The Washington PostDemocracy Dies in Darkness

Inside the mind of the Pentagon’s “Yoda”

Review of "The Last Warrior: Andrew Marshall and the Shaping of Modern American Defense Strategy" by Andrew Krepinevich and Barry Watts

Andrew Marshall, director of the Pentagon’s Office of Net Assessment, in 2002. (Photo by Sandro)

THE LAST WARRIOR: Andrew Marshall and the Shaping of Modern American Defense Strategy

By Andrew Krepinevich and Barry Watts. Basic. 305 pp. $29.99

The Last Warrior” is an important book that should not have been written.

It’s important because its subject, 93-year-old Andrew Marshall, spent four decades running a Pentagon think tank with a direct line to America’s defense secretaries. Credited with anticipating the Soviet economic collapse and criticized for treating a future conflict with China as nearly inevitable, Marshall has long been a mysterious Washington presence, nicknamed “Yoda” for his cryptic pronouncements and fanatical followers. With Marshall’s retirement this month, a look inside his Office of Net Assessment — a perfect name for an opaque bureaucratic outfit — is certainly worthwhile.

So why wish it away? I’ll let the authors, Andrew Krepinevich and Barry Watts, explain. “We have endeavored to be as unbiased and objective as we could,” they write. “Yet neither of us can claim to be disinterested observers. We both have a long history with Marshall.” That history includes serving on Marshall’s staff, participating in subsequent studies sponsored by Marshall and currently leading an outside think tank that receives funding from Marshall’s office. “Once a member of Marshall’s coterie of trusted former net assessors and outside defense experts — ‘St. Andrew’s Prep’ — always a member,” they write with unabashed chumminess.

Kudos for the full disclosure. But when the authors lionize Marshall as “one of America’s most influential and enduring strategic thinkers,” who is talking — the respected analysts or the devoted acolytes who dedicated this very book to him? It’s a question that hangs over every page, making it hard to take the book’s arguments at face value.

But for the moment, let’s go ahead. Krepinevich and Watts emphasize that they’ve written not a biography but an intellectual history. We come to know Marshall more through reports and research papers than anecdotes or interviews. And despite the authors’ veneration of their mentor, the mind that emerges seems dissatisfied, distant, disengaged — a Cold War thinker fixated on finding better measures of Soviet strength, reluctant to wage the battles needed to influence policy and uninterested in the two most traumatic security challenges of the past half-century: Vietnam and 9/11.

Self-taught as a boy in the 1930s, Marshall devoured Dostoevsky, Tolstoy and Toynbee at his local Detroit library, eventually enrolling in the University of Chicago’s graduate program in economics. Drawn to statistics, he soon left school to join a new research institute in California that would become a leading source for Cold War nuclear strategy. Starting in 1949, the Rand Corp. became Marshall’s intellectual home for 22 years.

His early work reflected the mind-set of the time. In a study titled “The Deterrence and Strategy of Total War, 1959-1961,” Marshall spun harrowing scenarios for nuclear war. If the Soviet Union struck first, he concluded, the best response would be for the United States to attack counterforce targets in the U.S.S.R. “using high yield weapons with ground bursts so as to produce extensive fallout.” (The civilian casualties this approach would inflict were considered a “bonus.”)

As the superpowers settled into stalemate, Marshall began exploring how the United States could compete long-term against the U.S.S.R. He challenged “rational-actor” models of Soviet behavior, believing instead that decisions reflected the competing interests of rival policymakers and bureaucracies. And he argued that the best way to compete with the Soviets was to boost those areas of defense where America was already stronger, forcing its rival to overcompensate and overspend.

In 1969, Henry Kissinger, then national security adviser to President Richard Nixon, asked Marshall to evaluate the intelligence the White House was receiving. It was supposed to be a brief Washington assignment, but Marshall never left. Two years later, Nixon established a Net Assessment Group in the National Security Council, responsible for “producing net assessments of US capabilities vis-a-vis those of foreign governments constituting a threat to US security.” Marshall became director, and soon the office moved to the Pentagon, where he conducted assessments of U.S. vs. Soviet ground forces, the strategic nuclear balance and NATO vs. the Warsaw Pact, among other subjects.

Here we encounter a significant conceptual problem, both in Marshall’s work and the book: The practice of net assessment — with which Marshall has become synonymous — is only vaguely defined. Krepinevich and Watts call it “an analytic framework for comprehending the fundamental character of a competitive situation.”

How about Marshall? In 1972, he wrote a memo that the authors hail as the “definitive, enduring vision” of net assessment. Yet in that document, they explain, Marshall “exhibited a reluctance — which he would demonstrate time and again over the next four decades — to produce anything approaching a methodology” for the effort. Instead, he emphasized that the assessments of U.S. weapons, forces and policies compared with those of other countries needed to be “careful” and “comprehensive,” and required “sustained intellectual effort” (all things that apply to, say, beekeeping). More baffling, Marshall did not share the memo with his staff; it was discovered three decades later, “filed away in a binder in Marshall’s office.”

Adding to the murkiness, Defense Secretary James Schlesinger and Marshall agreed that net assessment was too complex to suffer meddling from other government agencies; ONA’s work would remain protected in the Pentagon’s A-ring. They also agreed, remarkably, that there would be no set schedule for the studies. “The emphasis would be on getting them done well rather than fast,” the authors explain. “Marshall would be able to devote as much time as necessary to produce full-blown assessments.” No interference or deadlines — who wouldn’t stay for 40 years?

We constantly hear Marshall and the authors complain about how difficult net assessment is, lament the lack of quality data to conduct it or question the interpretations of others. At Schlesinger’s behest, for example, Marshall began what became a long-running debate with the CIA over how much of the Soviet economy was dedicated to defense. The agency put it as low as 6 percent in the mid-1970s, whereas Marshall had it above 30 percent in the late 1980s. His pursuit of an accurate estimate of the Soviet defense burden, Krepinevich and Watts write, was “his most important contribution to U.S. strategy during the Cold War’s final decade.”

Oddly, it was during the ’80s, with a Cold Warrior president focused on the nuclear balance with the Soviets, that Marshall’s influence waned. He met with Defense Secretary Caspar Weinberger only a handful of times and “did little to promote ONA’s work to Weinberger,” the authors note. It is not clear why, except that Marshall didn’t care for the boss. On the day Weinberger’s portrait was placed on a wall at the Pentagon, a colleague asked Marshall why he had not attended the ceremony. “I might have gone if they were hanging Weinberger instead of his picture,” he answered.

In a rare criticism, Krepinevich and Watts suggest that Marshall didn’t make it easy for other officials to understand his work. “He did not go out of his way to allay mistrust,” they write, and his influence suffered. As they put it, a bit painfully: “He could lead the policy-making horses to water, but he could not make them drink — or, perhaps, think.” But hold your policymaking horses there: If you’ve deliberately insulated yourself from the broader policy process, whose fault is it when decisions don’t go your way?

Marshall is proudest not of particular assessments but of the network of scholars he developed, calling it his “major achievement.” Watts and Krepinevich highlight prominent Marshall disciples such as Graham Allison, Eliot Cohen and, yes, themselves, resulting in some awkward self-praise. A paper Krepinevich wrote in 1992 on the “military-technical revolution,” a focus of the ONA’s work at the time, “would have an enduring influence on the post-Cold War defense debate, both within the United States and abroad,” the authors write. If they say so.

During his final decade at the Pentagon, Marshall zeroed in on the rise of China, which had long concerned him. The authors point to President Obama’s pivot to Asia and the rise of an “Air-Sea Battle” military framework to counter China in the Pacific — a strategy some have called needlessly menacing — as vindication of Marshall, who “once again managed to peer further into the future than most others in the US government.”

But sometimes you look so far ahead that you miss what’s right in front of you. Marshall was so intent on discerning long-term threats that he seemed to ignore the Vietnam War. And he regarded 9/11 as something of an imposition, because the attacks “limited the time and energy [Donald] Rumsfeld was able to devote to advancing an advantage-based defense strategy,” which Marshall had proposed in early 2001. Yes, too bad George W. Bush’s defense secretary was preoccupied by 9/11.

The notion of Marshall as a transformative security sage is more stated than proved in “The Last Warrior.” At times he reminds me less of Yoda and more of Chauncey Gardiner, Peter Sellers’s servant turned presidential adviser in the 1979 film “Being There,” whose banal utterances are miscast as great wisdom. “As in a chess game, one has to think two and three moves ahead,” Marshall told Rumsfeld during the Ford administration. True, as far as that goes.

When more of Marshall’s work is declassified, his influence may become clearer. But I worry that “The Last Warrior” will discourage a more objective biography of Andrew Marshall. It’s a book that should be written. If it is, the author should interview Krepinevich and Watts at length.

After all, they’ve worked closely with Marshall.