—Donovan's fiefdom—was not preserved but dissolved, unceremoniously, after the war. Donovan was then excluded from the various planning forums and committees constructed by the federal government to deal with postwar intelligence questions.Footnote 33 If there was any consolation for “Wild Bill,” it came from the fact that Truman sided, in broad terms, with his initial vision. Donovan had formally written Truman within three weeks of the dropping of the first atomic bomb, proposing a World-Wide Intelligence Service “prohibited from carrying on clandestine activities within the United States,” and that would be “forbidden the exercise of any police functions either at home or abroad.”Footnote 34 When examining the final formula decided upon by the White House and passed into law by Congress, Donovan must have been happier than Hoover.
Hoover faced a larger threat from Harold Smith and the Budget Bureau. Smith corresponded directly and frequently with President Truman, and his bureaucrats were capable of producing detailed, dispassionate, and potentially bruising appraisals of the postwar intelligence plans. Painfully for Hoover, one such appraisal, shared in October 1945, subtly skewered the FBI plan, noting that “other than stating that the plan is similar to that in operation in South America and supplying a chart,” its descriptive document “only contains three paragraphs outlining the plan itself.” Further, “these three paragraphs provide for a joint operation in every country of the world without stating how joint operation is to be achieved; for a top group to determine basic policy, and for an operational committee, without stating either what basic policy is or what the operational committee would do.” Finally, “in using the South American experience as the basis for planning a world-wide system the proposal fails to consider the vast difference between the two situations”—namely, that “in South America, our operations were not directed primarily at the countries in which they were conducted. Our operation there was not secret in the sense that it would have to be in the big league.”Footnote 35
The Budget Bureau's staff also spent “months” building common ground with the State Department, allowing Director Smith to present a State-backed “progress report” to the president in October 1945 that both described steps already taken “to readjust the Government's intelligence activities to a post-war basis” and laid out a future model of discrete internal and external “coordinating authorities.” The eventual State/Budget Bureau plan was highly complex, containing not just these two authorities, but also a common “executive secretariat,” two “advisory groups,” two deputy secretaries, and numerous subcommittees and planning groups. With President Truman yet to provide his own template for postwar intelligence, Smith felt bold enough to tell his commander-in-chief, in a November 28 memo, that this plan was the one that “you” (that is, Truman) had in mind.Footnote 36
In short, even though the bureaucracy literature suggests that war empowers national security bureaucrats, Hoover found himself facing stiff competition from the seemingly innocuous Budget Bureau—in cahoots with the State Department—after the Second World War. Harold Smith always praised the FBI for its wartime role in Latin America, but pivoted to highlight its limitations in the postwar world. Budget's cutting analysis of the FBI intelligence plan also exposed a weakness in Hoover's favored method of bureaucratic jostling: a brief proposal in memo form with the aim of testing the waters, before providing something more substantial.Footnote 37 What, to Hoover, was an effective tool of influence, Smith's men painted as sloppiness and complacency.
Along with Harold Smith, the FBI director was up against President Truman's military advisors, who had taken on an increasingly prominent policymaking role during the Second World War. Their proposed intelligence plan, presented to the president with the JCS imprimatur, shared State/Budget's separation of “intelligence” and “policing,” but one of its key sponsors, Missourian Admiral Sidney Souers, suggested in a December 1945 memo that “differences between the two plans are far greater, and more fundamental, then they appear to be on the surface.” Above all, Souers argued, the State/Budget plan would allow the secretary of State to “determine the character of intelligence furnished to the president,” whereas the JCS plan envisaged a combined role for the Navy, War, and State Departments in providing information to a director of Central Intelligence, who—temptingly—would be appointed by the president.Footnote 38
By early 1946, these arguments had won over Truman. “My inclination,” he reflected, “was to favor the plan worked out by the Army and the Navy, with the aid of Admiral Souers.” His precise reasoning is unclear, although it is probable that he was attracted to the comparative simplicity of the JCS plan, presenting, as it did, “an agency which was assisted by a board,” rather than “one or two authorities directing an interdepartmental staff, which, assisted by two advisory groups, directed numerous committees.”Footnote 39
The full realization of the plan would still have to await congressional approval in July 1947, while the interim intelligence bodies would include the interdepartmental staffing arrangements desired by State/Budget. To this extent, there was some give-and-take between the two plans that ended up on Truman's desk in late 1945 and early 1946. Crucially, however, neither of these were authored by Hoover. The Budget Bureau and the JCS had successfully sidelined the FBI director. The war, in other words, created both opportunities and challenges for this formidable national security bureaucrat. On the one hand, Hoover had more money, staff, and formal responsibilities than ever before. On the other, he faced bureaucrats, diplomats, and generals determined to extend their wartime roles. In this context, the reach of Hoover's autonomy was severely constrained.
Still, the underlying reason for Truman's appetite for anything-but-Hoover remains unclear. Indeed, it seems somewhat puzzling given the degree to which Hoover's plan promised to empower the president. From the beginning, Hoover pitched his plan not only as an extension of effective wartime espionage, but also as a politically convenient option that would remove “the necessity for any legislative enactment creating operating agencies or empowering them to act.” He continued: “If, on the other hand, the General Donovan plan or even the plan presently under consideration by the State Department is accepted, it will be necessary to seek Congressional authority for the program and to obtain funds which will be earmarked for and otherwise identified as being for the operation of an international espionage organization.”Footnote 40 If Truman's principal motives were based on self-aggrandizement, he should have leapt at this prospect of a peacetime intelligence apparatus constructed without congressional meddling.
That he did not is partly due to conflicting legal advice underlining the need for congressional approval, yet records of Truman's personal reflections and private conservations point to a deeper fear of centralized intelligence.Footnote 41 Harold Smith's papers document a meeting discussing postwar intelligence as early as May 1945, when the president insisted “with considerable vigor that he was ‘very much against building up a gestapo.’” Similarly, when Hoover sent his subordinate, Morton B. Chiles, to meet with Truman and push Hoover's plan in October, Chiles reported back that although the president “agreed that there could be no satisfactory separation in the handling of domestic intelligence and foreign intelligence,” he also “expressed concern regarding the possibility that a World Wide Intelligence organization would gain the reputation of a ‘Gestapo.’”Footnote 42
Hoover had anticipated this concern in his initial planning memo. “A hazard in intelligence operation,” he granted, “is the possibility of a charge being made that the organization is a ‘Gestapo.’ Also, a police agency which engages in intelligence operation may be called a ‘political police.’ Both charges are obnoxious to American citizens.” In an attempt at reassurance, Hoover argued that “the set-up operating in the Western Hemisphere throughout the war has engaged in both police and intelligence activities and its record of protecting civil liberties has been highly praised even by the American Civil Liberties Union.”Footnote 43
Unfortunately for Hoover, his rivals were already tarring his plan with the label of Nazi-like centralization. The Budget Bureau's deputy director played on Truman's “gestapo” fears by describing Hoover's intelligence plan as “permit[ting] FBI to have all responsibility for secret intelligence under only the mildest kind of direction.” Similarly, Harold Smith told Truman that plans like Hoover's had a “backbone” built on “centralized secret operations.” Donovan, meanwhile, had long argued that a “central agency” would not inevitably “mean the establishment of a Gestapo,” so long as it would have “no powers of arrest either at home or abroad,” have “no authority to exercise surveillance at home,” and deal “only with intelligence in foreign affairs.” As Donovan and other opponents of Hoover well knew, Hoover's proposed plan for an all-encompassing intelligence agency could not offer these assurances. If any plan promised a gestapo, Truman heard, it was Hoover's.Footnote 44
Without much subtlety, Truman appeared to associate Hoover and the FBI with the excessive secrecy and dangerous power of totalitarian spy agencies. In an April 1946 Special Conference with the American Society of Newspaper Editors, Truman was asked why he did not integrate the FBI more fully into his CIG. “Well,” he responded, “you want to be very careful in any of these things. What we have to guard against is a Gestapo, in this instance, and a military dictatorship, in the setting up of a national defense program … I have got no business giving you a lecture on free government, however [laughter].” Later, Truman wrote his wife, Bess, vowing that “if I can prevent [it] there'll be no NKVD or Gestapo in this country. Edgar Hoover's organization would make a good start toward a citizen spy system. Not for me.” Another presidential confidant, Clark Clifford, concluded that Truman was “very strongly anti-FBI” because he was deeply “afraid of a ‘Gestapo.’”Footnote 45
The “gestapo” charge resonated beyond the White House, especially among congressional critics of the National Security Act. “I am very much interested in seeing the United States have as fine a foreign military and naval intelligence as they can possibly have,” Congressman Clarence Brown (OH-R) stated in the House hearings on the new legislation. “But I am not interested in setting up here in the United States any particular central policy agency under any President, and I do not care what his name may be, and just allow him to have a gestapo of his own if he wants to have it.” When Admiral Inglis of the JCS was questioned by House Expenditures Committee members in closed-door hearings, he underlined how the separation of domestic and international intelligence was the main safeguard against a gestapo. “It is imperative,” he argued, “not only for the production of good intelligence, but for the defense of the American principle of government, that there be no confusion between the pursuit of intelligence abroad and police powers at home.” “It is significant,” he continued, “that the merging of these two fields is characteristic of totalitarian states. Domestic security and foreign intelligence were controlled by the same hands in the last years of the Nazi state; they have always been in the same hands in the Soviet Union.”Footnote 46
As David Barrett has demonstrated in his history of the early CIA, congressional debates over the National Security Act resembled a battle of analogies: opponents and skeptics of peacetime intelligence agencies invoking the WWII Gestapo, and supporters invoking the specter of another Pearl Harbor—a catastrophe that had been blamed, by Congress itself, on the lack of robust intelligence. Lawmakers like Senator Edward Robertson of Wyoming, who predicted “an American gestapo” regardless of how Congress finessed the statutory language, ultimately lost the fight. Yet their colleagues, who were willing to tolerate some form of permanent spy agency, but remained anxious about gestapos and NKVDs (Soviet secret police), could find comfort in the National Security Act's clear division between foreign and domestic realms.Footnote 47
Analogies, of course, are often post hoc justifications, rather than genuine heuristic devices. The use and abuse of “the lessons of Munich” in support of military force is a prime example, vindicating A. J. P. Taylor's memorable adage, “Men use the past to prop up their prejudices.”Footnote 48 Although this possibility cannot be ruled out in the debates surrounding the foundation of the American intelligence bureaucracy, it is hard to pin down any obvious ulterior motives harbored by lawmakers or by President Truman that might have been concealed with “gestapo” references.
It would, undoubtedly, be naïve to conclude that Truman and other opponents of Hoover's intelligence plan were driven by solemn fidelity to civil liberties or the constitution. Still, it is striking to observe that, while opponents of Hoover disagreed on multiple issues—including civil-military relations under the National Security Act, term limits for CIA directors, staffing, and funding—they were unified in their revulsion toward a single intelligence regime encompassing domestic and international realms. The eccentric “Wild Bill” Donovan, the bookish Harold Smith, the straight-talking Harry Truman, lawyerly congressional committee chairs, generals, and admirals all lined up against the FBI director, even though most respected and admired him. Much of this common ground was founded on self-interest: budgets, turf, and prestige. However, especially in the most important place—the White House—the recent memory of the world's most notorious spy agency played no small part in ensuring that the American intelligence bureaucracy would not be based on Hoover's vision.
4. Hoover in historical perspective
Hoover might have lost the battle and won the war. Not only did he continue to amass power through the FBI in the second half of the twentieth century, but his vision—that “foreign and domestic civil intelligence are inseparable and constitute one field of operation”—also has been widely embraced by the national security bureaucracy since September 11, 2001. Seventeen intelligence agencies are now spread throughout cabinet departments under the authority of a Director of National Intelligence, and as the revelations of Edward Snowden have shown, these agencies have often paid scant respect to formal distinctions between domestic and foreign jurisdictions.Footnote 49
Perhaps mainstream political aversion to centralized intelligence agencies in the United States has diminished over time. It is almost comical to imagine George W. Bush invoking the gestapo when pondering the creation of a new, enormous, integrated Department of Homeland Security in 2002. Even Democratic members of Congress who labeled President Trump an aspiring dictator were willing to grant his security agencies generous budget increases and expansive surveillance powers.Footnote 50
Hoover, it seems, was stymied by fears of a police state that were specific to the post-WWII context. Should scholars of the national security bureaucracy therefore relegate Hoover's abortive “World-Wide Intelligence Service” to the category of historical curiosity? Although this conclusion is tempting, it overlooks the potential for future scholars to examine the wider role of analogies in American bureaucratic development. The gestapo analogy was both an instrument exploited by Hoover's rivals and an underlying driver of Truman's opposition to Hoover. How and under what conditions have other analogies shaped key struggles over the national security bureaucracy? Have some presidents—possibly Truman-like “accidental” presidents—proved more open to analogical reasoning than others?Footnote 51
Beyond these questions, Hoover's postwar defeat is a prime example of presidential control over national security bureaucrats. Hoover's domestic and international responsibilities had grown considerably with the aid of President Roosevelt. It would not be a stretch to argue that the FBI's bureaucratic autonomy stemmed primarily from presidential support. Equally, however, President Truman kept this autonomy in its box. Congress, of course, still had to wrangle over the final plan for a peacetime global intelligence bureaucracy, but by that point Truman had already pushed Hoover's plan way off the table. To the degree that Hoover retained bureaucratic autonomy, he did so only within his existing policy domain.
Few scholars of the American national security bureaucracy would be shocked by the prevalence of presidential power in this case study. Instead, the original value of this finding lies more in its challenge to existing scholarly accounts of Hoover. Most of these studies emphasize the seemingly inexorable expansion of Hoover's power throughout his half-century career. However, the failure of his postwar intelligence plan provides a useful case study of how his power was limited. It underlines in a stark light how bureaucratic autonomy is relative and domain-specific.
This article has also demonstrated the crucial role of Hoover's rivals in orchestrating his defeat. The Budget Bureau, State Department, and JCS were particularly significant actors. The war had propelled the JCS to the forefront of political decision making, while State and Budget constituted other formidable barriers to Hoover's global ambitions. The APD and comparative politics literature have explored various avenues through which bureaucrats accumulate power during wartime, but here was a case of war leading to a more crowded field of bureaucratic competitors. By demonstrating the negative consequences of this competition for Hoover, this article has invited future scholars of the national security state to examine not only how postwar bureaucratic jostling creates path-dependent inefficiencies and flaws, as demonstrated most notably by Amy Zegart, but also how this jostling affects the autonomy of individual bureaucrats.
Thus, even if Hoover's postwar defeat was a distinctive historical event, it carries important implications for studies of the national security bureaucracy. The case itself may not be easily generalizable, but its core themes—presidential power, fears of centralized intelligence, and bureaucratic competition—are relevant and testable beyond the scope of this article. By documenting the limits of Hoover's autonomy, I have also placed his career in a more measured historical perspective. Ultimately, the sad fate of Hoover's World-Wide Intelligence Service is an enduring testament to the stubborn constraints faced by one of this nation's most accomplished bureaucratic empire builders.
Acknowledgments
I would like to thank the anonymous reviewers and Dr. Susan Woodward for their helpful comments on the manuscript.
