The Last Warrior

by Andrew F. Krepinevich

  Schlesinger and Marshall next discussed what assessments should receive top priority. They quickly agreed that ONA would initially concentrate on three areas. First and foremost would be the strategic nuclear balance between the United States and the Soviet Union. Marshall had done extensive work in this area for several decades. More important, the Soviet nuclear arsenal posed the only immediate existential threat to America’s survival.

  The second assessment would address the military balance between NATO and the Warsaw Pact, an alliance comprising the Soviet Union and its Eastern European satellite states.* This, too, was an obvious choice. Most of the Soviet Union’s conventional air and ground forces were opposite NATO in central Europe, and most Soviet reinforcements were located just beyond Poland in the USSR’s Western Military Districts. The threat these forces posed to Western Europe was both immediate and clear, as was the need to defend the United States’ NATO allies, particularly France, Great Britain, and West Germany.† Were the Soviets to seize Western Europe, the overall military balance between the United States and the USSR would be fundamentally altered in the latter’s favor. In addition, a conventional war in central Europe was the most likely trigger for nuclear escalation.

  Marshall and Schlesinger decided that the third area would be the maritime balance between the two superpowers. To begin with they viewed the maritime balance as a catchall for strategic mobility and overseas power projection, areas of military capability in which the US military was well ahead.23 But the Soviet Navy was now receiving greater priority, raising the possibility that the Soviets might be looking to move into a new “business area” and begin competing more directly with the United States in the maritime domain. Finally, this assessment would also enable Marshall to examine possible Soviet efforts to engage in power-projection operations far from their borders.

  Not long after Schlesinger and Marshall agreed upon these three assessments, the defense secretary added a fourth. Recalling his disagreement with the CIA’s estimate of the USSR’s military burden when he was DCI, he asked Marshall to pressure the CIA to reconsider its burden estimate. This led to a series of US-Soviet investment balances with particular emphasis on each side’s future military capabilities. In terms of American budget categories, this meant focusing on spending for research and development (R&D), procurement, and military construction.24 In Marshall’s lexicon, this assessment would be neither a functional balance like the US-Soviet strategic nuclear competition nor a regional balance such as the NATO–Warsaw Pact competition. Instead it sought to compare both sides’ resource allocations to their evolving military forces and capabilities.

  Schlesinger’s intense interest in the issue of Soviet military expenditures was understandable as the answer had profound implications for the long-term competition between the superpowers. If in fact the USSR could generate substantially more military capability than the United States with an economy half the size, all while maintaining the defense burden at 6 or 7 percent of gross national product, the prognosis for the United States would be bleak indeed. But despite pressure from ONA, the CIA persisted with its 6–7 percent burden estimate into the mid-1970s.25

  Schlesinger later recalled that while he was DCI he had framed the burden issue for some of the Agency’s economists as follows: “Look at those guys [the Soviets]. According to the CIA they’re spending 6 percent of the GDP or GNP on the military, and look at all this stuff [they are producing]! . . . [A]re they miracle workers—or is there something wrong with the analysis?”26 If the US economy was twice the size of the Soviet economy and both were spending 6 percent of their economy on defense, then the US defense budget should be twice the size of the Soviet defense budget. But CIA’s efforts to cost Soviet military programs suggested the opposite. Schlesinger told his staff:

  [W]e are producing . . . 180 tanks a year . . . And look at these guys, they are producing 3,000 tanks a year—or 2,800. Probably ours are better. They may not be. But if you look at their military production, that’s pretty impressive. They are producing “X” number of aircraft a year. We are producing a small fraction thereof. Ours have advanced electronics but, you know, look at the numbers that are coming out over there. . . . They’ve got to be spending in dollar terms about 160 percent of what we’re spending. . . . They are devoting a major part of their industrial capability and their economic activity to defense.27

  Schlesinger and Marshall believed the Soviet buildup was placing far more strain on the USSR’s economy than the CIA was willing to admit. Rather than being “miracle workers,” the Soviets were devoting a cripplingly large share of their GNP and their industrial base to military purposes, an effort they could not hope to sustain over the long term. The crucial bottom line: If Schlesinger and Marshall were right, then time was on America’s side—it was the Soviet Union that would eventually find itself on the ropes, not the United States.

  Schlesinger’s tenure at the CIA had lasted only five months, not long enough to make sufficient headway against its bureaucracy. After getting his feet on the ground as defense secretary, however, he personally told Edward Proctor, the deputy director for intelligence at the CIA, that he remained wholly unconvinced that the Soviet defense burden was as small a percentage of the USSR’s GNP as the Agency persisted in maintaining.28

  While Schlesinger asked Marshall to press the CIA to reexamine its estimates of the USSR’s military burden, its economists proved persistently loath to rethink their position.29 Despite Marshall’s persistent efforts to get them to do so, he would have no better luck in getting the CIA to rethink its estimate of the USSR’s military burden than Schlesinger had had as DCI.30 For the remainder of the Cold War, the Agency consistently asserted that the military burden on the Soviet economy was substantially less than Marshall and Schlesinger believed it to be.

  After agreeing on the four key areas of the US-Soviet military competition in which ONA would conduct major assessments, Marshall next needed to determine how best to structure them. In 1976, after several iterations, he converged on a basic four-part structure for the balances. The first—the basic assessment—would provide the reader with an overview of the competition under examination: How was the United States faring in the competition? Was its position improving relative to the past, or not? Given current conditions, was its position likely to improve over time, or not? The second section would identify key asymmetries in the competition: Where did the two competitors differ in significant and important ways, particularly in how they were pursuing the competition? These asymmetries could vary widely, to include (but not be limited to) objectives, doctrine, force structure, force posture (e.g., basing of forces), allies, logistics, and modernization efforts. The assessment would then analyze the key asymmetries’ significance in terms of their influence on the competition. The third section would identify and discuss major uncertainties that could exert a significant bearing on the conclusions reached in the basic assessment. The fourth and final part would address emerging problem areas in the competition and, equally important, key opportunities, both of which might be exploited to improve the United States’ competitive position.31

  Schlesinger gave Marshall and his office a remarkable degree of autonomy—although this did not free them from dependence on the Pentagon or the intelligence community. Marshall realized that he would need the support of both to conduct his assessments. They had data that he could not obtain from other sources. For example, the CIA was the primary source for order-of-battle data on Soviet forces—everything from nuclear weapons to tank armies to air regiments and naval forces—as well as of dollar-cost estimates of Soviet defense expenditures and the size of the USSR’s overall economy. As for the US military services, they were the primary source of historical order-of-battle data on US forces, and could also provide important insights as to how new weapon systems in development would likely be employed. Initially Marshall expected to benefit from his close association with Schlesinger by using the defense secretary’s name to ga
in access to the desired data held by these organizations. Nevertheless, like most large organizations they could be reluctant to share information, particularly when the highest levels of security classification were involved or when the data might be used to support analyses that would challenge their parochial interests.

  As it turned out Marshall very much needed Schlesinger’s support, or “top cover,” to use the parlance of the Pentagon. The Office of Net Assessment was viewed warily by other elements of the government, particularly those that felt its director would challenge their prerogatives or analyses and in so doing undermine their influence. The fact that Marshall already had a track record of challenging the accepted wisdom only increased their anxiety. The general response from OSD’s analytic organizations, such as PA&E, as well as the military services was that they either already did, or could, conduct net assessments. In their eyes, there was simply no need for what Marshall proposed to do. This was especially true in the case of PA&E’s systems analysts who wanted to fold net assessment’s work into their annual Defense Department budgeting and planning cycle.

  These organizations did not understand that in both Marshall and Schlesinger’s eyes net assessment was intended not to compete with systems analysis but to provide broader, more comprehensive analyses of military issues that looked beyond near-term choices of weapons systems. If they had any doubts, they only had to listen to Marshall talk of his absolute determination that net assessment must not engage in making programmatic recommendations. If his enterprise had been subsumed into the Pentagon’s planning, programming, and budgeting process it would have almost certainly lost its broad, long-term focus and been pushed toward the reductionist, near-term prescriptive focus that characterized systems analysis. As for the notion that net assessment would duplicate regular intelligence production processes, as some suggested, this ignored the belief of Marshall, Schlesinger, Kissinger, and many others, including President Nixon, that an independent net assessment group was needed to both evaluate existing national intelligence products and provide balanced, objective assessments of US and allied capabilities relative to those of competitors and adversaries.

  While ONA would have received a chilly reception from its bureaucratic cousins—and self-anointed rivals—under almost any circumstances, their lack of comprehension of Marshall’s intentions made it downright icy. In the Pentagon (and in other large organizations) there always seems to be someone advancing a new method, a new process, a new way of doing business that will cure what ails the bureaucracy. In nearly all instances these purported fixes prove to be illusory. But until they are rooted out, these new practices often play havoc with the institution’s traditional ways of doing business. Armed with these institutional memories and having little if any understanding of what Marshall had in mind, the bureaucracy proved more than willing to give full reign to its fears.

  Some of the blame for this rests with Marshall himself, for he did not go out of his way to allay mistrust. With rare exceptions he was extremely reticent to define or to elaborate his views on what a “net assessment” was or would accomplish. While his 1972 paper on “The Nature and Scope of Net Assessments” provided a fairly clear conception of what he had in mind, it was never widely circulated, even among his own staff. In fact, not until 2002 was the paper discovered filed away in a binder in Marshall’s office.

  While Marshall’s reticence was unhelpful, his reasons for holding back were understandable. He was literally in the process of trying to create a new analytic discipline. Consequently there was little to gain, and much to lose, were he to lock himself into specific ways of bounding, formally defining, or conducting net assessments that he might later regret.

  Marshall had other reasons for his sphinxlike demeanor. Reflecting the enormous value he had received from his own efforts at self-education, he felt that the military officers and civilian assistants on his staff would be better off working out for themselves the details of how to approach their own particular net assessment topic. This attitude would prove intensely frustrating to many of the military officers he would employ. They were accustomed to being given a clear mission to accomplish and provided with guidance—oftentimes detailed in the extreme—as to how to proceed. For those who had grown comfortable operating in such an environment, Marshall’s general guidance and light hand when it came to oversight left many feeling as though they had been abandoned in an analytic no-man’s-land. He viewed the challenge of producing a good assessment as comparable to writing a doctoral dissertation. As with a PhD dissertation, there was no cookie-cutter formula that could be applied to crafting a net assessment.32 ONA staff members undertaking assessments were expected to break new ground in the area being examined—to provide some new and original insights that made a significant contribution to the existing body of knowledge. As one of Marshall’s military assistants later remarked to a frustrated colleague: “He [Marshall] wants you to tell him things that are worthwhile that he doesn’t already know. If he knew what those things are, he wouldn’t need us.”33

  In addition to developing an analytic framework for ONA’s assessments, Marshall needed good data to draw upon in crafting them. Frustrated by the limits of what the intelligence community and military services could—or would—provide, he began developing his own databases. Toward this end he eventually hired BDM’s Phillip Karber to take the lead in building a historical database on NATO and Warsaw Pact forces. In the aftermath of the NSSM 186 Phase I assessment of US and Soviet ground forces, it was decided to conduct a Phase II assessment comparing NATO and Warsaw Pact tactical air forces. Karber agreed to lead the analysis, even though it was burdened by an interagency structure and process similar to that used in Phase I. Karber was provided with a staff consisting mainly of senior field-grade military officers nearing retirement, individuals who were not well suited for the data-gathering drudgery and analytic rigor required for the assessment. The Phase II study began in the summer of 1974 but was not completed until November 1975.

  Because Karber was a civilian contractor, Robin Pirie on the ONA staff was assigned to supervise the tactical air power assessment. Toward the end of the Phase II study a frustrated Pirie suggested to Karber that he modify the contract so that he could bring young BDM analysts on board to begin replacing the military officers who were due to depart once Phase II was completed. Marshall agreed, and decided to use his research budget to break free from bureaucratic constraints by establishing his own independent research effort on the European military balance.

  Marshall’s new approach became known as Project 186, or simply P-186, the “186” a reference to NSSM 186. P-186 was designed as a long-term research project funded and administered by Marshall’s office. It would be run by Karber at BDM. Karber began assembling a small team to support the effort. There would be no more assessments done through an interagency process.

  Project 186 would continue to the end of the Cold War. Its principal focus was to assess the military capabilities of NATO and Warsaw Pact forces in central Europe, an area that included Belgium, the Netherlands, Luxembourg, West and East Germany, Poland, and Czechoslovakia.34 Under ONA, the P-186 effort was shaped and guided by a number of considerations. First and foremost, Karber was to provide Marshall with studies on NATO and Warsaw Pact forces that satisfied the latter’s desire for new insights into the military competition in Europe. Frustrated by the lack of data support from the intelligence community and US military, Marshall decided to develop an independent and integrated database on NATO and Warsaw Pact forces, and inserted contractual language that enabled the database work to continue year in and year out. Finally, the effort was informed by guidelines that emerged from a discussion Karber had had with Schlesinger after briefing the defense secretary during the Phase II work.

  Schlesinger told Karber that he was looking for three things from an assessment. First, he wanted trend data on the opposing forces over time, so that he could see where the two sides had been and how the competition
was developing, as opposed to getting a single snapshot of the current situation. By examining the trends it might be possible to see shifts in the USSR’s relative emphasis on various kinds of military forces and thus insights into their intentions and strategy. Second, Schlesinger wanted a sense of the qualitative aspects of the competition affecting the US or NATO position—factors that went beyond simple tabulations or “bean counts” of tanks, aircraft, artillery pieces, numbers of divisions, and such. Both of these points echoed Marshall’s desire for the effort to mine for new insights on the character of the two alliances. Third, Schlesinger wanted more meaningful force comparisons. For example, instead of comparing NATO and Warsaw Pact strike aircraft “side by side” with one another, as Phase II assessment had done, Schlesinger wanted each side’s tactical aircraft compared against the other’s integrated air defenses, which represented their respective real-world opponents in the event of war.35

  Over time the database work on NATO and Warsaw Pact conventional and theater-nuclear forces proved foundational. As Schlesinger and Marshall knew, comparing the two sides at a single point in time provided no insight into how the competition had gotten to its present state, or where it might be headed in the future. The Project 186 database on NATO and Warsaw Pact forces eventually extended from the mid-1960s through the late 1980s and included reinforcements from the United States and the western military districts of the USSR in addition to the forces stationed in Central Europe. It also acquired enough detail on NATO and Warsaw Pact mobilization and reinforcement capabilities to be able to depict how force ratios would change over time as the alliances began mobilizing prior to war. And, as Schlesinger and Marshall intended, it facilitated more operationally meaningful comparisons between NATO and Warsaw Pact forces, such as armor versus antiarmor and conventional air-strike capabilities versus opposing air defenses.36 This enabled Marshall and his staff to gain a more complete picture of the military balance than had been possible up to that time. For example, if during a month-long period of mobilization there occurred a few days in which the military balance shifted greatly in the Warsaw Pact’s favor, this represented a NATO weakness that needed to be brought to the defense secretary’s attention. After all, if the Soviets saw the same weakness, it could encourage them to exploit it by launching their attack at that point in time, and deterrence would have failed.