AGF Study, NO. 7: Provision of Enlisted Replacements



In 1942, as has been said, the apportionment of total RTC capacity among arms and services, and among individual jobs in each arm and service, had been based on needs for initial filling of units, not on anticipation of losses. When combat developed on a significant scale, after November 1942, the requirement for loss replacements in the combat arms immediately mounted. Replacement needs in the services, except engineers and medical, were little affected by combat. In May 1943 the War Department took steps to reorient the replacement centers more definitely toward the production of loss replacements, estimating that 655,000 replacements would be needed in the ground arms in 1944.42 At a conference on 18 May, between representatives of G-3 WDGS, the Army Service Forces and the Army Ground Forces, it was decided to reduce annual ASF replacement capacity by 140,000, and to increase AGF annual capacity by the same amount.43 The two commands were instructed to determine the actual capacities (number of men in training at a given time) required by each of their respective arms or services to produce the annual totals estimated as necessary for 1944. Immediate compliance was impossible for the Ground Forces, since the actual capacity needed to produce a given number of replacements in a year depended entirely on the length of the training cycle, which was then in doubt.44

When the seventeen-week training program was adopted, the Army Ground Forces estimated that an actual capacity of 277,800 would be needed to meet requirements estimated by the War Department.45 This represented an increase of about 75,000 over the capacity currently in effect. Instructions were issued on 25 July to the Replacement and School Command, and to the Armored and Antiaircraft Commands, apportioning the new capacity to individual replacement centers of the various arms. It was then decided that replacements returning from pre-embarkation furloughs should report directly to the new AGF replacement depots instead of to the replacement training centers from which they came. Less housing was required at replacement centers, and the figure of 277,800 was cut by the War Department to 220,000.46 Before corrective orders could be prepared for the field, the War Department produced new estimates adjusted to the revised Troop Basis of 1 July 1943. The new Troop Basis canceled ten divisions from the mobilization program. With fewer units for which to plan replacements, the War Department scaled down its earlier estimates, and on 23 August 1943 prescribed an actual trainee strength of 203,000 for the AGF replacement centers. ASF centers were drastically reduced to 81,000.47


With a trainee strength of 203,000 (to be in effect by 1 February 1944), the actual capacity set for AGF replacement training centers was almost identical with the actual capacity in effect since the earlier part of 1943 (see Table I). This capacity, since it turned over fewer times per year under a 17-week than under a 13-week cycle, was capable of producing annually about 135,000 fewer replacements than before.48 At the very time when ground operations began in Europe with the invasion of Sicily, and mounting casualties in the ground arms had consequently to be expected, manpower conditions in the United States were such that production of replacements in the ground arms was curtailed. The War Department, in its directive of 23 August, observed that replacement center graduates must be used only for loss replacements (not for new units), "to guard against a breakdown of the replacement system." At the same time, theaters were instructed, as a necessary measure of economy, not to carry over 5% of their total strength in replacements.49 Hence the growth of a reserve of replacements, prerequisite to accurate assignment, was held in check. Actually, in 1944, the theaters built up concealed overstrengths of replacements.50

Restrictions on total numbers made it the more important that distribution of output between arms and jobs should conform as exactly as possible to the incidence of actual losses. The readjustment of AGF replacement training center capacities in the summer of 1943, while not enlarging total AGF capacity, changed the proportion among the arms to meet anticipated combat losses more closely. Infantry suffered the highest proportion of casualties. The proportion of AGF RTC trainees trained in infantry, formerly 37%, was projected in September 1943 to reach 67% by 1 February 1944 (see Table I). But since no clear figures on actual casualties incurred to data by each arm were as yet available to the War Department, these projected capacities were subject to further change.

There remained the question of training the right number of men in individual specialties within each arm. As early as 12 March 1943 the Army Ground Forces, to clarify replacement planning, had requested the War Department to supply new requirements tables.51 These tables showed, for each arm, the number of enlisted men per 1000 required for each job according to specification serial numbers — SSN's. The SSN rates in current tables followed T/O requirements, making no allowance for casualties, since the system had been geared to needs of mobilization. Replacements were being trained in the various specialties, such as SSN 745 Rifleman and SSN 060 Cook, without regard to the fact that battle losses were far higher in some specialties than in others. According to the oversea replacement procedure as codified by the War Department on 26 March 1943, theater commanders,, in requisitioning enlisted replacements, normally specified only the number required in each arm or service without regard to SSN's.52 The WD, in filling the requisition, included specialists according to rates per 1000 based on T/Os. Riflemen and cooks were replaced on the same basis. If a theater commander needed more or fewer replacements of certain SSNs than the tables prescribed, and if he knew his needs, he could specify SSN requirements in his requisition. Under this procedure, however, it was impossible to train the right number of men in the various SSNs in advance.

The WD was unable to comply with the AGF request of 12 March to supply requirements tables. In June tables were provided for various theaters, but they still took no account of casualties. Gen. McNair believed them inadequate as a basis for guiding replacement center training.53 On 26 July, in connection with planning the seventeen-week program, AGF again requested new requirements tables.54 Tables for each theater were desired, and a weighted consolidated table reflecting the total of overseas needs. Should these not be available, figures were requested on the casualties in infantry and armored divisions in Tunisia, broken down by arm or service and by SSN, and on other aspects of the actual losses in North Africa. It was learned from G-1 WDGS that such figures were not immediately available.55


Up to this time, AGF's efforts had been directed entirely at securing revisions of requirements tables which would reflect differential attrition ratios for the various specialties. In June 1943 it was discovered that the preparation of realistic requirements tables, suitable as a basis for planning replacement training, presented another and more fundamental aspect. The crucial fact now noticed was that AGF replacement centers did not train men in all the specialties listed in unit T/Os. Hundreds of different specialties - SSNs - occurred in the T/Os of each arm. Only a small fraction of these - the most basic specialties - were trained by the RTCs. Current requirements tables listed all SSNs used in each arm, giving a percentage requirement for each. Such tables, even if revised to include up-to-date estimates of casualty rates, would be almost useless in planning replacement training, because there would be no correspondence between the multifarious categories of demand and the restricted categories of supply.

One solution to the problem would have been to bring training into conformity with unit requirements, by instituting in the RTCs training in all specialties found in the tables of organization. For reasons set forth in Study No. 31, such complication of replacement training was undesirable and impracticable. The alternative was to bring requirements into conformity with the types of training being conducted, by basing requirement tables on the limited number of SSNs trained. This solution required a scheme of translation by which SSNs required overseas, but not trained in the RTCs, could be regarded as derivatives from the basic SSNs being trained. After much study, involving both a determination of what training was being conducted and an analysis of all T/Os to determine relationships among specialties, groupings of all SSNs of each arm were evolved. Around each basic SSN trained in RTCs were grouped all the other SSNs of the arm, not produced by RTC training, into which a man with the basic training could be expected to develop after appropriate on-the-job training and experience. Thus SSN 745, Rifleman, trained in the infantry RTCs, was parent to the following specialties:



Ammunition Handler



Mortar Gunner



Ammunition NCO



Platoon Sergeant






Section Leader, Gun



Duty NCO



Squad Leader



First Sergeant









Automatic Rifleman



Light Machine Gunner

A demand for an orderly, SSN 695, or for an automatic rifleman, SSN 746, was, so far as the infantry training centers were concerned, a demand for a rifleman. Similar groupings were constructed for all other SSNs.

Preparation of these SSN groupings, or conversion tables, was carried out during June - September 1943 by the Classification and Replacement Division, AGF, in collaboration with representatives of The Adjutant General's Office. On 29 September Army Ground Forces requested new replacement requirement tables, to be based on the SSN groupings, which were now ready.56 The Adjutant General's Office declared itself willing to comply, but unable for want of personnel.57 Army Ground Forces detailed one captain, one warrant officer, and six enlisted men to compile data in the office of The Adjutant General.58 Preparation of requirements tables was now a relatively simple matter. Figures were available showing the rate per 1,000 at which men would be needed for each SSN in a combat arm. Adding together the separate rates for all SSNs in a group and the rate for the basic specialty gave the requirement per 1,000 for training in that basic specialty in the RTCs. By the end of November new requirements tables for infantry were ready. These were soon followed by tables for the other ground arms.59 The machinery was established for planning replacement training in anticipation of future needs and in conformity with the training system in effect in the centers.


In November 1943 the SSN groupings discussed above were published by the War Department as Circular 283, as a guide to commanders in requisitioning RTC-trained men to fill vacancies in their units.60

By the end of 1943, as a result of these measures, not only had the training of replacements been lengthened from 13 to 17 weeks and otherwise made more thorough, but administrative changes had been introduced to assure that replacements arriving in the theaters should be properly qualified and ready for use, and be divided among arms and jobs in proportion to actual needs of the theaters. It was hoped that they would therefore be assigned to positions for which they were trained. With the adoption of the measures described above, which would be fully in effect by the early months of 1944, replacement production in the ground arms may be said to have shifted from a basis of mobilization to a basis of overseas loss replacement requirements. Reports from Italy at the end of 1943 stated that replacements were generally satisfactory, except for occasional defects, such as physical softness due to time in transit, which could be readily corrected overseas. It was believed at the headquarters of the Army Ground Forces, in January 1944, that the grounds for earlier complaints concerning overseas replacements had been removed, that replacements were being satisfactorily trained and processed, and that except possibly for the question of supply in adequate numbers the replacements issue was settled. This belief was destined to be abruptly upset.61

Those who held that replacements should have six months of training, including experience in tactically organized units, continued to urge their views. On 11 August 1943, G-1 WDGS suggested withdrawing 1,000 men from each of twenty or more divisions in the United States for overseas replacements, and using replacement center men to replenish these divisions.62 The Army Ground Forces repeated the arguments against such a system.63 It was pointed out in addition that the number of men obtained in this way would be only a small fraction of all replacements required in the ground arms, and that divisions in the United States had already suffered from extreme turnover of personnel. G-1 was converted; G-3 agreed with the Army Ground Forces.64 As a training policy, the depleting of divisions to obtain replacements was rejected. But what was rejected as a policy became necessary as an expedient. By January 1944 26,000 infantrymen had been withdrawn by the Army Ground Forces from unalerted infantry divisions, leaving most such divisions about 2,000 understrength.65 This was necessary because the productive capacity of infantry replacement centers was insufficient.



Go to:

Next Chapter

Return to Table of Contents

Search CMH Online
Last updated 17 October 2005