The Last Warrior

by Andrew F. Krepinevich

  Carter appointed Harold Brown to succeed Rumsfeld as defense secretary. Brown came to office with imposing academic credentials and a wealth of national security experience. He had earned a PhD in physics from Columbia University at the age of twenty-one, joined the Lawrence Radiation Laboratory at Livermore in California in 1952, becoming its director in 1960, and was president of the California Institute of Technology (Caltech) from 1969 to 1977. His defense experience included serving as McNamara’s director of defense research and engineering from 1961 to 1965, and as secretary of the Air Force from October 1965 to February 1969. Brown was also the first scientist to become secretary of defense.

  Unlike either Schlesinger or Rumsfeld, who often preferred to discuss issues face-to-face, Brown was a reader. He preferred to address strategic issues by reading extended papers, annotating those he found useful with handwritten observations, comments, and guidance. One of those was Marshall and Roche’s 1976 paper “Strategy for Competing with the Soviets in the Military Sector of the Continuing Political-Military Competition,” which they had originally written for Rumsfeld. Like Rumsfeld, Brown found himself very much in agreement with the paper’s recommendation that rather than simply attempting to parry Soviet threats, as the military services were prone to do—especially when arguing for larger budgets—far greater effort should be made to develop ways to exploit Soviet weaknesses. This approach would force the USSR to react to what the United States was doing in the competition, rather than the United States relinquishing the initiative to its opponent.31

  Many years later Marshall shared this paper with the business strategist Richard Rumelt. At the time Rumelt was a UCLA professor and one of the most influential gurus on management strategy in the private sector.32 Rumelt understood the fundamental power of leveraging a firm’s comparative advantages to impose out-of-proportion costs on its rivals, and saw in Marshall’s paper the same approach.33 “Marshall and Roche’s idea,” he later wrote, “was a break with the budget-driven balance-of-forces logic of 1976. It was simple. The United States should actually compete with the Soviet Union, using its strengths to good effect and exploiting Soviet weaknesses. There were no complex charts or graphs, no abstruse formulas, no acronym-jammed buzz speak: just an idea and some pointers on how it might be used—the terrible simplicity of the discovery of hidden power in a situation.”34

  If anything, Marshall’s influence on both the defense secretary and US strategy was greater under Harold Brown during his tenure from 1977 to 1981 than it had been under Rumsfeld. These years were also among ONA’s most prolific in terms of completing net assessments. Marshall forwarded eleven assessments to Brown during these four years as compared with eight reports to Brown’s next three successors, Caspar Weinberger, Frank Carlucci, and Dick Cheney from 1982 to 1991, and only four to defense secretaries Les Aspen, William Perry, and William Cohen from 1992 to 2001.

  Harold Brown had no hesitation about retaining Marshall and his net assessment program when he took over as defense secretary. However, he had over thirty people in OSD reporting directly to him, which he felt was too many. So he decided to create a new undersecretary of defense position for policy [USD(P)] and put Marshall’s office under it.35 This had little effect on ONA. Brown agreed that Marshall’s assessments would still be written for him, go directly to him, and not be coordinated with others in the Pentagon.

  With exception of Schlesinger, Harold Brown may well have been the defense secretary who most deeply understood and appreciated the Office of Net Assessment. As he recalled some two decades later, the assessments and analyses he received from Marshall were fundamentally different from those generated by other Pentagon staff elements, primarily because they started from a different premise: that the United States was in a long-term competition with the Soviet Union in which success over time depended on identifying areas of comparative US advantage and using them to exploit areas of comparative Soviet weakness or disadvantage. The kinds of questions embodied by ONA’s assessments were: What are we good at compared to the Soviets? What are we trying to do? How can we use what we are good at to gain advantage despite the countervailing actions of the other side?36 Brown clearly recognized what Marshall had long realized: that the Pentagon’s many bureaucracies and power centers were too consumed by the internal competition with one another over budget shares and their own agendas to focus dispassionately on these larger strategic questions that formed the core of ONA’s assessments. In Brown’s eyes Marshall and his office provided the kind of longer-term strategic focus he needed.

  A good example of this can be seen in the strategic rationale Marshall had provided in 1976 for fielding the B-1 bomber. Marshall argued that the Soviet General Staff had been obsessed with the USSR’s air defense ever since its air force had lost some four thousand planes during the opening week of Operation Barbarossa—Germany’s invasion of the USSR in June 1941.37 Consequently, in the late 1940s the USSR’s National Air Defense Troops (PVO Strany) were established as a separate military service and began developing the world’s most extensive air defense network. The US U-2 reconnaissance over flights of Soviet territory in the late 1950s only served to reinforce the Soviet military’s sense of vulnerability and determination to accord high priority to territorial air defenses. From 1945 to the early 1960s the Soviets spent a great deal more on air defenses than on their nuclear forces.38 Therefore, Marshall concluded, there was good reason to expect that fielding the nuclear-armed B-1 designed to penetrate Soviet airspace would reinforce the USSR’s propensity to continue investing heavily in territorial air defenses along borders that stretched across eleven time zones.

  FIGURE 6.2.United States versus Soviet Bomber Defenses.

  SOURCE: NIE 11-3/8-76 in Steury, Intentions and Capabilities: Estimates on Soviet Strategic Forces, 1950–1983 (Washington, DC: Center for the Study of Intelligence, CIA, 1996), p. 357.

  The United States, by contrast, had no such tendency. To be sure, in the 1950s the Air Force and the Army had fielded both interceptor squadrons and surface-to-air missile (SAM) batteries, respectively, to defend North America against a Soviet bomber force that was expected to number in the hundreds. But by the late 1960s it was evident the USSR’s bomber force would remain relatively small, whereas the Soviets were rapidly expanding their ICBM force. Because neither Air Force interceptors nor Army Nike-Hercules SAMs could defend against Soviet ICBMs, the United States began reducing its investment in its air defenses.39 Marshall’s support for the B-1 was rooted in his belief that fielding the plane would, at a relatively modest cost, enable the United States to continue posing a penetrating nuclear bomber threat to the USSR. Doing so would encourage the Soviets to continue investing far more in territorial air defenses than it would cost to field and maintain a fleet of B-1s. Moreover, the United States could avoid similar costs by abandoning the costly air defense “business,” thereby giving the United States “the most leverage” over Soviet defense expenditures.40

  Marshall’s advice fell on deaf ears. In July 1977 the Carter administration stunned the defense establishment by canceling the B-1 in favor of arming B-52s with air-launched cruise missiles (ALCMs). The president argued that his decision constituted a cheaper option and would give the United States “just as good a capability of penetrating Soviet air defenses” as the B-1.41 Nonetheless, ONA’s B-1 memorandum epitomized the kind of “competitive strategies” thinking that impressed Harold Brown. It would later resurface during the administration of President Ronald Reagan.

  Marshall’s influence was also significantly enhanced by President Carter’s restructuring of the National Security Council’s staff and its organization. Carter was determined to rectify what he considered the excessive power over foreign affairs that Henry Kissinger had acquired during his years as both national security adviser and secretary of state. The new president’s concern was that Kissinger’s accumulation of power had shielded Nixon and (particularly) Ford from competing viewpoints within the foreign policy establ
ishment. Carter intended to be the final arbiter of the nation’s foreign policy. He set out to maintain access to a broad spectrum of views and information by more fully engaging his NSC cabinet officers in the formulation of national security policy.

  The Polish-American political scientist Zbigniew Brzezinski had been Carter’s principal foreign policy adviser during his campaign for the presidency. Brzezinski was also an outspoken critic of what he viewed as Nixon and Kissinger’s overreliance on détente to slow or constrain the arms race between the United States and the USSR, particularly in nuclear forces. Carter selected Brzezinski as his national security adviser because he wanted an assertive intellectual at his side, one who would provide him with day-to-day advice. The president envisaged the NSC’s main functions to be policy coordination and research. Upon taking office he reduced its staff by 50 percent and cut the number of standing NSC committees from eight to two: the Policy Review and Special Coordinating Committees.42 Brzezinski always chaired the Special Coordinating Committee, but the Policy Review Committee was generally chaired by the department most directly concerned with the issue at hand. By allowing Brzezinski to chair only one of these two committees, Carter believed that his national security adviser and the NSC staff would be precluded from dominating the policy-making process.

  Carter also decided to undertake a major reassessment of US security needs. In mid-February 1977, shortly after taking office, he signed Presidential Review Memorandum/NSC-10 (PRM/NSC-10), “Comprehensive Net Assessment and Military Force Posture Review,” which directed two studies to be done concurrently. The Policy Review Committee, chaired by Harold Brown, was instructed to “define a wide range of alternative military strategies and construct alternative military force postures” to support each of the alternatives. The Special Coordination Committee under Brzezinski was to conduct “a dynamic net assessment” that reviewed and compared “overall trends in the political, diplomatic, economic, technological, and military capabilities of the United States, its allies, and potential adversaries.”43 Initial results from both efforts were to be completed by early July 1977, although work on PRM/NSC-10 continued into 1978.

  Marshall’s office was soon drawn into both PRM/NSC-10 reviews. He and Roche were designated the Pentagon’s staff leads for Brzezinski’s net assessment. This effort was led by Samuel Huntington, a Harvard political scientist Brzezinski had brought in to oversee his portion of the review. Huntington was known for his seminal analysis of US civil-military relations in his 1957 The Soldier and the State. Later, in 1993, he presciently predicted in Foreign Affairs that the central aspect of global politics would become a civilizational clash between the West and Confucian-Islamic states,44 a theme he expanded in his 1996 The Clash of Civilizations and the Remaking of the World Order. These publications were the bookends of a long and distinguished career by a scholar considered by many to be the preeminent political scientist of his day.

  Huntington was assisted in his efforts by Richard Betts and Catherine Kelleher, both also political scientists, and Brzezinski’s military assistant, Army colonel William Odom, who had come to him from West Point’s Department of Social Sciences. To give a sense of the scale of this assessment, Brzezinski’s portion of the enterprise was based on the reports of eleven separate panels as well as special analyses. Drawing on these various reports Huntington, Odom, Betts, and Kelleher generated an overview of some three hundred pages.45

  Marshall and Roche worked with Huntington and Odom on the Comprehensive Net Assessment portion of PRM/NSC-10. Fortunately, by early 1977 ONA had completed full-blown assessments of the US-Soviet strategic nuclear and antisubmarine warfare (ASW) balances. Brown reviewed and commented on the strategic nuclear assessment at the end of March 1977, liked it, and suggested an abbreviated version be sent to Brzezinski. Brown followed this up by forwarding versions of ONA’s ASW, maritime, and NATO–Warsaw Pact balances to Brzezinski shortly thereafter.46 These balances, along with conversations Marshall and Roche had with the NSC staff, proved helpful to Brzezinski’s NSC team. Odom found that listening to Marshall yielded strategic “gems,” while Roche provided Huntington with new ways of thinking about power projection.47

  The PRM/NSC-10 force posture review was done by OSD. This review was led by Lynn Davis, deputy assistant secretary of defense for policy plans, and the NSC’s Victor Utgoff. Its aim was to elicit from Carter policy guidance on national military strategy.48 After being completed in June 1977, Brown’s forwarding memorandum stated that the posture review had been useful in focusing attention on the value of developing a strategy to guide the evolution of US military forces over the next decade. But he did not think that any of the study’s various alternative military strategies could provide “a sufficient basis for specific decisions on US military force structures or force planning.”49

  As an outside observer Marshall had similar qualms about the force posture exercise, particularly concerning US objectives. He had long argued that US national objectives had never been clearly or precisely articulated, and one of the limitations highlighted by the study was that it was not based on “overall US national objectives because no agreed set of national objectives exists.”50 Were Marshall to have put his concerns in colloquial terms, he might have said, “If you don’t know where you want to go, any path will take you there.”

  After reviewing both the force posture and comprehensive net assessment studies, Carter agreed with Brown’s reservations about the force posture recommendations. In Presidential Directive/NSC-18, “US National Strategy,” the president concluded that US-Soviet relations would be characterized by both competition and cooperation for the foreseeable future. The national strategy of the United States, he directed, would be to exploit relative US advantages in economic strength, technological superiority, and popular political support. On the military side, US nuclear forces would be needed to deter attacks on the “United States, upon our forces, our allies, and others whose security Washington deemed important to the United States and, if deterrence failed, to inflict [an] appropriate retaliatory response on the Soviet Union.”51

  In light of these judgments, Carter ordered his secretary of defense to undertake a nuclear targeting study. Walter Slocombe, the deputy assistant secretary of defense for international affairs, and Marshall were directed to oversee this effort. Leon Sloss, the assistant director of the US Arms Control and Disarmament Agency, was tasked with leading the effort. Once the targeting study was done, Marshall and Slocombe held a series of meetings with Harold Brown and air force general David Jones, the chairman of the Joint Chiefs of Staff, to help refine its conclusions. Their findings were forwarded to the White House, where they were reviewed and edited by Brzezinski and Odom.

  The targeting study led to Presidential Decision/NSC-59, “Nuclear Weapons Employment Policy,” which Carter signed on July 25, 1980. PD/NSC-59 retained the Single Integrated Operational Plan (SIOP) as a set of preplanned options for nuclear strikes against the Soviet Union, its allies, and its forces.52 However, PD/NSC-59 also called for creating the capability to design nuclear employment plans on short notice that would integrate “strategic force employment with theater nuclear force employment and general purpose force employment.” Their stated aim was to achieve “theater campaign objectives and other national objectives” when SIOP options were not judged suitable.53 In the early 1970s Schlesinger had sought to develop options for limited nuclear strikes against the Soviet Union with an eye toward being able to stop nuclear use short of an all-out nuclear exchange, thereby limiting the damage to the United States. But as Odom later emphasized, even nuclear strikes against the USSR involving a small number of weapons seemed likely to trigger a massive nuclear response.

  PD/NSC-59’s more fundamental change in targeting was aimed at enhancing deterrence. During Sloss’s targeting study, the Defense Intelligence Agency discovered that the USSR had constructed extensive facilities intended to protect its leadership in the event of a nuclear exchange. Marshall, am
ong others, pushed the White House to exploit this knowledge. PD/NSC-59 did so, directing that deterrence could be strengthened by making it clear to Soviet leaders that their efforts to survive a nuclear attack by building facilities to protect themselves could not succeed.54 Carter’s revised targeting policy not only put the economic and social structures of the Soviet state and its external empire at risk, it explicitly targeted the USSR’s leaders. Secretary Brown directed and personally cleared certain articles and discussions of PD/NSC-59 to drive this point home to the Kremlin.55

  Another presidential directive, PD/NSC-18, explicitly called for the US to maintain a posture of “essential equivalence” with the USSR, adding that the United States would not allow its nuclear forces to be inferior to those of the Soviet Union.56 However, Carter’s insistence on essential equivalence raised the perennial question of what analytic measures should be used to determine whether US nuclear forces had met this requirement, either at present or in the foreseeable future.

  Marshall had addressed this issue before. In his August 1976 memo to Rumsfeld he had argued against relying on existing static and dynamic measures because they did not adequately capture the evolving relationship between the two sides’ nuclear forces. Beyond indicating general trends, static measures such as the numbers of US and Soviet deployed warheads, ICBMs, SLBMs and heavy bombers, he thought, were at best “extraordinarily crude.”57 The same was true of more sophisticated static metrics, such as total ballistic missile throw-weight, equivalent megatonnage and countermilitary potential.* As for calculations of the numbers of US and Soviet warheads likely to survive the initial exchange, or of US and Soviet fatalities, Marshall felt that these metrics ignored too many complexities and uncertainties. Thanks to his collaboration with Joseph Loftus at RAND, Marshall also knew that US intelligence had very limited insight into Soviet assessments of the strategic nuclear competition. If the overriding US objective was to deter a Soviet nuclear attack, he argued, the United States needed to know what metrics and calculations the Soviets were employing to assess the risks of nuclear war, for the simple reason that they were the objects of US deterrence strategy. Marshall noted that US and Soviet decision makers did not appear to be using the same metrics. Why else would the Soviets continue building up their ICBM forces despite US restraint when both sides had achieved a secure second-strike capability?