One of the thorniest problems of strategic planning for the war against Japan was to decide whether the principal objective of drives that had brought the Allies into the western Pacific should be Luzon or Formosa. The decision was made by the U.S. Joint Chiefs of Staff, since the Pacific was an American area of strategic responsibility. They made it after long debate and careful study of the views of the commanders in the Central and Southwest Pacific theaters. Among the considerations that determined their choice when they finally made it, logistical factors played the major role, but here, as in other connections, they had to take into account the commitments and progress of the Allies in other theaters, and particularly in Europe. It was in this sense a decision in global strategy.
In January 1945, after more than three years of war, United States forces returned to Luzon Island in the Philippines, where in 1942 American troops had suffered a historic defeat.  The loss of
 This essay is essentially Chapter 1 of Triumph in the Philippines, by Robert Ross Smith, a forthcoming volume in UNITED STATES ARMY IN WORLD WAR II. Additional background information is to be found in The Approach to the Philippines (Washington, 1953), by the same author; M. Hamlin Cannon, Leyte: The Return to the Philippines (Washington, 1954); and Philip A. Crowl and Edmund G. Love, Seizure of the Gilberts and Marshalls (Washington, 1953), all in UNITED STATES ARMY IN WORLD WAR II. Maurice Matloff, Strategic Planning for Coalition Warfare, 1943-1944 (Washington, 1959) in that series, deals at length with the broad strategic aspects of the decision.
the Philippines in May of that year, after the disaster that befell the U.S. Pacific Fleet at Pearl Harbor, had rendered obsolete the inoperable American prewar plans for action in the Pacific in the event of war with Japan.  By the late spring of 1943 the U.S. Joint Chiefs of Staff (who, by agreement of the U.S.-British Combined Chiefs o Staff, were responsible for the conduct of the war in the Pacific) had developed a new strategic plan for the defeat of Japan. The plan was neither sacrosanct nor immutable-it was not intended to be-but its underlying concepts governed the planning and execution of operations in the Pacific during a year and a half of debate over the relative priority of Luzon and Formosa as primary objectives of an Allied drive into the western Pacific. 
The plan was premised upon the concept that the Allies might very well find it necessary to invade Japan in order to end the war in the Pacific. (See Map VIII, inside back cover.) The Joint Chiefs of Staff foresaw that intensive aerial bombardment of the Japanese home islands would be prerequisite to invasion, and that such bombardment would have to be co-ordinated with combined air, surface, and submarine operations aimed at cutting Japan's overwater lines of communication to the rich territories she had seized in the Netherlands Indies and southeastern Asia. The Joint Chiefs believed that the Allies could best undertake the necessary bombardment of Japan from airfields in eastern China. They decided that to secure and develop adequate air bases in China, Allied forces would have to seize at least one major port on the south China coast. The Allies would require such a port to replace the poor overland and air routes from India and Burma as the principal means of moving men and materiel into China.
To secure a port on the China coast, and simultaneously to cut Japan's line of communication to the south, the Allies would have to gain control of the South China Sea. Gaining this control, the Joint Chiefs realized, would in turn involve the seizure and development of large air, naval, and logistical bases in the strategic triangle formed by the south China coast, Formosa, and Luzon. But before they could safely move into this triangle, the Joint Chiefs decided, the
 See Louis Morton, The Fall of the Philippines (Washington, 1953), UNITED STATES ARMY IN WORLD WAR II, for the opening phases of Japan's attack in the Pacific and a description of prewar plans with especial reference to the Philippines. Morton's general volume on the Pacific theaters, Strategy and Command: Turning the Tide, 1941-1943, will cover prewar plans in more detail.  See JCS 287/1, 8 May 43, and JPS 67/4, 29 Apr 43, both entitled Strategic Plan for the Defeat of Japan, and associated papers in OPD ABC 381 Japan (8-27-42) Secs. 1 and 2.
Allies would have to secure air bases in the southern or central Philippines from which to neutralize Japanese air power on Luzon. The Allies might also need staging bases in the southern and central Philippines from which to mount amphibious attacks against Luzon, Formosa, and the China coast.
In accordance with these 1943 plans, Allied forces in the Pacific had struck westward toward the strategic triangle along two axes of advance. Air, ground, and naval forces of the Southwest Pacific Area, under General Douglas MacArthur, had driven up the north coast of New Guinea to Morotai Island, lying between the northwestern tip of New Guinea and Mindanao, southernmost large island of the Philippine Archipelago. Simultaneously, Admiral Chester W. Nimitz, commander of the Pacific Ocean Areas, had directed the forces of the Central Pacific Area in a drive through the Gilberts, Marshalls, and Marianas to the Palau Islands, some 500 miles east of Mindanao.
The Importance of Formosa
Studying various plans for Allied entry into the strategic triangle, the Joint Chiefs and their subordinate advisory committees concluded that Formosa constituted the most important single objective in the target area.  The island possessed so many obvious advantages and was located in such a strategically important position that most planners in Washington believed the Allies would have to seize it no matter what other operations they conducted in the western Pacific. Until they seized Formosa, the Allies would be unable to establish and secure an overwater supply route to China. Formosa, therefore, seemed a necessary steppingstone to the China coast. Moreover, Allied air and naval forces could sever the Japanese lines of communication to the south much more effectively from Formosa than from either Luzon or the south China coast alone. Furthermore, from fields in northern Formosa, the Army Air Forces' new B-29's could carry heavier bomb loads against Japan than from more distant Luzon. 
Many planners considered Formosa such a valuable strategic prize that they devoted considerable attention to the possibility of bypassing all the Philippines in favor of a direct descent upon Formosa. Discussion of this proposal waxed and waned in Washington during much of 1943 and 1944 despite the fact that the strategic outline
 Nimitz' Pacific Ocean Areas included the North, Central, and South Pacific Areas.  See the sources cited in note 1, above, and also JCS 713, 16 Feb 44, Strategy in the Pacific; JCS 713/1, 10 Mar 44, Future Opns in the Pacific, and associated sources in OPD ABC 384 Pacific (1-17-43).  Northern Formosa, affording some good airfield sites, lies 300-odd nautical miles closer to Tokyo than the best airfield areas of northern Luzon.
plan for the defeat of Japan called for the seizure of bases in the southern or central Philippines before going on into the Luzon-Formosa-China coast triangle. Such discussions found the War and Navy Departments internally divided. Admiral Ernest J. King, Chief of Naval Operations and Navy member of the Joint Chiefs of Staff, was a leading advocate of plans to bypass the Philippines. On the other hand, Admiral Nimitz and other ranking naval commanders in the Pacific favored at least reoccupying the southern or central Philippines before striking on toward Formosa. These officers believed it would be impossible to secure the Allied line of communications to Formosa until Allied land-based aircraft from southern Philippine bases had neutralized Japanese air power on Luzon. 
General George C. Marshall, Chief of Staff and Army member of the Joint Chiefs, played a relatively inactive part in the debate until late 1944, but at one time at least seemed inclined toward bypassing both the Philippines and Formosa in favor of a direct invasion of Kyushu in southern Japan. Some officers high in Army councils, including Lt. Gen. Joseph T. McNarney, Deputy Chief of Staff, strongly advocated bypassing the Philippines on the way to Formosa. General Henry H. Arnold, Army Air Forces member of the Joint Chiefs, also appears to have maintained through much of 1943 and 1944 that it might prove desirable to bypass the Philippines.  Other Army planners, including those of the chief logistician, Lt. Gen. Brehon B. Somervell, commander of the Army Service Forces, favored taking the entire Philippine Archipelago before making any move toward Formosa or the China coast. In the field, General MacArthur stood adamant against bypassing any part of the Philippines, a stand in which he had the support of most other ranking Army officers in the Pacific. 
 Memo, King for Marshall, 8 Feb 44, sub: CINCSWPA Despatch [sic] C- 121702 Feb 44, and other documents in OPD ABC 384 Pacific (28 Jun 43), JCS Memo for Info No. 200, 7 Mar 44, sub: Sequence and Timing of Opns CenPac Campaign [a rpt by Nimitz] and associated sources in OPD ABC 384 Pacific (1-17-43) Secs. 3-A and 4; Supplementary Minutes, JCS 145th and 150th Mtgs, 8 Feb and 7 Mar 44; Minutes, JCS 151st Mtg, 11 Mar 44; Minutes, JPS 125th Mtg, 2 Feb 44; Rad, Nimitz to King and MacArthur, 4 Jul 44, CM-IN 2926.  Memo, Marshall for King 10 Feb 44, no sub, in OPD ABC 384 Pacific (28 Jun 43); Memo, Col Charles K. Gailey, Jr. (Exec O OPD) for Maj Gen Thomas T. Handy (ACofS OPD), 22 Feb 44, no sub [reporting McNarney remarks], and associated materials in OPD ABC 384 Pacific (1-17-43) Sec. 3-A;JPS 418/1, 23 Mar 44, Basic Decision Which Will Give Strategic Guidance for . . . the War in the Pacific, in OPD ABC 384 Pacific (8 Mar 44); Rad, Marshall to MacArthur, 23 Jun 44, CM-OUT 55718, Supplementary Minutes, JCS 150th Mtg, 7 Mar 44.  Memo, Somervell for Handy, 15 Jul 44, sub: JCS 924, and associated papers in OPD ABC 384 Pacific (1-17-43) Sec. 3-A, Rad, MacArthur to Marshall, C-3302, 20 Jun 43 CM-IN 13149; GHQ SWPA, Estimates of the Situation and Rough Draft RENO Plan [RENO I]. 25 Feb 43, photostat copy in OCMH files; Minutes, JPS 134th, 157th, and 159th Mtgs, 8 Mar, 28 Jun, and 26 Jul 44. See also below, p. 468.
In March 1944 the Joint Chiefs had directed MacArthur to be ready to move into the southern Philippines before the end of the year and to make plans to invade Luzon during February 1945. Simultaneously, they had ordered Nimitz to prepare plans for an assault against Formosa in February 1945.  These directives, which left in abeyance the relative priority of Luzon and Formosa, ostensibly settled the question of re-entry into the Philippines, but in mid-June the Joint Chiefs of Staff reopened the question of bypassing the archipelago.
Developments in the Pacific, Asia, and Europe between mid-March and mid-June 1944 tended to support those planners who wanted to bypass the Philippines. The U.S. Army had acquired new intelligence indicating that the Japanese were rapidly reinforcing their bastions throughout the western Pacific, including Formosa. Thus, the longer the Allies delayed an attack on Formosa, the more the operation would ultimately cost. Army planners suggested that the Allies might be able to reach Formosa during November 1944 if the Joint Chiefs immediately decided to bypass the Philippines. Moreover, the Joint Chiefs were beginning to fear an imminent collapse of Chinese resistance-some planners felt that the only way to avert such an eventuality would be the early seizure of Formosa and a port on the China coast without undertaking intermediary operations in the Philippines.  The Joint Chiefs were probably also stimulated by the success of the invasion of Normandy in early June and by the impending invasion of the Marianas in the Central Pacific, set for 15 June. At any rate, on 13 June, seeking ways and means to accelerate the pace of operations in the Pacific, and feeling that the time might be ripe for acceleration, the Joint Chiefs asked Admiral Nimitz and General MacArthur to consider the possibilities of bypassing all objectives already selected in the western Pacific, including both the Philippines and Formosa. 
Neither Nimitz nor MacArthur gave the Joint Chiefs any encouragement. Both declared that the next major step in the Pacific after the advance to the Palaus-Morotai line would have to be the seizure of air bases in the southern or central Philippines. The Joint Chiefs' subordinate committees, examining the theater commanders' replies and undertaking new studies of their own, reaffirmed the concept that the Allies would have to move into the central or southern Philippines before advancing to either Formosa or Luzon. Like MacArthur
 JCS 713/4, 12 Mar 44, Future Opns in the Pacific, in OPS ABC 384 Pacific (1-17-43) Sec. 3-A. See also Smith, Approach to the Philipigines, Ch. I.  JCS 713/8, 13 Jun 44, Future Opns in the Pacific, in OPD ABC 384 Formosa (8 Sep 43) Sec. 1-C; Rad, JCS to MacArthur and Nimitz, 13 Jun 44, CM-OUT 50007: Rad, Marshall to MacArthur, 23 Jun 44, CM-OUT 55718.  Rad, JCS to MacArthur and Nimitz, 13 Jun 44, CM-OUT 50007.
and Nimitz, the advisory bodies saw no possibility of a direct jump to Japan. The Joint Chiefs of Staff, apparently with some reluctance, agreed. 
Meeting with President Franklin D. Roosevelt in a conference at Pearl Harbor in late July 1944, both MacArthur and Nimitz again emphasized that MacArthur's forces would have to be firmly established in the southern or central Philippines before any advance to either Formosa or Luzon could take place-on this point almost everyone was agreed. MacArthur then argued persuasively that it was both necessary and proper to take Luzon before going on to Formosa, while Nimitz expounded a plan for striking straight across the western Pacific to Formosa, bypassing Luzon. Apparently, no decisions on strategy were reached at the Pearl Harbor conferences.  The Formosa versus Luzon debate continued without let-up at the highest planning levels for over two months, and even the question of bypassing the Philippines entirely in favor of a direct move on Formosa again came up for serious discussion.  The net result of the debate through July 1944 was reaffirmation of the decision to strike into the southern or central Philippines before advancing to either Formosa or Luzon. The Joint Chiefs still had to decide whether to seize Luzon or Formosa, or both, before executing any other major attacks against Japan.
The Views Presented
General MacArthur was a most vigorous adherent of the view that the Allies would have to secure Luzon before moving any farther toward Japan. Contrary to the views held by the Joint Chiefs of Staff, MacArthur believed that Luzon was a more valuable strategic prize
 Rad, MacArthur to Marshall, CX-13891, 18 Jun 44, CM-IN 15058; Rad, Nimitz to King and MacArthur, 4 Jul 44, CM-IN 2926; Rad, Marshall to MacArthur, 23 Jun 44, CM-OUT 55718; Minutes, JPS 157th, 158th, and 159th Mtgs, 28 Jun and 12 and 21 Jul 44; JPS 404/5, 23 Jun 44, Future Opns in the Pacific, and related papers in OPD ABC 384 Formosa (8 Sep 43) Sec. 1-C and ABC 384 Pacific (1-17-43) Sec. 4; see also Smith, Approach to the Philippines, pp. 451-52.  No evidence that strategic decisions were reached at Pearl Harbor is to be found in contemporary sources. See Rad, MacArthur to Marshall, C-15589, 1 Aug 44, CM-IN 496; Memo, King for Marshall and Arnold, 9 Aug 44, no sub [quoting parts of a letter on the Pearl Harbor Conference from Nimitz to King, dated 31 Jul 44], in OPD ABC 384 Pacific (1-17-43) Sec. 4; Ltr, Lt Gen Robert C. Richardson, COMGENPOA, to Marshall 1 Aug 44, no sub, in OPD Personal File on Gen Marshall. See also Fleet Admiral William D. Leahy, I Was There, (New York: Whittlesey House, 1950), pp. 247-52. Leahy also participated in the conferences, and Richardson was MacArthur's host in Hawaii).  See, for example, Minutes, JPS 160th Mtg, 2 Aug 44.
than Formosa. He declared that the Allies would need to reoccupy the entire Philippine Archipelago before they could completely sever Japan's lines of communication to the south. MacArthur also believed that an invasion of Formosa would prove unduly hazardous unless he provided air and logistical support from Luzon. Finally, he suggested, if the Allies took Luzon first they could then bypass Formosa and strike for targets farther north, thus hastening the end of the war. The Luzon-first course of action, he averred, would be the cheaper in terms of time, men, and money. 
In addition, MacArthur considered that bypassing part of the Philippines would have the "sinister implication" of imposing a food blockade upon unoccupied portions of the archipelago. (His meaning here is not clear, inasmuch as his own plans called for seizing a foothold in southeastern Mindanao, jumping thence to Leyte in the east-central Philippines, and then going on to Luzon, initially bypassing the bulk of Mindanao, the Sulu Archipelago, and most of the Visayan Islands.)  MacArthur had a more cogent argument, and one that was bound to have some influence upon planning in Washington. The reoccupation of the entire Philippine Archipelago as quickly and early as possible, was, MacArthur said, a national obligation and political necessity. To bypass any or all the islands, he declared, would destroy American honor and prestige throughout the Far East, if not in the rest of the world as well.
Just as General MacArthur was the most vigorous proponent of Luzon, so Admiral King was the most persistent advocate of the Formosa-first strategy. King believed that the seizure of Luzon before Formosa could only delay the execution of more decisive operations to the north. He also argued that the capture of Formosa first would greatly facilitate the subsequent occupation of Luzon. Moreover, King pointed out, the Allies could not secure and maintain a foothold on the China coast until they had seized Formosa. Finally, he suggested, if the Allies should bypass Formosa, then the principal objective in the western Pacific should be Japan, itself, not Luzon. 
MacArthur believed that the plans to bypass Luzon were purely Navy-inspired.  Actually, the War and Navy Departments were as
 Rad, MacArthur to Marshall, C-3302, 20 Jun 43, CM-IN 13139; Rad, MacArthur to Marshall, CX-13891, 18 Jun 44 CM-IN, 15058; Rad, MacArthur to Marshall, C-15689, 3 Aug 44, CM-IN 2479; RENO I, 25 Feb 43; GHQ SWPA, Basic Outline Plan for MUSKETEER (Philippine) Opns [MUSKETEER I], 10 Jul 44.  MUSKETEER I, 10 Jul 44; MUSKETEER II, 29 Aug 44; MUSKETEER III, 26 Sep 44.  See the sources cited in note 6, above, and also JCS 713/10, 4 Sep 44 [memo from King for the JCS], and associated papers in OPD ABC 384 Pacific (1-17-43) Sec. 5; Minutes,JCS 171st and 172d Mtg, 1 and 5 Sep 44.  Rad, MacArthur to Marshall, C-15689, 3 Aug 44, CM-IN 2479.
internally split during the Luzon versus Formosa debate as they had been earlier over the question of bypassing all the Philippines. For example, at least until mid-September 1944 General Marshall favored the Formosa-first strategy and like Admiral King had expressed the opinion that Japan itself, rather than Luzon, should be considered the substitute for Formosa. Most Army members of the Joint Chiefs' subordinate committees held similar views, and until September consistently pressed for an early decision in favor of Formosa. Army Air Forces planners, during the summer of 1944, expressed their interest in Formosa as a site for B-29 bases. 
Admiral Nimitz, the ranking naval officer in the Pacific, went on record until late September as favoring Formosa first. However, there are indications that his views were not enthusiastically shared by his staff, and there are grounds to believe that Nimitz grew steadily more lukewarm toward the idea of seizing Formosa. Nimitz had been at variance with Admiral King on the question of bypassing the entire Philippine Archipelago, and it is possible that his support of the Formosa-first strategy stemmed at least in part from deference to King's judgment. A hint of Nimitz' attitude is apparent in the fact that his staff was preparing plans to seize Okinawa, as a substitute for Formosa, well before such an operation gained serious consideration among high-level planners in Washington. 
The next ranking naval officer in the Pacific, Admiral William F. Halsey, commander of the Third Fleet (and until 15 June 1944 commander of the South Pacific Area as well), steadfastly opposed the Formosa-first plan. He wanted to go to Luzon and bypass Formosa in favor of seizing Okinawa. In this connection Halsey relates a classic story concerning a discussion between his chief of staff, Vice Adm. Robert B. Carney, and Admiral King. King, propounding his Formosa plan to Carney, who was arguing in favor of Luzon, asked, "Do you want to make a London out of Manila?" Carney's reply was: "No sir, I want to make an England out of Luzon." 
 JPS 414/10, 29 Jun 44, Future Opns in the Pacific, and associated sources in OPD ABC 384 Formosa (8 Sep 43) Sec. 1-C; JCS 713/14, 7 Sep 44, Proposed Directive, and connected materials in OPD ABC 384 Pacific (1-17-43) Sec. 5; Minutes, JCS 171st-173d Mtgs, 1, 5, and 8 Sep 44; Minutes, JPS 160th, 162d, 163d, 165th, and 167th Mtgs, 2, 10, 16, and 28 Aug and 2 Sep 44.  Rads, Nimitz to King, 18 and 24 Aug 44, CM-IN 16755 and CM-IN 22182; Rad, Nimitz to Arnold, 5 Sep 44, CM-IN 4996; Memo [unsigned but prepared by Col William L. Ritchie of OPD, who had just returned to Washington after talking with most of the ranking Army and Navy commanders in the Pacific], n.d. [circa 15 Aug 44], sub: Notes for Discussion With General Marshall [hereafter cited as Ritchie Notes for Marshall], and related sources in OPD 384 Pacific (1-17-43) Sec. 5; Fleet Admiral William F. Halsey, USN, and Lieutenant Commander J. Bryan, III, USNR, Admiral Halsey's Story (New York: McGraw-Hill Book Company, Inc., 1947), p. 195.  Halsey and Bryan, Halsey's Story, p. 195.
Most of the other senior Army and Navy officers on duty in the Pacific also favored the Luzon-first strategy and advocated bypassing Formosa. Lt. Gen. Robert C. Richardson, commanding U.S. Army Forces, Pacific Ocean Areas, strongly advised against Formosa. So, too, did MacArthur's air commander, Lt. Gen. George C. Kenney, and the Southwest Pacific Area's naval commander, Vice Adm. Thomas C. Kinkaid. But among the Joint Chiefs of Staff during the summer and early fall of 1944 only Admiral William D. Leahy, the President's Chief of Staff, favored going to Luzon instead of Formosa, and this stand represented a reversal of Leahy's earlier thinking on the subject. 
It is noteworthy that, with the possible exception of Nimitz, the ranking Army and Navy commanders in the Pacific-the men responsible for executing or supporting the operation-were opposed to the seizure of Formosa. In general, they favored a program calling for the capture of Luzon and a subsequent jump to Okinawa or Japan. In the face of this opinion of commanders on the spot, the consensus of most high-ranking Army and Navy planners in Washington-with Leahy and General Somervell as outstanding exceptions-was that the Formosa-first course of action was strategically the sounder and, therefore, the most desirable course for the Allies to follow in the western Pacific.
The Washington planners, however, had to give careful consideration to many factors other than ideal strategy. Study of these factors brought the Luzon versus Formosa debate to a climax in late September 1944.
Tactical and Logistical Problems
Perhaps the most influential event helping to precipitate the climax was a drastic change in the target date for the initial invasion of the Philippines. Until mid-September 1944, General MacArthur's plans had called for the first entry into the Philippines to take place in southeastern Mindanao on 15 November, while the major assault into the archipelago would occur at Leyte on 20 December. On 15 September, with the approval of the Joint Chiefs of Staff, MacArthur canceled preliminary Mindanao operations in favor of a direct jump from the Palaus-Morotai line to Leyte on 20 October. 
Soon after this change of schedule, MacArthur informed the Joint Chiefs that he could push on from Leyte to Luzon on 20 December, two months earlier than the date currently under consideration for an attack on either Luzon or Formosa. This new plan, MacArthur
 Ritchie Notes for Marshall; George C. Kenney, General Kenney Reports, A Personal History of the Pacific War, (New York: Duell, Sloan and Pearce, 1949), p. 371; Leahy, I Was There, p. 259; Rad, Richardson to Marshall, R-28617, 22 Aug 44, CM-IN 19958.  For the events leading up to this change in plans, see Cannon, Leyte: The Return to the Philippines, Ch. I.
suggested, would permit the Allies to execute the Formosa operation on the date already selected, but, he went on, the prior seizure of Luzon would render unnecessary the occupation of Formosa. 
MacArthur's new schedule contained much to recommend it to the Joint Chiefs of Staff. His proposed sequence of operations-Leyte on 20 October, Luzon on 20 December, and Formosa, possibly, on 20 February 1945-would permit the Allies to maintain steady pressure against the Japanese. Should the Allies drop Luzon out of the sequence, the Japanese would have ample time to realign their defenses during the interval between the Leyte and Formosa operations. Moreover, dropping out Luzon could in no way accelerate the advance to Formosa-logistical problems would make it impossible for the Allies to mount an invasion of Formosa under any circumstances before late February 1945.
While MacArthur's proposals were gaining some favor in Washington, especially among Army planners, Nimitz' proposals for advancing to Formosa and the south China coast were losing ground.  Plans developed in Washington had long called for the seizure of all Formosa, after which amphibious forces would strike on westward to secure a port on the mainland. But Nimitz' latest plans provided for simultaneous assaults on southern Formosa and in the Amoy area of the China coast. Nimitz proposed to occupy the bulk of Formosa only if such a step proved necessary and feasible after he had established a firm bridgehead at Amoy.
Army planners quickly decided that Nimitz' new plans possessed major drawbacks. The Japanese would hardly allow Allied forces to sit unmolested in southern Formosa. Instead, the Japanese would mount strong counterattacks from northern Formosa with troops already on the island and with reinforcements staged in from China. Occupying and defending one beachhead on southern Formosa and another at Amoy would involve problems far different from those the Allies had encountered previously in the Pacific. So far during the war, the Japanese had usually been hard put to move air and ground reinforcements against the island perimeters Allied amphibious task forces had seized. In the southern Formosa-Amoy area, on the other hand, the Allies would not have the protection of distance from
 Rad, MacArthur to Marshall, C-18103, 21 Sep 44, CM-IN 19803.  The discussion of tactical and logistical problems in the remainder of this subsection is based generally upon: Minutes, JPS 162d, 165th, and 167th Mtgs, 10 and 28 Aug and 2 Sep 44; OPD, Draft Appreciation of a Plan of Campaign, n.d. [circa 1 Sep 44], and associated sources in OPD 381 Strategy Section Papers (4 Sep 44); Memo, Handy for Marshall, n.d. [circa 5 Sep 44], sub: Opns in the Western Pacific, and related documents in OPD ABC 384 Pacific (1-17-43) Sec. 5; Minutes, JCS 171st and 172d Mtgs, 1 and 5 Sep 44.
major Japanese bases they had enjoyed in earlier campaigns. The Allies did not have sufficient aircraft in the Pacific to keep neutralized all existing Japanese airfields within range of southern Formosa and Amoy. In addition, experience in the Pacific had demonstrated that Allied air and naval forces could not be expected to forestall all Japanese efforts to move strong reinforcements across the narrow strait between China and Formosa.
Having considered these factors, Army planners swung to the opinion that a southern Formosa-Amoy operation would be impractical. They believed that it would inevitably lead to protracted, costly campaigns to secure all Formosa and large areas of the adjacent China mainland as well. Major ground campaigns of such scope could only delay progress toward Japan and would prove an unacceptable drain upon Allied manpower resources.
Further study of manpower needed for the southern Formosa-Amoy operation revealed additional difficulties. Army intelligence estimates of Japanese strength in Formosa-Amoy region, for example, were far higher than those Nimitz' staff had produced. Army planners therefore believed that the southern Formosa-Amoy campaign would require many more combat units than Nimitz was planning to employ. Furthermore, according to various estimates made during September, Nimitz would lack from 77,000 to 200,000 of the service troops needed for the campaign he proposed.
Planners studied a number of suggestions for securing the necessary service forces. One thought, originating with the Navy, which was seeking ways to accelerate the Formosa target date, proposed taking service units from the Southwest Pacific area. But MacArthur's command was already short of service troops. To remove any from his area might jeopardize the success of the Leyte operation and would certainly immobilize his forces in the central Philippines until long after Nimitz had secured the southern Formosa-Amoy region. Although the southern Formosa-Amoy and Luzon operations would each require about the same number of U.S. combat troops in the assault phase, MacArthur could count upon hundreds of thousands of loyal Filipinos to augment both his service and his combat strength. No similar source of friendly manpower would be available on Formosa.
By mid-September 1944 so few service units were available in the United States that the only way Army planners could see to solve the service troop shortage for Nimitz' proposed operation was to await redeployment from Europe. Army planners and the Joint Logistic Committee both estimated that Nimitz could launch the southern Formosa-Amoy campaign even as early as 1 March 1945 only if the war in Europe ended by 1 November 1944, thereby permitting timely
redeployment of service units to the Pacific. And even if the Allies could effect such redeployment from Europe, logistical planners still felt that Nimitz would be unable to move against Formosa by 1 March 1945 unless the Joint Chiefs of Staff immediately decided to cancel the Luzon operation, thus providing for an early and unbroken build-up of the resources required to execute Nimitz' campaign. On the other hand, the logistical experts were convinced that MacArthur could move to Luzon before the end of 1944 regardless of developments in Europe. Army planners, not as optimistic as they had been a few months earlier about an early end to the war in Europe, pointed out that it would be unsound to schedule the southern Formosa-Amoy operation on the presumption of a German collapse by 1 November 1944. Events were to prove this argument sound.
Army planners saw other combined logistical-tactical disadvantages in Nimitz' plan. They believed, for instance, that the campaign would tie down so many troops, ships, landing craft, and planes that an invasion of Luzon, assuming Formosa came first, could not take place until November 1945. By the same token any other major step toward Japan, such as the seizure of Okinawa, would be equally delayed. A hiatus of this length would be unacceptable for tactical reasons alone. In addition, the Luzon-first course, it appeared, would be far safer logistically than the southern Formosa-Amoy undertaking. As Army Service Forces planners pointed out, the Allied lines of communication to Luzon would be shorter and easier to protect than those to Formosa. The logisticians predicted that the Allies would find it especially difficult to safeguard the lines of communication to Formosa if Luzon remained in Japanese hands.
Other aspects of the logistical problems attained disturbing overtones. Admiral Leahy, for example, believed that although the Formosa-first course of action might ultimately hasten the end of the war in the Pacific, capturing Luzon and bypassing Formosa would prove far cheaper in terms of lives and other resources. By mid-September he, as well as most Army planners, was favoring what promised to be the longer course at the lesser cost. General MacArthur, meanwhile, expressed the opinion that the Formosa-first strategy would cost not only more lives but also more time. He was prepared to guarantee to the Joint Chiefs that he could secure the most strategically important area of Luzon-the Central Plains-Manila Bay region-within four to six weeks after initial landings on the island.
General Marshall also began to show misgivings about the cost of the southern Formosa-Amoy operation vis-a-vis Luzon, although he remained convinced that the Formosa-first course was strategically the more desirable. Admiral Nimitz expressed no strong opinion on
the relative cost of the two campaigns, but, "backing" into the problem, stated that the occupation of Luzon after Formosa need not delay the pace of the war in the Pacific. If Formosa came first, Nimitz pointed out, MacArthur's task on Luzon would be considerably eased and, presumably, less costly. Admiral King, however, declared himself convinced that the Formosa-first course would save time and, therefore, reduce casualties over the long run. By late September 1944 King alone among the upper-level planners seems to have retained a strong conviction along these lines.
While the discussions over tactical and logistical problems continued in Washington, the Allied position in China had been steadily deteriorating. In mid-September General Joseph W. Stilwell, commanding U.S. Army forces in China, Burma, and India, and Allied Chief of Staff to Generalissimo Chiang Kai-shek, reported to the Joint Chiefs that Japanese offensives in eastern and southeastern China were overrunning the last air bases from which the China-based U.S. Fourteenth Air Force could effectively support invasions of either Luzon or Formosa. Chiang's armies were unable to either hold or recapture the air bases. 
This news had an obvious impact upon the thinking of both the ground and the air planners in Washington. The Army Air Forces had intended to expand its airfields in eastern China as staging bases for B-29's flying against targets in Japan, Korea, Manchuria, and Formosa, and to base on these fields much of the tactical bombardment preceding the actual invasion of Japan. The east China fields now appeared irretrievably lost, and the Allies could not afford to expend the manpower necessary to retake and hold them. The need for seizure and development of a port on the China coast was therefore deprived of much of its urgency since the Allies had needed such a port primarily to open a good supply route into China for the development of air bases. By the same token, one of the principal reasons for seizing Formosa-to secure a steppingstone to the China coast-became much less compelling.
This line of thinking forced naval planners to reconsider the southern Formosa-Amoy plan. To most Navy planners a move to Formosa without the concomitant seizure of a mainland port would prove unsound, because Formosa lacked the anchorages and ports required for the large fleet and logistical bases the Allies needed in the western Pacific. Inevitably the question arose: If it was no longer feasible or desirable to seize and develop a port on the south China coast, was
 Rad, Stilwell to Marshall and MacArthur, CFBX-22674, 16 Sep 44. CM- IN 15768. See also, Charles F. Romanus and Riley Sunderland, Stilwell's Command Problem. (Washington, 1956), in UNITED STATES ARMY IN WORLD WAR II.
it feasible or desirable to occupy any part of Formosa. Since early September 1944 Army planners had been answering that question with an emphatic "no." 
The loss of existing and potential air base sites in eastern China, together with the limitations inherent in Nimitz' plans to occupy only southern Formosa, weighed heavily with Army Air Forces planners. There was no question but that B-29'S could operate more effectively against Japan from northern Formosa than they could from northern Luzon, the Mariana Islands, or western China, but the big bombers could accomplish little more from southern Formosa than they could from the other base areas. Indeed, Saipan and Tinian in the Marianas lay closer to Tokyo than Nimitz' proposed base area in southern Formosa, and the two islands of the Marianas were secure from Japanese air attack. Even northern Luzon, some 200 miles farther from Tokyo than southern Formosa, had some advantages over southern Formosa-it had more room for B-29 fields and as safer from air attack. Finally, assuming that Nimitz could meet the most optimistic target date for the invasion of southern Formosa-1 March 1945-B-29'S could not begin operations from that island until the late spring or early summer. The Army Air Forces was already planing to initiate B-29 operations from the Marianas before the end of 1944. In brief, by mid-September, the Army Air Forces had lost interest in Formosa and had begun to see eye to eye With other Army elements on the disadvantages and drawbacks of the southern Formosa-Amoy scheme.
An obvious political consideration may have had a bearing on the ultimate decision in the Luzon versus Formosa debate. General Mac-Arthur's argument that it would be disastrous to United States prestige to bypass any part of the Philippines could not be dismissed. Perhaps more important, Admiral Leahy took the same point of view. By virtue of his intimate contact with President Roosevelt, it must be presumed that his colleagues of the Joint Chiefs of Staff gave Leahy's opinion careful consideration.
Whatever the political implication involved, the Formosa versus Luzon question was decided primarily upon its military merits. By the end of September 1944 almost all the military considerations-especially the closely interrelated logistical problems concerning troops and timing-had weighted the scales heavily in favor of seizing Luzon,
 Memo, Hull for Handy, 2 Sep 44, sub Pacific Strategy; OPD, Draft Appreciation of a Plan of Campaign, n.d. [circa 1 Sep 44], both, with associated sources, in OPD 384 Pacific (1-17-43) Sec. 5; Minutes, JCS 172d Mtg, 5 Sep 44.
bypassing Formosa, forgetting about a port on the China coast, and jumping on to Okinawa. Admiral King was the only member of the Joint Chiefs of Staff, if not the only prominent military figure as well, who still maintained a strong stand in favor of bypassing Luzon and executing the southern Formosa-Amoy operation.
Realizing that the military and political factors had undermined his position, King took a new, negative tack in the debate by raising objections to the Luzon operation per se. He argued that the Luzon campaign as MacArthur had planned it would tie up all the Pacific Fleet's fast carrier task forces for at least six weeks for the purposes of protecting the Luzon beachhead and Luzon-bound convoys and neutralizing Japanese air power on both Luzon and Formosa. To pin down the carriers for so long would be unsound, King averred, and he therefore declared MacArthur's plan unacceptable to the U.S. Navy. 
Alerted by his deputy chief of Staff (Maj. Gen. Richard J. Marshall, then in Washington on official business), General MacArthur was able to provide Army planners with ammunition to counter King's last-ditch arguments.  MacArthur informed the Joint Chiefs that his only requirement for carriers after the initial assault on Luzon would be for a small group of escort carriers to remain off the island for a few days to provide support for ground operations until his engineers could ready a field for land-based planes at the invasion beaches. MacArthur continued by pointing out that only the first assault convoys would be routed through dangerous waters north of Luzon and consequently require protection from the fast carrier task forces. Resupply and reinforcement convoys would come through the central Philippines under an umbrella of land-based aircraft from Mindoro Island, south of Luzon, and would require no carrier-based air cover. Thus, MacArthur declared he would have no long-term requirement for the fast carrier task forces, which he could quickly release so that Nimitz could employ them elsewhere. MacArthur concluded with the counterargument that the fast carriers would be tied down to a specific area much longer during the proposed southern Formosa-Amoy operation, especially if Luzon remained in Japanese hands, than would be the case for the Luzon invasion. 
 Memo, King for Marshall, 23 Sep 44, no sub, in OPD ABC 384 Pacific (1-17 43) Sec. 5.  Rads, R. J. Marshall to MacArthur, 26 Sep 44, CM-OUT 37000 and 37001. The first radio informed MacArthur of the nature of King's arguments, told MacArthur what Army planners needed to counter King's objections, and cautioned MacArthur to make no reference to the first radio in replying to the second. The second radio, signed by R. J. Marshall, was actually a formal request for information sent by the Joint Chiefs of Staff to MacArthur.  Rad, MacArthur to Marshall, C-18496, 28 Sep 44, CM-IN 26358.
This exchange took much of the wind out of King's sails. Next, Admiral Nimitz withdrew whatever support he was still giving the Formosa plan. He had concluded that sufficient troops could not be made available for him to execute the southern Formosa-Amoy campaign within the foreseeable future. Accordingly, at the end of September, he threw the weight of his opinion behind the Luzon operation, proposing that plans to seize Formosa be at least temporarily dropped. Simultaneously, Nimitz presented for Admiral King's consideration a planned series of operations designed to maintain steady pressure against the Japanese and carry Allied forces speedily on toward Japan: MacArthur's forces would initiate the Luzon campaign on 20 December 1944; Central Pacific forces would move against Iwo Jima, in the Volcano Islands some 650 miles south of Tokyo, late in January 1945; and the Central Pacific would next attack Okinawa, 850 miles southwest of Tokyo, and other targets in the Ryukyu Islands, beginning on 1 March 1945. 
King accepted Nimitz' recommendations, with one last reservation. King felt that the hazards involved in routing the Luzon assault convoys into the waters between Luzon and Formosa were so great that approval for such action should come directly from the Joint Chiefs of Staff. He raised similar objections to plans for having the Pacific Fleet's fast carrier task forces operate in the same restricted waters. The other three members of the Joint Chiefs of Staff, however, agreed to leave the decision on these problems up to Nimitz and MacArthur, a settlement that King finally accepted. 
After King's eleventh-hour change of position, the Joint Chiefs were able to attain the unanimity that their major strategic decisions required. On 3 October 1944 they directed General MacArthur to launch the invasion of Luzon on or about 20 December and instructed Admiral Nimitz to execute the Iwo Jima and Okinawa operations on the dates he had proposed. Nimitz would provide naval cover and support, including fast and escort carriers, for the invasion of Luzon; MacArthur would provide Nimitz with as much air support as he could from Luzon for the attack on Okinawa. The two commanders would co-ordinate their plans with those of B-29 units in the Pacific
 Conf Notes, Rear Adm Forrest P. Sherman (Nimitz' planning chief) and Rear Adm Charles M. Cooke (King's deputy chief of staff), 27 Sep 44, in OPD Exec Files 17, Binder 3; JCS 713/18, 2 Oct 44. Future Opns in the Pacific [a memo by King to the JCS], in OPD 384 Pacific (1-17-43), Sec. 5. Nimitz personally presented his views to King at a secret conference in San Francisco over the weekend of 29 September-1 October 1944.  JCS 713/18, 2 Oct 44; Rad, JCS to MacArthur, Nimitz, and Stillwell, 3 Oct 44, CM-OUT 40782.
and India and with the plans of General Stilwell and the Fourteenth Air Force in China. 
The Joint Chiefs of Staff did not formally cancel the Formosa operation. Instead, they left in abeyance a final decision on the seizure of that island, but thereafter the occupation of Formosa as an operation of World War II never came up for serious consideration at the higher levels of Washington planning councils.
The Joint Chiefs had not reached their decision to take Luzon, bypass Formosa, and, in effect, substitute Okinawa for Formosa, either lightly or easily. From the beginning of the Luzon versus Formosa debate they had believed the seizure of Formosa and a port on the south China coast-bypassing Luzon-to be the best strategy the Allies could follow in the western Pacific. In the end, however, the Joint Chiefs had had to face the fact that the Allies could not assemble the resources required to execute that strategy, at least until after the end of the war in Europe. They could not seriously consider delaying the progress of the war in the Pacific until Germany collapsed. In the last analysis then, logistical considerations alone would have forced the Joint Chiefs to the decision they reached in favor of Luzon, although other military realities, and possibly political factors as well, had some influence upon the outcome of strategic planning for operations in the western Pacific.
For the Allied forces of the Pacific theaters, the Joint Chiefs' directive of 3 October 1944 ended months of uncertainty. The die was cast. Luzon would be taken; Formosa would be bypassed. United States forces would recapture the entire Philippines Archipelago in a consecutive series of advances, just as General MacArthur had been planning ever since he had left Corregidor in March 1942.
 Ibid. The B-29's operated under the direct control of the JCS, with General Arnold acting as the JCS executive agent.
ROBERT Ross SMITH, Historian with OCMH since 1947. B.A. and M.A. in history, Duke University. Historical Officer, General Headquarters, Southwest Pacific Area and U.S. Army Forces, Pacific, World War II. Major, Infantry, USAR. Author: The Approach to the Philippines (Washington, 1953), Triumph in the Philippines (in preparation), and Southern France and Alsace (in preparation), UNITED STATES ARMY IN WORLD WAR II.