Andrew Krepinevich and Barry Watts, The Last Warrior: Andrew Marshall and the Shaping of Modern American Defense Strategy (New York: Basic Books, 2015), 336 pp., $29.99.

I FIRST met Andrew Marshall, the longtime director of the Pentagon’s Office of Net Assessment (ONA), in the mid-1990s. The occasion was one of the late Harvard professor Samuel P. Huntington’s “Strategy and National Security” conferences at the Wianno Club on Cape Cod. A number of Huntington’s students, including Eliot Cohen, Aaron Friedberg and my Olin Institute for Strategic Studies colleague Stephen Rosen, were also Marshall protégés—alumni of St. Andrew’s Prep, as they referred to themselves—having spent some of their careers under his tutelage in the ONA.

The conference gave me my first taste of the reverence with which they held Marshall. Rosen had arranged to recognize him at the dinner for the thirty or so national-security experts in attendance and had carefully selected a memento to mark the event. It was a handsomely framed print of Jean-Léon Gérôme’s painting of François Leclerc du Tremblay, the Capuchin monk who was Cardinal Richelieu’s alter ego. Du Tremblay was so influential that French courtiers referred to him as his “Gray Eminence,” in deference to the authority he reputedly exercised behind the scenes belied only by the color of his humble friar’s habit.

It thus does not come as a surprise that in their new book The Last Warrior, two more St. Andrew’s alums, Andrew Krepinevich and Barry Watts, adopt a similarly reverential tone. They trace the arc of Marshall’s career, beginning with his early days at the newly established RAND Corporation through his founding in 1973 and long-term directorship of the ONA, from which he is slated to retire in 2015 after sixty-five years of U.S. government service. He was, they tell us, “an intellectual giant comparable to such nuclear strategists as Bernard Brodie, Herman Kahn, Henry Kissinger, James Schlesinger, and Albert Wohlstetter,” and was one of the most visionary thinkers of the post–Cold War era by virtue of his advocacy of the “revolution in military affairs” (RMA).

With their worshipful tone and appropriation of the parochial-school designation for Marshall’s coterie, the authors invite comparisons with the doctrines of the Catholic Church. To my well-catechized eye, it all resembles the panegyrics in The Lives of the Saints that I dutifully absorbed as a child. Unlike the biographies by the Roman pagan Plutarch in his Lives of the Noble Greeks and Romans upon which they are modeled, The Lives of the Saints are short, easily digestible biographies of the holy men and women of the church that offer tutelage in moral sanctity rather than a searching chronicle. They accomplish this hortatory task by emphasizing the saints’ virtues and glossing over, or passing over altogether, the less edifying aspects of their lives.

Here the case for canonization rests on two grounds. First, Krepinevich and Watts aver that Marshall has been involved in nearly every momentous national-security decision since the Manhattan Project, beginning as an economics graduate student laboring alongside physicist Enrico Fermi to repair the University of Chicago’s cyclotron and later staying up until all hours with his RAND colleague Herman Kahn running the Monte Carlo simulations of radiation flows that “paid off” in the 1952 Eniwetok H-bomb explosion. His strategic acuity, we are instructed, was such that early on he diagnosed the weaknesses of Warsaw Pact conventional forces and challenged the “exaggerated threat” to NATO in the intelligence community’s “worst-case” assessments. He also presciently anticipated the easy U.S. victory over Iraq in 1991, harbored doubts about the George W. Bush administration’s plans for war in Iraq in 2003 and espoused a rebalancing toward Asia to counter growing Chinese military power. Such a record of strategic sagacity demonstrates that Marshall was endowed with the ability to “peer further into the future than most others in the U.S. government.”

Second, unlike the rest of the national-security bureaucracy, which is prisoner to a “conventional, but often self-serving, wisdom,” Marshall was a heroic figure who rose above parochial interest and perceptively limned the contours of the security challenges the United States has faced since the beginning of the Cold War. Marshall “was—and remains—a pragmatic strategist who has consistently looked further into the future and focused more on the first-order problems of long-term competition in peacetime than those around him.” Given that, Krepinevich and Watts maintain that “Pentagon decision makers would do well to take Marshall’s views to heart. His observations on such matters over three score years reveal a brilliant strategic mind with an uncanny ability to peer into the long-term future and see the situation ‘plain’—for what it is—more clearly than most around him.”

Given the general tenor of this work, there is another Roman Catholic institution that it seems appropriate to invoke for a, well, net assessment of Marshall himself. Until quite recently, the Vatican team analyzing the case for the canonization of a potential saint included not only its proponents but also a canon lawyer whose job it was to make the best case possible in opposition. Alas, the institution of the advocatus diaboli is now largely defunct. Nonetheless, I propose that, at least for the purposes of reviewing Marshall’s career, we resurrect it and ask a pertinent question: How would the devil’s advocate view it?

MARSHALL’S LEGENDARY reputation begins with his silences. Marshall hardly ever speaks in public. He said little at the Wianno conference we attended together, which, as Krepinevich and Watts report, is par for the course given his predilection for what they loftily refer to as a “sphinxlike demeanor.” (As the advocatus might ask, what if he just doesn’t have much to contribute?) That does not mean they think he is silent where it counts in the corridors of power. From the beginning, Marshall insisted that his office report directly to the secretary of defense. He and his followers have fiercely resisted periodic efforts to disconnect the ONA’s direct line to the secretary of defense and place it under a lower-ranking official, as in 1985, when it reported to Under Secretary of Defense for Policy Fred Iklé.

Moreover, “because most of the assessments produced by ONA have been highly classified and written primarily for the use of the secretary of defense,” Krepinevich and Watts explain, “understanding of what Marshall’s office has accomplished over the last four decades has been limited, even within the Defense Department.” Indeed, of the roughly twenty-five net assessments his office has produced over the years, only one—the 1992 Military Technical Revolution report—has been declassified. Marshall’s followers attribute his indirect approach to his modesty and lack of ego. A skeptic might add that it also deflects criticism toward others.

To get a balanced perspective of Marshall’s standing, I consulted Bruce L. R. Smith’s classic history of RAND from its formative years through the mid-1960s (which roughly parallels Marshall’s tenure there). His name is not mentioned. What’s more, a review of the memoirs of the secretaries of defense whom Marshall subsequently served turns up only two references to him: a perfunctory mention by Harold Brown, and a fulsome embrace by Donald Rumsfeld, who “valued Marshall’s thinking” on the growing Soviet threat in the 1970s and the increasing military role of technology in the 2000s.

Krepinevich and Watts recount a vignette from 2001 in which Rumsfeld invited Marshall to lunch in the Pentagon dining room to telegraph his support for the director of the ONA. Of course, as Krepinevich and Watts realize, Marshall’s embrace by Rumsfeld, one of the architects of the Iraq debacle, is a mixed blessing for their cause because it is difficult to separate Rumsfeld’s enthusiasm for high-tech warfare—an enthusiasm that Marshall’s faith in the RMA helped encourage—from the Bush administration’s fatally bungled occupation of post-Saddam Iraq.

WHAT, PRECISELY, was Marshall’s distinctive approach to the art of “net assessment”? Krepinevich and Watts note that an answer to this question is not straightforward because Marshall “was extremely reticent to define or to elaborate his views on what a ‘net assessment’ was or would accomplish.” It is clear that what he envisioned was the antithesis of the dominant RAND approach: systems analysis. This technique sought to combine highly quantitative operations research and cost-benefit analysis to formulate higher-level U.S. national-security policy. But a positive definition remains elusive.

Marshall used the term “net assessment” in a different manner than many military professionals do—that is, as an analysis of the likely outcome of a major military campaign or war. “Estimating [an opponent’s] current posture is a traditional military intelligence problem, and is largely concerned with finding out as much as possible about specified countries’ current and past military posture, past R&D and procurement decisions, etc.,” Marshall explained in a 1966 paper. What Marshall had in mind was “estimating an opponent’s military posture, and, in particular, his future military posture.”

Krepinevich and Watts commend Marshall’s notion of net assessment because it has proven so durable. But even they concede that it is long on diagnosis but short on concrete prescriptions. They recount, for example, that Marshall offered only indirect guidance to policy makers, once advising Rumsfeld that he needed to think about the U.S.-Soviet competition as a chess game but never telling him what moves to make. It could be the case that Marshall’s approach has survived precisely because it is so oracular and nebulous.

In fact, Marshall consistently hedged his strategic bets. In the foreword to a 1999 RAND volume on information warfare, after almost twenty years of studying the RMA, he conceded that we still “do not have an analytic framework to get at [the effect of information warfare], and we certainly do not have adequate models.” Krepinevich and Watts suggest that “the work outside ONA by some members of St. Andrew’s Prep whom Marshall mentored provides perhaps the best window currently available into how net assessment worked in practice.” It certainly makes sense to gauge Marshall’s influence by proxy in the work of his protégés such as Rosen, Friedberg, Cohen, James Roche, Krepinevich and Watts. The authors hold up the major debate on net assessment and the state of the Cold War conventional balance in Europe in the pages of Harvard’s International Security in 1989 as one such exemplar. Readers should peruse for themselves the various contributions. They might ask whether St. Andrew’s Prep alum Eliot Cohen’s similarly Delphic approach to net assessment offered a compelling critique of Cold War conventional-balance optimists who did not think NATO was hopelessly outgunned by the Soviet Union. Indeed, the U.S. military’s stunning performance in the ground war against Iraq in 1991 seems more in line with the optimists’ expectations for future combat.

Unfortunately, this is not an exercise that can be repeated across the board, because the ONA’s relationship with the handful of external contractors and individuals that Marshall relies upon to provide alternative expertise and perspectives to those of the Pentagon bureaucracy is, as Krepinevich and Watts delicately put it, “opaque.” During the Cold War, one of the major recipients of ONA largesse was Phillip Karber of the BDM Corporation, a for-profit defense research organization in Washington, DC. Today, one of the largest recipients of ONA alms is none other than—you guessed it—Krepinevich and Watts’s own Center for Strategic and Budgetary Assessments, which reportedly gets almost $3 million per year (or 40 percent of its budget) from Marshall’s office. That kind of largesse can inspire an awful lot of pious devotion, even in a city as cynical and opportunistic as the capital of the free world.

DESPITE MARSHALL’S admission that net assessment could not provide clear predictions or concrete advice, he nonetheless had a clear and consistent policy agenda: the United States needed to gird itself for a long-term rivalry with the Soviet Union in which it should be prepared to use military force to win. In a 1972 RAND report, for example, Marshall maintained that “the United States is now, and will continue to be, in an extended, continuing strategic arms competition with the Soviet Union. . . . The competition will be prolonged—indeed, for planning purposes, endless [emphasis added].”

Marshall was enamored with nuclear weapons. He was a long-standing proponent of nuclear war-fighting strategies for the United States. Rejecting “mutual assured destruction” and the arms-control community’s “overstated” preoccupation with strategic stability, along with the Eisenhower administration’s “New Look” policy of threatening “massive retaliation” to deter Soviet aggression, Marshall consistently pursued different nuclear strategies. Beginning with his participation in RAND’s “Strategic Objectives Committee,” which advocated that the United States develop a nuclear force capable of executing a first strike that could eliminate the Soviet Union’s strategic arsenal (or a “counterforce” strike, as it was known), and continuing in his intellectual collaboration with Herman Kahn, a model for the title character in Stanley Kubrick’s Dr. Strangelove, Marshall never showed any hesitation about thinking the unthinkable: fighting and winning thermonuclear war.

During the late 1950s and early 1960s, Marshall worked together with former U.S. Air Force targeteer and intelligence analyst Joseph Loftus. Through Loftus, Marshall gained access to sensitive intelligence about the Soviet nuclear arsenal that was off-limits to other RAND analysts. It persuaded him that a successful counterforce strike against the Soviet Union was feasible. Marshall and other proponents of counterforce were not, so far as we can tell, advocating a preventive nuclear war. Instead, their attraction to counterforce was its potential for limiting damage to the United States through preemptive strikes should war with the Soviet Union seem unavoidable. However, this policy was a tough sell to other RAND analysts in the era of the “missile gap.”

Whether Marshall’s privileged access to sensitive intelligence about the actual state of the Soviet nuclear arsenal justified his confidence in counterforce is questionable. But having this detailed information about the rather limited size and scale of the Soviet Union’s nuclear stockpile should certainly have revealed to him that the missile gap was a myth. Yet despite his access to this information, Marshall apparently failed to convince the Gaither Committee, which played a major role in peddling alarmism and which he served as a consultant. Nor did it stop him from moonlighting for the Kennedy campaign, along with a number of other RAND luminaries. John F. Kennedy, of course, won the 1960 election in part by preaching the false gospel that the passivity of the Eisenhower administration had allowed Washington to fall dangerously behind Moscow in the nuclear sphere.

Nevertheless, there was considerable intellectual consistency in Marshall’s embrace of counterforce. His access to intelligence convinced him that Soviet nuclear strategy was oriented toward war fighting. Given that, Marshall thought that the United States should embrace a similar doctrine. This would be the first of several occasions in which Marshall would use Soviet military doctrine as his guide for the United States, sometimes by emulating it, other times by exploiting its weaknesses. Given what he saw as the imperative to meet the Soviet challenge on its own terms, Marshall consistently advocated hardware and doctrinal developments that would make counterforce and nuclear war fighting possible for the United States. This would be a major theme, for example, of the preposterous Team B exercise, which granted hard-line anti-Soviet ideologues access to sensitive intelligence to see if they came to different conclusions from CIA analysts (spoiler alert: they did), to which Marshall contributed, and which offered a far gloomier assessment of the strategic balance than did the U.S. intelligence community.

IN KREPINEVICH and Watts’s account, Marshall made his most important contributions to U.S. national security in three particular areas: in forcing a much-needed debate about the CIA’s supposedly flawed assessments of the Soviet economy and defense spending; in identifying before anyone else in the Pentagon the outlines of an imminent RMA flowing from new technologies and innovations; and in predicting the rise of China as a “peer competitor” for the United States after the Cold War.

Krepinevich and Watts suggest that Marshall’s “pursuit of an accurate estimate of the Soviet defense burden, enabled by both ONA’s charter and Marshall’s long tenure . . . may well represent his most important contribution to US strategy during the Cold War’s final decade.” Marshall’s experience on the Gaither Committee first alerted him to the difficulty of determining the size of the Soviet economy as well as the extent of its defense burden and sparked his dissatisfaction with the CIA’s approach. But his concern needs to be seen in the larger context of growing conservative dissatisfaction with CIA estimates of the Soviet Union in general, to which a chorus of voices contributed, including neoconservatives and the various other ideological hard-liners associated with Team B and the Committee on the Present Danger.

With the benefit of hindsight, however, it is far from obvious that Marshall was right. For one thing, it is not at all clear that the CIA got things as wrong in its estimates of the Soviet economy and its defense burden as Krepinevich and Watts maintain. While the state of the art in the 1950s and 1960s may have suffered from serious flaws, as a declassified study in the CIA’s Studies in Intelligence notes, “by the mid-1970s, it was clear from the Agency’s publications that the fears of the USSR’s becoming the foremost world economic power were unfounded.” Marshall, though, consistently took the pessimistic view that the Soviets were using their strategic resources more efficiently than the United States. He was wrong. Krepinevich and Watts, though, seek to shield the sage from error: they reproach the CIA for some mistakes of Marshall’s in estimating the size of the Soviet defense effort.

Whatever private reservations Marshall may have had about the CIA’s Soviet economic data, his public line in a 1991 letter to Deputy Director for Intelligence John Helgerson was that the CIA’s economic estimates of Soviet defense programs have “been the major foundation for my office’s work on a military investment balance report that top defense officials have found very useful in characterizing broad trends in our security situation.” In short, when it suited his purposes, Marshall was perfectly content to rely upon CIA figures. Krepinevich and Watts argue that it was Marshall’s dissatisfaction with CIA estimates that led him to rely upon outside experts, but it is equally plausible that he did this because the political interpretations CIA analysts drew from this data were at variance with his own.

Marshall’s second major contribution, and the one most outside observers are most likely to credit him with, is his advocacy of the notion that we are in the midst of another RMA. As with Marshall’s interest in counterforce, it was Soviet military writings in the early 1970s about what they called the Military Technical Revolution (MTR) that first alerted him to this possibility. Indeed, a retired flag-rank officer who was on active duty throughout the RMA debate told me that Pentagon wags used to joke about the rechristening of the MTR as the RMA to camouflage its Soviet pedigree by imitating a herd of sheep bleating “MTR = Baaaaaad, RMA = goooood.” But it was no laughing matter to Marshall, who, as usual, viewed Soviet interest in the direst of terms: “It is clear that the Soviets have been engaged in a well-financed, steadily growing, increasingly effective effort to catch-up, across the board, in military technology.”

Marshall’s final signal contribution, we are often told, was his anticipation of the rise of China as the United States’ primary “peer competitor” in the twenty-first century. In Marshall’s vision, a rising China was embracing elements of the RMA to develop a strategy and force posture that would increasingly make it impossible for the United States to operate in the western Pacific littoral.

While there is no doubt that the rise of China constitutes the United States’ major strategic challenge in the new century, Marshall’s role in identifying this threat needs to be put in perspective. First, there were plenty of other strategists who came to the same conclusion about China independent of Marshall’s net assessments. Second, China’s strategy is hardly an example of the mature RMA in action that he envisioned during the later stage of the Cold War. Indeed, it is probably better regarded as an example of the sort of “asymmetric warfare” that inferior powers have historically embraced to counter a technologically superior foe. And finally, it’s difficult not to conclude that Marshall’s preoccupation, even obsession, with the RMA and China in the 1990s and early 2000s led him to ignore the very real threat of terrorism that thrust itself onto our national-security agenda so dramatically on September 11, 2001.

THROUGHOUT, KREPINEVICH and Watts also don’t address important inconsistencies in Marshall’s strategic thinking. Despite having helped create the “bureaucratic politics” framework, Marshall never consistently applied it. If Soviet strategic decisions and weapons-acquisition policies were principally the result of internal processes, then why should we have concluded that they reflected the most malign intentions and used them as our strategic polestar? Furthermore, Marshall clearly wanted to have it both ways on the issue of whether and how much U.S. actions could influence adversary behavior. In his early work in the 1950s and 1960s, he was quite critical of action/reaction models of arms racing, arguing that Soviet behavior would be driven largely by internal dynamics. But later he would take a more bullish view of the action/reaction thesis, declaring in 1999 that “every action creates a reaction; every new weapon spurs the creation of a new defense.” So which was it?

If Marshall really knew that the Soviet economy was in deep trouble and the Red Army was a house of cards, why did he feel such urgency about waging a long-term (perhaps endless) competition? In Krepinevich and Watts’s account, Marshall believed that the rationale for détente with the Soviets was animated by a misplaced hopelessness about growing Soviet power (in some ways reminiscent of some discussion of China today) that turned out to be too pessimistic. But Marshall was reportedly not so pessimistic about the conventional balance after sponsoring interviews of Soviet émigrés that revealed deep ethnic cleavages and serious demographic problems in the Red Army. He also understood that the Soviets were themselves pessimistic about their side of the strategic nuclear balance. Given this, why was he so exercised about the Soviet threat that he thought we needed to arms race them into submission, with all the attendant costs and risks this entailed?

FINALLY, KREPINEVICH and Watts fail to discuss Marshall’s close professional and personal relationship with Daniel Ellsberg, of Pentagon Papers fame, in any depth. As a young Harvard graduate student, Ellsberg interned at RAND in the summer of 1958, serving as rapporteur for a strategy group that included such luminaries as Albert Wohlstetter, Herman Kahn, Harry Rowen and Marshall himself. Once Ellsberg joined RAND full-time, Marshall and his first wife Mary became very close to Ellsberg and his first wife Carol. Tom Wells’s biography of Ellsberg, Wild Man, draws heavily upon interviews with the Marshalls—the title coming from Mary’s description of him.

In Wells’s account, it is clear that the Marshalls were intimately familiar with Ellsberg’s character flaws—his reckless womanizing with RAND employees or wives of RAND colleagues and his desultory work ethic, about which Mary complained that the RAND leadership was “always excusing him, indulging him.” Indeed, it was to Mary that Carol Ellsberg first confessed her suspicions that her husband was the source of the Pentagon Papers leaks. The two of them tried to alert RAND president Harry Rowen about it, to no avail. After Ellsberg’s arrest and release on bail, it was from the Marshalls’ Los Angeles house that Ellsberg called his children.

We should not, of course, judge people by the behavior of their close friends; standing with them in their moment of tribulation is in many respects admirable, and the Marshalls were as dismayed and angered by their close friend’s breach of trust as anyone else at RAND. But the absence of this story from Marshall’s professional and personal life seems strange. It would be interesting to know what Marshall learned from his experience with the brilliant but complicated strategic wunderkind he took under his wing. Krepinevich and Watts themselves open the door to pondering this episode with their insistence that Andy and Mary Marshall were such shrewd judges of his colleagues.

AT THE end of the day, Marshall was not a saint but rather a public official who served too long and whose record was more mixed than his incense burners are prepared to admit. He consistently pushed his insights to their most hawkish extremes and endorsed expensive, even fantastic, weapons schemes, which may be why he is so beloved among the neocons. But that is no cause for celebration, and it is why any discussion of him is inextricably linked to the fate of his church.

In the Roman Catholic Church, canonizations have increasingly become as much about larger institutional political struggles as they are about the character of the saint. For example, the church has canonized two recent popes from very different ends of the political spectrum. Liberals love John XXIII, the convener of the Second Vatican Council; conservatives, John Paul II, who they think reined in its excesses. Similarly, St. Andy’s canonization is also about a larger political agenda: institutionalizing his mission.

But the fate of the Office of Net Assessment ought not to be determined just by the credulous devotion of its flock. Instead, a systematic scrutiny of its work is long overdue because so much of what it was supposed to have accomplished over the years remains officially classified. If such an inquisition concludes that there is true merit in the cause, then Marshall’s crusade will be the better for it in the future. If not, perhaps St. Andrew’s Prep, like so many now-superfluous parochial schools, should close its doors.

Michael C. Desch is professor and chair of the Department of Political Science at the University of Notre Dame.