Keep From All Thoughtful Men

by James G. Lacey

  Had Marshall strongly backed a cross-Channel invasion for 1943 at Casablanca, as the official historians contended, there were several instances during the meeting where Marshall remained silent when, if he had still been advocating a 1943 invasion, he most definitely would have been on his feet. For instance, at one point Sir Alan Brooke stated that with the limited ground forces available he did not believe that the Allies could muster a sufficient threat to northern France to force the Germans to divert much from the Eastern Front. To this Marshall made no rejoinder. When Brooke continued that the Germans still had sufficient strength to overwhelm us on the ground or perhaps hem us in to such an extent that any expansion of the bridgehead would be impossible, Marshall remained silent. Marshall even let slide by without comment a statement by one of the American chiefs, Hap Arnold, to the effect that based on discussions so far it looked very much as though no continental operations on any scale were in prospect before spring 1944. Arnold even went so far as to include the suspension of any further consideration of the small invasion of Brest without attracting Marshall’s ire. Finally, when Admiral King said it was necessary to accept that the Allies could do nothing in France before April 1944, Marshall failed to protest.51

  What the Allied chiefs did agree on was a firm commitment to land substantial forces on the north coast of France in 1944, and Marshall seemed content with that result. If, as the evidence indicates, Marshall was in agreement with a 1944 cross-Channel invasion date, one still has to explain what turned his mind from his previous fixation on reentering the European continent at the earliest possible date. Earlier in the war Marshall had pushed hard for an invasion in 1942, long before American forces were even remotely capable of such a feat. When that failed, he pushed strongly for a 1943 invasion. But by November 1942 he was at best lukewarm to the idea, and by the end of the year he appears to have given up on a 1943 invasion entirely. Prior to November 1942 Marshall fought tooth-and-nail against anything he perceived as a diversion from that goal to the point of threatening to divert the entire war effort to the Pacific if his plans for a rapid invasion of Europe were not adhered to. In all of this, he had the full support, if somewhat grudgingly given, of the Joint Chiefs.

  Sometime between Torch and the Casablanca Conference, however, Marshall abandoned his single-minded crusade for a second front in 1943 and supported a major post-Torch diversion of resources to further Mediterranean operations. In this, too, the remainder of the Joint Chiefs joined him. The Americans were not overawed or overwhelmed by superior British negotiating skills or staff procedures, as historians have often suggested they were. Rather, they had simply changed their minds about the wisdom of a major 1943 invasion, though they do appear to have been more than a bit reticent about announcing their change of heart and thereby admitting that British strategists had been right from the beginning.

  But why the change of heart? Part of the answer is that after Torch the American chiefs had sobered up to just how difficult a contested landing in northern France against a determined enemy might be. Eisenhower’s doubling of what he considered the minimum number of divisions necessary for a successful invasion had shocked Marshall, as evidenced by his comments on the new requirement for twelve divisions as a minimum several times during the Casablanca Conference and again when meeting separately with the president. Running over second-rate Vichy French divisions had cost the American combat forces involved a surprising number of casualties; how much worse would it be facing the battle-hardened Wehrmacht?

  Even more important, though, was that just about the time of Torch, the production agencies had informed the Joint Chiefs that they were not going to provide nearly the amount of munitions the chiefs were expecting to equip the Army’s formations. Moreover, Marshall discovered that virtually all these cuts were going to fall on the Army while the Navy and air forces would escape almost unscathed. In fact, the day after Torch, Marshall sent a bitter memo to the president over troop estimates for 1943 in which he informed Roosevelt that this would cause a reduction of fourteen transportable divisions in 1943, and that such a reduction would preclude any chance of a 1943 cross-Chanel invasion.52 Though this reduction turned out to be a misunderstanding that was eventually straightened out after a flurry of notes between Marshall and Roosevelt, Marshall still would face 1943 with significantly fewer divisions than he had assumed, because the production lines simply could not produce enough equipment.

  CHAPTER 10

  Why Marshall Changed His Mind

  Questions remain. Why did Marshall hold onto his hope of a major 1943 invasion, and what caused his rapid position reversal? A presentation given by a staff officer who served on the Joint Staff’s logistics planning section gives the short answer to these questions: “The very date of the landing in Normandy was based on studies of the availability of resources.” 1 In essence, when Marshall learned that the force he expected to have on hand in June or July of 1943 would not be available until June of 1944, the decision was made for him. Even a man as determined as Marshall could see that an invasion was impossible if the forces required did not exist.

  The second question is, Why did it take Marshall so long to come to this conclusion? The blame must fall squarely on Somervell’s shoulders. 2 Either because he was absolutely convinced the civilian production experts were wrong, or possibly because he, like almost every other Army officer, never wanted to disappoint Marshall, Somervell failed to provide his boss with any indication of looming production problems. Even when shortfalls in 1942 production were becoming obvious, Somervell continued to present Marshall with optimistic reports.

  As late as 24 August 1942, Somervell approved an estimate for Army strength that Marshall forwarded to the other Joint Chiefs. It stated, “This plan provides for a ground Army of 111 combat divisions and an air force of 224 combat groups with a total mobilization of 7 and a half million men by December 31,1943.This is a modest establishment as compared with the actual and potential armies of our enemies. . . . This force represents an increase over the full 1942 Army in the amount of 37 division and 109 air groups.”3 Admiral King opposed this Army estimate because he believed the Army should grow only at the rate the Navy could move it and sustain it overseas.4 This argument reached Roosevelt’s ears, and the president ordered Marshall and King to take “a fresh and realistic look at 1943 over-all requirements from a strategic point of view. This implies that certain assumptions must be made on the estimated requirements for the United States forces of all types in the various probable theaters of war, by number and general composition, are determined.”5

  As a result of this memorandum, Marshall requested that Somervell reexamine the 1943 mobilization plan and report back to him on whether it was feasible. Somervell had his staff undertake this project at the same time Kuznets was delivering his final feasibility report to the Planning Committee. By mid-September, Somervell’s analysis was complete; he then confirmed to Marshall that the strength of the Army to be mobilized by the end of 1943 had been “tentatively established” at approximately 7.5 million, adding, “The composition and strength of this force has been primarily dictated by our proposed strategic operational plans, which in turn have been integrated with the all important factors of available shipping and production forecasts of the required munitions.”6

  Somervell made this last statement having already read and replied to Kuznets’ feasibility study. He thus was fully aware that the civilian production experts did not believe the Army’s 1943 plans were achievable. Disregarding their views, he continued by stating, “The latest production analysis indicates that the objectives of the Army Supply Program for 1943 will be met . . . I feel confident that adequate munitions will be available to meet the full operational requirements for that portion of the total 1943 Army destined for overseas or for the defense of the Continental United States.” Somervell concluded by recommending that “no change be made at this time in the Army Supply Program, which represents the munitions required to sustain the plan
ned operations of the Army of 7,500,000 men by 31 December 1943” (emphasis added).7

  So, a full two weeks after Kuznets had informed Somervell that U.S. industry could not meet the 1943 munitions objectives and less than a month before he would have to reverse his position and admit the Planning Committee was right, Somervell was telling Marshall that the Army could have everything planned for in 1943. Even more inexplicably, even after Somervell had conceded the Planning Committee’s arguments, he could not bring himself to give the bad news to Marshall.

  Instead, he continued to insist that the civilians were wrong and that Marshall could still get what he wanted. In a memorandum to Marshall dated 30 October 1942, Somervell related that Nelson and his statisticians believed this dollar amount was impossible to achieve. Somervell then attacked this position by claiming that because actual production costs to date were less than estimated, the discrepancy between what the Army programmed and what the statisticians said was possible was not that great. He continued by questioning the validity of comparing requirements on a dollar basis and claimed that the real production choke point was a lack of raw materials. For this, Somervell blamed the War Production Board (WPB). Remarkably, he supported his arguments by claiming that Leon Henderson shared his opinion: that U.S. output could be materially increased by better management. He then blamed the WPB’s failure to properly allocate materials as the greatest single cause of mismanagement. Somervell finally concluded that the current Army program would be met because he expected, “a marked improvement both in management and in the vigor with which the production of raw materials would be handled.”8

  Here, Somervell had grossly misstated Henderson’s position, while at the same time demonstrating either that he still did not understand the statistical concepts that underpinned the feasibility study or that he was willing to take advantage of Marshall’s lack of education in this regard to paint a rosy picture of the production program he knew was not true. Which case is true cannot be known, although other documentation in Somervell’s papers indicate that he fully understood how all requirements could be translated in dollar values. He even used this method within his own office to explain requirements to logistics officers. Whatever the truth, it was still a remarkable memorandum to send to Marshall while the Joint Chiefs of Staff (JCS) planners were already hard at work reducing munitions requirements. In fact, only ten days later the JCS planners would inform Marshall that they would have to cut his Army supply program by more than 20 percent.

  The net result of this reduction of military objectives was that the Army reduced its 111-division goal for 1943 to 100 divisions. Moreover, the Army immediately cancelled the organization of fourteen divisions that Marshall had counted on being available by June 1943.9 Even this reduction (along with a reduction in the strength of units already organized) was not sufficient to make sure newly organized divisions could be properly equipped. As a result, in January 1943 the Army slowed the activation schedules for many of its planned divisions. For instance, it deferred three divisions planned for activation in May, June, and August 1943 to the last months of 1943.10

  In summary, without even taking into account shipping constraints or the drain of units sent to the Pacific and other operational theaters, the production experts had forced the Army to cut by more than half the number of troops Marshall thought would be in the United Kingdom by June 1943. This reduction occurred at the same time Marshall was beginning his preparations for Casablanca. Thus, the Army’s chief of staff had no choice but to postpone his plans to invade northern Europe in 1943 for at least a year—not because of any change in the strategic situation or in his own ideas of how to win the war, but because the economists had forced him to do so.

  APPENDIX 1

  The Feasibility Concept

  As Robert Nathan socialized the idea of feasibility within various policy circles, he asked Simon Kuznets to prepare an in-depth memorandum he could leave with people that further developed the concepts, a memorandum he often presented in outline form in conversation. Since these concepts were the underlying basis for the “great feasibility dispute of 1942” it is profitable to outline them as understood by the disputes’ participants. Below is a digest version of Kuznets’ memorandum to Nathan on feasibility.1

  1. The Concept of Feasibility

  The feasibility of a set of production goals is far from an unequivocal concept. It is important to distinguish the different variants and raise the question as to which concept of feasibility we wish to be guided by.

  Three variants may be employed:a. We may ask what production goals are feasible under conditions as they exist now or are definitely and securely in view. This, for our purposes, is the concept of minimum feasibility since it measures potential output without taking into account the intensification of the war production effort that may be expected. If applied, it would tell us what military or other production can be expected under a set-up identical with the present and including changes only insofar as they have already been enacted.

  b. We may ask what production goals are feasible with an all-out effort, the latter developing not in an ideal situation but with human inertia, frailties, ignorance, selfish interests, etc., exercising a retarding effect. This concept of realistic maximum feasibility would yield an approximation to output to be produced under conditions in which an increasingly all-out effort encounters various resistances that might be expected.

  c. Finally, we may ask what production goals are feasible under ideal conditions—i.e., under conditions in which every member of the community devotes himself whole-heartedly to the war effort in complete disregard of private interests, and in which such devotion is most intelligently and effectively applied.

  2. Why Worry about Feasibility

  The basic reason why feasibility is considered is that an attempt to attain goals that are unattainable is likely to result in a situation which, for practical purposes, is worse than trying to attain more moderate but feasible goals. An attempt to secure unattainable goals is likely to result in an unbalanced performance in the sense that facilities will be built for which there will not be enough raw materials; that semi-finished products will be produced which it will be impossible to finish; and that finished products will be produced for which indispensable complements will be lacking [emphasis in original]. If this criterion of balance were the only consideration, the best thing to do would be to set minimum goals. But, on the other hand, goals set will determine the magnitude of the productive effort eventually attained. Setting the goals low may, therefore, assure a fully balanced performance in that all equipment and inter-related final products will be turned out at a rate permitting their full use. But, the balance will be secured at a level of production far short of one that could be attained. The real problem in determining feasibility is therefore to set your goals as high as possible, compatible with the attainment of a balance that is needed to assure usability of equipment and final products [emphasis in original].

  This reasoning can now be applied to a choice among the three concepts of feasibility defined above. If the criterion of balance is paramount it will be safest to apply the minimum concept of feasibility. But it obviously involves the sacrifice of possibly a substantial part of the production effort. The real choice is, therefore, one between the realistic maximum and the ideal maximum criteria of feasibility.

  The advantage of using the former is that is permits the setting of goals with fair expectation of balance in the results. Even this inference must be qualified since uncertainties attach to the forecast of what an all-out effort can attain in the face of various obstacles. Both the effort and the obstacles are intangibles; and fluctuations in the fortunes of war and various other unforecastable elements may introduce a sizeable error in the total estimated to be feasible. On the other hand, this concept of feasibility lacks the advantages of the ideal maximum in that it may not set the goals high enough. In particular, it omits the possible influence on the realized productive effo
rt of the incentives provided by maximum goals.

  This important point can be made more specific. Let us assume that a realistic consideration of what an all-out effort in the production in munitions can attain in the face of the usual type of difficulties involves the consumption of X million tons of copper. With X million tons needed for military production, T million tons may be left for non-military consumption. The goal thus set does not recognize the possible influence on the demand for or allocation of raw material or the setting of a higher goal for military production. If a higher goal were set, let us say one that would require X plus A million tons for military production, and a shortage developed, the pressure of that demand might make it possible.

  a. To divert to military production not the full amount of X plus A, but perhaps X plus one-half of A

  b. To force a revision of specifications by which a larger total of finished military equipment would be turned out with A million tons of copper.

  In the realistic situation in which the diversion of economic activities to war production occurs, the setting of higher goals for military production is an important factor in accelerating diversion toward the war effort; and the setting of low goals may quite effectively retard our war effort. It may be argued that the slowness with which the various restrictive mechanisms have been applied to civilian consumption in this and other countries and the delay that has occurred in the conversion of a number of industries that are potentially important for military production have been due largely to delay in formulating high enough goals for military objectives.